Split (1)

train · ~9.11M rows (showing the first 93.5k)

identifier stringlengths 11 32	pdf_url stringlengths 17 4.62k ⌀	lang stringclasses 120 values	error stringclasses 1 value	title stringlengths 2 500 ⌀	source_name stringlengths 1 435 ⌀	publication_year float64 1.9k 2.02k	license stringclasses 3 values	word_count int64 0 1.64M	text stringlengths 1 9.75M
https://openalex.org/W2114095858	https://curis.ku.dk/ws/files/186525668/cp_11_153_2015.pdf	English	null	The WAIS Divide deep ice core WD2014 chronology – Part 1: Methane synchronization (68–31 ka BP) and the gas age–ice age difference	Climate of the past	2,015	cc-by	22,268	Citation for published version (APA): Buizert, C., Cuffey, K. M., Severinghaus, J. P., Baggenstos, D., Fudge, T. J., Steig, E. J., Markle, B. R., Winstrup, M., Rhodes, R. H., Brook, E. J., Sowers, T. A., Clow, G. D., Cheng, H., Edwards, R. L., Sigl, M., McConnell, J. R., & Taylor, K. C. (2015). The WAIS Divide deep ice core WD2014 chronology - Part 1: Methane synchronization (68-31 kaBP) and the gas age-ice age difference. Climate of the Past, 11(2), 153-173. https://doi.org/10.5194/cp-11-153-2015 university of copenhagen The WAIS Divide deep ice core WD2014 chronology - Part 1 Methane synchronization (68-31 kaBP) and the gas age-ice age difference Buizert, C.; Cuffey, K. M.; Severinghaus, J. P.; Baggenstos, D.; Fudge, T. J.; Steig, E. J.; Markle, B. R.; Winstrup, M.; Rhodes, R. H.; Brook, E. J.; Sowers, T. A.; Clow, G. D.; Cheng, H.; Edwards, R. L.; Sigl, M.; McConnell, J. R.; Taylor, K. C. P bli h d i university of copenhagen The WAIS Divide deep ice core WD2014 chronology - Part 1 Methane synchronization (68-31 kaBP) and the gas age-ice age difference Buizert, C.; Cuffey, K. M.; Severinghaus, J. P.; Baggenstos, D.; Fudge, T. J.; Steig, E. J.; Markle, B. R.; Winstrup, M.; Rhodes, R. H.; Brook, E. J.; Sowers, T. A.; Clow, G. D.; Cheng, H.; Edwards, R. L.; Sigl, M.; McConnell, J. R.; Taylor, K. C. The WAIS Divide deep ice core WD2014 chronology - Part 1 Methane synchronization (68-31 kaBP) and the gas age-ice age difference Buizert, C.; Cuffey, K. M.; Severinghaus, J. P.; Baggenstos, D.; Fudge, T. J.; Steig, E. J.; Markle, B. R.; Winstrup, M.; Rhodes, R. H.; Brook, E. J.; Sowers, T. A.; Clow, G. D.; Cheng, H.; Edwards, R. L.; Sigl, M.; McConnell, J. R.; Taylor, K. C. Published in: Climate of the Past Published in: Climate of the Past DOI: 10.5194/cp-11-153-2015 DOI: 10.5194/cp-11-153-2015 Publication date: 2015 Document version Publisher's PDF, also known as Version of record Document version Publisher's PDF, also known as Version of record Download date: 24. Oct. 2024 university of copenhagen The WAIS Divide deep ice core WD2014 chronology - Part 1 Methane synchronization (68-31 kaBP) and the gas age-ice age difference Buizert, C.; Cuffey, K. M.; Severinghaus, J. P.; Baggenstos, D.; Fudge, T. J.; Steig, E. J.; Markle, B. R.; Winstrup, M.; Rhodes, R. H.; Brook, E. J.; Sowers, T. A.; Clow, G. D.; Cheng, H.; Edwards, R. L.; Sigl, M.; McConnell, J. R.; Taylor, K. C. university of copenhagen The WAIS Divide deep ice core WD2014 chronology - Part 1 Methane synchronization (68-31 kaBP) and the gas age-ice age difference Buizert, C.; Cuffey, K. M.; Severinghaus, J. P.; Baggenstos, D.; Fudge, T. J.; Steig, E. J.; Markle, B. R.; Winstrup, M.; Rhodes, R. H.; Brook, E. J.; Sowers, T. A.; Clow, G. D.; Cheng, H.; Edwards, R. L.; Sigl, M.; McConnell, J. R.; Taylor, K. C. The WAIS Divide deep ice core WD2014 chronology – Part 1: Methane synchronization (68–31 kaBP) and the gas age–ice age difference Past Discuss.: 28 August 2014 Revised: 24 December 2014 – Accepted: 7 January 2015 – Published: 5 February 2015 Revised: 24 December 2014 – Accepted: 7 January 2015 – Published: 5 February 2015 Revised: 24 December 2014 – Accepted: 7 January 2015 – Published: 5 February 2015 Abstract. The West Antarctic Ice Sheet Divide (WAIS Di- vide, WD) ice core is a newly drilled, high-accumulation deep ice core that provides Antarctic climate records of the past ∼68 ka at unprecedented temporal resolution. The upper 2850 m (back to 31.2 ka BP) have been dated using annual-layer counting. Here we present a chronology for the deep part of the core (67.8–31.2 ka BP), which is based on stratigraphic matching to annual-layer-counted Greenland ice cores using globally well-mixed atmospheric methane. We calculate the WD gas age–ice age difference (1age) us- ing a combination of ﬁrn densiﬁcation modeling, ice-ﬂow modeling, and a data set of δ15N-N2, a proxy for past ﬁrn column thickness. The largest 1age at WD occurs during the Last Glacial Maximum, and is 525 ± 120 years. Internally consistent solutions can be found only when assuming lit- tle to no inﬂuence of impurity content on densiﬁcation rates, contrary to a recently proposed hypothesis. We synchro- nize the WD chronology to a linearly scaled version of the layer-counted Greenland Ice Core Chronology (GICC05), which brings the age of Dansgaard–Oeschger (DO) events into agreement with the U / Th absolutely dated Hulu Cave speleothem record. The small 1age at WD provides valuable opportunities to investigate the timing of atmospheric green- house gas variations relative to Antarctic climate, as well as the interhemispheric phasing of the “bipolar seesaw”. Document license: CC BY Citation for published version (APA): Buizert, C., Cuffey, K. M., Severinghaus, J. P., Baggenstos, D., Fudge, T. J., Steig, E. J., Markle, B. R., Winstrup, M., Rhodes, R. H., Brook, E. J., Sowers, T. A., Clow, G. D., Cheng, H., Edwards, R. L., Sigl, M., McConnell, J. R., & Taylor, K. C. (2015). The WAIS Divide deep ice core WD2014 chronology - Part 1: Methane synchronization (68-31 kaBP) and the gas age-ice age difference. Climate of the Past, 11(2), 153-173. https://doi.org/10.5194/cp-11-153-2015 Citation for published version (APA): Buizert, C., Cuffey, K. M., Severinghaus, J. P., Baggenstos, D., Fudge, T. J., Steig, E. J., Markle, B. R., Winstrup, M., Rhodes, R. H., Brook, E. J., Sowers, T. A., Clow, G. D., Cheng, H., Edwards, R. L., Sigl, M., McConnell, J. R., & Taylor, K. C. (2015). The WAIS Divide deep ice core WD2014 chronology - Part 1: Methane synchronization (68-31 kaBP) and the gas age-ice age difference. Climate of the Past, 11(2), 153-173. https://doi.org/10.5194/cp-11-153-2015 Download date: 24. Oct. 2024 Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ doi:10.5194/cp-11-153-2015 © Author(s) 2015. CC Attribution 3.0 License. Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ doi:10.5194/cp-11-153-2015 © Author(s) 2015. CC Attribution 3.0 License. 1 Introduction Deep ice cores from the polar regions provide high- resolution climate records of past atmospheric composition, aerosol loading and polar temperatures (e.g., NGRIP com- munity members, 2004; EPICA Community Members, 2006; Wolff et al., 2006; Ahn and Brook, 2008). Furthermore, the coring itself gives access to the ice sheet interior and bed, allowing investigation of glaciologically important processes such as ice deformation (Gundestrup et al., 1993), folding (NEEM community members, 2013), crystal fabric evolution The WAIS Divide deep ice core WD2014 chronology – Part 1: Methane synchronization (68–31 kaBP) and the gas age–ice age difference C. Buizert1, K. M. Cuffey2, J. P. Severinghaus3, D. Baggenstos3, T. J. Fudge4, E. J. Steig4, B. R. Markle4, M. Winstrup4, R. H. Rhodes1, E. J. Brook1, T. A. Sowers5, G. D. Clow6, H. Cheng7,8, R. L. Edwards8, M. Sigl9, J. R. McConnell9, and K. C. Taylor9 C. Buizert1, K. M. Cuffey2, J. P. Severinghaus3, D. Baggenstos3, T. J. Fudge4, E. J. Steig4, B. R. Markle4, M. Winstrup4, R. H. Rhodes1, E. J. Brook1, T. A. Sowers5, G. D. Clow6, H. Cheng7,8, R. L. Edwards8, M. Sigl9, J. R. McConnell9, and K. C. The WAIS Divide deep ice core WD2014 chronology – Part 1: Methane synchronization (68–31 kaBP) and the gas age–ice age difference Taylor9 1College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, Corvallis, OR 97331, USA 2Department of Geography, University of California, Berkeley, CA 94720, USA 3 1College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, Corvallis, OR 97331, USA 2Department of Geography, University of California, Berkeley, CA 94720, USA 3 cripps Institution of Oceanography, University of California, San Diego, La Jolla, CA 92093, USA Scripps Institution of Oceanography, University of California, San Diego, La Jolla, CA 92093, USA 4Quaternary Research Center and Department of Earth and Space Sciences, University of Washington, Seattle WA 98195, USA 4Quaternary Research Center and Department of Earth and Space Sciences, University of Washington, Seattle, WA 98195, USA 4Quaternary Research Center and Department of Earth and Space Sciences, University of Washington, Seattle, WA 98195, USA 4Quaternary Research Center and Department of Earth and Space Sciences, University of Washing WA 98195, USA 5Department of Geosciences and Earth and Environmental Systems Institute, Pennsylvania State U University Park, PA 16802, USA 5Department of Geosciences and Earth and Environmental Systems Institute, Pennsylvania State University, University Park, PA 16802, USA 5Department of Geosciences and Earth and Environmental Systems Institute, Pennsylvania State University, University Park, PA 16802, USA 6US Geological Survey, Boulder, CO 80309, USA nstitute of Global Environmental Change, Xi’an Jiaotong University, Xi’an 710049, China Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, USA 7Institute of Global Environmental Change, Xi’an Jiaotong University, Xi’an 710049, China 8Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, USA 9Desert Research Institute, Nevada System of Higher Education, Reno, NV 89512, USA 7Institute of Global Environmental Change, Xi’an Jiaotong University, Xi’an 710049, China 8Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, USA 9Desert Research Institute, Nevada System of Higher Education, Reno, NV 89512, USA Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, USA 9Desert Research Institute, Nevada System of Higher Education, Reno, NV 89512, USA 9Desert Research Institute, Nevada System of Higher Education, Reno, NV 89512, USA Correspondence to: C. Buizert (buizertc@science.oregonstate.edu) Correspondence to: C. Buizert (buizertc@science.oregonstate.edu) Received: 1 August 2014 – Published in Clim. Past Discuss.: 28 August 2014 Received: 1 August 2014 – Published in Clim. C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 154 (Gow et al., 1997), and geothermal heat ﬂow (Dahl-Jensen et al., 1998). Having a reliable ice core chronology (i.e., an age–depth relationship) is paramount for the interpretation of the climate records and comparison to marine and terrestrial paleoclimate archives. et al., 2010). This method is particularly suited to WD be- cause of the small gas age–ice age difference (1age, Sect. 3) and the high-resolution, high-precision CH4 record available (Sect. 2.1). The method has three main sources of uncer- tainty: (i) the age uncertainty in the records one synchro- nizes to, (ii) 1age of the ice core being dated, and (iii) the interpolation scheme used in between the CH4 tie points. We present several improvements over previous work that reduce and quantify these uncertainties: (i) we combine the layer- counted Greenland Ice Core Chronology (GICC05) and a re- cently reﬁned version of the U / Th-dated Hulu speleothem record (Edwards et al., 2015; Reimer et al., 2013; Southon et al., 2012) to obtain a more accurate estimate of the (ab- solute) ages of abrupt Dansgaard–Oeschger (DO) events (Sect. 4.4); (ii) we combine ﬁrn densiﬁcation modeling, ice- ﬂow modeling, a new WD δ15N-N2 data set that spans the entire core, and a Monte Carlo sensitivity study to obtain a re- liable 1age estimate (Sect. 3); and (iii) we compare four dif- ferent interpolation schemes to obtain an objective estimate of the interpolation uncertainty (Sect. 4.5). p The West Antarctic Ice Sheet Divide (WAIS Divide, WD) ice core (79.48◦S, 112.11◦W; 1766 m above sea level; −30 ◦C present-day mean annual temperature) was drilled and recovered to 3404 m depth (WAIS Divide Project Mem- bers, 2013). Drilling was stopped 50 m above the estimated bedrock depth to prevent contamination of the basal hy- drology. Due to high accumulation rates of 22 cmicea−1 at present and ∼10 cmicea−1 during the Last Glacial Maxi- mum (LGM), the WD core delivers climate records of un- precedented temporal resolution (Steig et al., 2013; Sigl et al., 2013) as well as gas records that are only minimally affected by diffusive smoothing in the ﬁrn column (Mischler et al., 2009; Mitchell et al., 2011, 2013; Marcott et al., 2014). C. Buizert et al.: The WAIS Divide chronology – Part 1 The combination of high accumulation rates and basal melt- ing at the WD site results in ice near the bed that is rela- tively young (∼68 ka) compared to cores drilled in central East Antarctica. This work is the ﬁrst part in a series of two papers de- scribing the WD2014 chronology for the WD core in de- tail. The second part describes the development of the annual layer count from both multi-parameter chemistry and elec- trical conductivity measurements. The WD2014 chronology is currently the recommended gas and ice timescale for the WD deep core, and as such it supersedes the previously pub- lished WDC06A-7 chronology (WAIS Divide Project Mem- bers, 2013). In WD, annual layers can be identiﬁed reliably for the upper 2850 m of the core, reaching back to 31.2 kaBP (thousands of years before present, with present deﬁned as 1950 CE). Below 2850 m depth an alternative dating strat- egy is needed. Several methods have been employed previ- ously at other deep ice core sites. First, orbital tuning via δO2 / N2 has been applied successfully to several Antarctic cores (Bender, 2002; Kawamura et al., 2007). However, an age span of only ∼3 precessional cycles in WD, in combi- nation with the low signal-to-noise ratio of δO2 / N2 data, makes this technique unsuitable for WD. The uncertainty in the orbital tuning is about one-fourth of a precessional cycle (∼5 ka), making it a relatively low-resolution dating tool. Second, in Greenland, ice-ﬂow modeling has been used to extend layer-counted chronologies (e.g., Johnsen et al., 2001; Wolff et al., 2010). This method requires assumptions about past accumulation rates, ice ﬂow, and ice sheet elevation. Particularly for the oldest WD ice, the resulting uncertainty would be substantial. Third, several radiometric techniques have been proposed to date ancient ice. Radiocarbon (14C) dating of atmospheric CO2 trapped in the ice is unsuitable as it suffers from in situ cosmogenic production in the ﬁrn (Lal et al., 1990), and the oldest WD ice dates beyond the reach of 14C dating. Other absolute (radiometric) dating techniques, such as recoil 234U dating (Aciego et al., 2011), 81Kr dating (Buizert et al., 2014a), or atmospheric 40Ar buildup (Bender et al., 2008), currently suffer from uncertainties that are too large (≥20 ka) to make them applicable at WD. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. 2.2 Firn densiﬁcation model description Air exchange with the overlying atmosphere keeps the inter- stitial air in the porous ﬁrn layer younger than the surround- ing ice matrix, resulting in an age difference between po- lar ice and the gas bubbles it contains, commonly referred to as 1age (Schwander and Stauffer, 1984). Here we use a coupled ﬁrn–densiﬁcation–heat–diffusion model to calcu- late 1age back in time (Barnola et al., 1991; Goujon et al., 2003; Schwander et al., 1997; Rasmussen et al., 2013), con- strained by measurements of δ15N of N2, a proxy for past ﬁrn column thickness (Sowers et al., 1992). The model is based on a dynamical description of the Herron–Langway model formulated in terms of overburden load (Herron and Lang- way, 1980), which is solved in a Lagrangian reference frame. This model has been applied previously to the Greenland NEEM, NGRIP, and GISP2 cores (Rasmussen et al., 2013; Seierstad et al., 2015; Buizert et al., 2014b), where it gives a good agreement to the Goujon densiﬁcation model (Ras- mussen et al., 2013; Goujon et al., 2003). The model allows for the inclusion of softening of ﬁrn in response to impurity loading (Horhold et al., 2012), following the mathematical description of Freitag et al. (2013a). The equations govern- ing the model densiﬁcation rates are given in Appendix A. This model has been applied previously to the Greenland NEEM, NGRIP, and GISP2 cores (Rasmussen et al., 2013; Seierstad et al., 2015; Buizert et al., 2014b), where it gives a good agreement to the Goujon densiﬁcation model (Ras- mussen et al., 2013; Goujon et al., 2003). The model allows for the inclusion of softening of ﬁrn in response to impurity loading (Horhold et al., 2012), following the mathematical description of Freitag et al. (2013a). The equations govern- ing the model densiﬁcation rates are given in Appendix A. Measurements of δ15N. Atmospheric N2 isotopic composi- tion (δ15N) was measured at Scripps Institution of Oceanog- raphy, University of California. Air was extracted from ∼ 12 gram ice samples using a melt–refreeze technique, and collected in stainless steel tubes at liquid-He temperature. δ15N was analyzed using conventional dual-inlet isotope ra- tio mass spectrometry (IRMS) on a Thermo Finnigan Delta V mass spectrometer. Results are normalized to La Jolla (Cal- ifornia, USA) air, and routine analytical corrections are ap- plied (Sowers et al., 1989; Petrenko et al., 2006; Severing- haus et al., 2009). 2.2 Firn densiﬁcation model description Duplicates were not run for most δ15N data in this study, but the pooled standard deviations of Holocene WD δ15N data sets with duplicate analyses are 0.003 ‰ (Orsi, 2013). We conservatively adopt an analytical uncer- tainty of 0.005 ‰ for this data set to allow for other sources of error. g g pp The model uses a 2-year time step and 0.5 m depth reso- lution down to 1000 m, the lower model boundary. A thick model domain is needed because of the long thermal mem- ory of the ice sheet. At WD, downward advection of cold surface ice is strong due to the relatively high accumulation rates, and the geothermal gradient does not penetrate the ﬁrn column (Cuffey and Paterson, 2010). We further use a lock- in density that equals the mean close-off density (Martinerie et al., 1994) minus 17.5 kgm−3 (as in Blunier and Schwan- der, 2000) and an empirical parameterization of lock-in gas age based on ﬁrn air measurements from 10 sites (Buizert et al., 2012, 2013). Measurements of [Ca]. Ca concentrations in the ice were measured at the Ultra Trace Chemistry Laboratory at the Desert Research Institute via continuous ﬂow analysis. Lon- gitudinal samples of ice (approximately 100cm × 3.3cm × 3.3 cm) were melted continuously on a melter head that di- vides the meltwater into three parallel streams. Elemental measurements were made on meltwater from the innermost part of the core with ultra-pure nitric acid added to the melt stream immediately after the melter head; potentially con- taminated water from the outer part of the ice is discarded. Elemental analysis of the innermost meltwater stream is per- formed in parallel on two inductively coupled plasma mass spectrometers (ICPMS), each measuring a different set of elements; some elements were analyzed on both. The dual ICPMS setup allows for measurement of a broad range of 30 elements and data quality control (McConnell et al., 2002, 2007). Precision of the Ca measurements in WD glacial ice is estimated to be ±3 %, with a lower detection limit of 0.15 ngg−1. Continuous Ca and CH4 measurements are done on the same ice, and are exactly co-registered in depth. We furthermore use the steady-state version of the Herron–Langway model (Herron and Langway, 1980) in per- forming sensitivity studies (Sect. 2.2 Firn densiﬁcation model description 3.2) and the dynamical Ar- naud model (Arnaud et al., 2000; Goujon et al., 2003) to val- idate our 1age solution. C. Buizert et al.: The WAIS Divide chronology – Part 1 155 Project Members, 2013). The second data set is a contin- uous CH4 record measured by coupling a laser spectrom- eter to a continuous ﬂow analysis setup (Stowasser et al., 2012; Rhodes et al., 2013; Chappellaz et al., 2013), and was measured jointly by Oregon State University and the Desert Research Institute (Rhodes et al., 2015). The continu- ous data set is used to identify the abrupt DO transitions, as it provides better temporal resolution and analytical precision. Both records are reported on the NOAA04 scale (Dlugo- kencky et al., 2005). Analytical precision in the CH4 data (2σ pooled standard deviation) is around 3.2 and 14 ppb for the discrete data from Oregon State University and Pennsylvania State University, respectively, and 3 to 8 ppb for the continu- ous CH4 data, depending on the analyzer used (Rhodes et al., 2015); the 14 ppb stated for the PSU discrete data may be an overestimation, as depth-adjacent (rather than true replicate) samples were used in the analysis. 2.1 Data description Measurements of water stable isotopes. Water isotopic com- position (δ18O and δD = δ2H) was measured at IsoLab, Uni- versity of Washington. Procedures for the deep section of the core are identical to those used for the upper part of the core reported in WAIS Divide Project Members (2013) and Steig et al. (2013). Measurements were made at 0.25 to 0.5 m depth resolution using laser spectroscopy (Picarro L2120-i water isotope analyzer), and normalized to VSMOW-SLAP (Vi- enna Standard Mean Ocean Water – Standard Light Antarctic Precipitation). The precision of the measurements is better than 0.1 and 0.8 ‰ for δ18O and δD, respectively. Measurements of CH4. Two CH4 data sets were used for WD. The ﬁrst is from discrete ice samples, and was mea- sured jointly at Pennsylvania State University (0–68 ka, 0.5– 2 m resolution) and Oregon State University (11.4–24.8 ka, 1–2 m resolution). Air was extracted from ∼50 g ice samples using a melt–refreeze technique, and analyzed on a standard gas chromatograph equipped with a ﬂame-ionization detec- tor. Corrections for solubility, blank size and gravitational enrichment are applied (Mitchell et al., 2011; WAIS Divide Instead, at WD we use stratigraphic matching to well- dated Greenland ice cores using globally well-mixed atmo- spheric methane (CH4) mixing ratios (Blunier et al., 1998; Blunier and Brook, 2001; Blunier et al., 2007; Petrenko et al., 2006; EPICA Community Members, 2006; Capron Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ 2.3 Temperature reconstruction and ice-ﬂow model 0 10 20 30 40 50 60 70 WD2014 age (ka BP) 0 10 20 30 40 50 60 70 −42 −40 −38 −36 −34 −32 −30 −28 Temperature ( oC) 0.1 0.15 0.2 0.25 0.3 Acc rate (m ice a−1) 0.3 0.35 0.4 0.45 δ15N (‰) 200 300 400 500 600 Δage (years) 0 0.2 0.4 0.6 0.8 1 Thinning fraction (a) (b) (c) (d) (e) Figure 1. Modeling 1age for WAIS Divide. (a) Past temperatures reconstructed from water δD, calibrated to the borehole temperature proﬁle. (b) Past accumulation rates as reconstructed by the ﬁrn densiﬁcation inverse model (red), and from the annual-layer count (black). (c) δ15N data (black dots) with densiﬁcation model output (green). (d) 1age calculated using the densiﬁcation model (orange) and using the Parrenin 1depth method (black) with constant 4 m thick convective zone and no correction for thermal δ15N fractionation. (e) Modeled thinning function from ice-ﬂow model (solid), and a simple Nye strain model for comparison (dashed); the Nye thinning function, which has a uniform strain rate as a function of depth, is given as fλ(z) = (H −z)/H with H the ice sheet thickness (Cuffey and Paterson, 2010, p. 616). temperatures in the ice sheet. A coupled 1-D ice-ﬂow–heat- diffusion model converts surface T (t) into a depth proﬁle for comparison to measured borehole temperatures. The 1- D ice-ﬂow model calculates the vertical ice motion, taking into account the surface snow accumulation, the variation in density with depth, and a prescribed history of ice thickness. Vertical motion is calculated by integrating a depth proﬁle of strain rate and adding a rate of basal melt. As in the model of Dansgaard and Johnsen (1969), the strain rate maintains a uniform value between the surface and a depth equal to 80 % of the ice thickness, and then varies linearly to some value at the base of the ice. This basal value is deﬁned by the “basal stretching parameter” fb, the ratio of strain rate at the base to strain rate in the upper 80 % of the ice column. The basal ice is melting, so part of the ice motion likely occurs as slid- ing. The along-ﬂow gradient in such sliding is unknown and thus so too is the parameter fb. 2.3 Temperature reconstruction and ice-ﬂow model Our temperature reconstruction (Fig. 1a) is based on water δD, a proxy for local vapor condensation temperature, cal- ibrated using a measured borehole temperature proﬁle (fol- lowing Cuffey et al., 1995; Cuffey and Clow, 1997) and, for the last 31.2 ka, adjusted iteratively to satisfy constraints on ﬁrn thickness provided by δ15N and by the observed layer thickness λ(z). Using δ18O rather than δD in the temper- ature reconstruction leads to differences that are negligibly small. This borehole temperature calibration approach is pos- sible at WD because the large ice thickness and relatively high accumulation rates help to preserve a memory of past www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 156 C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 156 C. Buizert et al.: The WAIS Divide chronology Part 1 0 10 20 30 40 50 60 70 WD2014 age (ka BP) 0 10 20 30 40 50 60 70 −42 −40 −38 −36 −34 −32 −30 −28 Temperature ( oC) 0.1 0.15 0.2 0.25 0.3 Acc rate (m ice a−1) 0.3 0.35 0.4 0.45 δ15N (‰) 200 300 400 500 600 Δage (years) 0 0.2 0.4 0.6 0.8 1 Thinning fraction (a) (b) (c) (d) (e) Figure 1. Modeling 1age for WAIS Divide. (a) Past temperatures reconstructed from water δD, calibrated to the borehole temperature proﬁle. (b) Past accumulation rates as reconstructed by the ﬁrn densiﬁcation inverse model (red), and from the annual-layer count (black). (c) δ15N data (black dots) with densiﬁcation model output (green). (d) 1age calculated using the densiﬁcation model (orange) and using the Parrenin 1depth method (black) with constant 4 m thick convective zone and no correction for thermal δ15N fractionation. (e) Modeled thinning function from ice-ﬂow model (solid), and a simple Nye strain model for comparison (dashed); the Nye thinning function, which has a uniform strain rate as a function of depth, is given as fλ(z) = (H −z)/H with H the ice sheet thickness (Cuffey and Paterson, 2010, p. 616). 3.1 The WD2014 1age reconstruction The ﬁrn densiﬁcation forward model uses past surface tem- perature T (t) and accumulation A(t) as model forcings, and provides 1age(t) and δ15N(t) as model output. For the past 31.2 ka, WD has an annual-layer-counted chronology; for this period the annual-layer thickness λ(z) provides a constraint on past accumulation rates via λ(z) = A(z) × fλ(z). WD accumulation reconstructed from λ(z) is plotted in black in Fig. 1b. p g Prior to 31.2 ka we have no such constraint on A(t), and an alternative approach is needed. We use the densiﬁcation model as an inverse model, where we ask the model to ﬁnd the A(t) history that minimizes the root-mean-square (rms) deviation between measured and modeled δ15N, given the T (t) forcing. The δ15N data and model ﬁt are shown in Fig. 1c, the A(t) history that optimizes the δ15N ﬁt is shown in Fig. 1b (red), and the modeled 1age is shown in Fig. 1c (orange). The optimal A(t) history is estimated in two steps. First, we make an initial estimate Ainit(t) for the past accumulation history. Second, we adjust the A(t) forcing by applying a smooth perturbation ξ(t) such that A(t) = [1 + ξ(t)] × Ainit(t); an automated algorithm is used to ﬁnd the curve ξ(t) that optimizes the model ﬁt to the δ15N data. For the last 31.2 ka we obtain a good agreement be- tween A obtained from λ(z) and the modeled fλ(z) (Fig. 1b, black) and A obtained from the inverse method (red). The so- lution we present here is therefore fully internally consistent, i.e., the A and T histories used in the ﬁrn densiﬁcation mod- eling are the same as those used in the ice-ﬂow modeling, and they provide a good ﬁt to both the δ15N data and borehole temperature data. WD does not suffer from the δ15N model– data mismatch that is commonly observed for East Antarctic cores during the glacial period (Landais et al., 2006; Capron et al., 2013). Figure 2. Reconstructing A(t) and 1age from δ15N: the choice of accumulation template. (a) Ainit(t) based on λ(z) from the annual- layer count (0–31.2 kaBP) and Clausius–Clapeyron scaling of wa- ter stable isotope data (34.2–68 ka) in orange; Ainit = 0.22 ma−1 in blue; and for comparison the ﬁnal A(t) solution (Fig. 1b) in grey. For the orange curve we have used the Nye thinning func- tion (Fig. 3.1 The WD2014 1age reconstruction 1e); the ﬁnal ice-ﬂow model optimizes the agreement be- tween A(t) obtained from λ(z) and from the inverse ﬁrn modeling approach. The WD2014 chronology uses the orange Ainit(t) sce- nario. (b) A(t) found in the inverse ﬁrn modeling approach using both Ainit(t) scenarios; color coding as in panel (a). The function ξ(t) is found as follows. We use control points at 1500-year inter- vals (blue dots); the algorithm has the freedom to change the value of ξ(t) at each of these points. In between the control points, ξ(t) is found via linear interpolation. (c) Modeled 1age using both Ainit(t) scenarios; color coding as in panel (a). strain is represented by the thinning function fλ(z) (Cuf- fey and Paterson, 2010), the ratio of annual-layer thickness at depth in the ice sheet to its original ice-equivalent thick- ness at the surface when deposited. The modeled thinning function is shown in Fig. 1e (solid line). In the deep part of the ice sheet, fλ(z) becomes increasingly uncertain as the unknown basal melt rate and fb become the dominant con- trols. Here we optimize the model by comparing accumu- lation rates derived from fλ(z) with those implied by a ﬁrn densiﬁcation model and the measured δ15N of N2 (Sect. 3.1). While this has little effect on the temperature history recon- struction, it provides an important constraint on calculated basal melt rate, an interesting quantity for ice dynamics stud- ies. Our analysis of basal melt rates and further details of the temperature optimization process and 1-D ﬂow modeling will be presented elsewhere. We base our Ainit values on λ(z) for the past 31.2 ka; prior to that we use the common assumption that A follows δ18O (i.e., Clausius–Clapeyron scaling); the ﬁt to the δ15N data is optimized for A = 24.2×exp[0.1263×δ18O]. To test the validity of the Clausius–Clapeyron assumption, we addi- tionally run the scenario Ainit(t) = 0.22 ma−1 (i.e., constant accumulation at present-day level). The A(t) and 1age re- constructed under both Ainit scenarios are similar at multi- millennial timescales (Fig. 2). In the layer-counted interval (< 31.2 kaBP), A obtained from λ(z) and δ18O is signiﬁ- cantly coherent at all timescales longer than 3000 years, but not at higher frequencies. This is equivalent to the variability resolved in the Ainit(t) = 0.22 ma−1 scenario above. We con- clude that the WD δ15N data support the idea that A follows δ18O on multi-millennial timescales. 2.3 Temperature reconstruction and ice-ﬂow model We overcome this problem by making both the current ice thickness and the basal melt rate free parameters when optimizing models with respect to measured borehole temperatures. Because the basal melt rate and fb affect the vertical velocities in similar fashion, the optimization constrains a combination of melt rate and fb that is tightly constrained by the measured temperatures. Thus we ﬁnd that varying fb through a large range, from 0.1 to 1.5, changes the reconstructed LGM temperature by less than 0.2 ◦C. Effects of the prescribed ice-thickness his- tory are likewise minor; assuming a 150 m thickness increase from the LGM to 15 ka changes the reconstructed LGM tem- perature by less than 0.2 ◦C compared to a constant thick- ness. Note that the 1-D ﬂow model used here is simpler than the one used by Cuffey and Clow (1997) in that it does not at- tempt to calculate changes in the shape of the strain rate pro- ﬁle; the unknown basal sliding motion at the WD site negates the usefulness of such an exercise. One output of the 1-D ﬂow model is the strain history of ice layers as a function of depth and time. The cumulative Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 0 10 20 30 40 50 60 70 WD2014 age (ka BP) 0 10 20 30 40 50 60 70 0.1 0.2 0.3 Ainit (m ice a−1) 0.1 0.2 0.3 Afinal (m ice a−1) 200 300 400 500 Δage (year) (a) (b) (c) Figure 2. Reconstructing A(t) and 1age from δ15N: the choice of accumulation template. (a) Ainit(t) based on λ(z) from the annual- layer count (0–31.2 kaBP) and Clausius–Clapeyron scaling of wa- ter stable isotope data (34.2–68 ka) in orange; Ainit = 0.22 ma−1 in blue; and for comparison the ﬁnal A(t) solution (Fig. 1b) in grey. For the orange curve we have used the Nye thinning func- tion (Fig. 1e); the ﬁnal ice-ﬂow model optimizes the agreement be- tween A(t) obtained from λ(z) and from the inverse ﬁrn modeling approach. The WD2014 chronology uses the orange Ainit(t) sce- nario. (b) A(t) found in the inverse ﬁrn modeling approach using both Ainit(t) scenarios; color coding as in panel (a). The function ξ(t) is found as follows. 2.3 Temperature reconstruction and ice-ﬂow model We use control points at 1500-year inter- vals (blue dots); the algorithm has the freedom to change the value of ξ(t) at each of these points. In between the control points, ξ(t) is found via linear interpolation. (c) Modeled 1age using both Ainit(t) scenarios; color coding as in panel (a). 157 C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 0 10 20 30 40 50 60 70 WD2014 age (ka BP) 0 10 20 30 40 50 60 70 0.1 0.2 0.3 Ainit (m ice a−1) 0.1 0.2 0.3 Afinal (m ice a−1) 200 300 400 500 Δage (year) (a) (b) (c) www.clim-past.net/11/153/2015/ 3.1 The WD2014 1age reconstruction However, there may not be a strong relationship at timescales less than a few thousand years, as is clear from the abrupt A increase around 12 ka seen in λ(z) that is not reﬂected in δ18O (Fig. 1a and b). For www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 158 C. Buizert et al.: The WAIS Divide chronology – Part 1 consistency between the upper and deeper part of the core we use the 1age values obtained with the inverse densiﬁcation model for the entire core. thickness). In both methods, whenever CZ values are selected that are smaller than 0 m, the CZ thickness is set to 0 m in- stead. For each of the 1000 model runs in the sensitivity study we randomly selected either of the two methods. Recently, another δ15N-based approach has been sug- gested that uses 1depth, rather than 1age, in reconstruct- ing gas chronologies (Parrenin et al., 2012). This method removes the dependence on T (t) and replaces this with a dependence on the thinning function fλ(z). Note that this method is very successful in the upper part of an ice core, where fλ(z) is well constrained, but not very reliable near the base, where fλ(z) is highly uncertain. Therefore, the ﬁrn densiﬁcation modeling approach should be considered to be more reliable at WD during marine isotope stages (MIS) 2 through 4. Results from the 1depth method are plotted in black in Fig. 1c, and generally show good agreement with the ﬁrn modeling approach. A notable exception is the 60– 65 ka interval, where the 1depth method overestimates the 1age due to the fact that we have to compress λ(z) strongly in order to ﬁt age constraints derived from DO 18 (Sect. 4.5). Surface density. In the WD2014 model run we use past surface densities (ρ0) as given by the parameterization of Kaspers et al. (2004). In the sensitivity study we add a con- stant offset to the Kaspers values, the magnitude of which is drawn from a Gaussian distribution of zero mean and a 2σ width of 60 kgm−3. This range corresponds to the full range of observed ρ0 variability in Kaspers et al. (2004). Past temperatures. Model temperature forcing is con- strained by δD and measured borehole temperatures. 3.1 The WD2014 1age reconstruction There is, however, a range to the solutions allowed by the bore- hole temperature and ice-ﬂow model; here we use the up- per and lower extremes of this range, determined by Monte Carlo analysis using uncertainties of input variables. The sce- narios were chosen to provide the maximum T range for the glacial period rather than for the Holocene, because we are interested in the uncertainty in the methane synchroniza- tion (68–31.2 kaBP). In the sensitivity study we use T (t) = Toptimal(t) + κ × 1T (t), where Toptimal is the forcing used in the WD2014 model run (Fig. 1a), 1T (t) is half the difference between the maximum-T and minimum-T scenarios, and κ is drawn from a Gaussian distribution of zero mean and unit 2σ width (giving 95 % probability that T (t) is within the ex- treme range identiﬁed from the borehole, Fig. 3a). 15 Last, we want to point out that the δ15N data support an early warming at WD, as reported recently (WAIS Di- vide Project Members, 2013). WD δ15N starts to decrease around 20.5 kaBP, suggesting a thinning of the ﬁrn column. The λ(z) (as derived from the layer count) shows that ac- cumulation did not change until 18 kaBP, at which point it started to increase (which would act to increase the ﬁrn thickness). The most plausible explanation for the δ15N de- crease around 20.5 kaBP is therefore an early onset of West Antarctic deglacial warming, in agreement with increasing δ18O around that time. The warming enhances the densiﬁca- tion rate of polar ﬁrn, thereby decreasing its thickness (e.g., Herron and Langway, 1980). δ15N uncertainty. We conservatively adopt an analytical uncertainty of 0.005 ‰ for this data set; in addition, the inter- pretation of δ15N in terms of ﬁrn thickness is subject to fur- ther uncertainty due to irregular ﬁrn layering and the stochas- tic nature of bubble trapping, as was observed for other atmo- spheric gases such as CH4 (Etheridge et al., 1992; Rhodes et al., 2013). For each run of the sensitivity study, we there- fore perturb each of the individual δ15N data points by adding an offset that is drawn from a Gaussian distribution of zero mean and a 2σ width of 0.015 ‰. www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 159 0 10 20 30 40 50 60 WD2014 Age (ka BP) 0 10 20 30 40 50 60 −44 −42 −40 −38 −36 −34 −32 −30 Temperature ( oC) 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Accumulation (m ice a−1) 100 200 300 400 500 600 700 800 Δage (years) (d) (e) (f) All ± 2σ ± 1σ 200 250 300 350 0 50 100 150 200 262 ± 50 year Frequency Δage (year) 250 300 350 400 450 50 100 150 200 351 ± 73 year Δage (year) 400 500 600 700 50 100 150 200 521 ± 120 year Δage (year) (a) (b) (c) (d) (e) (f) 3. 1age sensitivity study. Shades of blue give the conﬁdence intervals as marked; the black curves represent the values used in the 14 chronology; the red curve gives an alternative 1age solution using the Arnaud densiﬁcation model. (a) Temperature forcing of the cation model. (b) Reconstructed accumulation rates. (c) Reconstructed 1age; note the reversed scale. Histograms of 1age distribution wn for (d) 60 kaBP, (e) 40 kaBP, and (f) 20 kaBP. Distribution mean and 2σ uncertainty bound is stated in each panel. 0 10 20 30 40 50 60 WD2014 A (k BP) 0 10 20 30 40 50 60 −44 −42 −40 −38 −36 −34 −32 −30 Temperature ( oC) 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Accumulation (m ice a−1) 100 200 300 400 500 600 700 800 Δage (years) (d) (e) (f) All ± 2σ ± 1σ (a) (b) (c) (b) 20 30 40 50 60 WD2014 Age (ka BP) 200 250 300 350 0 50 100 150 200 262 ± 50 year Frequency Δage (year) 250 300 350 400 450 50 100 150 200 351 ± 73 year Δage (year) 40 50 100 150 200 (d) (e) 400 500 600 700 521 ± 120 year Δage (year) (f) Figure 3. 1age sensitivity study. Shades of blue give the conﬁdence intervals as marked; the black curves represent the values used in the WD2014 chronology; the red curve gives an alternative 1age solution using the Arnaud densiﬁcation model. (a) Temperature forcing of the densiﬁcation model. (b) Reconstructed accumulation rates. (c) Reconstructed 1age; note the reversed scale. C. Buizert et al.: The WAIS Divide chronology – Part 1 Histograms of 1age distribution are shown for (d) 60 kaBP, (e) 40 kaBP, and (f) 20 kaBP. Distribution mean and 2σ uncertainty bound is stated in each panel. histograms of 1age at 20 kyr intervals (Fig. 3d–f). Based on the sensitivity study, we estimate the WD 1age to be 521 ± 120 years (2σ uncertainty) at the LGM (∼20 kaBP). The 1age value of 351 ± 73 years at 40 kaBP gives a rep- resentative 1age for MIS 3; 1age at 60 kaBP is 262 ± 50 years. (about 7 %) smaller than 1age from the Herron–Langway model. The root-mean-square (rms) difference between both solutions is 35 years, corresponding to 0.63 times the 2σ un- certainty found in the sensitivity study. Both solutions are thus found to be in good agreement. The Herron–Langway approach is preferred because the internally consistent solu- tion of temperature, accumulation, and ice ﬂow associated with it is in better agreement with borehole temperature data than the solutions associated with the Arnaud model. Fur- thermore, the Herron–Langway model is more successful in simulating the magnitude of the δ15N response to the accu- mulation anomaly at 12 ka (not shown), suggesting it has a more realistic sensitivity to accumulation variability. Additionally, we have repeated our 1age reconstruction using the ﬁrn densiﬁcation physics described by Arnaud et al. (2000) rather than the Herron–Langway description used so far; the Arnaud model provides the physical basis for the commonly used ﬁrn densiﬁcation model of Goujon et al. (2003). More details on the implementation of the Arnaud model are given in Appendix A. 1age found using the Ar- naud model is plotted in red in Fig. 3c. Averaged over the entire core, 1age found with the Arnaud model is 19 years Clim. Past, 11, 153–173, 2015 3.2 1age sensitivity study Besides A and T there are several model parameters that have the potential to inﬂuence the model outcome; these are the convective zone (CZ) thickness (Sowers et al., 1992; Kawa- mura et al., 2006), surface density (ρ0), and sensitivity to ice impurity content. In this section we evaluate the sensitivity of the model output to all of these parameters. We performed 1000 model runs in which the model parameters were ran- domly perturbed. The spread in 1age model results is used to calculate the WD2014 age uncertainty. Impurity-enhanced densiﬁcation. Following recent work we include the possibility that increased glacial impurity loading could have enhanced densiﬁcation rates (Horhold et al., 2012; Freitag et al., 2013a). We use the mathematical formulation of Freitag et al. (2013a), in which the activation energy of the sintering process is a function of the Ca concentration in the ﬁrn. The value of β, the sensitivity to Ca, is drawn from a Gaussian distribution with 0.0015 mean and a 2σ width of 0.0015. The topic of impurity-enhanced densiﬁcation is discussed in detail in Sect. 3.3. Convective zone thickness. In the WD2014 model run (Sect. 3.1) we use a constant 3.5 m CZ, corresponding to the present-day situation (Battle et al., 2011). In the sensitivity study we vary the CZ by one of two methods: (1) we let the CZ be constant in time; its thickness is set by drawing from a Gaussian distribution with 3.5 m mean and 3.5 m 2σ width (i.e., 95 % probability of drawing a value in the 0–7 m range). (2) We let the CZ be a function of accumulation rate (Drey- fus et al., 2010), CZ = 3.5 + k × (A–0.22); we draw k from a Gaussian distribution with mean of −10 and a 2σ width of 40 (at an LGM A of 10 cma−1 this gives a CZ of 0–10 m The A and 1age scenarios found in the sensitivity study are shown in Fig. 3b and c, respectively. The shaded areas in Fig. 3b and c give the total range of solutions, as well as the ±2σ and ±1σ conﬁdence intervals. Note that the total range of solutions will depend on the number of model runs (here 1000) but that the position of the ±2σ and ±1σ envelopes will not. To investigate the distribution of values, we include Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 160 0 5 10 15 20 25 0.1 0.15 0.2 0.25 0.3 Accumulation (m ice a −1) 0 5 10 15 20 25 100 200 300 400 500 600 ∆age (years) WD2014 age (ka BP) (a) (b) Figure 4. Impurity enhancement of densiﬁcation rates at WD. Den- siﬁcation modeling results for (a) accumulation rates and (b) 1age. We use Ca sensitivities β = 0 (red) through β = 1×10−2 (blue), in steps of 2.5 × 10−3 (shades of deep purple). Black curves give A and 1age from ice-ﬂow modeling and λ(z). 0 5 10 15 20 25 0.1 0.15 0.2 0.25 0.3 Accumulation (m ice a −1) 0 5 10 15 20 25 100 200 300 400 500 600 ∆age (years) WD2014 age (ka BP) (a) (b) ing 1age. Hence the model simulations with large β (blue) give a higher A and smaller 1age. For the past 31.2 ka we have an independent A estimate from λ(z) that we can compare to the solutions from the ﬁrn model (Fig. 4, black curve). We also plotted 1age re- constructed via the 1depth method of Parrenin et al. (2012). Remarkably, we ﬁnd consistent solutions only when using a Ca sensitivity β ≤2.5×10−3, i.e., less than one-quarter of the sensitivity suggested by Freitag et al. (2013a). The best ﬁt to the independent LGM (25–20 kaBP) A and 1age esti- mates is obtained for β = 0. We conclude that WD does not provide any evidence for impurity (or, more speciﬁcally, Ca) enhancement of densiﬁcation rates. An important caveat is that our model uses 10-year- average [Ca] values, and therefore cannot resolve effects of interannual layering within the ﬁrn. Explicitly modeling the layering would require centimeter-scale resolution in the dy- namical ﬁrn model, which is prohibitive from a computa- tional point of view. Furthermore, [Ca] data at the required sub-annual resolution are difﬁcult, if not impossible, to mea- sure for the deepest part of the core, where λ(z) is below 1 cma−1. Increased ﬁrn layering and enhanced bulk densiﬁ- cation affect the ﬁrn thickness in a similar manner; both lead to a shallower lock-in depth, and thereby a reduced δ15N. Therefore, in order to reconcile our WD results with the im- purity hypothesis of Horhold et al. C. Buizert et al.: The WAIS Divide chronology – Part 1 (2012), one would need to invoke a strong reduction in LGM ﬁrn layering relative to the present day to compensate for the impurity-driven increase in bulk densiﬁcation rates. Recent work on the EDML core suggests that ﬁrn density layering may have been more pro- nounced during glacial times (Bendel et al., 2013); including ﬁrn layering is therefore likely to only exacerbate the prob- lem. Figure 4. Impurity enhancement of densiﬁcation rates at WD. Den- siﬁcation modeling results for (a) accumulation rates and (b) 1age. We use Ca sensitivities β = 0 (red) through β = 1×10−2 (blue), in steps of 2.5 × 10−3 (shades of deep purple). Black curves give A and 1age from ice-ﬂow modeling and λ(z). Work on present-day ﬁrn has provided support for ﬁrn softening by impurity loading (Horhold et al., 2012; Freitag et al., 2013a, b). More work is needed to understand how densiﬁcation rates are linked to impurity content in a mecha- nistic, rather than purely empirical, way. Perhaps such a mi- croscopic description could provide an explanation why ﬁrn densiﬁcation rates at WD, to ﬁrst order, do not appear to be affected by order-of-magnitude variations in [Ca] loading. One possible explanation could be that densiﬁcation rates are controlled by some parameter that co-varies with Ca in modern day ﬁrn yet does not change appreciably over glacial cycles (Fujita et al., 2014). 3.3 Impurity softening of ﬁrn? Recent work suggests a link between densiﬁcation rates and impurity content (for which [Ca2+] is used as a proxy) in po- lar ﬁrn (Horhold et al., 2012; Freitag et al., 2013a). Here we measured total [Ca] by ICP-MS, but at WD nearly all Ca is in the form of Ca2+. The inﬂuence of the impurity sensitivity β (see Eq. A6 in the Appendix) on 1age at WD is shown in Fig. 4. The sensitivity recommended by Freitag et al. (2013a) from investigating present-day ﬁrn packs is β = 1×10−2. We reconstructed A and 1age with the ﬁrn densiﬁcation inverse model using ﬁve values of β ranging from β = 0 (red) to β = 1 × 10−2 (blue) in steps of 2.5 × 10−3. Average [Ca] is around 0.8 ngg−1 in the early Holocene and around 9 ngg−1 in the LGM; a change to about an order of magnitude. Fol- lowing Freitag et al. (2013a) we use the total [Ca] rather than non-sea-salt Ca. If densiﬁcation rates are sensitive to impu- rity loading (large β, blue curves), this results in increased ﬁrn compaction during the LGM. The densiﬁcation model, which is trying to match the δ15N data, will compensate by increasing the A forcing, which in turn results in a decreas- 4.1 Annual layer count (0–31.2 ka) dated Northern Hemisphere (NH) climate records of abrupt DO variability using the WD record of globally well-mixed CH4 (Fig. 5). This process consists of several steps: in this work, WD2014, uses an updated layer count for the upper 2850 m, based on new data and analyses that have be- come available since publication of WDC06A-7. These up- dates are as follows: 1. Determine the midpoint of the abrupt DO transitions in WD CH4, NGRIP δ18O, and Hulu speleothem δ18O. 1. a reassessment of the dating in the upper 577 m (2.4 ka) using high-resolution multi-parameter chemistry data in combination with automated layer detection algorithms (Winstrup et al., 2012); 2. Assign a gas age to the WD CH4 tie points (i.e., the DO transitions). 3. Apply the WD 1age (Sect. 3) to ﬁnd the corresponding ice age at the depth of the CH4 tie points. 2. a reassessment of the dating between 577 and 2300 m (2.4–15.3 ka) using high-resolution multi-parameter chemistry data in combination with ECM; 4. Interpolate between the ice age constraints to ﬁnd the WD depth–age relationship. 4. Interpolate between the ice age constraints to ﬁnd the WD depth–age relationship. 3. a reassessment of the dating between 2300 and 2800 m (15.3–29.5 ka) using ECM and dust particle measure- ments, with the ECM having increasing importance with depth; 5. Redo the 1age calculations on the new ice age scale. 5. Redo the 1age calculations on the new ice age scale. 6. Repeat steps 3–5 iteratively until the depth–age relation- ship is stable within 1 year. At WD this happened after three iterations. 6. Repeat steps 3–5 iteratively until the depth–age relation- ship is stable within 1 year. At WD this happened after three iterations. 4. an extension of the annual-layer dating between 2800 and 2850 m (29.5–31.2 ka) using ECM. These steps are described in more detail in the following sec- tions. These steps are described in more detail in the following sec- tions. These steps are described in more detail in the following sec- tions. Details on the updated WD layer count and the layer counting methodology are presented in part 2 of the WD2014 papers. Details on the updated WD layer count and the layer counting methodology are presented in part 2 of the WD2014 papers. 4.2 Methane synchronization (31.2–68 ka) The procedure for determining the midpoint of the abrupt DO warming transitions is depicted in Fig. 6. For each of For the deep part of the core where an annual-layer count is not available, we date WD by synchronization to well- 4.1 Annual layer count (0–31.2 ka) A ﬁrst layer-counted chronology for the upper 2800 m of the WD core based on electrical conductivity measurements (ECM), named WDC06A-7, was presented by WAIS Divide Project Members (2013). The WAIS chronology presented Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 161 30 35 40 45 50 55 60 WD2014 Age (ka BP) 30 35 40 45 50 55 60 −9 −8 −7 −6 −5 Hulu δ18O (‰) −46 −44 −42 −40 −38 −36 NGRIP δ18O (‰) 350 400 450 500 550 600 650 WD CH4 (ppb) 2 3 4 5.1 5.2 6 7 8 9 10 11 12 13 14 15 16 17 Figure 5. Records of abrupt DO climate variability, (a) revised Hulu Cave speleothem δ18O record on Hulu chronology with U / Th ages above the time series (red dots), (b) NGRIP ice core δ18O on 1.0063×GICC05 chronology, and (c) WD CH4 on WD2014 (discrete data). DO numbering is given in the bottom of the ﬁgure following Rasmussen et al. (2014). White dots denote the midpoints of the stadial–interstadial transitions; the orange vertical lines show the timing of the NGRIP tie points (on 1.0063 × GICC05). For DO 3, 4, and 5.1 the WD2014 chronology is based on annual-layer counting, and minor timing differences exist between WD and NGRIP. 30 35 40 45 50 55 60 WD2014 Age (ka BP) 30 35 40 45 50 55 60 −9 −8 −7 −6 −5 Hulu δ18O (‰) −46 −44 −42 −40 −38 −36 NGRIP δ18O (‰) 350 400 450 500 550 600 650 WD CH4 (ppb) 2 3 4 5.1 5.2 6 7 8 9 10 11 12 13 14 15 16 17 30 Figure 5. Records of abrupt DO climate variability, (a) revised Hulu Cave speleothem δ18O record on Hulu chronology with U / Th ages above the time series (red dots), (b) NGRIP ice core δ18O on 1.0063×GICC05 chronology, and (c) WD CH4 on WD2014 (discrete data). DO numbering is given in the bottom of the ﬁgure following Rasmussen et al. (2014). White dots denote the midpoints of the stadial–interstadial transitions; the orange vertical lines show the timing of the NGRIP tie points (on 1.0063 × GICC05). For DO 3, 4, and 5.1 the WD2014 chronology is based on annual-layer counting, and minor timing differences exist between WD and NGRIP. 4.4 Synchronizing WD to a NGRIP–Hulu hybrid chronology Red dots give the midpoint (50 %) of the DO transition; the blue dots give the 25 and 75 % marks in the DO transitions. The DO transition at 58.35 ka was not used in Hulu, where it is much more gradual than in the other records (pos- sibly because calcite sampling was not perfectly perpendicular to the stalagmite isochrones, or because growth rates were variable in between the U / Th ages). the transitions we manually determine pre-event and post- event averages, as indicated by the orange lines. The averag- ing time is set to 150 and 50 years for stadial and interstadial periods, respectively; this difference in duration is used be- cause (i) several of the interstadials are of short duration and (ii) Greenland δ18O is more variable during stadial climates, requiring longer averaging. For DO 16.1, the duration of the pre-event stadial baseline climate was shorter than 150 years, and the averaging time was reduced to 100 years (Fig. 6). the transitions we manually determine pre-event and post- event averages, as indicated by the orange lines. The averag- ing time is set to 150 and 50 years for stadial and interstadial periods, respectively; this difference in duration is used be- cause (i) several of the interstadials are of short duration and (ii) Greenland δ18O is more variable during stadial climates, requiring longer averaging. For DO 16.1, the duration of the pre-event stadial baseline climate was shorter than 150 years, and the averaging time was reduced to 100 years (Fig. 6). In both the NGRIP and Hulu Cave δ18O records we have determined the ages of the midpoints of the DO transitions (Fig. 6; Table 1); a plot of their difference (Hulu age mi- nus NGRIP age) is shown in Fig. 7, where the error bars denote the root sum square of the NGRIP and Hulu mid- point determination uncertainty (Sect. 4.3). The Hulu ages are systematically older than the NGRIP ages, and the age difference increases going further back in time. Note that the Hulu–NGRIP age difference is smaller than the stated GICC05 counting uncertainty (832 to 2573 years) but larger than the Hulu age uncertainty (92 to 366 years). 4.4 Synchronizing WD to a NGRIP–Hulu hybrid chronology Abrupt DO variability is expressed clearly in a great num- ber of NH climate records (Voelker, 2002). For the pur- pose of methane synchronization, our interest is in high- resolution records that express the abrupt DO events very clearly, and are furthermore exceptionally well dated. We here use a combination of two such NH records (Fig. 5), namely the Greenland NGRIP δ18O record (NGRIP com- munity members, 2004), and a reﬁned version of the Hulu Cave speleothem δ18O record (Edwards et al., 2015; Reimer et al., 2013; Southon et al., 2012) with improved resolution and additional dating constraints (see Wang et al., 2001, for the original, lower resolution Hulu δ18O record). The DO events are resolved most clearly in the NGRIP δ18O record, which is available at 20-year resolution. We use the GICC05- modelext chronology for this core, which is based on annual- layer counting back to 60 kaBP and ice-ﬂow modeling for ice older than 60 ka (Rasmussen et al., 2006; Svensson et al., 2006; Wolff et al., 2010). While annual-layer counting pro- vides accurate relative ages (e.g., the duration of DO inter- stadials), it provides relatively inaccurate absolute ages due to the cumulative nature of counting uncertainty (Table 1). The reﬁned Hulu δ18O record also shows the abrupt DO events in high temporal resolution (Fig. 6). The speleothem chronology is based on U / Th radiometric dating, providing much smaller uncertainty in the absolute ages than GICC05 (Table 1). The reason for selecting this record over other speleothem records is the large number of U / Th dates, the low detrital Th at the site, and the high sampling resolution of the δ18O record (Wang et al., 2001). In the Hulu data, as in other records of DO variability, the interstadial onsets are more pronounced and abrupt than their terminations. We therefore only use the timing of the former as age constraints, as they can be established more reliably. The onset of NH interstadial periods as expressed in Hulu δ18O is given in Ta- ble 1. Figure 6. Determining the midpoint for the abrupt warming phases of (from oldest to youngest) DO 17.2, 17.1, 16.2, and 16.1 in (a) NGRIP δ18O (on 1.0063×GICC05), (b) WD CH4 (continuous data, on WD2014), and (c) Hulu δ18O with U / Th ages beneath the time series (red dots with error bars). C. Buizert et al.: The WAIS Divide chronology – Part 1 162 58 58.5 59 59.5 60 −44 −42 −40 −38 58 58.5 59 59.5 60 450 500 550 600 650 WD2014 age (ka BP) −8.5 −8 −7.5 −7 −6.5 (a) (b) (c) NGRIP δ18O (‰) Hulu δ18O (‰) WD CH4 (ppb) Figure 6. Determining the midpoint for the abrupt warming phases of (from oldest to youngest) DO 17.2, 17.1, 16.2, and 16.1 in (a) NGRIP δ18O (on 1.0063×GICC05), (b) WD CH4 (continuous data, on WD2014), and (c) Hulu δ18O with U / Th ages beneath the time series (red dots with error bars). Red dots give the midpoint (50 %) of the DO transition; the blue dots give the 25 and 75 % marks in the DO transitions. The DO transition at 58.35 ka was not used in Hulu, where it is much more gradual than in the other records (pos- sibly because calcite sampling was not perfectly perpendicular to the stalagmite isochrones, or because growth rates were variable in between the U / Th ages). 58 58.5 59 59.5 60 −44 −42 −40 −38 58 58.5 59 59.5 60 450 500 550 600 650 WD2014 age (ka BP) −8.5 −8 −7.5 −7 −6.5 (a) (b) (c) NGRIP δ18O (‰) Hulu δ18O (‰) WD CH4 (ppb) The midpoints of abrupt interstadial terminations were de- termined in the same fashion (WD CH4 and NGRIP only). Tables 1 and 2 give the results for NH warming and NH cool- ing, respectively. www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 C. Buizert et al.: The WAIS Divide chronology – Part 1 Overview of CH4 tie points for NH warming events. WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies. DO 17.2 2420.35 59 386 2573 15 59 772 DO 18 2465.84 64 049 2611 30 0 10 20 30 40 50 60 70 0 100 200 300 400 500 600 GICC05 age (ka BP) Hulu−NGRIP age (years) 0.0063 × GICC05 Figure 7. Hulu–NGRIP age offset at the midpoint of the DO δ18O transitions. The error bars denote the root sum square of the mid- point determination uncertainty in NGRIP and Hulu δ18O (Table 1). The GICC05 ages are placed on the BP 1950 scale rather than the b2k scale (years prior to 2000 CE). 0 10 20 30 40 50 60 70 0 100 200 300 400 500 600 GICC05 age (ka BP) Hulu−NGRIP age (years) 0.0063 × GICC05 superior absolute ages of the Hulu chronology. Second, it is very simple to convert between the WD2014 and GICC05 chronologies (CH4-synchronized section of the chronology only); one simply needs to divide WD2014 ages by 1.0063 (and add 50 years to convert to the b2k reference date). Third, it still allows for direct synchronization of WD CH4 to the NGRIP δ18O record, providing more tie points than di- rect synchronization to the Hulu record would. Note that the GICC05 × 1.0063 target chronology respects the Hulu age constraints in an average sense only; the age of individual events differs between Hulu and our target chronology by up to 180 years. Our approach therefore represents only a ﬁrst- order correction of a growing offset between GICC05 and Hulu; nonlinear temporal changes in the counted dating er- ror may exist from one tie point to the next (Fleitmann et al., 2009). Hulu−NGRIP age (years) Figure 7. Hulu–NGRIP age offset at the midpoint of the DO δ18O transitions. The error bars denote the root sum square of the mid- point determination uncertainty in NGRIP and Hulu δ18O (Table 1). The GICC05 ages are placed on the BP 1950 scale rather than the b2k scale (years prior to 2000 CE). The exercise of ﬁnding the transition midpoints and deter- mining the GICC05–Hulu scaling factor was performed by two of the authors (J. P. Severinghaus and C. Buizert), inde- pendently of each other. C. Buizert et al.: The WAIS Divide chronology – Part 1 Hulu–NGRIP age offset at the midpoint of the DO δ18O transitions. The error bars denote the root sum square of the mid- point determination uncertainty in NGRIP and Hulu δ18O (Table 1). The GICC05 ages are placed on the BP 1950 scale rather than the b2k scale (years prior to 2000 CE). 0.0063 × GICC05 age, suggesting that the GICC05 annual- layer count on average misses 6.3 out of every 1000 layers. Because of this observation we use a linearly scaled version of the GICC05 chronology (GICC05 × 1.0063) as the target superior absolute ages of the Hulu chronology. Second, it is very simple to convert between the WD2014 and GICC05 chronologies (CH4-synchronized section of the chronology only); one simply needs to divide WD2014 ages by 1.0063 (and add 50 years to convert to the b2k reference date). Third, it still allows for direct synchronization of WD CH4 to the NGRIP δ18O record, providing more tie points than di- rect synchronization to the Hulu record would. Note that the GICC05 × 1.0063 target chronology respects the Hulu age constraints in an average sense only; the age of individual events differs between Hulu and our target chronology by up to 180 years. Our approach therefore represents only a ﬁrst- order correction of a growing offset between GICC05 and Hulu; nonlinear temporal changes in the counted dating er- ror may exist from one tie point to the next (Fleitmann et al., 2009). The exercise of ﬁnding the transition midpoints and deter- mining the GICC05–Hulu scaling factor was performed by two of the authors (J. P. Severinghaus and C. Buizert), inde- pendently of each other. The scaling factors obtained were 1.0063 and 1.0064, respectively, showing that, to ﬁrst order, this result is insensitive to (subjective) judgment in identi- fying the transitions. The difference between the Hulu ages and 1.0063 × GICC05 ages are all within the stated Hulu 2σ dating error (Table 1). Consequently, our chronology is not in violation of any Hulu constraint as it respects the Hulu 2σ error at all of the tie points In deriving the scaling we have Table 1. Overview of CH4 tie points for NH warming events. WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies. NGRIP Hulu WD Depth Age Age uncert. Midpoint Hulu age Age uncert. C. Buizert et al.: The WAIS Divide chronology – Part 1 Midpoint Depth Gas age Ice age Midpoint (m) (years BP) (years) (years) (years BP) (years) (years) (m) (years BP) (years BP) (years) YD-PB 1490.89 11 619 98 23 1983.02 11 546 11 740 33 OD-BA 1604.05 14 628 185 15 2259.40 14 576 14 804 29 DO 3 1869.00 27 728 832 12 27 922 95 39 2755.74 27 755 28 144 19 DO 4 1891.27 28 838 898 14 29 134 92 21 2797.92 29 011 29 397 22 DO 5.1 1919.48 30 731 1023 11 30 876 255 37 2848.38 30 730 31186 22 DO 5.2 1951.66 32 452 1132 15 32 667 236 21 2885.44 32 631 33 051 17 DO 6 1974.48 33 687 1213 19 34 034 337 36 2913.01 33 874 34 283 18 DO 7 2009.62 35 437 1321 16 35 532 299 20 2958.64 35 636 35 982 20 DO 8 2069.88 38 165 1449 13 38 307 155 19 3021.37 38 381 38 681 33 DO 9 2099.50 40 104 1580 13 40 264 241 42 3066.52 40 332 40 690 19 DO 10 2123.98 41 408 1633 14 41 664 310 27 3094.17 41 643 41 980 18 DO 11 2157.58 43 297 1736 17 43 634 144 26 3130.44 43 544 43 866 15 DO 12 2221.96 46 794 1912 21 47 264 153 20 3195.25 47 064 47 335 16 DO 13 2256.73 49 221 2031 17 49 562 251 52 3237.65 49 506 49 836 19 DO 14 2345.39 54 164 2301 11 3311.09 54 480 54 747 13 DO 15.1 2355.17 54 940 2349 16 3322.24 55 261 55 564 11 DO 15.2 2366.15 55 737 2392 26 3329.72 56 063 56 381 14 DO 16.1 2398.71 57 988 2498 11 3350.44 58 328 58 610 9 DO 16.2 2402.25 58 210 2510 12 58 545 226 22 3352.59 58 552 58 848 14 DO 17.1 2414.82 59 018 2557 15 59 364 366 18 3360.02 59 364 59 627 17 DO 17.2 2420.35 59 386 2573 15 59 772 254 23 3363.42 59 735 59 997 25 DO 18 2465.84 64 049 2611 30 3388.73 64 428 64 773 15 0 10 20 30 40 50 60 70 0 100 200 300 400 500 600 GICC05 age (ka BP) Hulu−NGRIP age (years) 0.0063 × GICC05 Figure 7. C. Buizert et al.: The WAIS Divide chronology – Part 1 The scaling factors obtained were 1.0063 and 1.0064, respectively, showing that, to ﬁrst order, this result is insensitive to (subjective) judgment in identi- fying the transitions. The difference between the Hulu ages and 1.0063 × GICC05 ages are all within the stated Hulu 2σ dating error (Table 1). Consequently, our chronology is not in violation of any Hulu constraint as it respects the Hulu 2σ error at all of the tie points. In deriving the scaling we have assumed that the abrupt DO transitions observed in NGRIP and Hulu are simultaneous, which is not necessarily true. The 0.0063 × GICC05 age, suggesting that the GICC05 annual- layer count on average misses 6.3 out of every 1000 layers. Because of this observation we use a linearly scaled version of the GICC05 chronology (GICC05 × 1.0063) as the target chronology for methane synchronization. This approach has several advantages. First, it respects both the superior rela- tive ages (i.e., interval durations) of GICC05, as well as the Clim. Past, 11, 153–173, 2015 C. Buizert et al.: The WAIS Divide chronology – Part 1 Hulu–NGRIP age offset at the midpoint of the DO δ18O transitions. The error bars denote the root sum square of the mid- point determination uncertainty in NGRIP and Hulu δ18O (Table 1). The GICC05 ages are placed on the BP 1950 scale rather than the b2k scale (years prior to 2000 CE). 0.0063 × GICC05 age, suggesting that the GICC05 annual- layer count on average misses 6.3 out of every 1000 layers. Because of this observation we use a linearly scaled version of the GICC05 chronology (GICC05 × 1 0063) as the target superior absolute ages of the Hulu chronology. Second, it is very simple to convert between the WD2014 and GICC05 chronologies (CH4-synchronized section of the chronology only); one simply needs to divide WD2014 ages by 1.0063 (and add 50 years to convert to the b2k reference date). Third, it still allows for direct synchronization of WD CH4 to the NGRIP δ18O record, providing more tie points than di- rect synchronization to the Hulu record would. Note that the GICC05 × 1.0063 target chronology respects the Hulu age constraints in an average sense only; the age of individual events differs between Hulu and our target chronology by up to 180 years. Our approach therefore represents only a ﬁrst- order correction of a growing offset between GICC05 and Hulu; nonlinear temporal changes in the counted dating er- ror may exist from one tie point to the next (Fleitmann et al., 2009). The exercise of ﬁnding the transition midpoints and deter- mining the GICC05–Hulu scaling factor was performed by two of the authors (J. P. Severinghaus and C. Buizert), inde- pendently of each other. The scaling factors obtained were 1.0063 and 1.0064, respectively, showing that, to ﬁrst order, this result is insensitive to (subjective) judgment in identi- fying the transitions. The difference between the Hulu ages and 1.0063 × GICC05 ages are all within the stated Hulu 2σ dating error (Table 1). Consequently, our chronology is not in violation of any Hulu constraint as it respects the Hulu 2σ oints for NH warming events. WD ages printed in boldface are assigned as part of the CH4 synchronization; endent chronologies. Table 1. Overview of CH4 tie points for NH warming events. WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies. Table 1. 4.4 Synchronizing WD to a NGRIP–Hulu hybrid chronology A linear ﬁt through these data, forced to intersect the origin, is given by After determining the pre- and post-event averages, we use linear interpolation of the time series to ﬁnd the time at which the variable of interest had completed 25, 50, and 75 % of the total transition (Fig. 6). We use the 50 % marker (red) as the midpoint of the transition, which is used in the methane syn- chronization. The 25 and 75 % markers (blue) are used as the ±1σ uncertainty estimate. In rare cases the time series contain inversions within the transitions that lead to ambigu- ity in the timing of the markers; for these events we ﬁnd the markers using a monotonic spline ﬁt to the data. Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 163 C. Buizert et al.: The WAIS Divide chronology – Part 1 163 Table 1. Overview of CH4 tie points for NH warming events. WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies. NGRIP Hulu WD Depth Age Age uncert. Midpoint Hulu age Age uncert. C. Buizert et al.: The WAIS Divide chronology – Part 1 Midpoint Depth Gas age Ice age Midpoint (m) (years BP) (years) (years) (years BP) (years) (years) (m) (years BP) (years BP) (years) YD-PB 1490.89 11 619 98 23 1983.02 11 546 11 740 33 OD-BA 1604.05 14 628 185 15 2259.40 14 576 14 804 29 DO 3 1869.00 27 728 832 12 27 922 95 39 2755.74 27 755 28 144 19 DO 4 1891.27 28 838 898 14 29 134 92 21 2797.92 29 011 29 397 22 DO 5.1 1919.48 30 731 1023 11 30 876 255 37 2848.38 30 730 31186 22 DO 5.2 1951.66 32 452 1132 15 32 667 236 21 2885.44 32 631 33 051 17 DO 6 1974.48 33 687 1213 19 34 034 337 36 2913.01 33 874 34 283 18 DO 7 2009.62 35 437 1321 16 35 532 299 20 2958.64 35 636 35 982 20 DO 8 2069.88 38 165 1449 13 38 307 155 19 3021.37 38 381 38 681 33 DO 9 2099.50 40 104 1580 13 40 264 241 42 3066.52 40 332 40 690 19 DO 10 2123.98 41 408 1633 14 41 664 310 27 3094.17 41 643 41 980 18 DO 11 2157.58 43 297 1736 17 43 634 144 26 3130.44 43 544 43 866 15 DO 12 2221.96 46 794 1912 21 47 264 153 20 3195.25 47 064 47 335 16 DO 13 2256.73 49 221 2031 17 49 562 251 52 3237.65 49 506 49 836 19 DO 14 2345.39 54 164 2301 11 3311.09 54 480 54 747 13 DO 15.1 2355.17 54 940 2349 16 3322.24 55 261 55 564 11 DO 15.2 2366.15 55 737 2392 26 3329.72 56 063 56 381 14 DO 16.1 2398.71 57 988 2498 11 3350.44 58 328 58 610 9 DO 16.2 2402.25 58 210 2510 12 58 545 226 22 3352.59 58 552 58 848 14 DO 17.1 2414.82 59 018 2557 15 59 364 366 18 3360.02 59 364 59 627 17 DO 17.2 2420.35 59 386 2573 15 59 772 254 23 3363.42 59 735 59 997 25 DO 18 2465.84 64 049 2611 30 3388.73 64 428 64 773 15 0 10 20 30 40 50 60 70 0 100 200 300 400 500 600 GICC05 age (ka BP) Hulu−NGRIP age (years) 0.0063 × GICC05 Figure 7. C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 variations in monsoon intensity represented by Hulu δ18O are commonly explained by meridional movement of the In- tertropical Convergence Zone (ITCZ) and tropical rainfall belts (Wang et al., 2001, 2006; Kanner et al., 2012); model- ing work suggests such atmospheric readjustments occur on decadal timescales in response to NH high-latitude forcing (Chiang and Bitz, 2005; Broccoli et al., 2006; Cvijanovic and Chiang, 2013). Moreover, CH4 emission changes are near- synchronous with Greenland δ18O variations, which they lag by only a few decades on average (Huber et al., 2006; Baum- gartner et al., 2014; Rosen et al., 2014). Since both CH4 emissions and Hulu δ18O are closely linked to tropical hy- drology, timing lags between NGRIP and Hulu are also ex- pected to be on decadal timescales. The uncertainty in the NGRIP–Hulu phasing is therefore probably small (decadal) relative to the correction we apply (up to 400 years). ice age. These assigned ages are printed in boldface in Ta- bles 1 and 2. To obtain a continuous depth–age relationship between these ice age constraints, we have to apply an interpolation strategy. This task amounts to estimating the annual-layer thickness λ(z) along the deep part of the core. The simplest approach is to assume a constant accumulation rate in be- tween the age constraints; this is shown in Fig. 8b for the case where we use the age constraints from NH warming events only (black) or the age constraints from both NH warming and cooling events (red). The disadvantage of this approach is that it results in discontinuities in λ(z) (the ﬁrst derivative of the depth–age relationship), which we consider highly un- realistic. A more realistic approach is therefore to assume that λ(z) is continuous and smooth (Fudge et al., 2014); Fig. 8b shows two scenarios in which we use a spline func- tion to estimate λ(z), where again we have applied age con- straints from NH warming events only (orange) or age con- straints from both NH warming and cooling events (blue). pp y p y Rather than synchronizing WD CH4 to Greenland CH4 records, we have chosen to synchronize directly to NGRIP δ18O, which varies in phase with CH4 (but with a nearly constant time lag). C. Buizert et al.: The WAIS Divide chronology – Part 1 We let the midpoint in the CH4 transi- tions lag the midpoint in the NGRIP δ18O transition by 25 years, as suggested by studies of Greenland δ15N-CH4 phas- ing (Huber et al., 2006; Baumgartner et al., 2014; Rasmussen et al., 2013; Kindler et al., 2014; Rosen et al., 2014). The rationale behind this approach is threefold. First, through- out MIS 3 the NGRIP δ18O record has both better precision and higher temporal resolution than any available Greenland CH4 record (Baumgartner et al., 2014; Brook et al., 1996; Blunier et al., 2007). Second, the dating of Greenland gas records depends on the highly variable 1age function, which is not equally well constrained for all DO events (Schwander et al., 1997; Rasmussen et al., 2013). This reliance on Green- land 1age would introduce an additional source of uncer- tainty. The NGRIP δ18O record, on the other hand, is accu- rately dated through the GICC05 layer count. Third, Green- land CH4 records are more strongly impacted by ﬁrn smooth- ing than the WD CH4 record, because glacial accumulation is lower in Greenland (Greenland glacial 1age is about 2–3 times as high as WD 1age during that time). In summary, our approach circumvents the uncertainties associated with using Greenland CH4 as an intermediary, or, to state this an- other way, the uncertainty in the phasing between CH4 and Greenland δ18O is smaller than the uncertainty in the Green- land 1age. For comparison, past A obtained from the ﬁrn densiﬁca- tion model (Sect. 3) is plotted in green (Fig. 8b). While the δ15N-based A follows the synchronization-based A estimates broadly, the millennial-scale details do not agree. We want to point out that this is not unexpected, since both methods have their imperfections. In particular, any errors in the (stretched) GICC05 age model or in our modeled thinning function fλ(z) will strongly impact the synchronization-based A es- timates in Fig. 8b. The discrepancy is pronounced between 60 and 65 ka, where we have to strongly reduce λ(z) in order to ﬁt the age constraint(s) from DO 18, while δ15N provides no evidence for low A during this interval. g For the WD2014 chronology we have applied the smooth λ(z) interpolation scheme using all age constraints (i.e., both NH warming and cooling events). C. Buizert et al.: The WAIS Divide chronology – Part 1 The midpoint detection uncertainty is comparable for all events and systematically smaller at the start of interstadial periods than at the termi- nations (Tables 1 and 2). For short interstadials (e.g., DO 9) this leads to a large relative uncertainty in the event dura- tion, and thereby a large uncertainty in the implied accumu- lation rates (Fig. 8b). We force the interpolation to ﬁt all NH warming constraints perfectly, yet relax this requirement for NH cooling constraints to prevent large swings in λ(z) for the short-duration events. The WD2014 chronology ﬁts the NH warming and NH cooling age constraints with a 0- and 16-year rms offset, respectively. Because the duration of (in- ter)stadial periods is well constrained in the layer-counted GICC05 chronology, using both the NH warming and NH cooling tie points results in a more robust chronology. The duration of (inter)stadial periods is 0.63 % longer in WD2014 than in GICC05, which is well within the stated GICC05 counting error of 5.4 % (31.2–60 ka interval). www.clim-past.net/11/153/2015/ 164 www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 165 350 400 450 500 550 600 650 WD CH4 (ppb) 5.1 5.2 10 11 12 13 14 15 16 17 18 6 7 8 9 30 35 40 45 50 55 60 65 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24 Accumulation (m ice a−1) constant ’warming’ constant ’all’ firn model smooth ’warming’ smooth ’all’ 10 1 10 2 10 3 Age uncertainty (years) Δage unc. Interp. unc Midpoint unc. Abs. age unc. 30 35 40 45 50 55 60 65 WD2014 ice age (ka BP) (a) (b) (c) Figure 8. Interpolating between the CH4 age constraints. (a) WD discrete CH4 record with the abrupt stadial–interstadial transitions marked. DO numbering given at the top of the panel. (b) Different annual-layer thickness scenarios, converted to an accumulation rate for comparison to the δ15N-based ﬁrn model reconstructions. The interpolation strategy is to use either constant accumulation rates between tie points (“constant”) or a smoothly varying λ(z) (“smooth”); the age constraints used are either only the NH warming events (“warming”), or both the NH warming and cooling events (“all”). (c) Estimated 2σ uncertainties in the WD2014 chronology due to 1age, choice of interpolation scheme, midpoint detection, and the absolute age constraints used in the synchronization. Total absolute ice age uncertainty plotted in solid black; relative age uncertainty (i.e., with absolute age uncertainty in the Hulu–GICC05 master chronology withheld) plotted in dashed black. 4.5 Interpolation between age constraints We can assign a gas age to each of the depths where an abrupt WD CH4 transitions occurs; we do this for DO 4.1 through DO 18, i.e., the events prior to 31.2 kaBP (the onset of the WD layer count). The gas age we assign is equal to 1.0063 times the GICC05 age for the same event, with 25 years sub- tracted to account for the slight CH4 lag behind Greenland δ18O. By adding 1age (Sect. 3) to this gas age we assign an www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 C. Buizert et al.: The WAIS Divide chronology – Part 1 165 350 400 450 500 550 600 650 WD CH4 (ppb) 5.1 5.2 10 11 12 13 14 15 16 17 18 6 7 8 9 30 35 40 45 50 55 60 65 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24 Accumulation (m ice a−1) constant ’warming’ constant ’all’ firn model smooth ’warming’ smooth ’all’ 10 1 10 2 10 3 Age uncertainty (years) Δage unc. Interp. unc Midpoint unc. Abs. age unc. 30 35 40 45 50 55 60 65 WD2014 ice age (ka BP) (a) (b) (c) Figure 8. Interpolating between the CH4 age constraints. (a) WD discrete CH4 record with the abrupt stadial–interstadial transitions marked. DO numbering given at the top of the panel. (b) Different annual-layer thickness scenarios, converted to an accumulation rate for comparison to the δ15N-based ﬁrn model reconstructions. The interpolation strategy is to use either constant accumulation rates between tie points (“constant”) or a smoothly varying λ(z) (“smooth”); the age constraints used are either only the NH warming events (“warming”), or both the NH warming and cooling events (“all”). (c) Estimated 2σ uncertainties in the WD2014 chronology due to 1age, choice of interpolation scheme, midpoint detection, and the absolute age constraints used in the synchronization. Total absolute ice age uncertainty plotted in solid black; relative age uncertainty (i.e., with absolute age uncertainty in the Hulu–GICC05 master chronology withheld) plotted in dashed black. C. Buizert et al.: The WAIS Divide chronology – Part 1 WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies Table 2. Overview of CH4 tie points for NH cooling events. WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies points for NH cooling events. WD ages printed in boldface are assigned as part of the CH4 synchronization; endent chronologies provides the opportunity for obtaining very detailed climatic records. The uncertainties (2σ values) are plotted in Fig. 8c (log scale). We assume these four uncertainties to be independent, and use their root sum square as the total uncertainty estimate on the WD2014 ice age scale (Fig. 8c, black curve). Note that the fourth source of uncertainty is only relevant when con- sidering absolute ages; when evaluating relative ages (e.g., between WD ice and WD gas phase, or between WD and NGRIP), this last contribution does not need to be consid- ered. For the deepest WD ice (3404 m depth) we thus ﬁnd an age of 67.7 ± 0.9 kaBP. High accumulation rates also result in a small 1age. Fig- ure 9 compares 1age between several Antarctic cores (note the logarithmic scale). 1age at WD is approximately one- third of the 1age at EPICA DML (EDML) and Talos Dome (TALDICE), and one-tenth of the 1age at EPICA Dome C (EDC), Vostok, and Dome Fuji. Because the uncertainty in the 1age (or 1depth) calculation is typically on the order of 20 %, a smaller 1age allows for a more precise interhemi- spheric synchronization with Greenland ice core records us- ing CH4. The small WD 1age uncertainty during MIS 3 allows for investigation of the phasing of the bipolar see- saw (Stocker and Johnsen, 2003) at sub-centennial precision (WAIS Divide Project Members, 2015). C. Buizert et al.: The WAIS Divide chronology – Part 1 166 166 C. Buizert et al.: The WAIS Divide chronology – Part 1 Table 2. Overview of CH4 tie points for NH cooling events. WD ages printed in boldface are assigned as part of the CH4 synchronization; all other ages are on their independent chronologies e 2. Overview of CH4 tie points for NH cooling events. WD ages printed in boldface are assigned as part of the CH4 synchronizati ther ages are on their independent chronologies NGRIP WD Depth Age Age uncert. Midpoint Depth Gas age Ice age Midpoint (m) (years BP) (years) (years) (m) (years BP) (years BP) (years) BA-YD 1524.21 12 775 136 81 2096.61 12 769 12 987 52 DO 3 1861.91 27 498 822 52 2747.25 27 520 27 905 38 DO 4 1882.59 28 548 887 17 2787.99 28 696 29 090 61 DO 5.1 1916.45 30 571 1010 70 2845.37 30 618 31 067 50 DO 5.2 1939.71 31 992 1108 13 2875.86 32 168 32 607 70 DO 6 1964.52 33 323 1192 37 2905.55 33 508 33 905 60 DO 7 1990.58 34 703 1286 13 2939.09 34 897 35 292 50 DO 8 2027.43 36 571 1401 21 2986.58 36 776 37 172 32 DO 9 2095.51 39 905 1572 42 3063.79 40 132 40 492 25 DO 10 2112.53 40 917 1621 44 3083.89 41 150 41 508 44 DO 11 2135.66 42 231 1685 27 3110.76 42 472 42 823 69 DO 12 2171.17 44 308 1783 41 3149.89 44 562 44 904 47 DO 13 2242.85 48 440 1996 27 3226.93 48 720 49 054 20 DO 14 2261.49 49 552 2052 20 3243.03 49 839 50 165 65 DO 15.1 2353.66 54 850 2339 18 3321.15 55 170 55 469 14 DO 15.2 2359.92 55 369 2370 55 3326.47 55 693 55 983 45 DO 16.1 2375.88 56 555 2435 49 3337.98 56 887 57 219 76 DO 16.2 2400.56 58 123 2508 15 3351.80 58 465 58 756 9 DO 17.1 2406.52 58 544 2530 35 3355.54 58 888 59 151 61 DO 17.2 2417.77 59 257 2570 18 3362.26 59 606 59 862 24 DO 18 2462.07 63 810 2611 14 3387.28 64 187 64 547 32 Table 2. Overview of CH4 tie points for NH cooling events. www.clim-past.net/11/153/2015/ 4.6 Age uncertainty last source of uncertainty is the age uncertainty in the hy- brid NGRIP–Hulu chronology that we synchronize to. We use the stated Hulu age uncertainty plus 50 years to account for possible leads or lags in the NGRIP–Hulu δ18O phasing, plus the absolute value of the offset between the Hulu ages and the 1.0063 × GICC05 ages. For DO events where we do not have reliable Hulu age estimates (Table 1), we set the uncertainty to the Hulu age uncertainty of the nearest event, plus the uncertainty in the interval duration speciﬁed by the GICC05 layer count. For example, for DO 14 we do not have a reliable Hulu age estimate, and we use the Hulu age un- certainty of DO 16.2 (226 years) plus the uncertainty in the DO 14 to DO 16.2 interval duration on GICC05 (209 years), giving a total of 226 + 209 = 435 years. The age uncertainty we assign to the deep part (> 2850 m) of the WD2014 chronology has four components. The ﬁrst source of uncertainty is the 1age calculation; we use the 2σ uncertainty obtained in the 1age sensitivity study (Sect. 3.2). The second source of uncertainty is the choice of interpolation scheme used to obtain a continuous chronol- ogy; here we use the standard deviation between the four different interpolation schemes of Fig. 8b as an uncertainty estimate. The third source of uncertainty is the difﬁculty in determining the timing of the abrupt events in the time se- ries; we use the uncertainty in the midpoint evaluation (root sum square of WD CH4 and NGRIP δ18O estimates). The Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 C. Buizert et al.: The WAIS Divide chronology – Part 1 167 0 10 20 30 40 50 60 70 100 200 300 400 500 600 700 800 900 1000 2000 3000 4000 5000 6000 7000 8000 Vostok EDC D Fuji EDML TALDICE WD Age (ka BP) Δage (years) Figure 9. Comparison of 1age for different Antarctic cores, plotted on the gas age scale. Dome Fuji 1age from Kawamura et al. (2007); WD from Sect. 3; all others from Bazin et al. (2013); Veres et al. (2013). 0 10 20 30 40 50 60 70 100 200 300 400 500 600 700 800 900 1000 2000 3000 4000 5000 6000 7000 8000 Vostok EDC D Fuji EDML TALDICE WD Age (ka BP) Δage (years) Fuji is about 4000 years old, and consequently the ﬁrn col- umn integrates over 4000 years of climate variability, thereby dampening the 1age response to millennial-scale climatic variability. At WD the glacial ﬁrn layer is only about 350 years old, and therefore the ﬁrn is in near equilibrium with the millennial-scale climate variations. This difference in re- sponse time is also obvious during the deglaciation, where WD 1age transitions from glacial to interglacial values be- tween 18 and 14.5 kaBP, while Dome Fuji takes more time (18–10 kaBP). Surprisingly, EDML 1age does not show a strong deglacial 1age response, unlike all the other cores. The relatively small 1age at WAIS Divide also allows for precise investigation of the relative timing of atmospheric greenhouse gas variations and Antarctic climate (Barnola et al., 1991; Pedro et al., 2012; Caillon et al., 2003; Par- renin et al., 2013; Ahn et al., 2012). Recent works suggest that during the last deglaciation the rise in atmospheric CO2 lagged the onset of pan-Antarctic warming by approximately 0 to 400 years (Pedro et al., 2012; Parrenin et al., 2013). This Antarctic warming around 18 kaBP is presumably driven by the bipolar seesaw, as it coincides with a reduction in At- lantic overturning circulation strength as seen in North At- lantic sediment records (McManus et al., 2004). The WD 1age at 18 ka (gas age) is 515±91 years (2σ), much smaller than at central East Antarctic sites such as EPICA Dome C, where 1age is approximately 3850 ± 900 years (Veres et al., 2013, with the 1age uncertainty taken to be the dif- ference between the gas age and ice age uncertainties). C. Buizert et al.: The WAIS Divide chronology – Part 1 The precision with which one can determine the relative phasing of climatic (i.e., δ18O of ice) and atmospheric signals is set by the uncertainty in 1age (or equivalently, the uncertainty in 1depth). High-resolution WD records of CO2 and CH4 (Marcott et al., 2014) place the onset of the deglacial rise in the atmospheric mixing ratio of these greenhouse gases on the WD2014 chronology at 18 010 and 17 820 years BP, re- spectively. However, evaluating the relative phasing of CO2 and Antarctic climate is complicated by the observation of asynchronous deglacial warming across the Antarctic con- tinent (WAIS Divide Project Members, 2013). Attempts to capture the climate–CO2 relationship in a single lead-lag value may be an oversimpliﬁcation of deglacial climate dy- namics. Figure 9. Comparison of 1age for different Antarctic cores, plotted on the gas age scale. Dome Fuji 1age from Kawamura et al. (2007); WD from Sect. 3; all others from Bazin et al. (2013); Veres et al. (2013). et al., 2011; Parrenin et al., 2013), and provide a more re- ﬁned pan-Antarctic picture of the climate–CO2 relationship. 5 Discussion While the WAIS Divide ice core does not extend as far back in time as deep cores from the East Antarctic Plateau, its relatively high temporal resolution (due to the high snow accumulation rate) makes it an ice core of great scientiﬁc value. WD accumulation rate during the LGM (∼10 cma−1 ice equivalent) is still higher than the present-day accumu- lation rate at the EPICA (European Project for Ice Coring in Antarctica) Dronning Maud Land core (7 cma−1), which is generally considered a high-accumulation core (EPICA Community Members, 2006). With 68 ka in 3404 m of core, the core average λ is 5 cma−1, at the onset of the last deglaciation (18 kaBP) λ is around 4 cma−1, and near the bed λ is around 0.4 cma−1. This high temporal resolution In comparing the shape of the 1age proﬁles, there are some interesting differences (Fig. 9). It is important to re- alize that not all the 1age histories shown were derived in the same way; WD and Dome Fuji 1age were derived using densiﬁcation models, and the other four were derived using the 1depth approach (Parrenin et al., 2012) and a Bayesian inverse method that includes a wide range of age markers (Veres et al., 2013). We will therefore focus on comparing the WD and Dome Fuji results. 1age at WD shows more pronounced variability than at Dome Fuji, particularly dur- ing MIS 3. The reason is that the glacial ﬁrn pack at Dome www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ Appendix A: Densiﬁcation physics The densiﬁcation rates used in this work are based on the empirical steady-state model by Herron and Langway (1980) (the H-L model). We use the H-L model with minor modiﬁ- cations that allow it to be run dynamically (i.e., with time- variable T and A) and to include the softening effect of im- purities following Freitag et al. (2013a). The H-L model di- vides the ﬁrn column in two stages, separated at the critical density ρc = 550 kgm−3, occurring at the critical depth zc. ECa = EHL × α 1 −β ln [Ca] [Ca]crit , (A6) (A6) where ECa and EHL are the Ca-modiﬁed and original H-L activation energies, respectively, [Ca]crit = 0.5 ngg−1 is the minimum concentration at which impurities affect densiﬁ- cation, and α and β are calibration parameters. Whenever [Ca](z) < [Ca]crit, we set [Ca](z) = [Ca]crit. For the upper ﬁrn (ρ ≤ρc, stage 1), the densiﬁcation rates are given by The parameter β sets the sensitivity to dust loading, and α is a normalization parameter that is included to account for the fact that the original H-L model was calibrated with- out the impurity effect. Consequently, if β > 0, one needs to compensate by setting α > 1 to preserve the original H- L calibration. The work by Freitag et al. (2013a) recom- mends β = 0.01 and α = 1.025 (which yields ECa = EHL at [Ca] = 5.73 ngg−1). dρ dt = k1A(ρice −ρ), (A1) (A1) with with k1 = 11exp −E1 RT , (A2) (A2) where E1 = 10.16 kJ mol−1 is the activation energy for stage 1 and R is the universal gas constant. Because both the sink- ing velocity of deposited layers (w = dz/dt) and the densiﬁ- cation rate scale linearly with A, the resulting density–depth proﬁle ρ(z) in stage 1 becomes independent of A, and sensi- tive to T variations only. Using the recommended value of α = 1.025 at WD pro- vides a poor ﬁt to observations of present-day ﬁrn density and close-off depth. The optimal ﬁt to present-day WD ob- servations is obtained using an activation energy equal to 1.007×EHL; this is in between the values suggested by Her- ron and Langway (1980) and Freitag et al. (2013a). In the ex- periment presented in Fig. C. Buizert et al.: The WAIS Divide chronology – Part 1 a given depth in Mgm−2: different time series; and (iv) the ages of the hybrid NGRIP– Hulu chronology we are synchronizing to. σz = z Z 0 ρ(z′)dz′/1000. (A5) σz = z Z 0 ρ(z′)dz′/1000. Due to the combination of a small 1age and a high- resolution methane record, the WAIS Divide ice core can be synchronized more precisely to Greenland records than any other Antarctic core to date. This is important when inves- tigating interhemispheric climate relationships such as the bipolar seesaw. The small WD 1age furthermore provides valuable opportunities for precise investigation of the rela- tive phasing of atmospheric greenhouse gas variations and Antarctic climate. (A5) Note that we divide by 1000 to convert from kgm−3 to Mgm−3, the units used by Herron and Langway (1980). We use the mathematical description by Freitag et al. (2013a) to include the hypothesized ﬁrn softening effect of impurities. In this approach an increasing Ca concentration, as a proxy for mineral dust content, lowers the activation en- ergy of ﬁrn, thereby enhancing densiﬁcation rates. This is tantamount to stating that dusty ﬁrn behaves as if it were “warmer” than its climatological temperature. The H-L ac- tivation energies of Eqs. (A2) and (A4) are modiﬁed by [Ca] in the following way: 6 Summary and conclusions We have presented a ﬁrst chronology for the deep (> 2850 m) section of the WAIS Divide ice core, which is based on stratigraphic matching to Greenland ice cores using glob- ally well-mixed methane. We use a dynamical ﬁrn densi- ﬁcation model constrained by δ15N data to calculate past 1age, and ﬁnd that 1age was smaller than 525 ± 120 years for all of the core. Using high-resolution WD records of at- mospheric CH4, we synchronize WD directly to Greenland NGRIP δ18O for the abrupt onset and termination of each of the DO interstadials. To each event we assign an age cor- responding to 1.0063 times its GICC05 age, which brings the ages in agreement with the high-resolution U / Th-dated Hulu speleothem record. The uncertainty in the ﬁnal chronol- ogy is based on the uncertainties in (i) the 1age calculations, as evaluated with a sensitivity study; (ii) the interpolation strategy, as evaluated by comparing four different interpo- lation methods; (iii) determining the timing of events in the An important next step will be to synchronize the WD chronology with other Antarctic cores via volcanic match- ing and other age markers (e.g., Severi et al., 2007; Sigl et al., 2014). Because of the annual-layer count and possi- bility of tight synchronization to Greenland ice cores, WD could contribute to an improved absolute dating of Antarctic cores, as well as improved cross-dating between cores. Such cross-dating could help inform the WD chronology as well, particularly in the deepest part of the core, where the ice is potentially highly strained, as suggested by the interpolation difﬁculties in the 60–65 ka interval (Fig. 8b). With a synchro- nized chronology, WD could improve the representation of West Antarctic climate in Antarctic ice core stacks (Pedro www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 168 References Aciego, S., Bourdon, B., Schwander, J., Baur, H., and Forieri, A.: Toward a radiometric ice clock: uranium ages of the Dome C ice core, Quaternary Sci. Rev., 30, 2389–2397, 2011. The difﬁculty in implementing this model is the following. The densiﬁcation rates of Eqs. (A7–A8) exhibit a discontinu- ity at the critical density D = Dc = 0.6 that cannot be reme- died with the scaling factor γ . On approaching Dc, densiﬁ- cation rates given by Eq. (A7) go to zero (due to the inclu- sion of the term (1 −5 3D), while densiﬁcation rates given by Eq. (A8) go to inﬁnity because the contact area a equals zero at D = Dc. Clearly neither equation gives a realistic result at D = Dc. Therefore, in our implementation of the Arnaud model we use the H-L densiﬁcation rates of Eq. (A1) instead of Eq. (A7) in stage 1. We take the onset of stage 2 to be the density at which Eqs. (A1) and (A8) intercept, thus avoiding the singularity in Eq. A8. This approach has the additional advantages of removing dependence on ad hoc scaling factor γ and introducing realistic temperature dependence for stage 1 densiﬁcation. Because stage 1 spans just the top 10–20 % of the ﬁrn column, the modiﬁcation has only a minor inﬂu- ence on the overall behavior of the Arnaud model. The Gou- jon model code avoids the singularity in Eq. (A8) by extend- ing stage 1 to Dc + ε (Anaïs Orsi, personal communication, 2014), a procedure not described in Goujon et al. (2003). Ahn, J. and Brook, E. J.: Atmospheric CO2 and climate on mil- lennial time scales during the last glacial period, Science, 322, 83–85, 2008. Ahn, J., Brook, E. J., Schmittner, A., and Kreutz, K.: Abrupt change in atmospheric CO2 during the last ice age, Geophys. Res. Lett., 39, L18711, doi:10.1029/2012gl053018, 2012. Alley, R. B.: Firn densiﬁcation by grain-boundary sliding – a 1st model, Journal De Physique, 48, 249–256, 1987. Arnaud, L., Barnola, J. M., and Duval, P.: Physical modeling of the densiﬁcation of snow/ﬁrn and ice in the upper part of polar ice sheets, in: Physics of Ice Core Records, edited by Hondoh, T., 285–305, 2000. Barnola, J. M., Pimienta, P., Raynaud, D., and Korotkevich, Y. References S.: CO2-climate relationship as deduced from the Vostok ice core: a re-examination based on new measurements and on a re- evaluation of the air dating, Tellus B., 43, 83–90, 1991. Battle, M. O., Severinghaus, J. P., Sofen, E. D., Plotkin, D., Orsi, A. J., Aydin, M., Montzka, S. A., Sowers, T., and Tans, P. P.: Controls on the movement and composition of ﬁrn air at the West Antarctic Ice Sheet Divide, Atmos. Chem. Phys., 11, 11007– 11021, doi:10.5194/acp-11-11007-2011, 2011. Appendix A: Densiﬁcation physics 4 we changed the dust sensitivity β; it is clear that we need to simultaneously change α to keep the model well-calibrated to present-day conditions. Due to the fact that the mean late Holocene WD [Ca] is around 0.8 ngg−1, we let α = 1.007/(1−β ln[0.8/0.5]) in the exper- iment of Fig. 4. This approach ensures that the present-day ECa is invariant with β, and equals ECa = 1.007×EHL. This means that whatever value we choose for β, we will obtain a good ﬁt to the present-day 1age, δ15N, and A values that are well known from direct observations (Battle et al., 2011). For the deeper ﬁrn (ρ > ρc, stage 2), we use Eq. (4c) from Herron and Langway (1980), which was ﬁrst derived by Sigfús J. Johnsen. This equation gives the densiﬁcation rate in terms of overburden load, which allows the model to be run dynamically. The stage 2 densiﬁcation rates are given by dρ dt = k2 2 (σz −σzc)(ρice −ρ) ln (ρice −ρc)/(ρice −ρ) , (A3) (A3) with To validate the H-L model 1age simulations, we repeated the ﬁrn modeling using the densiﬁcation physics of Arnaud et al. (2000), which is also the basis of the model by Gou- jon et al. (2003). Our implementation of the Arnaud model is based on the description in the latter paper, with one modiﬁ- cation at the critical density that we outline here. k2 = 575exp −E2 RT , (A4) (A4) where E2 = 21.4 kJmol−1 is the activation energy for stage 2 and σz denotes the ﬁrn overburden load at www.clim-past.net/11/153/2015/ Clim. Past, 11, 153–173, 2015 Clim. Past, 11, 153–173, 2015 169 C. Buizert et al.: The WAIS Divide chronology – Part 1 In the Arnaud model, densiﬁcation in the stage 1 follows the work of Alley (1987), and is given by in making the water-isotopologue measurements; Raytheon Polar Services for logistics support in Antarctica; the 109th New York Air National Guard for airlift in Antarctica; Anaïs Orsi for fruitful discussions; Dan Muhs for feedback on the manuscript; and the dozens of core handlers in the ﬁeld and at the National Ice Core Laboratory for the core processing. This work is funded by the US National Science Foundation through grants 0539232, 0537661 (to K. M. Cuffey), ANT05-38657 (to J. P. Severinghaus.), NSFC 41230524 (to H. Cheng. and R. L. Edwards), 0839093, 1142166 (to J. R. McConnell.), and 1043092 (to E. J. Steig.) and through grants 0230396, 0440817, 0944348, and 0944266 to the Desert Research Institute of Reno Nevada and University of New Hampshire for the collection and distribution of the WAIS Divide ice core and related tasks. We further acknowledge support by the NOAA Climate and Global Change fellowship program, administered by the University Corporation for Atmospheric Re- search (to C. Buizert.) and the Villum Foundation (to M. Winstrup.). dD dt = γ P D2 1 −5 3D , (A7) (A7) with D the relative density D = ρ/ρice, P the overburden pressure, and γ a scaling factor used to make the densiﬁca- tion rates continuous across the critical density Dc. Stage 2 densiﬁcation is given by with D the relative density D = ρ/ρice, P the overburden pressure, and γ a scaling factor used to make the densiﬁca- tion rates continuous across the critical density Dc. Stage 2 densiﬁcation is given by dD dt = kA D2Dc 1 3 a π 1 2 4πP 3aZD 3 , (A8) (A8) with kA = 4.182 × 104 exp −EA RT , (A9) (A9) Edited by: H. Fischer where a is the average contact area between the grains, Z is the coordination number, and EA is the activation en- ergy (60 kJmol−1). Arnaud densiﬁcation rates for stage 3 (D ≥0.9) are describe elsewhere (Goujon et al., 2003; Ar- naud et al., 2000). The Supplement related to this article is available online at doi:10.5194/cp-11-153-2015-supplement. Baumgartner, M., Kindler, P., Eicher, O., Floch, G., Schilt, A., Schwander, J., Spahni, R., Capron, E., Chappellaz, J., Leuen- berger, M., Fischer, H., and Stocker, T. F.: NGRIP CH4 concen- tration from 120 to 10 kyr before present and its relation to a δ15N temperature reconstruction from the same ice core, Clim. Past, 10, 903–920, doi:10.5194/cp-10-903-2014, 2014. Acknowledgements. We thank Mark Twickler and Joseph Souney Jr. of the WAIS Divide Science Coordination Ofﬁce; the Ice Drilling Design and Operations group at the University of Wisconsin for recovering the ice core; O. Maselli, N. Chellman, M. Grieman, J. D’Andrilli, L. Layman, and R. Grinstead for help in making the continuous CH4 and Ca measurements; A. J. Schauer, S. W. Schoenemann, P. D. Neff, and B. Vanden Heuvel for help Bazin, L., Landais, A., Lemieux-Dudon, B., Toyé Ma- hamadou Kele, H., Veres, D., Parrenin, F., Martinerie, P., Ritz, C., Capron, E., Lipenkov, V., Loutre, M.-F., Raynaud, D., Vinther, B., Svensson, A., Rasmussen, S. O., Severi, M., Blu- nier, T., Leuenberger, M., Fischer, H., Masson-Delmotte, V., Chappellaz, J., and Wolff, E.: An optimized multi-proxy, Clim. Past, 11, 153–173, 2015 Clim. Past, 11, 153–173, 2015 C. Buizert et al.: The WAIS Divide chronology – Part 1 170 multi-site Antarctic ice and gas orbital chronology (AICC2012): 120–800 ka, Clim. Past, 9, 1715–1731, doi:10.5194/cp-9-1715- 2013, 2013. forcing during the last deglaciation, Science, 345, 1177–1180, 2014b. forcing during the last deglaciation, Science, 345, 1177–1180, 2014b. Caillon, N., Severinghaus, J. P., Jouzel, J., Barnola, J.-M., Kang, J., and Lipenkov, V. Y.: Timing of atmospheric CO2 and antarctic temperature changes across termination III, Science, 299, 1728– 1731, doi:10.1126/science.1078758, 2003. Bendel, V., Ueltzhoffer, K. J., Freitag, J., Kipfstuhl, S., Kuhs, W. F., Garbe, C. S., and Faria, S. H.: High-resolution variations in size, number and arrangement of air bubbles in the EPICA DML (Antarctica) ice core, J. Glaciol., 59, 972–980, 2013. Capron, E., Landais, A., Chappellaz, J., Schilt, A., Buiron, D., Dahl- Jensen, D., Johnsen, S. J., Jouzel, J., Lemieux-Dudon, B., Louler- gue, L., Leuenberger, M., Masson-Delmotte, V., Meyer, H., Oerter, H., and Stenni, B.: Millennial and sub-millennial scale climatic variations recorded in polar ice cores over the last glacial period, Clim. Past, 6, 345–365, doi:10.5194/cp-6-345- 2010, 2010. Bender, M. L.: Orbital tuning chronology for the Vostok climate record supported by trapped gas composition, Earth Planet. Sc. Lett., 204, 275–289, 2002. Bender, M. L., Barnett, B., Dreyfus, G., Jouzel, J., and Por- celli, D.: The contemporary degassing rate of Ar-40 from the Solid Earth, P. Natl. Acad. Sci. USA, 105, 8232–8237, 2008. Capron, E., Landais, A., Buiron, D., Cauquoin, A., Chappellaz, J., Debret, M., Jouzel, J., Leuenberger, M., Martinerie, P., Masson- Delmotte, V., Mulvaney, R., Parrenin, F., and Prié, F.: Glacial– interglacial dynamics of Antarctic ﬁrn columns: comparison be- tween simulations and ice core air-δ15N measurements, Clim. Past, 9, 983–999, doi:10.5194/cp-9-983-2013, 2013. Blunier, T. and Brook, E. J.: Timing of millennial-scale climate change in Antarctica and Greenland during the last glacial pe- riod, Science, 291, 109–112, 2001. Blunier, T. and Schwander, J.: Gas enclosure in ice: age differ- ence and fractionation, in: Physics of Ice Core Records, edited by: Hondoh, T., Hokkaido University Press, Sapporo, 307–326, 2000. Chappellaz, J., Stowasser, C., Blunier, T., Baslev-Clausen, D., Brook, E. J., Dallmayr, R., Faïn, X., Lee, J. E., Mitchell, L. E., Pascual, O., Romanini, D., Rosen, J., and Schüpbach, S.: High- resolution glacial and deglacial record of atmospheric methane by continuous-ﬂow and laser spectrometer analysis along the NEEM ice core, Clim. Past, 9, 2579–2593, doi:10.5194/cp-9- 2579-2013, 2013. Blunier, T., Chappellaz, J., Schwander, J., Dallenbach, A., Stauf- fer, B., Stocker, T. C. Buizert et al.: The WAIS Divide chronology – Part 1 F., Raynaud, D., Jouzel, J., Clausen, H. B., Hammer, C. U., and Johnsen, S. J.: Asynchrony of Antarctic and Greenland climate change during the last glacial period, Nature, 394, 739–743, 1998. Blunier, T., Spahni, R., Barnola, J.-M., Chappellaz, J., Louler- gue, L., and Schwander, J.: Synchronization of ice core records via atmospheric gases, Clim. Past, 3, 325–330, doi:10.5194/cp- 3-325-2007, 2007. Chiang, J. C. H. and Bitz, C. M.: Inﬂuence of high latitude ice cover on the marine Intertropical Convergence Zone, Clim. Dynam., 25, 477–496, 2005. Broccoli, A. J., Dahl, K. A., and Stouffer, R. J.: Response of the ITCZ to Northern Hemisphere cooling, Geophys. Res. Lett., 33, L01702, doi:10.1029/2005gl024546, 2006. Cuffey, K. M. and Clow, G. D.: Temperature, accumulation, and ice sheet elevation in central Greenland through the last deglacial transition, J. Geophys. Res., 102, 26383–26396, 1997. Cuffey, K. M. and Paterson, W. S. B.: The Physics of Glaciers, 4th edn., Butterworth-Heinemann, Oxford, UK, 2010. Brook, E. J., Sowers, T., and Orchardo, J.: Rapid variations in at- mospheric methane concentration during the past 110,000 years, Science, 273, 1087–1091, 1996. Cuffey, K. M., Clow, G. D., Alley, R. B., Stuiver, M., Wadding- ton, E. D., and Saltus, R. W.: Large arctic temperature change at the Wisconsin-Holocene glacial transition, Science, 270, 455– 458, 1995. Buizert, C., Martinerie, P., Petrenko, V. V., Severinghaus, J. P., Trudinger, C. M., Witrant, E., Rosen, J. L., Orsi, A. J., Ru- bino, M., Etheridge, D. M., Steele, L. P., Hogan, C., Laube, J. C., Sturges, W. T., Levchenko, V. A., Smith, A. M., Levin, I., Con- way, T. J., Dlugokencky, E. J., Lang, P. M., Kawamura, K., Jenk, T. M., White, J. W. C., Sowers, T., Schwander, J., and Blunier, T.: Gas transport in ﬁrn: multiple-tracer characterisa- tion and model intercomparison for NEEM, Northern Greenland, Atmos. Chem. Phys., 12, 4259–4277, doi:10.5194/acp-12-4259- 2012, 2012. Cvijanovic, I. and Chiang, J. C. H.: Global energy budget changes to high latitude North Atlantic cooling and the tropical ITCZ re- sponse, Clim. Dynam., 40, 1435–1452, 2013. Dahl-Jensen, D., Mosegaard, K., Gundestrup, N., Clow, G. D., Johnsen, S. J., Hansen, A. W., and Balling, N.: Past tempera- tures directly from the greenland ice sheet, Science, 282, 268– 271, 1998. Buizert, C., Sowers, T., and Blunier, T.: Assessment of diffu- sive isotopic fractionation in polar ﬁrn, and application to ice core trace gas records, Earth Planet. Sc. Lett., 361, 110–119, doi:10.1016/j.epsl.2012.11.039, 2013. Dansgaard, W. www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 171 Kanner, L. C., Burns, S. J., Cheng, H., and Edwards, R. L.: High- latitude forcing of the south American summer monsoon during the last glacial, Science, 335, 570–573, 2012. Kong, X. G.: A Reﬁned Hulu and Dongge Cave Climate Record and the Timing of Climate Change during the Last Glacial Cycle, Earth Planet. Sci. Lett., in review, 2015. Earth Planet. Sci. Lett., in review, 2015. Kaspers, K. A., van de Wal, R. S. W., van den Broeke, M. R., Schwander, J., van Lipzig, N. P. M., and Brenninkmei- jer, C. A. M.: Model calculations of the age of ﬁrn air across the Antarctic continent, Atmos. Chem. Phys., 4, 1365–1380, doi:10.5194/acp-4-1365-2004, 2004. EPICA Community Members: One-to-one coupling of glacial cli- mate variability in Greenland and Antarctica, Nature, 444, 195– 198, 2006. Etheridge, D. M., Pearman, G. I., and Fraser, P. J.: Changes in tropospheric methane between 1841 and 1978 from a high accumulation-rate Antarctic ice core, Tellus B, 44, 282–294, 1992. Kawamura, K., Severinghaus, J. P., Ishidoya, S., Sugawara, S., Hashida, G., Motoyama, H., Fujii, Y., Aoki, S., and Nakazawa, T.: Convective mixing of air in ﬁrn at four po- lar sites, Earth Planet. Sc. Lett., 244, 672–682, 2006. Fleitmann, D., Cheng, H., Badertscher, S., Edwards, R. L., Mudelsee, M., Gokturk, O. M., Fankhauser, A., Pickering, R., Raible, C. C., Matter, A., Kramers, J., and Tuysuz, O.: Timing and climatic impact of Greenland interstadials recorded in sta- lagmites from northern Turkey, Geophys. Res. Lett., 36, L19707, doi:10.1029/2009gl040050, 2009. Kawamura, K., Parrenin, F., Lisiecki, L., Uemura, R., Vimeux, F., Severinghaus, J. P., Hutterli, M. A., Nakazawa, T., Aoki, S., Jouzel, J., Raymo, M. E., Matsumoto, K., Nakata, H., Mo- toyama, H., Fujita, S., Goto-Azuma, K., Fujii, Y., and Watan- abe, O.: Northern Hemisphere forcing of climatic cycles in Antarctica over the past 360 000 years, Nature, 448, 912–916, 2007. Freitag, J., Kipfstuhl, J., Laepple, T., and Wilhelms, F.: Impurity- controlled densiﬁcation: a new model for stratiﬁed polar ﬁrn, J. Glaciol., 59, 1163–1169, 2013a. Freitag, J., Kipfstuhl, S., and Laepple, T.: Core-scale radioscopic imaging: a new method reveals density-calcium link in Antarctic ﬁrn, J. Glaciol., 59, 1009–1014, 2013b. Kindler, P., Guillevic, M., Baumgartner, M., Schwander, J., Landais, A., and Leuenberger, M.: Temperature reconstruction from 10 to 120 kyr b2k from the NGRIP ice core, Clim. Past, 10, 887–902, doi:10.5194/cp-10-887-2014, 2014. Fudge, T. J., Waddington, E. D., Conway, H., Lundin, J. M. C. Buizert et al.: The WAIS Divide chronology – Part 1 D., and Taylor, K.: Interpolation methods for Antarctic ice-core timescales: application to Byrd, Siple Dome and Law Dome ice cores, Clim. Past, 10, 1195–1209, doi:10.5194/cp-10-1195-2014, 2014. Lal, D., Jull, A. J. T., Donahue, D. J., Burtner, D., and Nishi- izumi, K.: Polar ice ablation rates measured using in situ cos- mogenic C-14, Nature, 346, 350–352, 1990. Landais, A., Barnola, J. M., Kawamura, K., Caillon, N., Del- motte, M., Van Ommen, T., Dreyfus, G., Jouzel, J., Masson- Delmotte, V., Minster, B., Freitag, J., Leuenberger, M., Schwan- der, J., Huber, C., Etheridge, D., and Morgan, V.: Firn-air delta N-15 in modern polar sites and glacial-interglacial ice: a model- data mismatch during glacial periods in Antarctica?, Quaternary Sci. Rev., 25, 49–62, 2006. Fujita, S., Hirabayashi, M., Goto-Azuma, K., Dallmayr, R., Satow, K., Zheng, J., and Dahl-Jensen, D.: Densiﬁcation of layered ﬁrn of the ice sheet at NEEM, Greenland, J. Glaciol., 60, 905–921, doi:10.3189/2014JoG14J006, 2014. Goujon, C., Barnola, J. M., and Ritz, C.: Modeling the densiﬁca- tion of polar ﬁrn including heat diffusion: Application to close- off characteristics and gas isotopic fractionation for Antarc- tica and Greenland sites, J. Geophys. Res.-Atmos., 108, 18 pp., doi:10.1029/2002jd003319, 2003. Marcott, S. A., Bauska, T. K., Buizert, C., Steig, E. J., Rosen, J. L., Cuffey, K. M., Fudge, T. J., Severinghaus, J. P., Ahn, J., Kalk, M. L., McConnell, J. R., Sowers, T., Taylor, K. C., White, J. W. C., and Brook, E. J.: Centennial-scale changes in the global carbon cycle during the last deglaciation, Nature, 514, 616–619, 2014. Gow, A. J., Meese, D. A., Alley, R. B., Fitzpatrick, J. J., Anandakr- ishnan, S., Woods, G. A., and Elder, B. C.: Physical and struc- tural properties of the Greenland Ice Sheet Project 2 ice core: a review, J. Geophys. Res.-Oceans, 102, 26559–26575, 1997. Martinerie, P., Lipenkov, V. Y., Raynaud, D., Chappellaz, J., Barkov, N. I., and Lorius, C.: Air content paleo record in the Vos- tok ice core (Antarctica): A mixed record of climatic and glacio- logical parameters, J. Geophys. Res.-Atmos., 99, 10565–10576, 1994. Gundestrup, N. S., Dahl-Jensen, D., Hansen, B. L., and Kelty, J.: Bore-hole survey at Camp Century, 1989, Cold Reg. Sci. Tech- nol., 21, 187–193, 1993. Herron, M. M. and Langway, C. C.: Firn densiﬁcation: an empirical model, J. Glaciol., 25, 373–385, 1980. McConnell, J. R., Lamorey, G. W., Lambert, S. W., and Tay- lor, K. C. Buizert et al.: The WAIS Divide chronology – Part 1 and Johnsen, S.: A ﬂow model and a time scale for the ice core from Camp Century, Greenland, J. Glaciol., 8, 215– 223, 1969. Dlugokencky, E., Myers, R., Lang, P., Masarie, K., Crotwell, A., Thoning, K., Hall, B., Elkins, J., and Steele, L.: Conversion of NOAA atmospheric dry air CH4 mole fractions to a gravimet- rically prepared standard scale, J. Geophys. Res.-Atmos., 110, D18306, doi:10.1029/2005JD006035, 2005. Buizert, C., Baggenstos, D., Jiang, W., Purtschert, R., Pe- trenko, V. V., Lu, Z.-T., Muller, P., Kuhl, T., Lee, J., Severing- haus, J. P., and Brook, E. J.: Radiometric 81Kr dating identiﬁes 120,000-year-old ice at Taylor Glacier, Antarctica, P. Natl. Acad. Sci. USA, 111, 6876–6881, 2014a. Dreyfus, G. B., Jouzel, J., Bender, M. L., Landais, A., Masson- Delmotte, V., and Leuenberger, M.: Firn processes and δ15N: potential for a gas-phase climate proxy, Quaternary Sci. Rev., 29, 28–42, 2010. Buizert, C., Gkinis, V., Severinghaus, J. P., He, F., Lecavalier, B. S., Kindler, P., Leuenberger, M., Carlson, A. E., Vinther, B., Masson-Delmotte, V., White, J. W. C., Liu, Z., Otto-Bliesner, B., and Brook, E. J.: Greenland temperature response to climate Edwards, R. L., Cheng, H., Wang, Y. J., Yuan, D. X., Kelly, M. J., Severinghaus, J. P., Burnett, A., Wang, X. F., Smith, E., and Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 172 Spencer, M. K.: Carbon and hydrogen isotopic composition of methane over the last 1000 years, Global Biogeochem. Cy., 23, GB4024, doi:10.1029/2009gb003460, 2009. S. J., Fischer, H., Gkinis, V., Guillevic, M., Hoek, W. Z., Lowe, J. J., Pedro, J. B., Popp, T., Seierstad, I. K., Steffensen, J. P., J. J., Pedro, J. B., Popp, T., Seierstad, I. K., Steffensen, J. P., Svensson, A. M., Vallelonga, P., Vinther, B. M., Walker, M. J., Wheatley, J. J., and Winstrup, M.: A stratigraphic framework for abrupt climatic changes during the Last Glacial period based on three synchronized Greenland ice-core records: reﬁning and ex- tending the INTIMATE event stratigraphy, Quaternary Sci. Rev., 106, 14–28, 2014. Mitchell, L., Brook, E., Lee, J. E., Buizert, C., and Sowers, T.: Con- straints on the late holocene anthropogenic contribution to the atmospheric methane budget, Science, 342, 964–966, 2013. Mitchell, L. E., Brook, E. J., Sowers, T., McConnell, J. R., and Taylor, K.: Multidecadal variability of atmospheric methane, 1000–1800 C.E, J. Geophys. Res., 116, G02007, doi:10.1029/2010jg001441, 2011. Reimer, P., Bard, E., Bayliss, A., Beck, J., Blackwell, P., Ram- sey, C. B., Buck, C., Cheng, H., Edwards, R. L., Friedrich, M., Grootes, P., Guilderson, T., Haﬂidason, H., Hajdas, I., Hatté, C., Heaton, T., Hoffmann, D., Hogg, A., Hughen, K., Kaiser, K., Kromer, B., Manning, S., Niu, M., Reimer, R., Richards, D., Scott, E., Southon, J., Staff, R., Turney, C., and van der Plicht, J.: IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50 000 Years cal BP, Radiocarbon, 55, 1869–1887, 2013. Re NEEM community members: Eemian interglacial reconstructed from a Greenland folded ice core, Nature, 493, 489–494, 2013. Heaton, T., Hoffmann, D., Hogg, A., Hughen, K., Kaiser, K., Kromer, B., Manning, S., Niu, M., Reimer, R., Richards, D., NGRIP community members: High-resolution record of Northern Hemisphere climate extending into the last interglacial period, Nature, 431, 147–151, 2004. Scott, E., Southon, J., Staff, R., Turney, C., and van der Plicht, J.: IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50 000 Years cal BP, Radiocarbon, 55, 1869–1887, 2013. Orsi, A. J.: Temperature reconstruction at the West Antarctic Ice Sheet Divide, for the last millennium, from the combination of borehole temperature and inert gas isotope measurements, Ph. D. thesis, University of California, San Diego, 2013. Rhodes, R. H., Fain, X., Stowasser, C., Blunier, T., Chappellaz, J., McConnell, J. R., Romanini, D., Mitchell, L. E., and Brook, E. C. Buizert et al.: The WAIS Divide chronology – Part 1 J.: Continuous methane measurements from a late Holocene Green- land ice core: atmospheric and in-situ signals, Earth Planet. Sc. Lett., 368, 9–19, 2013. Parrenin, F., Barker, S., Blunier, T., Chappellaz, J., Jouzel, J., Landais, A., Masson-Delmotte, V., Schwander, J., and Veres, D.: On the gas-ice depth difference (1depth) along the EPICA Dome C ice core, Clim. Past, 8, 1239–1255, doi:10.5194/cp-8-1239- 2012, 2012. Rhodes, R. H. Brook, E. J., Chiang, J. C. H., Blunier, T., Maselli, O. J., McConnell, J. R., Romanini, D., Severing- haus, J. P.: Enhanced tropical methane production in response to iceberg discharge in the North Atlantic, in review, 2015. Parrenin, F., Masson-Delmotte, V., Kohler, P., Raynaud, D., Pail- lard, D., Schwander, J., Barbante, C., Landais, A., Wegner, A., and Jouzel, J.: Synchronous change of atmospheric CO2 and antarctic temperature during the last deglacial warming, Science, 339, 1060–1063, 2013. Rosen, J. L., Brook, E. J., Severinghaus, J. P., Blunier, T., Mitchell, L. E., Lee, J. E., Edwards, J. S., and Gkinis, V.: An ice core record of near-synchronous global climate changes at the Bolling transition, Nat. Geosci., 7, 459–463, 2014. Pedro, J. B., van Ommen, T. D., Rasmussen, S. O., Morgan, V. I., Chappellaz, J., Moy, A. D., Masson-Delmotte, V., and Del- motte, M.: The last deglaciation: timing the bipolar seesaw, Clim. Past, 7, 671–683, doi:10.5194/cp-7-671-2011, 2011. Schwander, J. and Stauffer, B.: Age difference between polar ice and the air trapped in its bubbles, Nature, 311, 45–47, 1984. Schwander, J., Sowers, T., Barnola, J. M., Blunier, T., Fuchs, A., and Malaize, B.: Age scale of the air in the summit ice: Impli- cation for glacial-interglacial temperature change, J. Geophys. Res.-Atmos., 102, 19483–19493, 1997. Pedro, J. B., Rasmussen, S. O., and van Ommen, T. D.: Tight- ened constraints on the time-lag between Antarctic temperature and CO2 during the last deglaciation, Clim. Past, 8, 1213–1221, doi:10.5194/cp-8-1213-2012, 2012. Seierstad, I., Abbott, P., Bigler, M., Blunier, T., Bourne, A., Brook, E., Buchardt, S. L., Buizert, C., Clausen, H. B., Cook, E., Dahl-Jensen, D., Davies, S., Guillevic, M., Johnsen, S., Peder- sen, D., Popp, T., Rasmussen, S. O., Severinghaus, J., Svens- son, A., and Vinther, B.: Consistently dated records from the Greenland GRIP, GISP2 and NGRIP ice cores for the past 104 ka reveal regional millennial-scale isotope gradients with possible Heinrich Event imprint, Quaternary Sci. Rev., 106, 29–46, 2014. Petrenko, V. V., Severinghaus, J. P., Brook, E. C. Buizert et al.: The WAIS Divide chronology – Part 1 C.: Continuous ice-core chemical analyses using induc- tively coupled plasma mass spectrometry, Environ. Sci. Technol., 36, 7–11, 2002. Horhold, M. W., Laepple, T., Freitag, J., Bigler, M., Fischer, H., and Kipfstuhl, S.: On the impact of impurities on the densiﬁcation of polar ﬁrn, Earth Planet. Sc. Lett., 325–326, 93–99, 2012. McConnell, J. R., Edwards, R., Kok, G. L., Flanner, M. G., Zender, C. S., Saltzman, E. S., Banta, J. R., Pasteris, D. R., Carter, M. M., and Kahl, J. D. W.: 20th-century industrial black carbon emissions altered arctic climate forcing, Science, 317, 1381–1384, 2007. Huber, C., Leuenberger, M., Spahni, R., Fluckiger, J., Schwan- der, J., Stocker, T. F., Johnsen, S., Landals, A., and Jouzel, J.: Iso- tope calibrated Greenland temperature record over Marine Iso- tope Stage 3 and its relation to CH4, Earth Planet. Sc. Lett., 243, 504–519, 2006. McManus, J. F., Francois, R., Gherardi, J. M., Keigwin, L. D., and Brown-Leger, S.: Collapse and rapid resumption of Atlantic meridional circulation linked to deglacial climate changes, Na- ture, 428, 834–837, 2004. Johnsen, S. J., Dahl-Jensen, D., Gundestrup, N., Steffensen, J. P., Clausen, H. B., Miller, H., Masson-Delmotte, V., Sveinbjörns- dottir, A. E., and White, J.: Oxygen isotope and palaeotemper- ature records from six Greenland ice-core stations: Camp Cen- tury, Dye-3, GRIP, GISP2, Renland and NorthGRIP, J. Quater- nary Sci., 16, 299–307, 2001. Mischler, J. A., Sowers, T. A., Alley, R. B., Battle, M., McConnell, J. R., Mitchell, L., Popp, T., Sofen, E., and Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 173 Veres, D., Bazin, L., Landais, A., Toyé Mahamadou Kele, H., Lemieux-Dudon, B., Parrenin, F., Martinerie, P., Blayo, E., Blu- nier, T., Capron, E., Chappellaz, J., Rasmussen, S. O., Severi, M., Svensson, A., Vinther, B., and Wolff, E. W.: The Antarctic ice core chronology (AICC2012): an optimized multi-parameter and multi-site dating approach for the last 120 thousand years, Clim. Past, 9, 1733–1748, doi:10.5194/cp-9-1733-2013, 2013. Sigl, M., McConnell, J. R., Toohey, M., Curran, M., Das, S. B., Ed- wards, R., Isaksson, E., Kawamura, K., Kipfstuhl, S., Krüger, K., Layman, L., Maselli, O. J., Motizuki, Y., Motoyama, H., Pas- teris, D. R., and Severi, M.: Insights from Antarctica on volcanic forcing during the Common Era, Nat. Clim. Change, 4, 693–697, 2014. Southon, J., Noronha, A. L., Cheng, H., Edwards, R. L., and Wang, Y.: A high-resolution record of atmospheric 14C based on Hulu Cave speleothem H82, Quaternary Sci. Rev., 33, 32–41, 2012. Voelker, A. H. L.: Global distribution of centennial-scale records for Marine Isotope Stage (MIS) 3: A database, Quaternary Sci. Rev., 21, 1185–1212, 2002. Sowers, T., Bender, M., and Raynaud, D.: Elemental and isotopic composition of occluded O2 and N2 in polar ice, J. Geophys. Res.-Atmos., 94, 5137–5150, 1989. WAIS Divide Project Members: Onset of deglacial warming in West Antarctica driven by local orbital forcing, Nature, 500, 440–444, 2013. Sowers, T., Bender, M., Raynaud, D., and Korotkevich, Y. S.: δ15N of N2 in air trapped in polar ice: A tracer of gas transport in the ﬁrn and a possible constraint on ice age-gas age differences, J. Geophys. Res.-Atmos., 97, 15683–15697, 1992. WAIS Divide Project Members: Precise interpolar phasing of abrupt climate change during the last ice age, in press, 2015. Wang, X. F., Auler, A. S., Edwards, R. L., Cheng, H., Ito, E., and Solheid, M.: Interhemispheric anti-phasing of rainfall during the last glacial period, Quaternary Sci. Rev., 25, 3391–3403, 2006. Steig, E. J., Ding, Q., White, J. W. C., Kuttel, M., Rupper, S. B., Neumann, T. A., Neff, P. D., Gallant, A. J. E., Mayewski, P. A., Taylor, K. C., Hoffmann, G., Dixon, D. A., Schoenemann, S. W., Markle, B. R., Fudge, T. J., Schneider, D. P., Schauer, A. J., Teel, R. P., Vaughn, B. H., Burgener, L., Williams, J., and Ko- rotkikh, E.: Recent climate and ice-sheet changes in West Antarc- tica compared with the past 2000 years, Nat. Geosci., 6, 372–375, 2013. Wang, Y. C. Buizert et al.: The WAIS Divide chronology – Part 1 J., Reeh, N., and Schaefer, H.: Gas records from the West Greenland ice mar- gin covering the Last Glacial Termination: a horizontal ice core, Quaternary Sci. Rev., 25, 865–875, 2006. Rasmussen, S. O., Andersen, K. K., Svensson, A. M., Stef- fensen, J. P., Vinther, B. M., Clausen, H. B., Siggaard- Andersen, M. L., Johnsen, S. J., Larsen, L. B., Dahl-Jensen, D., Bigler, M., Röthlisberger, R., Fischer, H., Goto-Azuma, K., Hansson, M. E., and Ruth, U.: A new Greenland ice core chronol- ogy for the last glacial termination, J. Geophys. Res., 111, D06102, doi:10.1029/2005jd006079, 2006. Severi, M., Becagli, S., Castellano, E., Morganti, A., Traversi, R., Udisti, R., Ruth, U., Fischer, H., Huybrechts, P., Wolff, E., Par- renin, F., Kaufmann, P., Lambert, F., and Steffensen, J. P.: Syn- chronisation of the EDML and EDC ice cores for the last 52 kyr by volcanic signature matching, Clim. Past, 3, 367–374, doi:10.5194/cp-3-367-2007, 2007. Rasmussen, S. O., Abbott, P. M., Blunier, T., Bourne, A. J., Brook, E., Buchardt, S. L., Buizert, C., Chappellaz, J., Clausen, H. B., Cook, E., Dahl-Jensen, D., Davies, S. M., Guille- vic, M., Kipfstuhl, S., Laepple, T., Seierstad, I. K., Severing- haus, J. P., Steffensen, J. P., Stowasser, C., Svensson, A., Val- lelonga, P., Vinther, B. M., Wilhelms, F., and Winstrup, M.: A ﬁrst chronology for the North Greenland Eemian Ice Drilling (NEEM) ice core, Clim. Past, 9, 2713–2730, doi:10.5194/cp-9- 2713-2013, 2013. Severinghaus, J. P., Beaudette, R., Headly, M. A., Taylor, K., and Brook, E. J.: Oxygen-18 of O2 records the impact of abrupt cli- mate change on the terrestrial biosphere, Science, 324, 1431– 1434, 2009. Sigl, M., McConnell, J. R., Layman, L., Maselli, O., McGwire, K., Pasteris, D., Dahl-Jensen, D., Steffensen, J. P., Vinther, B., Ed- wards, R., Mulvaney, R., and Kipfstuhl, S.: A new bipolar ice core record of volcanism from WAIS Divide and NEEM and im- plications for climate forcing of the last 2000 years, J. Geophys. Res.-Atmos., 118, 1151–1169, 2013. Rasmussen, S. O., Bigler, M., Blockley, S. P., Blunier, T., Buchardt, S. L., Clausen, H. B., Cvijanovic, I., Dahl-Jensen, D., Johnsen, Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ C. Buizert et al.: The WAIS Divide chronology – Part 1 J., Cheng, H., Edwards, R. L., An, Z. S., Wu, J. Y., Shen, C. C., and Dorale, J. A.: A high-resolution absolute-dated late pleistocene monsoon record from Hulu Cave, China, Sci- ence, 294, 2345–2348, 2001. Winstrup, M., Svensson, A. M., Rasmussen, S. O., Winther, O., Steig, E. J., and Axelrod, A. E.: An automated approach for annual layer counting in ice cores, Clim. Past, 8, 1881–1895, doi:10.5194/cp-8-1881-2012, 2012. Stocker, T. F. and Johnsen, S. J.: A minimum thermodynamic model for the bipolar seesaw, Paleoceanography, 18, 1087, doi:10.1029/2003pa000920, 2003. Wolff, E. W., Fischer, H., Fundel, F., Ruth, U., Twarloh, B., Lit- tot, G. C., Mulvaney, R., Rothlisberger, R., de Angelis, M., Boutron, C. F., Hansson, M., Jonsell, U., Hutterli, M. A., Lambert, F., Kaufmann, P., Stauffer, B., Stocker, T. F., Stef- fensen, J. P., Bigler, M., Siggaard-Andersen, M. L., Udisti, R., Becagli, S., Castellano, E., Severi, M., Wagenbach, D., Bar- bante, C., Gabrielli, P., and Gaspari, V.: Southern Ocean sea-ice extent, productivity and iron ﬂux over the past eight glacial cy- cles, Nature, 440, 491–496, 2006. Stowasser, C., Buizert, C., Gkinis, V., Chappellaz, J., Schüpbach, S., Bigler, M., Faïn, X., Sperlich, P., Baumgartner, M., Schilt, A., and Blunier, T.: Continuous measurements of methane mixing ratios from ice cores, Atmos. Meas. Tech., 5, 999–1013, 2012, http://www.atmos-meas-tech.net/5/999/2012/. Becagli, S., Castellano, E., Severi, M., Wagenbach, D., Bar- bante, C., Gabrielli, P., and Gaspari, V.: Southern Ocean sea-ice extent, productivity and iron ﬂux over the past eight glacial cy- cles, Nature, 440, 491–496, 2006. Svensson, A., Andersen, K. K., Bigler, M., Clausen, H. B., Dahl-Jensen, D., Davies, S. M., Johnsen, S. J., Muscheler, R., Rasmussen, S. O., Rothlisberger, R., Steffensen, J. P., and Vinther, B. M.: The Greenland Ice Core Chronology 2005, 15– 42 ka. Part 2: Comparison to other records, Quaternary Sci. Rev., 25, 3258–3267, 2006. Wolff, E. W., Chappellaz, J., Blunier, T., Rasmussen, S. O., and Svensson, A.: Millennial-scale variability during the last glacial: the ice core record, Quaternary Sci. Rev., 29, 2828–2838, 2010. Clim. Past, 11, 153–173, 2015 www.clim-past.net/11/153/2015/ www.clim-past.net/11/153/2015/
https://openalex.org/W4235173660	https://www.qeios.com/read/P3HXZ1/pdf	English	null	Ras-Related Protein M-Ras	Definitions	2,020	cc-by	74	Qeios · Definition, February 8, 2020 Open Peer Review on Qeios Open Peer Review on Qeios Ras-Related Protein M-Ras National Cancer Institute National Cancer Institute Qeios ID: P3HXZ1 · https://doi.org/10.32388/P3HXZ1 Source National Cancer Institute. Ras-Related Protein M-Ras. NCI Thesaurus. Code C33955. Ras-related protein M-Ras (208 aa, ~24 kDa) is encoded by the human MRAS gene. This protein is involved in GTPase activity and the modulation of cell proliferation. Qeios ID: P3HXZ1 · https://doi.org/10.32388/P3HXZ1 1/1
https://openalex.org/W4220676565	https://www.researchsquare.com/article/rs-890328/latest.pdf	English	null	Assessing the reliability of zygomatic bone landmarks as guides to reach the inferior orbital fissure in orbitozygomatic osteotomy: anatomical study of 83 human skulls	Neurosurgical review	2,022	cc-by	5,203	Assessing The Reliability of Zygomatic Bone Landmarks as Guides to Reach The Inferior Orbital Fissure in Orbitozygomatic Osteotomy. Anatomical Study Of 83 Human Skulls. Fissure in Orbitozygomatic Osteotomy. Anatomica Study Of 83 Human Skulls. Semyon A Melchenko Burdenko National Medical Research Center of Neurosurgery Vasiliy A Cherekaev Burdenko National Medical Research Center of Neurosurgery Olga Yu Aleshkina Saratov State Medical University Gleb V Danilov Burdenko National Medical Research Center of Neurosurgery Gerald Musa (  gerryMD@outlook.com ) Rossijskij universitet druzby narodov https://orcid.org/0000-0001-8710-8652 Uliya V Strunina Burdenko National Medical Research Center of Neurosurgery Denis A Golbin Burdenko National Medical Research Center of Neurosurgery Nikolay V Lasunin Burdenko National Medical Research Center of Neurosurgery Alexander A Zaychenko Saratov State Medical University R h A ti l Semyon A Melchenko Burdenko National Medical Research Center of Neurosurgery Vasiliy A Cherekaev Burdenko National Medical Research Center of Neurosurgery Olga Yu Aleshkina Saratov State Medical University Gleb V Danilov Burdenko National Medical Research Center of Neurosurgery Gerald Musa (  gerryMD@outlook.com ) Rossijskij universitet druzby narodov https://orcid.org/0000-0001-8710-8652 Uliya V Strunina Burdenko National Medical Research Center of Neurosurgery Denis A Golbin Burdenko National Medical Research Center of Neurosurgery Nikolay V Lasunin Burdenko National Medical Research Center of Neurosurgery Alexander A Zaychenko Saratov State Medical University Research Article Keywords: Orbitozygomatic, craniotomy, inferior orbital fissure, anatomic landmarks Posted Date: September 24th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-890328/v1 Method Eighty-three adult skulls were used in this study. The IOF dimensions and the relationship with the malar eminence, the point superior to the malar eminence, the zygomaticofacial foramen and 3 alternative points (E, C, F) were analyzed. Results. The malar eminence was unacceptable for use as a guide to the IOF. The point superior to the malar eminence was also unacceptable as a guide as only 9.4% and 10.9% were in the projection of the IOF on the right and left, respectively. 59.7% of the total zygomaticofacial foramina fell in the IOF projection. The point F fell in the projection of the IOF in 98.8% and 100.0% on the right and left, respectively. Conclusion. The use of the malar eminence as a guide to reach the IOF is unreliable in one third of cases as it is not easily identified intraoperatively in these cases. The zygomaticofacial foramen cannot be considered a reliable surgical landmark to reach the IOF. The authors recommend using the point “F” which is reliable in 98.8-100% of cases. Research Article Page 1/15 Version of Record: A version of this preprint was published at Neurosurgical Review on January 13th, 2022. See the published version at https://doi.org/10.1007/s10143-021-01726-8. Version of Record: A version of this preprint was published at Neurosurgical Review on January 13th, 2022. See the published version at https://doi.org/10.1007/s10143-021-01726-8. Version of Record: A version of this preprint was published at Neurosurgical Review on January 13th, 2022. See the published version at https://doi.org/10.1007/s10143-021-01726-8. Page 2/15 Objectives To perform an adequate orbito-zygomatic craniotomy, it is very important that the bone cut which passes through the body of the zygoma reaches the inferior orbital fissure (IOF). To reach the IOF, two surface landmarks on the body of the zygoma are described: a point located directly superior to the malar eminence and the zygomaticofacial foramen. The article explores the reliability of these landmarks and three other alternative points to reach the IOF. Introduction The orbitozygomatic approach was proposed in 1984 [6]. It is used for a wide range of pathology, with a rather variable both intra-and extracranial localization. The orbitozygomatic craniotomy has been shown to offer the best angle of approach and working area compared to other craniotomies in this region [5]. It allows a surgeon access to the middle fossa, sellar region, the anterior fossa and the posterior fossa [11]. The advantages of the two-flap over the single flap orbitozygomatic craniotomy [12] are undoubted. Recent modification by Alvaro et al. includes a 3-piece approach [2]. The most commonly used version is the two-flap orbitozygomatic craniotomy, which was described in 1998. In an article by J.M. Zabramski, the author gives a detailed description of performing an orbitozygomatic bone flap [14]. For this method 6 bone cuts are performed (Fig. 1). Recent modification by Alvaro et al. includes a 3-piece approach [2]. The most commonly used version is the two-flap orbitozygomatic craniotomy, which was described in 1998. In an article by J.M. Zabramski, the author gives a detailed description of performing an orbitozygomatic bone flap [14]. For this method 6 bone cuts are performed (Fig. 1). Page 3/15 According to Zabramski et al., one of the most difficult osteotomies is the third one, which begins on the body of the zygoma extending perpendicular to the inferolateral border of the orbit and then runs along the lateral wall of the orbit ending in the anterolateral edge of the inferior orbital fissure (IOF). The importance of this bone cut is mentioned in the anatomy works of C. Martins et al. [10]. In order to perform an orbitozygomatic craniotomy with as little retraction on the contents of the orbit as possible, the osteotomy should be extended to reach the IOF (Fig. 2). Zabramski et al. suggested that the starting point of the third bone cut should lie immediately above the malar eminence. From this point a line is drawn perpendicular to the inferolateral border of the orbit towards the IOF [14]. A number of authors also propose using the zygomaticofacial foramen as a guide. According to classical anatomy works, there are usually two zygomaticofacial foramina on each side through which the zygomaticofacial nerve and artery pass. The nerve is a branch of the zygomatic nerve, one of the terminal branches of the maxillary nerve (V2). Objectives 1. To examine the effectivity of the zygomaticofacial foramen and malar eminence when performing an adequate orbitozygomatic craniotomy 1. To examine the effectivity of the zygomaticofacial foramen and malar eminence when performing an adequate orbitozygomatic craniotomy 2. To introduce and examine 3 alternative points E, C, F identified on the body of the zygoma when performing an orbitozygomatic craniotomy 2. To introduce and examine 3 alternative points E, C, F identified on the body of the zygoma when performing an orbitozygomatic craniotomy Introduction Its zone of innervation overlaps with other branches of the trigeminal nerve, therefore sensory disorders do not occur when it is damaged [10, 7]. We did not find data in the literature on the possibility of using the above-mentioned landmarks depending on the width of the IOF. Also, there is no data on the distance from the inferior edge of the orbit to the anterolateral border of the IOF or the value of such landmarks as malar eminence and zygomaticofacial foramen in relation with each other and their variations. Materials And Methods The morphology of the orbitozygomatic complex was studied on 83 adult human skulls from the fundamental anatomic museum at the Department of Human Anatomy of Saratov State Medical University named after I. I. Razumovsky. All the skulls belonged to natives of Saratov and Saratov region who lived in this territory in the 20th century. The main inclusion criteria in the study were preservation of bone structures and unaltered skull anatomy. Measurements were made using nearly 0.01 mm scale spreading calipers according to the conventional methods of craniology. All skulls were divided according to age, sex, skull shape and basilar angle type. The anterolateral IOF margin width (Fig. 3a) and the distance from the anterolateral point of the IOF along the inferolateral edge of the orbit extending to the inferior margin of the orbit was measured (Fig. 3b). Other variables included: presence and prominence of the malar eminence, number of zygomaticofacial foramina, their diameter, and the distance between each zygomaticofacial foramen and the inferior margin of the orbit. The projection of the anterolateral margin of the IOF on the body of the zygoma is the Page 4/15 Page 4/15 zone located between the two vectors. The first one begins from the superior edge of the anterolateral edge of the IOF, continues vertical to the inferolateral edge of the orbit and crossing the body of the zygoma. The second line begins from the inferior edge of the anterolateral part of the IOF and continues vertical to the inferolateral edge of the orbit, and then crosses the body of the zygoma (Fig. 3c). The possibility of finding the malar eminence and zygomaticofacial foramina in the area of IOF projection on the body of the zygoma was assessed (Fig. 3d). The use of three alternative points – E, C, F (identified on the body of the zygoma when performing the orbitozygomatic craniotomy) to reach the IOF was examined. A line was drawn across EC on the body of the zygoma. The EC segment is drawn on a line marked “b” an extension of a line from the lower edge of the zygomatic process onto the body of the zygoma. This line runs parallel to the other line “a” which extends from the superior edge of the zygomatic process. Results The distance from the lateral point of the IOF to the lateral edge of the orbit was 16.2 ± 2.1 mm (median 16 mm, range of 11–22 mm), and 16.1 ± 2.1 mm (median 16 mm, range of 11–21 mm) on the right and left respectively. The width of the IOF on the right side was 5.2 ± 2.4 mm (median 5 mm, range 2–18 mm), and on the left side 5.1 ± 2.6 mm (median 4 mm, range 2–19 mm) (Fig. 5). We evaluated malar eminence for its use as a guide to the third bone cut. If the malar eminence was easy to identify on the body of the zygoma and no difficulties arose when identifying the place of the third bone cut, then such an eminence was labeled an acceptable landmark. Conversely, an absent, barely palpable or very wide eminence was considered unacceptable (Fig. 6). The malar eminence was an unacceptable landmark in 36,1%(N = 30) on the right side and 33,7%(N = 28) on the left side. In cases with acceptable malar eminence, the point located immediately superior to it did not fall into the projection of the IOF on the right side in 9.4%(N = 5) and in 10.9%(N = 6) on the left side (Fig. 7). The number of zygomaticofacial foramina ranged from 0 to 4 small foramina. One foramen was present on the right in 42,2%(N = 35) and 44,6%(N = 37) on the left. Two foramina were found in 37.3%(N = 31) on the right side and in 36.1% (N = 30) on the left side, three foramina on the right – in 12.0% N = 10 and on the left - in 8.4% (N = 7), and four foramina on the right in 3.6%(N = 3) and on the left in 4.8%(N = 4). The zygomaticofacial foramen was not visualized on the right in 4,8%(N = 4) and on the left – in 6,0%(N = 5). For skulls with one zygomaticofacial foramen, its average diameter was 1.7 ± 0.8 mm (median 2 mm, range 0.2-3 mm) on the right and 1.9 ± 0.8 mm (median 2 mm, range 0.2–3.5 mm) on the left. Materials And Methods From point B, located on the elevation of the lateral edge of the frontal process of the zygoma, a vertical line is drawn to intersect the line “b” at point C i.e., line BC. Another line perpendicular to line "b" is drawn from point D which is the point located on the medial edge of the frontal process of the zygoma, to the point E on the body of the zygoma i.e., line “DE”. A midpoint between E and C is marked point F (Fig. 4). Results In cases with three foramina on the right, the average diameter of the upper foramina was 1.7 ± 1.6 mm (media 1.25 mm, range 0.3-6 mm), middle foramina 1.3 ± 0.7 mm (median 1.25 mm, range 0.3–2.5 mm), lower foramina 1.1 ± 0.7 mm (median 1 mm, range 0.3-2 mm), and on the left, the upper foramina was 1.4 ± 0.8 mm (median 1 mm, range 0.3–2.5 mm), middle foramina 0.8 ± 0.5 mm (median 0.5 mm, range 0.3–1.5 mm), the lower foramina 1.1 ± 0.9 mm (median 1 mm, range 0.3-3 mm). In case of four foramina on the right, the average diameter of the upper foramina was 2.2 ± 0.3 mm (median 2 mm, range 2-2.5 mm), the middle upper foramina 1 ± 0 mm, the middle upper foramina 1.6 ± 0.7 mm (median 2 mm, range 0.8- 2mm) and the lower foramina 1.3 ± 1.1 mm (median 0.8 mm, range 0.5–2.5 mm) distribution. On the left the average diameter of the upper foramina was 1.8 ± 0.5 mm (median 2 mm, range 1–2 mm), the middle upper foramina 1.3 ± 1.2 mm (median 0.85 mm, range 0.5-3 mm), the middle lower foramina 1.3 ± 0.7 mm (median 1.5 mm, range 0.3-2 mm) and the lower foramina 0.7 ± 0.4 mm (median 0.75 mm, range 0.3- 1 mm). There was no statistically significant difference in the size of the foramina. Of the total number of zygomaticofacial foramina (N = 273, right – 139, left – 134) 59.7%(N = 163, right – 83, left – 80) were in the projection of the IOF. In cases with one foramen, it was in the projection of the IOF on the right in 60% and on the left in 56.8% of cases. With two foramina, the lower foramen commonly falls in the projection of the IOF. On the right, the upper and lower foramen were in the IOF projection in 29% and 87.1%, respectively, and on the left – in 36.7% and 83.3%, respectively. With three zygomaticofacial foramina, we did not find statistically significant results considering the location of the foramina in the projection of the IOF. On the right, the upper, middle and lower foramina were in the projection of the IOF in 20%, 60% and 70%, respectively and in 28.6%, 71.4% and 28.6% on the left respectively. Results With four zygomaticofacial foramina, we did not find statistically significant results considering location of the foramina in the projection of the IOF. The upper, upper-middle, lower-middle and lower foramina were in the projection of the IOF in 33.3%, 100%, 100%, and 66.7% on the right, respectively and in 25%, 75%, 100% and 50% on the left, respectively (Fig. 8). For the proposed alternative landmarks E, F and C, the probability of correspondence with the IOF projection was calculated. The point E fell within the projection of the IOF in 52.4%(N = 45) on the right and in 49.3%(N = 41) – on the left; the point C in 73.4%(N = 61) on the right and in 68.6%(N = 57) on the left; the point F – in 98.8%(N = 82) on the right and in 100%(N = 83) on the left. There was a direct relationship between points E and C and the width of the IOF. The wider the IOF is, the more likely is that points E and C will fall in the projection of the IOF (p < 0.005) relationship between points E and C and the width of the IOF. The wider the IOF is, the more likely is that points E and C will fall in the projection of the IOF (p < 0.005) Results In cases with two zygomaticofacial foramina on the right, the average diameter of the upper foramina was 1.8 ± 0.5 mm (median 2 mm, range 1-3.5 mm) and the lower – 1.6 ± 0.6 mm (median 2 mm, range 0.3-3mm), Page 5/15 and on the left the average diameter of the upper foramina was 1.5 ± 0.6 mm (median 1.5 mm, range 0.5- 3 mm) and the lower foramina 1.5 ± 0.7 mm (media 1.5 mm, range 0.2-3 mm). In cases with three foramina on the right, the average diameter of the upper foramina was 1.7 ± 1.6 mm (media 1.25 mm, range 0.3-6 mm), middle foramina 1.3 ± 0.7 mm (median 1.25 mm, range 0.3–2.5 mm), lower foramina 1.1 ± 0.7 mm (median 1 mm, range 0.3-2 mm), and on the left, the upper foramina was 1.4 ± 0.8 mm (median 1 mm, range 0.3–2.5 mm), middle foramina 0.8 ± 0.5 mm (median 0.5 mm, range 0.3–1.5 mm), the lower foramina 1.1 ± 0.9 mm (median 1 mm, range 0.3-3 mm). In case of four foramina on the right, the average diameter of the upper foramina was 2.2 ± 0.3 mm (median 2 mm, range 2-2.5 mm), the middle upper foramina 1 ± 0 mm, the middle upper foramina 1.6 ± 0.7 mm (median 2 mm, range 0.8- 2mm) and the lower foramina 1.3 ± 1.1 mm (median 0.8 mm, range 0.5–2.5 mm) distribution. On the left the average diameter of the upper foramina was 1.8 ± 0.5 mm (median 2 mm, range 1–2 mm), the middle upper foramina 1.3 ± 1.2 mm (median 0.85 mm, range 0.5-3 mm), the middle lower foramina 1.3 ± 0.7 mm (median 1.5 mm, range 0.3-2 mm) and the lower foramina 0.7 ± 0.4 mm (median 0.75 mm, range 0.3- 1 mm). There was no statistically significant difference in the size of the foramina. and on the left the average diameter of the upper foramina was 1.5 ± 0.6 mm (median 1.5 mm, range 0.5- 3 mm) and the lower foramina 1.5 ± 0.7 mm (media 1.5 mm, range 0.2-3 mm). Discussion When performing orbito-zygomatic craniotomy, a surgeon is not always able to identify the lateral edge of the IOF. This is due to the fact that good exposure of the IOF requires traction of the contents of the orbit which can lead to damage to these structures and can induce an orbito-cardiac parasympathetic reflex which may cause bradycardia and asystole [13]. In addition, this requires considerable traction on the Page 6/15 myofascial flap which can damage the frontal branch of the facial nerve [3, 8]. The average distance from the orbital edge to the lateral point of the IOF is 16 mm, ranging from a minimum of 11 mm to a maximum of 22 mm. This knowledge allows a surgeon to plan the depth of osteotomy using surgical instruments with a minimal risk of damage to the structures of the IOF. According to literature review, natural landmarks on the body of the zygoma that have been used to perform an osteotomy in the direction of IOF include malar eminence and the zygomaticofacial foramen. However, to be considered a reliable surgical landmark, the anatomical structure must be easily identifiable and relatively constant [10]. But according to our data, the malar eminence in 1/3 of cases is difficult to identify (33.7–36.1%). In some cases where the eminence was easily palpable, the point located directly superior to it did not fall in the projection of the IOF (9.4–10.9%). The zygomaticofacial foramina were absent on the body of the zygoma in some cases (4.8-6.0%). In our findings one zygomaticofacial foramen was seen in slightly less than half of the skulls (42-44.6%) and two foramina were seen in one third of cases (36.1–37.3%), three foramina in 8,4–12,0% and four foramina in 3,6 − 4,8%. In cases with one zygomaticofacial foramen, it was a reliable guide in 56.8–60%. And when two zygomaticofacial foramina were identified, the reliable guide was the lower foramen as it fell in the projection of the IOF in 83.3–87.1%, while the upper foramen was only reliable in 29-36.7%. When three and four foramina were identified, reliability as a landmark ranged from 20–100%. This result was not statistically significant because very few cases with more than 2 foramina were seen. Whether or not the zygomaticofacial foramen falls in the projection of the IOF is not dependent on its diameter. Discussion The use of the zygomaticofacial foramen as a landmark to reach the IOF is controversial and impractical, given the low probability to reach the IOF. Many authors have found this landmark unreliable, considering variability of positions, the number of foramina and their frequent absence [10, 1, 4, 9]. This is in agreement with the above results. The use of the malar eminence as a surgical landmark increases the probability of reaching the IOF during surgery. However, in about one third of patients the protuberance is difficult to identify intraoperatively, making it unreliable as a landmark for orbitozygomatic craniotomy. The greater the width of the IOF, the more likely a surgeon will get into the projection of the IOF when performing the craniotomy. Hence, the need for a cautious CT analysis preoperatively cannot be overemphasized. The use of point F that we have proposed, allows the surgeon to reach the IOF in 98.8–100%. This point is calculated individually on each side and is directly dependent on the anatomic structure of the orbitozygomatic complex. It is easily identified intraoperatively. As far as we know, this is the first description of this point. The major limitation of our research was a relatively small sample size – 83 human skulls. It may have reduced the significance of our results. In the future, including skulls from different geographical regions and different races would help to increase the sample size and reduce the selection bias. Abbreviations IOF inferior orbital fissure Conclusions Page 7/15 Page 7/15 Any surgeon planning to perform an orbitozygomatic craniotomy should be well aware of the anatomical structure of the orbitozygomatic complex. During preoperative planning, it is important to pay attention to the width of the IOF on skull CT. In most cases, it is enough to drive the bone saw to a depth of about 16 mm in order to reach the IOF. The use of the malar eminence as a landmark to reach the IOF is unreliable in one third of cases, as it is difficult to identify intraoperatively. The zygomaticofacial foramen, in our opinion, cannot be considered a reliable surgical landmark to reach the IOF due to its high anatomic variability and low probability of reaching the IOF (13% to 75%, on average 40%). The authors propose using point F which is easy to locate intraoperatively and which has the highest probability of reaching the IOF (98.8-100%). Declarations Funding No funds, grants, or other support was received. Conflicts of interest/Competing interests Not applicable Availability of data and material Not applicable Code availability Not applicable Funding No funds, grants, or other support was received. Conflicts of interest/Competing interests Not applicable Authors' contributions Semyon A. Melchenko: Conceptualization, methodology, Writing - Review & Editing, Supervision Vasiliy A. Cherekaev: Conceptualization, Writing - Review & Editing, Supervision Olga Yu. Alyoshkina: Writing - Review & Editing, Formal analysis Gleb V. Danilov: Writing - Review & Editing, Resources Page 8/15 Musa Gerald: Writing - Original Draft, Resources, Writing - Review & Editing, Uliya V. Strunina: Writing - Original Draft, Supervision, Data curation Denis A. Golbin: Writing - Original Draft, Formal analysis, Visualization Nikolay V. Lasunin: Writing - Review & Editing, Data curation Alexander A. Zaychenko: Writing - Review & Editing, Methodology Musa Gerald: Writing - Original Draft, Resources, Writing - Review & Editing, Uliya V. Strunina: Writing - Original Draft, Supervision, Data curation Denis A. Golbin: Writing - Original Draft, Formal analysis, Visualization Nikolay V. Lasunin: Writing - Review & Editing, Data curation Alexander A. Zaychenko: Writing - Review & Editing, Methodology Consent to participate Not applicable Consent to participate Not applicable Consent for publication Not applicable Consent for publication Not applicable References 1. Aksu F, Ceri NG, Arman C, Zeybek FG, Tetik S (2009) Location and incidence of the zygomaticofacial foramen: An anatomic study. Clinical Anatomy: The Official Journal of the American Association of Clinical Anatomists the British Association of Clinical Anatomists 22:559–562 1. Aksu F, Ceri NG, Arman C, Zeybek FG, Tetik S (2009) Location and incidence of the zygomaticofacial foramen: An anatomic study. Clinical Anatomy: The Official Journal of the American Association of Clinical Anatomists the British Association of Clinical Anatomists 22:559–562 2. Campero A, Martins C, Socolovsky M, Torino R, Yasuda A, Domitrovic L, Rhoton A Jr (2010) Three- piece orbitozygomatic approach. Operative Neurosurgery 66:ons-E119-ons-E120 2. Campero A, Martins C, Socolovsky M, Torino R, Yasuda A, Domitrovic L, Rhoton A Jr (2010) Three- piece orbitozygomatic approach. Operative Neurosurgery 66:ons-E119-ons-E120 3. Cohen-Gadol A (2020) The orbitozygomatic craniotomy and its judicious use. Operative Neurosurgery 18:559–569 3. Cohen-Gadol A (2020) The orbitozygomatic craniotomy and its judicious use. Operative Neurosurgery 18:559–569 4. Ferro A, Basyuni S, Brassett C, Santhanam V (2017) Study of anatomical variations of the zygomaticofacial foramen and calculation of reliable reference points for operation. Br J Oral Maxillofac Surg 55:1035–1041 4. Ferro A, Basyuni S, Brassett C, Santhanam V (2017) Study of anatomical variations of the zygomaticofacial foramen and calculation of reliable reference points for operation. Br J Oral Maxillofac Surg 55:1035–1041 5. Gonzalez LF, Crawford NR, Horgan MA, Deshmukh P, Zabramski JM, Spetzler RF (2002) Working area and angle of attack in three cranial base approaches: pterional, orbitozygomatic, and maxillary extension of the orbitozygomatic approach. Neurosurgery 50:550–557 5. Gonzalez LF, Crawford NR, Horgan MA, Deshmukh P, Zabramski JM, Spetzler RF (2002) Working area and angle of attack in three cranial base approaches: pterional, orbitozygomatic, and maxillary extension of the orbitozygomatic approach. Neurosurgery 50:550–557 6. Hakuba A, Liu S, Nishimura S (1986) The orbitozygomatic infratemporal approach: a new surgical technique. Surg Neurol 26:271–276. doi:10.1016/0090-3019(86)90161-8 6. Hakuba A, Liu S, Nishimura S (1986) The orbitozygomatic infratemporal approach: a new surgical technique. Surg Neurol 26:271–276. doi:10.1016/0090-3019(86)90161-8 7. Kodera T, Arishima H, Yamada S, Arai H, Akazawa A, Higashino Y, Kitai R, Iino S, Bertalanffy H, Kikuta K-i (2017) Orbitozygomatic craniotomy with modified zabramski's technique: a technical note and anatomic and clinical findings. World neurosurgery 97:49–57 8. Lemole GM, Henn JS, Zabramski JM, Spetzler RF (2003) Modifications to the orbitozygomatic approach. Journal of neurosurgery 99:924–930 8. Page 8/15 Ethics approval The Saratov State Medical University named after I. I. Razumovsky Research Ethics Committee has confirmed that no ethical approval is required. References Lemole GM, Henn JS, Zabramski JM, Spetzler RF (2003) Modifications to the orbitozygomatic approach. Journal of neurosurgery 99:924–930 9. Lone M, Telang A, Rajgopal L, Bhuiyan PS (2016) Location and incidence of the zygomatico-facial foramen in dry human skulls: An anatomical study. Journal of the Anatomical Society of India 65:164–166 9. Lone M, Telang A, Rajgopal L, Bhuiyan PS (2016) Location and incidence of the zygomatico-facial foramen in dry human skulls: An anatomical study. Journal of the Anatomical Society of India 65:164–166 10. Martins C, Li X, Rhoton AL Jr (2003) Role of the zygomaticofacial foramen in the orbitozygomatic craniotomy: anatomic report. Neurosurgery 53:168–172. doi:10.1227/01.neu.0000068841.17293.bb discussion 172 – 163. 11. Ruzevick J, Raza SM, Recinos PF, Chaichana K, Pradilla G, Kim JE, Olivi A, Weingart J, Evans J, Quinones-Hinojosa A (2015) Orbitozygomatic craniotomy using an ultrasonic osteotome for precise osteotomies. Clinical neurology neurosurgery 134:24–27 Page 9/15 Page 9/15 12. Tanriover N, Ulm AJ, Rhoton AL Jr, Kawashima M, Yoshioka N, Lewis SB (2006) One-piece versus two-piece orbitozygomatic craniotomy: quantitative and qualitative considerations. Neurosurgery 58:ONS-229-237; discussion ONS-237. doi:10.1227/01.Neu.0000210010.46680.B4 13. van Furth WR, Agur AMR, Woolridge N, Cusimano MD (2006) The orbitozygomatic approach. Neurosurgery 58:ONS-103-ONS-107 13. van Furth WR, Agur AMR, Woolridge N, Cusimano MD (2006) The orbitozygomatic approach. Neurosurgery 58:ONS-103-ONS-107 14. Zabramski JM, Kiriş T, Sankhla SK, Cabiol J, Spetzler RF (1998) Orbitozygomatic craniotomy. Technical note. J Neurosurg 89:336–341. doi:10.3171/jns.1998.89.2.0336 14. Zabramski JM, Kiriş T, Sankhla SK, Cabiol J, Spetzler RF (1998) Orbitozygomatic craniotomy. Technical note. J Neurosurg 89:336–341. doi:10.3171/jns.1998.89.2.0336 Figures Figure 1 Location of six the 6 bone cuts used to form an orbitozygomatic craniotomy. The figures indicate the order of placement of the cuts; a - Lateral view; b - Anterior view Figure 1 Figure 1 Location of six the 6 bone cuts used to form an orbitozygomatic craniotomy. The figures indicate the order of placement of the cuts; a - Lateral view; b - Anterior view Location of six the 6 bone cuts used to form an orbitozygomatic craniotomy. The figures indicate the order of placement of the cuts; a - Lateral view; b - Anterior view Page 10/15 Figure 2 A skull 3D CT reconstruction with emphasis on the third bone section when performing an orbitozygomatic craniotomy. a - correct bone section, b - incorrect bone section Figure 2 A skull 3D CT reconstruction with emphasis on the third bone section when performing an orbitozygomatic craniotomy. a - correct bone section, b - incorrect bone section A skull 3D CT reconstruction with emphasis on the third bone section when performing an orbitozygomatic craniotomy. a - correct bone section, b - incorrect bone section Page 11/15 Page 11/15 Figure 3 a- The width of the inferior orbital fissure, b- Distance from the anterolateral point of the inferior orbita fissure to the inferolateral edge of the orbit on the axis of the IOF. c- Projection of the anterolateral edg the IOF on the body of the zygoma. d- Natural landmarks found on the body of the zygoma; 1 - a point above the malar eminence, 2 - Zygomaticofacial foramen Figure 3 Figure 3 a- The width of the inferior orbital fissure, b- Distance from the anterolateral point of the inferior orbital fissure to the inferolateral edge of the orbit on the axis of the IOF. c- Projection of the anterolateral edge of the IOF on the body of the zygoma. d- Natural landmarks found on the body of the zygoma; 1 - a point above the malar eminence, 2 - Zygomaticofacial foramen Page 12/15 Page 12/15 Fi 4 Page 13/15 Figure 4 Points E, C and F on the body of the zygoma Figure 5 Figure 4 Figure 4 Points E, C and F on the body of the zygoma Figure 5 Points E, C and F on the body of the zygoma , y yg Figure 5 Page 13/15 Variations for the width of the IOF; moderate, very wide and narrow Figure 6 Variations of the malar eminence: a-Eminence on the body of the Zygoma; b - Eminence is not identified anterior view, c- Eminence not identified lateral view erate, very wide and narrow Variations for the width of the IOF; moderate, very wide and narrow Variations for the width of the IOF; moderate, very wide and narrow Figure 8 The relationship between the zygomaticofacial foramen and the IOF. On the zygoma the number of foramina located inside (black digits) and outside (red digits) of the IOF projection are indicated Figure 6 Variations of the malar eminence: a-Eminence on the body of the Zygoma; b - Eminence is not identified anterior view, c- Eminence not identified lateral view Variations of the malar eminence: a-Eminence on the body of the Zygoma; b - Eminence is not identified anterior view, c- Eminence not identified lateral view Figure 7 Page 14/15 The point above the malar eminence does not fall into the IOF projection Figure 8 The relationship between the zygomaticofacial foramen and the IOF. On the zygoma the number of foramina located inside (black digits) and outside (red digits) of the IOF projection are indicated Supplementary Files The point above the malar eminence does not fall into the IOF projection into the IOF projection point above the malar eminence does not fall into the IOF projection Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. ACTAChecklistforauthors161027.doc Page 15/15
https://openalex.org/W2005203135	https://www.nature.com/articles/cddis2013157.pdf	English	null	Activation of executioner caspases is a predictor of progression-free survival in glioblastoma patients: a systems medicine approach	Cell death and disease	2,013	cc-by	10,241	1Centre for Systems Medicine, Department of Physiology and Medical Physics, St. Stephen’s Green, Dublin, Ireland; 2Department of Experimental Neurosurgery, Centre for Neurology and Neurosurgery, Johann Wolfgang Goethe University Hospital, Frankfurt, Germany; 3Department of Neurosurgery, Centre for Neurology and Neurosurgery, Johann Wolfgang Goethe University Hospital, Frankfurt, Germany and 4Institute of Neurology (Edinger Institute), Goethe-University, Frankfurt, Germany Corresponding author: BM Murphy, Centre for Systems Medicine, Department of Physiology and Medical Physics, York House, St. Stephen’s Green, Dublin 2, Ireland. Tel: þ 353 1 402 2119; Fax: 353 1 402 2447; E-mail: bronamurphy@rcsi.ie Received 15.2.13; revised 22.3.13; accepted 03.4.13; Edited by A Stephanou Keywords: glioblastoma; systems medicine; apoptosis Abbreviations: Apaf-1, apoptotic protease activating factor-1; GBM, glioblastoma; MGMT, O6-methylguanine-DNA methyltransferase; MOMP, mitochondrial outer membrane permeabilization; PARP, poly (ADP-ribose) polymerase; PFS, progression-free survival; Smac/DIABLO, second mitochondrial-derived activator of caspase/ direct inhibitor of apoptosis protein-binding protein with low pI; TMZ, temozolomide; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand; XIAP, X-linked inhibitor of apoptosis protein; zVAD, n-benzyloxycarbonyl-Val-Ala-Asp(O-Me) Activation of executioner caspases is a predictor of progression-free survival in glioblastoma patients: a systems medicine approach A´ C Murphy1, B Weyhenmeyer1, J Schmid1, SM Kilbride1, M Rehm1, HJ Huber1, C Senft2,3, J Weissenb M Mittelbronn4, D Ko¨gel2, JHM Prehn1 and BM Murphy,1 Glioblastoma (GBM) is the most common and aggressive primary brain tumor in adults. GBM cells are highly resistant to apoptosis induced by antitumor drugs and radiotherapy resulting in cancer progression. We assessed whether a systems medicine approach, analysing the ability of tumor cells to execute apoptosis could be utilized to predict the response of GBM patients to treatment. Concentrations of the key proapoptotic proteins procaspase-3, procaspase-9, Smac and Apaf-1 and the antiapopotic protein XIAP were determined in a panel of GBM cell lines and GBM patient tumor resections. These values were used as input for APOPTO-CELL, a systems biological based mathematical model built to predict cellular susceptibility to undergo caspase activation. The modeling was capable of accurately distinguishing between GBM cells that die or survive in response to treatment with temozolomide in 10 of the 11 lines analysed. Importantly the results obtained using GBM patient samples show that APOPTO-CELL was capable of stratifying patients according to their progression-free survival times and predicted the ability of tumor cells to support caspase activation in 16 of the 21 GBM patients analysed. Calculating the susceptibility to apoptosis execution may be a potent tool in predicting GBM patient therapy responsiveness and may allow for the use of APOPTO-CELL in a clinical setting. Citation: Cell Death and Disease (2013) 4, e629; doi:10.1038/cddis.2013.157 & 2013 Macmillan Publishers Limited All rights reserved 2041-4889/13 Citation: Cell Death and Disease (2013) 4, e629; doi:10.1038/cddis.2013.157 & 2013 Macmillan Publishers Limited All rights reserved 2041-4889/13 OPEN OPEN www.nature.com/cddis Activation of executioner caspases is a predictor of progression-free survival in glioblastoma patients: a systems medicine approach Therefore we asked if a systems medicine approach examining the expression of Apaf-1, procaspase-9 and procaspase-3, Smac and XIAP could be utilized to determine GBM patient therapeutic response. We adopted the previously established mathematical systems model, APOPTO-CELL22,23 for our study. This model is based on the biochemical reactions between these critical components of the intrinsic mitochondrial pathway22 and enabled us to predict how the interactions between these apoptotic proteins impacted on apoptosis induction in GBM cell lines and more importantly GBM patient resections. Our study successfully demonstrates that a systems medicine approach analysing the interplay between these ﬁve key apoptotic proteins has clinical applicability and can be utilized to determine the progression-free survival (PFS) of GBM patients. (100%), MZ-51 (93%), MZ-256 (100%), MZ-294 (93%), MZ-304 (100%) and MZ-327 (100%; Figure 1). The second group of cells, which displayed signiﬁcantly reduced survival rates (Po0.01), all o40% after 120 h exposure to TMZ, consisted of U87 (30%), U251 (10%), U343 (8%), U373 (39%) and A172 (35%; Figure 1). Thus, two distinct groups of cell lines displaying signiﬁcantly differing sensitivities to TMZ were evident. Furthermore, when we treated this panel of cell lines with the broad-spectrum kinase inhibitor staurosporine, we also observed a similar pattern of response and could identify the same two distinct groups of cells (Supplementary Figure). GBM cell lines display differential protein expression of apoptosis-related family members. We next sought to establish if the expression levels of key proteins in the mitochondrial apoptotic pathway could be utilized to predict the observed cellular response to treatment. Accordingly, we determined the amounts of ﬁve apoptosis-related family members, namely the proapoptotic proteins Apaf-1, procas- pase-3, procaspase-9 and Smac and the antiapoptotic protein XIAP in our panel of 11 GBM cell lines using quantitative western blotting against HeLa cell extracts, in which concentrations of each protein have previously been determined (Figure 2a).22 The expression of Apaf-1 and procaspase-9 were consistently lower in the GBM cell lines than in the HeLa control (Figures 2b and d). Differential expression of procaspase-3 was observed in the GBM cell line lysates with some of the cell lines expressing greater levels than the HeLa cell line and others expressing lower levels (Figure 2c). Activation of executioner caspases is a predictor of progression-free survival in glioblastoma patients: a systems medicine approach Cell Death and Disease (2013) 4, e629; doi:10.1038/cddis.2013.157; published online 16 May 2013 Subject Category: Experimental Medicine mitochondrial pathway of apoptosis to relay their death signals within their target population.10,11 Brieﬂy, in this pathway upon mitochondrial outer membrane permeabiliza- tion (MOMP), cytochrome c is released from the mitochondria into the cytosol, triggering formation of the apoptosome, which includes the apoptotic protease activating factor-1 (Apaf-1) and procaspase-9.12,13 Apoptosome formation enables acti- vation of the initiator caspase, procaspase-9 and subsequent activation of the executioner caspases, procaspase-3 and -7.13–15 Also released from the mitochondria is Smac/ DIABLO, which prevents the X-linked inhibitor of apoptosis protein (IAPs), XIAP from inhibiting its caspase targets.16,17 Once activated, the executioner caspases cleave key regulatory and structural proteins to produce most of the morphological and biochemical changes, characteristic of this type of cell death.13,15,18,19 Glioblastoma (GBM) is the most common and malignant primary brain tumor in adults.1,2 The current treatment regimen for GBM patients of surgical resection, radiotherapy and temozolomide (TMZ) chemotherapy, can prolong survival of patients for several months.3,4 However, the median survival rate for GBM patients remains dismal at B12 months.5 It is widely recognized that one of the major contributors to the observed resistance of GBMs to chemotherapy and radiotherapy is the dysregulation of apoptotic pathways.6,7 Indeed, numerous survival proteins have been shown to be overexpressed in GBM, and attenuated levels of several proapoptotic proteins have been reported.8,9 However, systematic analyses examining how such complex alterations may impact on the ability of GBMs to successfully evade death signaling pathways have rarely been performed. g g The anticancer treatment strategies of chemotherapy and radiotherapy typically attempt to engage the intrinsic or As the proteins, Apaf-1, procaspase-9 and procaspase-3, Smac and XIAP are all critical players in the intrinsic Systems biology to predict GBM patient survival A´C Murphy et al Systems biology to predict GBM patient survival A´C M h t l 2 2 mitochondrial pathway of apoptosis, their expression proﬁle has previously been analysed in GBM cell lines and tumor samples. Procaspase-9 has been reported to be over- expressed in brain tumors20 while XIAP overexpression has been associated with poor prognosis.21 However, few studies have considered how the interactions of these proteins impact on the execution of apoptosis within GBMs and more importantly determined if the ability of a tumor cell to undergo apoptosis can subsequently be related to patient response to therapy. Activation of executioner caspases is a predictor of progression-free survival in glioblastoma patients: a systems medicine approach Differential expression of Smac (Figure 2e) and XIAP (Figure 2f) was also noted in the GBM cell line lysates with the majority of samples expressing greater levels of Smac (Figure 2e) and varied levels of XIAP (Figure 2f) compared with the HeLa control. Statistical analysis of these protein expression levels relative to treatment outcome (Figure 1), indicated signiﬁcantly higher expression of procaspase-3 in the cell lines that responded to TMZ, than those which did not respond (Mann–Whitney U-test; P ¼ 0.03). No signiﬁcant differences were identiﬁed in the expression of Apaf-1 (P ¼ 0.93), procaspase-9 (P ¼ 0.93), XIAP (P ¼ 0.13) and Smac (P ¼ 0.25). Results Survival rates of individual GBM cell lines differ following treatment with TMZ. The response of GBM patients to TMZ can differ signiﬁcantly.3 To examine this range of responses we treated our panel of GBM cell lines with TMZ. Our panel of cell lines consisted of both commercially available cell lines, A172, U87, U251, U343 and U373 and a cohort of lines derived from patient GBMs, including both primary (MZ-18, MZ-51, MZ-294, MZ-327) and recurrent (MZ-256, MZ-304) tumors.24 Cell survival was analysed using MTT cell viability assays (Figure 1). Following incubation with 150 mM TMZ, a concentration, which has previously been demonstrated to correspond with the levels achieved in the serum of patients during treatment,25 one group of cells had survival rates of 90% or greater even after an incubation period of 120 h and consisted of MZ-18 Figure 1 Treatment with TMZ signiﬁcantly decreases cell survival in a subset of GBM cell lines. The panel of GBM cell lines was treated with TMZ (150 mM) and a cell viability assay was carried out at 24, 48, 72, 96 and 120 h post-treatment. Two groups of cells were evident from this analysis: TMZ tolerant and TMZ sensitive. Data are expressed as mean±S.E.M. from three independent experiments (Student’s t-test at end point, Po0.01) A systems medicine approach, analysing apoptosome- dependent caspase activation based on basal protein expression levels, predicts the ability of each GBM cell line to undergo apoptosis. To investigate if the abundance and the interaction of these apoptotic proteins could be utilized to calculate cell line response to apoptosis-inducing therapy we employed APOPTO-CELL, a systems biological mathematical model developed in our laboratory using HeLa cells.22 APOPTO-CELL is built on the input of quantitatively determined procaspase-3, procaspase-9, Smac, Apaf-1 and XIAP protein levels and its output is the prediction of cellular susceptibility to undergo caspase cleavage and activation. A systems medicine approach, analysing apoptosome- dependent caspase activation based on basal protein expression levels, predicts the ability of each GBM cell line to undergo apoptosis. Results To investigate if the abundance and the interaction of these apoptotic proteins could be utilized to calculate cell line response to apoptosis-inducing therapy we employed APOPTO-CELL, a systems biological mathematical model developed in our laboratory using HeLa cells.22 APOPTO-CELL is built on the input of quantitatively determined procaspase-3, procaspase-9, Smac, Apaf-1 and XIAP protein levels and its output is the prediction of cellular susceptibility to undergo caspase cleavage and activation. The quantities of the apoptosis-related proteins expressed in each GBM cell line (Table 3A; Supplementary Material) were used as input for the model system and the time to achieve 80% substrate cleavage, a level previously determined to be required for apoptosis to proceed Figure 1 Treatment with TMZ signiﬁcantly decreases cell survival in a subset of GBM cell lines. The panel of GBM cell lines was treated with TMZ (150 mM) and a cell viability assay was carried out at 24, 48, 72, 96 and 120 h post-treatment. Two groups of cells were evident from this analysis: TMZ tolerant and TMZ sensitive. Data are expressed as mean±S.E.M. from three independent experiments (Student’s t-test at end point, Po0.01) The quantities of the apoptosis-related proteins expressed in each GBM cell line (Table 3A; Supplementary Material) were used as input for the model system and the time to achieve 80% substrate cleavage, a level previously determined to be required for apoptosis to proceed Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al 3 3 Figure 2 GBM cell lines differ in their basal protein expression levels of apoptotic proteins. Untreated whole-cell lysates of GBM cell lines and a HeLa cell line were analysed by western blotting. Western blots showing basal Apaf-1, procaspase-3, procaspase-9, Smac and XIAP expression levels in 11 GBM cell lines and the HeLa cell line; b-actin was used as a loading control (a). Densitometry analysis of basal protein expression levels of Apaf-1 (b), procaspase-3; Mann–Whitney U-test, Po0.05 (c), procaspase-9 (d), Smac (e) and XIAP (f) in tolerant and sensitive GBM cell lines and the HeLa cell line (white bar). Protein expression levels are displayed as a percentage of HeLa protein expression (100%). Data are expressed as means±S.E.M. from three independent experiments Figure 2 GBM cell lines differ in their basal protein expression levels of apoptotic proteins. Untreated whole-cell lysates of GBM cell lines and a HeLa cell line were analysed by western blotting. Results Kaplan–Meier curve showing that the predicted times to 80% caspase-3 substrate cleavage vary signiﬁcantly between TMZ tolerant and sensitive cell lines ((c); log-rank P ¼ 0.0007) Systems biology to predict GBM patient survival A´C Murphy et al 4 O6-methylguanine-DNA methyltransferase (MGMT) in our cell lines (Table 1), as it has previously been shown that a methylated MGMT promoter is associated with chemosensi- tivity to TMZ.26–28 All sensitive cell lines displayed methylation of their MGMT promoter, explaining their ability to respond to TMZ, but so too did 50% of the cell lines, which were unresponsive to the drug (Table 1), giving an overall predictive rate of 73% (8 of 11 cell lines). Interestingly, therefore, our systems medicine approach was better able to explain the differing responses of the cell lines to TMZ than analysis of the methylation status of the MGMT promoter (predictive rates of 91 versus 73%, respectively). Figure 3 The APOPTO-CELL outputs for TMZ tolerant and sensitive GBM cel lines. The calculated protein expression levels of procaspase-3, Apaf-1 procaspase-9, Smac and XIAP in the 11 GBM cell lines were used as input fo the APOPTO-CELL model software. The time (min) taken to achieve 80% caspase-3 substrate cleavage after cytochrome c release was calculated fo the TMZ tolerant (a) and TMZ sensitive GBM cell lines (b). Kaplan–Meie curve showing that the predicted times to 80% caspase-3 substrate cleavage vary signiﬁcantly between TMZ tolerant and sensitive cell lines ((c); log-rank P ¼ 0.0007) Conﬁrming the induction of apoptosis in cell lines predicted by APOPTO-CELL as TMZ-sensitive. It has been reported that TMZ predominantly induces an apoptotic type of cell death in glioma cells.29 To verify that the GBM cell lines, predicted by APOPTO-CELL as capable of supporting caspase activation upon appropriate stimulation, were under- going apoptosis following exposure to TMZ in vitro, two representative cell lines from the sensitive group, U251 and U343 and two TMZ-tolerant cell lines, MZ294 and MZ304 were selected for further analysis (Figure 4a). Initially, the observed cell death following TMZ treatment was quantiﬁed using Hoechst staining (Figure 4b). The loss of viability in the sensitive cell lines was associated with the appearance of pyknotic nuclei and morphological changes characteristic of apoptosis (Figure 4b; Po0.05). To further examine this mode of cell death, the cells were treated with TMZ in the presence and absence of zVAD and stained with AnnexinV (Figure 4c). Results Western blots showing basal Apaf-1, procaspase-3, procaspase-9, Smac and XIAP expression levels in 11 GBM cell lines and the HeLa cell line; b-actin was used as a loading control (a). Densitometry analysis of basal protein expression levels of Apaf-1 (b), procaspase-3; Mann–Whitney U-test, Po0.05 (c), procaspase-9 (d), Smac (e) and XIAP (f) in tolerant and sensitive GBM cell lines and the HeLa cell line (white bar). Protein expression levels are displayed as a percentage of HeLa protein expression (100%). Data are expressed as means±S.E.M. from three independent experiments efﬁciently,22 was calculated (Figures 3a and b). The APOPTO-CELL output, accurately predicted the GBM cell line sensitivities (Figure 1) in 10 of the 11 cell lines, which corresponds to a predictive rate of 90.9%. All of the cell lines we observed as sensitive to TMZ treatment were also correctly predicted by APOPTO-CELL to achieve appropriate levels of caspase-3 activation within the 60 min simulation timeframe, with a median time of 26 min to achieve 80% substrate cleavage (Figure 3a). Of the cell lines, which failed to respond to TMZ in vitro, 83% also failed to reach sufﬁcient substrate cleavage within the simulation time of 60 min (Figure 3b). The time calculated by APOPTO-CELL to achieve sufﬁcient levels of caspase-3 activation for efﬁcient execution of apoptosis varied signiﬁcantly between the TMZ sensitive and TMZ tolerant groups of cells (log-rank, P ¼ 0.0007; Figure 3c). g ) We also undertook an examination of the methylation status of the promoter region of the DNA-repair enzyme, Cell Death and Disease Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al Figure 3 The APOPTO-CELL outputs for TMZ tolerant and sensitive GBM cell lines. The calculated protein expression levels of procaspase-3, Apaf-1, procaspase-9, Smac and XIAP in the 11 GBM cell lines were used as input for the APOPTO-CELL model software. The time (min) taken to achieve 80% caspase-3 substrate cleavage after cytochrome c release was calculated for the TMZ tolerant (a) and TMZ sensitive GBM cell lines (b). Results APOPTO-CELL outputs for two TMZ-tolerant cell lines, MZ294 and Z304, and two TMZ-sensitive cell lines, U251 and U343 (a). All cell lines were treated with TMZ (150 mM) for 72 h. TMZ-induced apoptosis was assessed by Hoechst and nnexinV positive staining and western blot analysis. Brightﬁeld and UV-light images ( 20 magniﬁcation) of TMZ-treated GBM cells following Hoechst staining (b), ondensed/fragmented nuclei were considered apoptotic, counted and expressed as a percentage of the total cell number ((b); Po0.05, n ¼ 3). The cell lines were treated th TMZ in the presence or absence of zVAD for 72 h and TMZ-induced apoptosis was determined by AnnexinV positive staining using ﬂow cytometry ((c); **Po0.001; ersus control treated cells, þ þ þPo0.001; versus TMZ-treated cells). Data are expressed as mean±S.E.M. from three independent experiments. Western blot analysis of e protein expression levels of procaspase-3, cleaved caspase-3, PARP and cleaved PARP in the cell lines following TMZ treatment for 72 h (d) Systems biology to predict GBM patient survival A´C Murphy et al Figure 4 U251 and U343 cells undergo caspase-dependent apoptosis following treatment with TMZ. APOPTO-CELL outputs for two TMZ-tolerant cell lines, MZ294 and MZ304, and two TMZ-sensitive cell lines, U251 and U343 (a). All cell lines were treated with TMZ (150 mM) for 72 h. TMZ-induced apoptosis was assessed by Hoechst and AnnexinV positive staining and western blot analysis. Brightﬁeld and UV-light images ( 20 magniﬁcation) of TMZ-treated GBM cells following Hoechst staining (b), condensed/fragmented nuclei were considered apoptotic, counted and expressed as a percentage of the total cell number ((b); Po0.05, n ¼ 3). The cell lines were treated with TMZ in the presence or absence of zVAD for 72 h and TMZ-induced apoptosis was determined by AnnexinV positive staining using ﬂow cytometry ((c); **Po0.001; versus control treated cells, þ þ þPo0.001; versus TMZ-treated cells). Data are expressed as mean±S.E.M. from three independent experiments. Western blot analysis of the protein expression levels of procaspase-3, cleaved caspase-3, PARP and cleaved PARP in the cell lines following TMZ treatment for 72 h (d) Figure 4 U251 and U343 cells undergo caspase-dependent apoptosis following treatment with TMZ. APOPTO-CELL outputs for two TMZ-tolerant cell lines, MZ294 and MZ304, and two TMZ-sensitive cell lines, U251 and U343 (a). All cell lines were treated with TMZ (150 mM) for 72 h. TMZ-induced apoptosis was assessed by Hoechst and AnnexinV positive staining and western blot analysis. Results Following treatment, TMZ-sensitive cells were AnnexinV positive in numbers reﬂective of those achieved by the previous Hoechst staining of drug-treated cells (Figures 4b and c). The expression of AnnexinV was signiﬁcantly (Po0.001) but not entirely blocked by zVAD (Figure 4c). In correlation with the results obtained from the MTT viability assay (Figure 1), neither Hoechst nor AnnexinV staining was evident in the treated MZ294 and MZ304 cells (Figures 4b and c). Apoptotic death of the sensitive lines following TMZ treatment was also supported by the activation of the executioner caspase-3, and cleavage of the caspase substrate PARP in the sensitive cell lines, but not in the TMZ-tolerant lines (Figure 4d). Collectively these data Figure 3 The APOPTO-CELL outputs for TMZ tolerant and sensitive GBM cell lines. The calculated protein expression levels of procaspase-3, Apaf-1, procaspase-9, Smac and XIAP in the 11 GBM cell lines were used as input for the APOPTO-CELL model software. The time (min) taken to achieve 80% caspase-3 substrate cleavage after cytochrome c release was calculated for the TMZ tolerant (a) and TMZ sensitive GBM cell lines (b). Kaplan–Meier curve showing that the predicted times to 80% caspase-3 substrate cleavage vary signiﬁcantly between TMZ tolerant and sensitive cell lines ((c); log-rank P ¼ 0.0007) Table 1 Cell line characteristics Cell line Response to TMZ Methylation status of MGMT promoter p53 status U87 Sensitive Methylated Wild-type42 U251 Sensitive Methylated Mutant42 U343 Sensitive Methylated Mutant43 U373 Sensitive Methylated Mutant42 A172 Sensitive Methylated Wild-type42 MZ-18 Tolerant Methylated Unknown MZ-51 Tolerant Methylated Unknown MZ-256 Tolerant Unmethylated Unknown MZ-294 Tolerant Methylated Unknown MZ-304 Tolerant Unmethylated Unknown MZ-327 Tolerant Unmethylated Unknown Abbreviation: TMZ, temozolomide. Figure 3 The APOPTO-CELL outputs for TMZ tolerant and sensitive GBM cell lines. The calculated protein expression levels of procaspase-3, Apaf-1, procaspase-9, Smac and XIAP in the 11 GBM cell lines were used as input for the APOPTO-CELL model software. The time (min) taken to achieve 80% caspase-3 substrate cleavage after cytochrome c release was calculated for the TMZ tolerant (a) and TMZ sensitive GBM cell lines (b). Kaplan–Meier curve showing that the predicted times to 80% caspase-3 substrate cleavage vary signiﬁcantly between TMZ tolerant and sensitive cell lines ((c); log-rank P ¼ 0.0007) Abbreviation: TMZ, temozolomide. Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al 5 igure 4 U251 and U343 cells undergo caspase-dependent apoptosis following treatment with TMZ. Results The concentrations of Apaf-1 (a), procaspase-3; Mann–Whitney U-test, Po0.05 (b), procaspase-9 (c), Smac (d) and XIAP (e) were determined in resections from GBM patients by western blotting and densitometry. These concentrations were compared with a HeLa standard control (white bar), data expressed as mean ±SEM for 3 independent experiments Systems biology to predict GBM patient survival A´C Murphy et al eath and Disease 6 Figure 5 Determination of apoptosis-associated protein concentrations in resections from GBM patients. The patient samples were grouped according to their PFS times. Those with PFS times of 4 7 months were deﬁned as long-term while those with PFS times of o 7 months were deﬁned as short-term. The concentrations of Apaf-1 (a), procaspase-3; Mann–Whitney U-test, Po0.05 (b), procaspase-9 (c), Smac (d) and XIAP (e) were determined in resections from GBM patients by western blotting and densitometry. These concentrations were compared with a HeLa standard control (white bar), data expressed as mean ±SEM for 3 independent experiments HeLa standard (Figure 5b). In contrast and consistent with the ﬁndings in cell lines, levels of procaspase-9 in all GBM resections were consistently lower than the HeLa standard control (Figure 5c). Differential expression of Smac and XIAP was observed in the patient samples with the majority of samples expressing greater levels of Smac (Figure 5d) but lower levels of XIAP expression compared with the HeLa control (Figure 5e). A statistical analysis of these protein expression levels relative to patient PFS times (Table 2), indicated signiﬁcantly higher expression of procaspase-3 in the patients with longer PFS times (47 months) than in those individuals with shorter PFS times (o7 months; Mann–Whitney U-test; P ¼ 0.004). No signi- ﬁcant differences were identiﬁed in the expression of Apaf-1 (P ¼ 0.59), procaspase-9 (P ¼ 0.34), XIAP (P ¼ 0.86) and Smac (P ¼ 0.051). Next, using the calculated concentrations of the apoptosis- related family of proteins as input (Table 3B; Supplementary Material), the APOPTO-CELL model was used to determine the susceptibility towards caspase activation for each patient. Figure 6a shows the APOPTO-CELL predictions for those patients surviving 7 months or longer without tumor reoccur- rence and all were predicted to be capable of supporting executioner caspase activation (Figure 6a). Of the eight patients who had suffered tumor reoccurrence within 7 months, three were also accurately predicted by APOPTO-CELL as being incapable of activating sufﬁcient levels of executioner caspases (Figure 6b). Results Brightﬁeld and UV-light images ( 20 magniﬁcation) of TMZ-treated GBM cells following Hoechst staining (b), condensed/fragmented nuclei were considered apoptotic, counted and expressed as a percentage of the total cell number ((b); Po0.05, n ¼ 3). The cell lines were treated with TMZ in the presence or absence of zVAD for 72 h and TMZ-induced apoptosis was determined by AnnexinV positive staining using ﬂow cytometry ((c); **Po0.001; versus control treated cells, þ þ þPo0.001; versus TMZ-treated cells). Data are expressed as mean±S.E.M. from three independent experiments. Western blot analysis of the protein expression levels of procaspase-3, cleaved caspase-3, PARP and cleaved PARP in the cell lines following TMZ treatment for 72 h (d) (Figures 4b–d) provide strong support for the role of the intrinsic apoptotic pathway as a critical mediator of cell death elicited by TMZ within the sensitive GBM cell lines. Although not a focus of this study, other forms of cell death may also make a contribution, however. (Figures 4b–d) provide strong support for the role of the intrinsic apoptotic pathway as a critical mediator of cell death elicited by TMZ within the sensitive GBM cell lines. Although not a focus of this study, other forms of cell death may also make a contribution, however. that this approach could be applied in a clinical setting to predict patient responses to therapy. Therefore, we analysed tumor resections isolated from 21 GBM patients by quanti- tative western blotting,22 to determine the levels of the proapoptotic proteins Apaf-1, procaspase-3, procaspase-9 and Smac and the antiapoptotic protein XIAP (Figure 5). The expression of Apaf-1 was comparable to the HeLa control (Figure 5a). Examination of the data showed signiﬁcantly greater expression of procaspase-3 in certain GBM patient samples compared with others and when compared with the Application of the systems model, APOPTO-CELL determines PFS of GBM patients. Having identiﬁed the potential use of a systems biology approach to predict the response of GBM cell lines to treatment, we hypothesized Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al Figure 5 Determination of apoptosis-associated protein concentrations in resections from GBM patients. The patient samples were grouped according to their PFS times. Those with PFS times of 4 7 months were deﬁned as long-term while those with PFS times of o 7 months were deﬁned as short-term. Results Analysis of the data showed that the predicted time to achieve 80% substrate cleavage was signiﬁcantly shorter for patients with longer PFS times than for those patients whose tumors had reoccurred within the same timeframe (6.3 min versus 24.6 min, log-rank test, P ¼ 0.0047; Figure 6c). Overall, the APOPTO-CELL model had a predicative rate of 76.2% as the model’s substrate cleavage proﬁles correlated with PFS times in 16 of the 21 patients analysed. We also undertook an analysis of the methylation status of the MGMT promoter within the patient tumor samples (Table 2). Amongst the patients with short PFS times (o7 months), six had unmethylated promoters and the remaining two had inconclusive results despite repeated analysis. For patients with longer PFS times (47 months), ﬁve expressed a methylated promoter and ﬁve expressed an unmethylated promoter. Three samples were again incon- clusive. Regarding patients with short PFS times, the MGMT promoter methylation analysis had a greater predictive rate (100% (6/6) versus 37.5% (3/8)), while for those patients with longer PFS times, the APOPTO-CELL model was more accurate (40% (4/10) versus 100% (13/13)), indicating perhaps that when apoptosis can occur in a system, the APOPTO-CELL model is better able to predict overall outcome. Taking both patient populations into consideration and excluding those with unknown status of their MGMT promoter, the APOPTO-CELL model was more accurate as a predictive determinant of patient PFS times (62.5% (10/16) versus 81.25% (13/16)). However, given the limited number of patient-derived samples and the retrospective analysis of clinical data that were accessible for this study, this result will need to be validated independently and in a larger clinical cohort. Figure 5 Determination of apoptosis-associated protein concentrations in resections from GBM patients. The patient samples were grouped according to their PFS times. Those with PFS times of 4 7 months were deﬁned as long-term while those with PFS times of o 7 months were deﬁned as short-term. The concentrations of Apaf-1 (a), procaspase-3; Mann–Whitney U-test, Po0.05 (b), procaspase-9 (c), Smac (d) and XIAP (e) were determined in resections from GBM patients by western blotting and densitometry. Discussion when compared with the HeLa control. It has previously been published that HeLa cell lines express levels of procaspase-9 that are similar to other cancer cell lines and its levels of Apaf- 1 are lower than other cancer cell lines.22 In contrast to our data, procaspase-9 has previously been shown to be over- expressed in brain tumors.20 There is conﬂicting evidence in the literature as to whether Apaf-1 is upregulated in GBM30 or downregulated.31 Our ﬁnding of low-procaspase-9 expression in GBM patient samples and both low procaspase-9 and Apaf-1 expression in GBM cell lines may account, in part, for the observed apoptosis resistance in GBM as it could result in reduced apoptosome formation after the release of cytochrome c. This study highlights the potential use of a systems medicine approach in determining PFS in GBM patients. We have validated our previous systems biology approach,22,23 and demonstrated that determining the interaction between apoptosis-associated proteins, namely Apaf-1, procaspase-3, procaspase-9, Smac and XIAP, can predict the susceptibility of GBM cell lines to undergo apoptosis following treatment with TMZ. Importantly, analysis of the simulated interplay between quantitatively determined expression levels of these proteins in resected tumor patient samples allowed for the retrospective stratiﬁcation of GBM patients with long or short- term PFS times, 47 months and o7months, respectively, thus delivering a unique and novel prognostic tool. Elucidating the basal expression levels of these proteins in GBM allowed us to examine the validity of using these proteins individually as discriminators of patient or cell line response to treatment. XIAP, a member of the IAPs, is known to inhibit apoptosis by binding to caspase-9 and downstream effector caspases 3 and 7.32 Previous studies have shown XIAP to be primarily highly expressed in GBM cell lines and its overexpression associated with poor prognosis.21 This study demonstrated predominantly low-levels of XIAP expression in patient samples and varied expression levels in the GBM cell lines and these basal levels did not correlate with the PFS times of GBM patients, nor with the susceptibility of GBM cell lines to undergo apoptosis. Similarly, basal levels of procas- pase-9, Apaf-1 or Smac were unable to discriminate between patients with short or long PFS times or GBM cell lines that were sensitive or tolerant to treatment with TMZ. Results These concentrations were compared with a HeLa standard control (white bar), data expressed as mean ±SEM for 3 independent experiments Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al 7 7 Table 2 Patient clinical characteristics Table 2 Patient clinical characteristics Patient number Tumor entity Age Sex KPS Treatment strategy before resection Steroid use Dose reduction PFS (months) Methylation status of MGMT promoter 1 GBM 48 F 100 None No No 17.2 Unmethylated 2 GBM 59 F 90 None No No 16.8 Unknown 3 GBM 22 F 100 None No No Ltf after 46.5 Unmethylated 4 GBM 42 F 90 None No No 21.5 Methylated 5 GBM 68 M 30 None No Unknown Ltf after 10 Unmethylated 6 GBM 66 M 80 None Yes No 11.1 Unknown 7 GBM 71 F 70 None No No 16.1 Unmethylated 8 GBM, recurrence 70 F 80 None No Unknown Ltf after 10 Methylated 9 GBM 59 F 80 None No Yes (thrombocytopenia) 8.4 Unknown 11 GBM 59 M 70 None No No 7.4 Unmethylated 12 GBM 61 M 70 None No No 2.6 Unmethylated 13 GBM 44 M 90 None No No 6.1 Unmethylated 14 GBM 55 F 70 None No No 5.4 Unknown 16 GBM 67 M 70 None Yes No 3.7 Unmethylated 17 GBM 64 F 60 None Yes No 8.9 Methylated 18 GBM, recurrence 53 F 70 Combined radiotherapy and TMZ-therapy before resection No Yes (pancytopenia) 6.9 Unmethylated 19 GBM, recurrence 66 M 80 Combined radiotherapy and TMZ-therapy before resection No No 2.7 Unmethylated 20 GBM 66 M 90 None No No 7.3 Unmethylated 21 GBM 55 F 90 None No No 3.3 Unknown 22 GBM 53 F 60 None No No 7.5 Methylated 23 GBM, recurrence 55 F 90 Combined radiotherapy and TMZ-therapy before resection Yes No 2.3 Unmethylated Abbreviations: GBM, glioblastoma, MGMT, O6-methylguanine-DNA methyltransferase; PFS, progression-free survival; KPS, karnofsky performance score; Ltf, lost to follow up. ma, MGMT, O6-methylguanine-DNA methyltransferase; PFS, progression-free survival; KPS, karnofsky performance score; Ltf, lost Figure 6 APOPTO-CELL outputs of resections from GBM patients. Concentra- tions of Apaf-1, procaspase-3, procaspase-9, Smac, and XIAP in GBM patient resections were entered into the APOPTO-CELL software. The time to 80% substrate cleavage following cytochrome c release was calculated for each sample. APOPTO-CELL outputs for long-term progression-free survivors (a) and for short- term progression-free survivors (b). Kaplan–Meier curve showing that the predicted times to 80% caspase-3 substrate cleavage are signiﬁcantly shorter for patients with long-term PFS times compared with patients with short-term PFS times ((c); median 6.3 min versus 24.6 min, log-rank test, P ¼ 0.0047) Discussion However, due to the lack of alternative therapeutic strategies for patients and the fact that even patients without MGMT promoter methylation can respond to treatment with TMZ,26 as was also highlighted by our methylation analysis of MGMT promoter amongst the patient cohort with longer PFS times, this biomarker analysis has no direct inﬂuence on treatment strategy options for the clinician. Regardless of MGMT promoter methylation status, the prognosis for GBM remains dismal, with the median survival rate for patients with these highly aggressive brain tumors being 12 months.4,34 As a result, new treatment paradigms are continually being explored. Targeting IAPs to release caspases and stimulate apoptosis is being investigated for various cancers, with some IAP-targeting drugs currently in preclinical trials.35 Interest- ingly, targeting XIAP through gene transfer of Smac or exposure to Smac peptides has been shown to sensitize GBM cells in vitro and in vivo to apoptosis induced by tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)36 However, the use of a statistical test as a method of stratifying patients into those likely to respond best to apoptosis-inducing therapies requires a large cohort of patient data to be statistically meaningful and, most crucially, is less able to predict how patients may respond to new therapeutics that interfere with this signaling pathway. Our systems biology approach, APOPTO-CELL, overcomes this obstacle by analyzing the dynamic interactions of ﬁve key apoptotic proteins in an individual patient sample. This method may become highly relevant in the clinical setting, as it may allow for a redeﬁnition of clinical phenotypes, and a stratiﬁcation of patients who may respond and may not respond to radio/ chemotherapy, or who may beneﬁt from novel chemother- apeutics such as Smac mimetics/IAP antagonists. TMZ is the current standard chemotherapeutic drug for GBM patients.3,4 p g p To date, the only signiﬁcant predictive marker of GBM patient response to TMZ is the DNA repair enzyme MGMT. Patients with MGMT promoter methylation have a much better prognosis with TMZ than those without it.27,28 We too observed that the tumors isolated from our patient cohort with short PFS times all expressed unmethylated MGMT promoters. Discussion However, the stratifying patient apoptosis-inducin data to be statistic able to predict how that interfere with t approach, APOP analyzing the dy proteins in an ind become highly rel for a redeﬁnition o patients who may chemotherapy, or apeutics such as S current standard c To date, the only s response to TMZ with MGMT pro prognosis with T observed that the with short PFS t promoters. Howev strategies for patie MGMT promoter TMZ,26 as was als MGMT promoter a times, this bioma treatment strateg MGMT promoter remains dismal, w these highly aggre a result, new tre explored. Targetin apoptosis is being IAP-targeting drug ingly, targeting X exposure to Sma GBM cells in vitro necrosis factor-re Figure 6 APOPTO-C tions of Apaf-1, procas resections were entere substrate cleavage follow APOPTO-CELL outputs term progression-free su times to 80% caspase-3 Cell Death and Disease p p p more comprehensive studies are required to investigate this caveat further. However, the use of a statistical test as a method of stratifying patients into those likely to respond best to apoptosis-inducing therapies requires a large cohort of patient data to be statistically meaningful and, most crucially, is less able to predict how patients may respond to new therapeutics that interfere with this signaling pathway. Our systems biology approach, APOPTO-CELL, overcomes this obstacle by analyzing the dynamic interactions of ﬁve key apoptotic proteins in an individual patient sample. This method may become highly relevant in the clinical setting, as it may allow for a redeﬁnition of clinical phenotypes, and a stratiﬁcation of patients who may respond and may not respond to radio/ chemotherapy, or who may beneﬁt from novel chemother- apeutics such as Smac mimetics/IAP antagonists. TMZ is the current standard chemotherapeutic drug for GBM patients.3,4 To date, the only signiﬁcant predictive marker of GBM patient response to TMZ is the DNA repair enzyme MGMT. Patients with MGMT promoter methylation have a much better prognosis with TMZ than those without it.27,28 We too observed that the tumors isolated from our patient cohort with short PFS times all expressed unmethylated MGMT promoters. Discussion Interestingly, This study collectively assessed the basal levels of Apaf-1, procaspase-3, procaspase-9, Smac and XIAP in GBM patient samples and a panel of human cell lines. Establishing the basal levels of these apoptotic proteins within GBMs is essential as it determines if these tumor cells have the molecular machinery necessary to undergo apoptosis follow- ing MOMP. After the release of cytochrome c from the mitochondria, it binds Apaf-1 resulting in the recruitment of procaspase-9 and formation of the apoptosome.13 We have shown that all patient samples and GBM cell lines express these proteins and therefore, in theory, have the ability to undergo apoptosis following MOMP. However, we observed low-levels of procaspase-9 in GBM patient samples and low-levels of procaspase-9 and Apaf-1 in the GBM cell lines, Cell Death and Disease Systems biology to predict GBM patient survival A´C M h t l Systems biology to predict GBM patient survival A´C Murphy et al 8 8 differential expression of procaspase-3 was observed in the GBM patient samples and in the GBM cell lines, and was found to correlate with TMZ sensitivity in the cell lines and PFS times in patients. High-expression levels of active caspase-3 have been previously shown to be associated with an apoptotic phenotype of cells in GBM patient samples,33 and therefore our data suggest that longer PFS times in patients may be associated with a higher rate of apoptosis within the tumors. Of note, three patients were exposed to chemo- radiation before surgery, and it cannot be excluded that this may have resulted in the altered expression levels of the apoptotic proteins examined for these individuals. In future, more comprehensive studies are required to investigate this caveat further. times in patients. High-expression levels of active caspase-3 have been previously shown to be associated with an apoptotic phenotype of cells in GBM patient samples,33 and therefore our data suggest that longer PFS times in patients may be associated with a higher rate of apoptosis within the tumors. Of note, three patients were exposed to chemo- radiation before surgery, and it cannot be excluded that this may have resulted in the altered expression levels of the apoptotic proteins examined for these individuals. In future, more comprehensive studies are required to investigate this caveat further. may be associate tumors. Of note, radiation before s may have resulte apoptotic proteins more comprehens caveat further. Materials and Methods GBM ll li Th After 4 h, the medium was aspirated and the dark blue crystals were dissolved in DMSO (200 ml). The absorbance of each sample was measured at 560nm using a microplate reader (GENios, Tecan, Mannedorf, Switzerland). The absorbance was considered proportional to the number of viable cells. For quantitative western blotting, standard curves from HeLa cell extracts (5–20 mg) were run concurrently with the GBM cell line and patient lysates (20 mg) to ensure linearity of the signal detection range. The protein concentrations of Apaf-1, procaspase-9, procaspase-3, Smac and XIAP within each GBM cell line and individual GBM patient resected tumor sample were determined by comparison to signals from HeLa cell extracts (Table 3; Supplementary Material). The concentrations of each protein in HeLa cells were previously determined and adapted from Rehm et al.,22 Special care was taken not to overexpose any pixels to guarantee accurate quantiﬁcations. Digital densitometry was performed using ImageJ software after background subtraction by quantifying signal intensities of the chemiluminescence signals and normalized to b actin. For all proteins at least three detections were evaluated. Data are expressed as mean±S.E.M. Patient samples. Patient samples were obtained from the tumor tissue bank at the University Cancer Centre, Frankfurt (Table 2). Their use was approved by the local ethics committee of the Frankfurt University Hospital. These resections were removed from patients before treatment with radiotherapy and TMZ chemotherapy except for patients 18, 19 and 23. PFS time is deﬁned as the period of time between surgical removal of the tumor and subsequent tumor regrowth. For classiﬁcation purposes, any patient who had a PFS time of greater than 7 months was classiﬁed a long-term progression-free survivor and any patient who had a PFS time of less than 7 months was classiﬁed a short-term progression-free survivor. This stratiﬁcation enabled an optimum number of patients (n) in each group (long-term PFS, n ¼ 13; short term PFS, n ¼ 8; Table 2) for meaningful statistical analysis. Flow cytometry. MZ-294, MZ304, U251 and U343 cell were pretreated with zVAD (50 mM) for 30 min and then treated with TMZ (150 mM) for 72 h. Following treatment, monolayer cells were harvested with trypsin-EDTA and washed with PBS. Cells were then incubated at room temperature in binding buffer (10 mM HEPES, 135 mM NaCl, 5 mM CaCl2), which contained an AnnexinV-FITC conjugate (1 ml/ml; BioVision, Mountain View, CA, USA) for 15 min. Materials and Methods GBM ll li Th Materials and Methods GBM cell lines. The commercially available GBM cell lines A172, U87, U251, U343 and U373, as well as the established GBM cell lines MZ-18, MZ-51, MZ-294, MZ-304, MZ-327 and MZ256 were kindly donated by Prof Donat Ko¨gel of Johann Wolfgang Goethe University Hospital, Frankfurt, Germany. The characterization of the commercially available cell lines is available from the American Type Culture Collection and the newly established cell lines have previously been characterized by immunostaining.24 The cells were grown in DMEM with 10% heat-inactivated fetal calf serum, 100 U/ml penicillin and 100 mg/ml streptomycin and were maintained in a humidiﬁed incubator at 37 1C and 5% CO2. HeLa cells were grown and maintained under the same conditions. Western blot analysis. Resected tumor tissue from GBM patients and GBM cell lines were homogenized in lysis buffer containing 0.5 mmol/l Tris-HCl (pH 6.8), 2% SDS (w/v), 10% glycerin (w/v) and protease and phosphatase inhibitor cocktails (Sigma Aldrich). After determining the protein concentration of the samples using a BCA protein assay (Pierce, Rockford, IL, USA), 20 mg samples were boiled in gel-loading buffer and separated on 10–15% SDS-PAGE gels. Proteins were transferred to nitrocellulose membranes using the iBlot gel transfer device (Life Technologies, Invitrogen, Paisley, Scotland). The membranes were incubated with the following antibodies: Smac, caspase-3, caspase-9 (Cell signalling Technology, Beverly, MA, USA), XIAP (BD Biosciences, Oxford, UK) and Apaf-1 (Chemicon, Hampshire, UK). For the GBM cell lines treated with TMZ (150 mM) for 72 h, the membranes were incubated with antibodies against caspase-3, cleaved caspase-3, PARP and cleaved PARP (Cell signalling Technology). Membranes were next incubated with horseradish peroxidase- conjugated secondary antibodies (Jackson ImmunoResearch, Plymouth, PA, USA) and protein bands were visualized using Supersignal West Pico Chemiluminescent Substrate (Pierce). Images were captured using a LAS-3000 Imager equipped with a cooled 12 bit digital CCD camera (FUJIFILM UK Ltd Systems, Shefﬁeld, UK). MTT assay. The GBM cell lines were plated in 96-well plates (500–2000 cells/ well), treated with TMZ (150 mM) and a cell viability assay was carried out 24, 48, 72, 96 and 120 h post-treatment. Following treatment, thiazolyl blue tetrazolium bromide (MTT, 5 mg/ml; Sigma Aldrich, Dublin, Ireland) was added to each well and incubated in the dark at 371C. MTT produces a yellowish solution that is converted to dark blue, water-insoluble MTT formazan by mitochondrial dehydrogenases of living cells. (III) MGMT-unmethylated forward primer: 50-TGTGTTTTTAGAATGTTTTGTGTT TTGAT-30; (IV) MGMT-unmethylated reverse primer: 50-CTACCACCATCCCAAAAAAAAAC TCCA-30. DNA from the glioma cell line LNT229 was used as positive control for a methylated MGMT promoter and DNA from healthy volunteer donors were used as positive control for an unmethylated MGMT promoter status. Water was used as negative control in each MSP run. After MSP, 20 ml of each sample was loaded on a 2% agarose-gel at 120 V for 35 min. In conclusion, few if any studies have examined the combination of these apoptosis-associated proteins or indeed taken into consideration their possible interactions, when investigating potential treatment outcomes. Consequently, our analysis incorporating the interactions amongst these proteins represents a novel strategy for predicting patient outcome in GBM. Considering the wealth of strategies targeting apoptosis in GBM currently in clinical trials,39,40 this systems medicine approach for determining GBM therapeutic response not only has validity for today’s patients but also will be highly valuable in the future treatment of GBM. Hoechst staining of nuclear chromatin. Hoechst staining was performed on both untreated control cells and cells treated with TMZ (150 mM) for 72 h. Cells were stained with Hoechst 33258 (1 mg/ml; Sigma Aldrich), incubated for 10 min at 37 1C and nuclear morphology was visualized using an Eclipse TE 300 inverted microscope (Nikon, Du¨sseldorf, Germany) using a 20 objective. A minimum of 300 cells were counted in three subﬁelds of each culture. Those cells with condensed/fragmented nuclei were deemed apoptotic, counted (using ImageJ software, National Institute of Mental Health, Bethesda, MD, USA) and expressed as a percentage of total cell number. Systems biology to predict GBM patient survival A´C Murphy et al 9 9 9 9 DNA was treated with sodium bisulﬁte using the EZ DNA Methylation-Gold Kit (Zymo Research, ORT, New York, NY, USA). PCR run was performed on the Thermocycler T3000 (Biometra, Gottingen, Germany). For PCR, 2 ml of sodium bisulﬁte-pretreated DNA was ampliﬁed by the following primer sets: and by g-irradiation.37 In GBM, where a decrease in apopto- some components may contribute to drug resistance, it may also prove beneﬁcial to trial agents that increase the expression of apoptosome components Apaf-1 and procaspase-9, for example, by utilizing the pan-histone deacetylase inhibitor LAQ824.38 Thus, APOPTO-CELL may be a useful tool for classifying patients who will beneﬁt from these novel therapies that target the intrinsic apoptotic pathway and may in the future help oncologists make better informed treatment decisions. (I) MGMT-methylated forward primer: 50-GTTTTTAGAACGTTTTGCGTTT CGAC-30; (II) MGMT-methylated reverse primer: 50-CACCGTCCCGAAAAAAAACT CCG-30; (III) MGMT-unmethylated forward primer: 50-TGTGTTTTTAGAATGTTTTGTGTT TTGAT-30; Discussion However, due to the lack of alternative therapeutic strategies for patients and the fact that even patients without MGMT promoter methylation can respond to treatment with TMZ,26 as was also highlighted by our methylation analysis of MGMT promoter amongst the patient cohort with longer PFS times, this biomarker analysis has no direct inﬂuence on treatment strategy options for the clinician. Regardless of MGMT promoter methylation status, the prognosis for GBM remains dismal, with the median survival rate for patients with these highly aggressive brain tumors being 12 months.4,34 As a result, new treatment paradigms are continually being explored. Targeting IAPs to release caspases and stimulate apoptosis is being investigated for various cancers, with some IAP-targeting drugs currently in preclinical trials.35 Interest- ingly, targeting XIAP through gene transfer of Smac or exposure to Smac peptides has been shown to sensitize GBM cells in vitro and in vivo to apoptosis induced by tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)36 Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al Systems biology to predict GBM patient survival A´C M h t l The authors declare no conﬂict of interest. The authors declare no conﬂict of interest. 26. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M et al. MGMT gene silencing and beneﬁt from temozolomide in glioblastoma. N Engl J Med 2005; 352: 997–1003. Acknowledgements. We would like to express our sincere thanks to the patients who kindly agreed to donate their tissue for this study. This research was generously supported by grants from the Health Research Board (RP/2008/69) and the RCSI Research Committee (GR 08-0155) to BMM, the Health Research Board (RP/2008/7) to MR, Science Foundation Ireland grants 08/INV1/1949 and 07/SK/ B1243a to JHMP and Deutsche Krebshilfe (Grant 108795) to DK. 27. Weller M, Felsberg J, Hartmann C, Berger H, Steinbach JP, Schramm J et al. Molecular predictors of progression-free and overall survival in patients with newly diagnosed glioblastoma: a prospective translational study of the German Glioma Network. J Clin Oncol 2009; 27: 5743–5750. 28. Esteller M, Garcia-Foncillas J, Andion E, Goodman SN, Hidalgo OF, Vanaclocha V et al. Inactivation of the DNA-repair gene MGMT and the clinical response of gliomas to alkylating agents. N Engl J Med 2000; 343: 1350–1354. y g g g 29. Roos WP, Batista LF, Naumann SC, Wick W, Weller M, Menck CF et al. Apoptosis in malignant glioma cells triggered by the temozolomide-induced DNA lesion O6-methylguanine. Oncogene 2007; 26: 186–197. 1. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A et al. The 2007 WHO classiﬁcation of tumours of the central nervous system. Acta Neuropathol 2007; 114: 97–109. y g g 30. Johnson CE, Huang YY, Parrish AB, Smith MI, Vaughn AE, Zhang Q et al. Differential Apaf-1 levels allow cytochrome c to induce apoptosis in brain tumors but not in normal neural tissues. Proc Natl Acad SciUSA 2007; 104: 20820–20825. 2. Maher EA, Furnari FB, Bachoo RM, Rowitch DH, Louis DN, Cavenee WK et al. Malignant glioma: genetics and biology of a grave matter. Genes Dev 2001; 15: 1311–1333. 3. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005; 352: 987–996. 31. Watanabe T, Hirota Y, Arakawa Y, Fujisawa H, Tachibana O, Hasegawa M et al. Frequent LOH at chromosome 12q22-23 and Apaf-1 inactivation in glioblastoma. Brain Pathol 2003; 13: 431–439. 4. Materials and Methods GBM ll li Th Cells were counted in a BD LSR II ﬂow cytometer (BD Biosciences). Excitation of AnnexinV-FITC was done with a 488 nm laser and ﬂuorescence emission was collected in the FL1 channel through a 520 nm band pass ﬁlter. A total of 1 104 gated cells were acquired for each sample and analysed using the Flowmax software (Partec, Munster, Germany). MGMT promoter methylation status assessed by methylation- speciﬁc PCR (MSP). Tumor specimens showing the largest amount of vital tumor tissue (goal: 470% vital tumor tissue) were selected for MSP. Four slides of 10-mm thickness were cut from each parafﬁn block. Slides were deparafﬁnized using xylene and 2 96% alcohol. DNA-Isolation was performed using DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). DNA was isolated from the cell lines using QIAamp DNA Mini Kit (Qiagen). Nucleic acid concentration was determined by UV spectrophotometric analysis using Nanodrop 1000 Spectrophotometer. Cell Death and Disease Systems biology to predict GBM patient survival A´C M h t l Systems biology to predict GBM patient survival A´C Murphy et al Systems biology to predict GBM patient survival A´C Murphy et al 10 Mathematical modeling of apoptosis execution. The model used is the biochemical reactions network, APOPTO-CELL, previously published.22,23 This model is based on a set of ordinary differential equations representing a reaction network of 53 reactions, 19 reaction partners and 75 reaction parameters and is implemented in MATLAB. The model was developed in HeLa cells and validated in cellulo and by adapting the model to cancer cell lines deﬁcient in apoptosis signaling proteins.22,41 The reaction network is initiated by cytochrome c-dependent apoptosome formation and mitochondrial Smac release while substrate cleavage by effector caspases is the model output.23 To generate substrate cleavage proﬁles for the GBM cell lines and GBM patient samples, protein expression values for procaspases-9 and -3, XIAP, Smac and Apaf-1, as determined by quantitative western blotting for each individual patient sample and cell line, were used as input for APOPTO-CELL and modeled over 60 min. At 60 min varied substrate cleavage proﬁles were predicted, ranging from o1 to 100%. Apoptosis susceptibility in GBM tumor resections and GBM cell lines was deﬁned as the ability to efﬁciently generate 480% substrate cleavage within this 60 min timeframe. A 480% substrate cleavage was chosen as it has previously been determined that this % enables apoptosis to proceed efﬁciently.22 12. Materials and Methods GBM ll li Th Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 1997; 90: 405–413. 13. Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, Alnemri ES et al. Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade. Cell 1997; 91: 479–489. p 14. Hu Y, Benedict MA, Ding L, Nunez G. Role of cytochrome c and dATP/ATP hydrolysis in Apaf-1-mediated caspase-9 activation and apoptosis. EMBO J 1999; 18: 3586–3595. 15. Zou H, Li Y, Liu X, Wang X. An APAF-1.cytochrome c multimeric complex is a functional apoptosome that activates procaspase-9. J Biol Chem 1999; 274: 11549–11556. 16. Du C, Fang M, Li Y, Li L, Wang X. Smac, a mitochondrial protein that promotes cytochrome c-dependent caspase activation by eliminating IAP inhibition. Cell 2000; 102: 33–42. c-dependent caspase activation by eliminating IAP inhibition. Cell 2000; 102: 33–42. 17. Chai J, Du C, Wu JW, Kyin S, Wang X, Shi Y. Structural and biochemical basis of apoptotic activation by Smac/DIABLO. Nature 2000; 406: 855–862. 18. Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA, Newmeyer DD et al. Ordering the cytochrome c-initiated caspase cascade: hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a caspase-9-dependent manner. J Cell Biol 1999; 144: 281–292. 19. Janicke RU, Sprengart ML, Wati MR, Porter AG. Caspase-3 is required for DNA fragmentation and morphological changes associated with apoptosis. J Biol Chem 1998; 273: 9357–9360. 20. Bodey B, Bodey V, Siegel SE, Nasir A, Coppola D, Hakam A et al. Immunocytochemical detection of members of the caspase cascade of apoptosis in high-grade astrocytomas. In Vivo 2004; 18: 593–602. Statistical analysis. Student’s t-test was used to analyse the response to TMZ in the cell lines. Mann–Whitney U-tests were performed to identify signiﬁcant differences in protein expression relative to b-actin between TMZ sensitive and resistant cell lines and between GBM patient long- and short-term PFS times. To compare the censored times to sufﬁcient caspase activation predicted by systems modeling between TMZ sensitive and resistant cell lines after apoptosis inducing treatment and between patients having long- and short-term PFS, we used log- rank tests on the Kaplan–Meier plots. A P-value of o0.05 was considered to be statistically signiﬁcant. 21. Wagenknecht B, Glaser T, Naumann U, Kugler S, Isenmann S, Bahr M et al. The authors declare no conﬂict of interest. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ, Janzer RC et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC- NCIC trial. Lancet Oncol 2009; 10: 459–466. 32. Deveraux QL, Roy N, Stennicke HR, Van Arsdale T, Zhou Q, Srinivasula SM et al. IAPs block apoptotic events induced by caspase-8 and cytochrome c by direct inhibition of distinct caspases. EMBO J 1998; 17: 2215–2223. 33. Gdynia G, Grund K, Eckert A, Bock BC, Funke B, Macher-Goeppinger S et al. Basal caspase activity promotes migration and invasiveness in glioblastoma cells. Mol Cancer Res 2007; 5: 1232–1240. 5. Ohgaki H, Kleihues P. Population-based studies on incidence, survival rates, and genetic alterations in astrocytic and oligodendroglial gliomas. J Neuropathol Exp Neurol 2005; 64: 479–489. 34. Johnson DR, O’Neill BP. Glioblastoma survival in the United States before and during the temozolomide era. J Neurooncol 2011; 107: 359–364. 6. Krakstad C, Chekenya M. Survival signalling and apoptosis resistance in glioblastomas: opportunities for targeted therapeutics. Mol Cancer 2010; 9: 135. 35. LaCasse EC, Mahoney DJ, Cheung HH, Plenchette S, Baird S, Korneluk RG. IAP-targeted therapies for cancer. Oncogene 2008; 27: 6252–6275. 35. LaCasse EC, Mahoney DJ, Cheung HH, Plenchette S, Baird therapies for cancer. Oncogene 2008; 27: 6252–6275. 7. Ziegler DS, Kung AL, Kieran MW. Anti-apoptosis mechanisms in malignant gliomas. J Clin Oncol 2008; 26: 493–500. therapies for cancer. Oncogene 2008; 27: 6252–6275. 36. Fulda S, Wick W, Weller M, Debatin KM. Smac agonists sensitize for Apo2L/TRAIL- or anticancer drug-induced apoptosis and induce regression of malignant glioma in vivo. Nat Med 2002; 8: 808–815. 8. Stegh AH, Kim H, Bachoo RM, Forloney KL, Zhang J, Schulze H et al. Bcl2L12 inhibits post-mitochondrial apoptosis signaling in glioblastoma. Genes Dev 2007; 21: 98–111. 9. Ruano Y, Mollejo M, Camacho FI, Rodriguez de Lope A, Fiano C, Ribalta T et al. Identiﬁcation of survival-related genes of the phosphatidylinositol 3’-kinase signaling pathway in glioblastoma multiforme. Cancer 2008; 112: 1575–1584. 37. Vellanki SH, Grabrucker A, Liebau S, Proepper C, Eramo A, Braun V et al. Small-molecule XIAP inhibitors enhance gamma-irradiation-induced apoptosis in glioblastoma. Neoplasia 2009; 11: 743–752. 10. Strasser A, Cory S, Adams JM. Deciphering the rules of programmed cell death to improve therapy of cancer and other diseases. EMBO J 2011; 30: 3667–3683. 38. 1. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A et al. The 2007 WHO classiﬁcation of tumours of the central nervous system. Acta Neuropathol 2007; 114: 97–109. Materials and Methods GBM ll li Th Expression and biological activity of X-linked inhibitor of apoptosis (XIAP) in human malignant glioma. Cell Death Differ 1999; 6: 370–376. 22. Rehm M, Huber HJ, Dussmann H, Prehn JH. Systems analysis of effector caspase activation and its control by X-linked inhibitor of apoptosis protein. EMBO J 2006; 25: 4338–4349. 23. Huber HJ, Rehm M, Plchut M, Dussmann H, Prehn JH. APOPTO-CELL—a simulation tool and interactive database for analyzing cellular susceptibility to apoptosis. Bioinformatics 2007; 23: 648–650. 24. Hetschko H, Voss V, Horn S, Seifert V, Prehn JH, Kogel D. Pharmacological inhibition of Bcl-2 family members reactivates TRAIL-induced apoptosis in malignant glioma. J Neurooncol 2008; 86: 265–272. 25. Hammond LA, Eckardt JR, Kuhn JG, Gerson SL, Johnson T, Smith L et al. A randomized phase I and pharmacological trial of sequences of 1,3-bis(2-chloroethyl)-1-nitrosourea and temozolomide in patients with advanced solid neoplasms. Clin Cancer Res 2004; 10: 1645–1656. 2. Maher EA, Furnari FB, Bachoo RM, Rowitch DH, Louis DN, Cavenee WK et al. Malignant glioma: genetics and biology of a grave matter. Genes Dev 2001; 15: 1311–1333. 3. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005; 352: 987–996. The authors declare no conﬂict of interest. Wang S, Yan-Neale Y, Cai R, Alimov I, Cohen D. Activation of mitochondrial pathway is crucial for tumor selective induction of apoptosis by LAQ824. Cell Cycle 2006; 5: 1662–1668. py 11. Kaufmann SH, Vaux DL. Alterations in the apoptotic machinery and their potential role in anticancer drug resistance. Oncogene 2003; 22: 7414–7430. Cell Death and Disease Systems biology to predict GBM patient survival A´C Murphy et al Systems biology to predict GBM patient survival A´C Murphy et al Systems biology to predict GBM patient survival A´C M h t l 11 43. Yin D, Zhou H, Kumagai T, Liu G, Ong JM, Black KL et al. Proteasome inhibitor PS-341 causes cell growth arrest and apoptosis in human glioblastoma multiforme (GBM). Oncogene 2005; 24: 344–354. 39. Kogel D, Fulda S, Mittelbronn M. Therapeutic exploitation of apoptosis and autophagy for glioblastoma. Anticancer Agents Med Chem 2010; 10: 438–449. g g 40. Eisele G, Weller M. Targeting apoptosis pathways in glioblastoma. Cancer Lett 2011; 332: 335–345. 41. O’Connor CL, Anguissola S, Huber HJ, Dussmann H, Prehn JH, Rehm M. Intracellular signaling dynamics during apoptosis execution in the presence or absence of X-linked- inhibitor-of-apoptosis-protein. Biochim Biophys Acta 2008; 1783: 1903–1913. Cell Death and Disease is an open-access journal published by Nature Publishing Group. This work is licensed under a Creative Commons Attribution-NonCommercial- NoDerivs 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/3.0/ 42. Mittelbronn M, Platten M, Zeiner P, Dombrowski Y, Frank B, Zachskorn C et al. Macrophage migration inhibitory factor (MIF) expression in human malignant gliomas contributes to immune escape and tumour progression. Acta Neuropathol 2011; 122: 353–365. accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis) Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis) Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis) Cell Death and Disease Cell Death and Disease
https://openalex.org/W2783658719	https://advancesindifferenceequations.springeropen.com/track/pdf/10.1186/s13662-017-1464-z	English	null	Hopf bifurcation analysis in a fractional-order survival red blood cells model and PDα$\mathit{PD}^{\alpha} $ control	Advances in difference equations	2,018	cc-by	6,971	© The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, pro- vided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. R ES EA RCH Open Access Hopf bifurcation analysis in a fractional-order survival red blood cells model and PDα control Qingshan Sun1, Min Xiao1, Binbin Tao1, Guoping Jiang1, Jinde Cao2,3, Fuchen Zhang4 and Chengdai Huang5 Correspondence: candymanxm2003@aliyun.com 1College of Automation, Nanjing University of Posts and Telecommunications, Nanjing, 210003, China Full list of author information is available at the end of the article Abstract In this paper, we put forward a fractional-order survival red blood cells model and study the dynamics through the Hopf bifurcation. When the delay transcends the threshold, a series of Hopf bifurcations occur at the positive equilibrium. Then, a fractional-order Proportional and Derivative (PDα) controller is applied to the proposed model for the Hopf bifurcation control. It is discovered that by setting proper parameters, the PDα controller can delay or advance the onset of Hopf bifurcations. Therefore the Hopf bifurcation of the fractional-order survival red blood cells model becomes controllable to achieve desirable behaviors. Finally, numerical examples are presented to demonstrate the theoretical analysis. Keywords: Hopf bifurcation; bifurcation control; time delays; PDα controller; survival red blood cells model Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 https://doi.org/10.1186/s13662-017-1464-z Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 https://doi.org/10.1186/s13662-017-1464-z 1 Introduction Fractional calculus was born in 1695 as an important branch of mathematics, almost si- multaneously with classical calculus. Compared with integer-order derivatives, it has been found that fractional derivatives have the superiority of accuracy and ﬂexibility when used to describe some non-classical phenomena in natural science and engineering applications such as neurons [1], ﬁnance systems [2], biological systems [3], and so on. Especially in biological systems, fractional calculus has more advantages than traditional integer-order calculus in describing molecular dynamics with memory characteristics and historical de- pendence [4, 5]. Fractional calculus accumulates the global information of the function in a weighted form, which is also called memory. A large number of examples show that the fractional calculus has a more universal meaning than the integer calculus. In recent years, there have been many papers about the stability and the Hopf bifurcation analysis of integer-order survival red blood cells models [6–9]. With the rapid develop- ment of biomedical and molecular biology, researchers have proposed some new research topics [10–12]. Many biological systems have shown the feature of fractal geometry, the characteristics of memory and diversity discharge activities which cannot be described accurately by using the classical calculus theory. It is known that the integer-order calcu- lus is only determined by the local character of the function, while the fractional-order Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 2 of 12 Page 2 of 12 one can gather the global information of the function in the weighted form [13]. In the biological ﬁeld, Magin argued that the fractional-order derivative can describe the activ- ities of the organism more accurately [10, 11]. In [12], the dynamics of a red blood cells model is fully described by linear fractional-order diﬀerential equations, and the theory of fractional calculus provides a concise way to describe and quantify the biomechanical behaviors of membranes, cells and tissues. However, the qualitative theory of bifurcations in a fractional-order system is still a problem, which has not been researched thoroughly. It is more meaningful to investigate the fractional-order survival red blood cells model instead of the integer-order counterpart. For getting better desirable stability domain, we can add some eﬀective controllers, such as hybrid controller [14, 15], state feedback controller [16, 17] and delayed feedback con- troller [18, 19], for dynamics control in nonlinear systems. As we know, the PID controller consists of the proportional unit P, the integrating unit I and the diﬀerential unit D. We can adjust the stability of the systems by setting the three control parameters kp, ki and kd in the PID controller. The PID controller is mainly applicable to systems with essentially linear and dynamic characteristics which are not changeable through time [20]. In consideration of the feature of fractional-order systems, we develop a fractional-order Proportional and Derivative (PDα) scheme to control the bifurcation of the fractional-order survival red blood cells model in this paper. It is worth mentioning that such a control strategy has not been reported in the control of bifurcation for fractional-order systems. Motivated by the above discussions, we investigate the problem of bifurcation and control for the delayed fractional-order survival red blood cells model in the present paper. 2 Model description 2 Model description 2 Model description In [21], Wazewska-Czyzewska and Lasota proposed the survival red blood cells model: dX dt = –aX(t) + be–cX(t–τ), a > 0,b > 0,c > 0,τ > 0,t ≥0, (2.1) (2.1) where X(t) represents the number of red blood cells in time t, a is the death rate of red blood cells, b and c describe the production of red blood cells per unit time and τ is the necessary time to produce a red blood cell. There are many deﬁnitions of fractional derivatives. The Grünwald-Letnikov deﬁni- tion, the Riemann-Liouville deﬁnition and the Caputo deﬁnition are usually used to deal with fractional-order systems. Since the Caputo derivative only requires the initial con- ditions which are based on integer-order derivative and represents well-understood fea- tures of physical situation, it is more applicable to real world problems. Hence, the Caputo fractional-order derivative is employed in this paper. The Caputo fractional-order derivative is deﬁned as follows: C e Dα t f (t) = 1 (d – α) t e (t – τ)d–α–1f (n)(τ)dτ, (2.2) (2.2) where d – 1 < α < d, d ∈N, and (·) is the gamma function. The symbol α denotes the value of the fractional order that is usually chosen in the range 0 < α ≤1. Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 3 of 12 in Diﬀerence Equations ( 2018) 2018:10 Page 3 of 12 Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 3 of 12 Page 3 of 12 The Laplace transformation of the Caputo fractional-order derivative is represented as L C 0 Dα t f (t) = sαF(s) – d–1 k=0 sα–k–1f (k)(0). (2.3) (2.3) If f (k)(0) = 0, k = 0,1,...,d – 1, then L{C 0 Dα t f (t)} = sαF(s). If f (k)(0) = 0, k = 0,1,...,d – 1, then L{C 0 Dα t f (t)} = sαF(s). 2 Model description Advances in Diﬀerence Equations ( 2018) 2018:10 Page 4 of 12 Then we get Then we get du(t) dt = –mu(t) + e–u(t–υ), t ≥0,m > 0,υ > 0. (2.6) du(t) dt = –mu(t) + e–u(t–υ), t ≥0,m > 0,υ > 0. (2.6) (2.6) In this paper, we focus on the dynamics of the following fractional-order survival red blood cells model with time delays: In this paper, we focus on the dynamics of the following fractional-order survival red blood cells model with time delays: dαu(t) dαt = –mu(t) + e–u(t–υ), t ≥0,m > 0,υ > 0. (2.7) (2.7) We can easily see that, for model (2.6), there is a unique u∗satisfying the following equa- tion: We can easily see that, for model (2.6), there is a unique u∗satisfying the following equa- tion: mu∗= e–u∗. (2.8) mu∗= e–u∗. (2.8) It can be seen that u∗> 1 if and only if 0 < m < 1/e. It can be seen that u∗> 1 if and only if 0 < m < 1/e. 2 Model description A class of n-dimensional linear fractional-order systems with multiple time delays can be represented in the following form [22]: dα1x1 dtα1 = a11x1(t – τ11) + a12x2(t – τ12) + ··· + a1nxn(t – τ1n), dα2x2 dtα2 = a21x1(t – τ21) + a22x2(t – τ22) + ··· + a2nxn(t – τ2n), (2.4) (2.4) dαnxn dtαn = an1x1(t – τn1) + an2x2(t – τn2) + ··· + annxn(t – τnn), with the characteristic equation det ⎛ ⎜⎜⎜⎜⎝ sα1 – a11e–sτ11 –a12e–sτ12 ··· –a1ne–sτ1n –a21e–sτ21 sα2 – a22e–sτ22 ··· –a2ne–sτ2n ... ... ... ... –an1e–sτn1 –an2e–sτn2 ... sαn – anne–sτnn ⎞ ⎟⎟⎟⎟⎠ = 0, (2.5) det ⎛ ⎜⎜⎜⎜⎝ sα1 – a11e–sτ11 –a12e–sτ12 ··· –a1ne–sτ1n –a21e–sτ21 sα2 – a22e–sτ22 ··· –a2ne–sτ2n ... ... ... ... –an1e–sτn1 –an2e–sτn2 ... sαn – anne–sτnn ⎞ ⎟⎟⎟⎟⎠ = 0, (2.5) (2.5) where 0 < αi ≤1 (i = 1,2,3,...,n), and dαi/dtαi is chosen as the Caputo fractional derivative (2.2). where 0 < αi ≤1 (i = 1,2,3,...,n), and dαi/dtαi is chosen as the Caputo fractional derivative (2.2). Theorem 2.1 ([22]) Given that all the roots of the characteristic equation (2.5) have nega- tive real parts, the zero solution of system (2.4) is Lyapunov globally asymptotically stable. Remark 2.1 Theorem 2.1 indicates that the stability boundary for the delayed fractional- order system (2.4) is the imaginary axis. Remark 2.2 If τij = 0, i,j = 1,...,n, then Theorem 2.1 converts into Matignon criterion [23]: if all the roots λs of the equation det(λI – A) = 0 satisfy \|arg(λ)\| > απ/2, then the zero solution of system (2.4) is Lyapunov globally asymptotically stable, where A = (aij)n×n is the coeﬃcient matrix and λ = sα. It can be seen that the stability boundary is described by \|arg(λ)\| = απ/2 (or \|arg(s)\| = π/2) for the fractional-order system (2.4) without delays. Remark 2.3 If all the eigenvalues λs of A satisfy \|arg(λ)\| > απ/2 and the characteristic equation (2.5) has no purely imaginary roots for any τij > 0, i,j = 1,...,n, then the zero solution of system (2.4) is Lyapunov globally asymptotically stable [22]. For model (2.1), we let m = a bc, υ = τbc, u(t) = cX t bc . nces in Diﬀerence Equations ( 2018) 2018:10 Page Sun et al. 3 Bifurcation analysis of the uncontrolled model 3 Bifurcation analysis of the uncontrolled model In this part, we investigate the stability of the fractional-order survival red blood cells model (2.7), and some existence conditions of Hopf bifurcations are addressed. In this part, we investigate the stability of the fractional-order survival red blood cells model (2.7), and some existence conditions of Hopf bifurcations are addressed. Let u(t) – u∗= y(t), then the linearized model (2.7) is Let u(t) – u∗= y(t), then the linearized model (2.7) is dαy dtα = –my(t) – mu∗y(t – υ), dαy dtα = –my(t) – mu∗y(t – υ), (3.1) (3.1) with the characteristic equation sq + m + mu∗e–sυ = 0. (3.2) sq + m + mu∗e–sυ = 0 (3 2) (3.2) sq + m + mu∗e–sυ = 0. (3.2) sq + m + mu∗e–sυ = 0. sq + m + mu∗e–sυ = 0. In the following, we investigate the roots distribution of equation (3.2) by regarding the time delay υ as the bifurcation parameter. In the following, we investigate the roots distribution of equation (3.2) by regarding the time delay υ as the bifurcation parameter. Let s = ω(cos π 2 + isin π 2 ) (ω > 0). Then equation (3.2) becomes ωq cos qπ 2 + ωq sin qπ 2 · i + m + m · u∗(cosωυ – isinωυ) = 0. (3.3) (3.3) Separating the real and imaginary parts gives Separating the real and imaginary parts gives ⎧ ⎨ ⎩ a2 cosωυ = a1 – ωq cos qπ 2 , a2 sinωυ = ωq sin qπ 2 , (3.4) (3.4) where a1 = –m, a2 = mu∗. It can be obtained from (3.4) that where a1 = –m, a2 = mu∗. It can be obtained from (3.4) that ω2q – 2a1ωq cos qπ 2 + a2 1 – a2 2 = 0. (3.5) ω2q – 2a1ωq cos qπ 2 + a2 1 – a2 2 = 0. (3.5) Denote Denote h1(ω) = ω2q – 2a1ωq cos qπ 2 + a2 1 – a2 2. in Diﬀerence Equations ( 2018) 2018:10 Page 5 of 12 Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 5 of 12 Lemma 3.1 For equation (3.2), we have the following: Lemma 3.1 For equation (3.2), we have the following: (i) If m ≥1/e, then all roots of the characteristic equation (3.2) have negative real parts. 3 Bifurcation analysis of the uncontrolled model (ii) If 0 < m < 1/e, then equation (3.2) has a pair of purely imaginary roots ±ω0 when υ = υj, j = 0,1,... , where υj = 1 ω0 arccos a1 – ω0 cos qπ 2 a2 + 2jπ , j = 0,1,2,..., (3.6) where ω0 is the unique positive zero of the function h1(ω). (ii) If 0 < m < 1/e, then equation (3.2) has a pair of purely imaginary roots ±ω0 when υ = υj, j = 0,1,... , where υj = 1 ω0 arccos a1 – ω0 cos qπ 2 a2 + 2jπ , j = 0,1,2,..., (3.6) (3.6) where ω0 is the unique positive zero of the function h1(ω). where ω0 is the unique positive zero of the function h1(ω). Proof (i) From \|a1\| > a2, then h1(0) > 0, and the symmetry axis is a1 cos qπ 2 < 0. Combining q > 0, we can see that equation (3.5) has no real root, so equation (3.2) has no purely imaginary root. This ﬁnishes the proof of (i). Proof (i) From \|a1\| > a2, then h1(0) > 0, and the symmetry axis is a1 cos qπ 2 < 0. Combining q > 0, we can see that equation (3.5) has no real root, so equation (3.2) has no purely imaginary root. This ﬁnishes the proof of (i). (ii) By means of \|a1\| < a2, it is easy to see that h1(0) < 0. Combining q > 0, there exists a unique positive number ω0 such that h1(ω) = 0. Then ω0 is a root of equation (3.5). Hence, for υj as deﬁned in (3.6), (ω0,υj) is a root of equation (3.3). It can be seen that ±ω0 is a pair of purely imaginary roots of equation (3.2), while υ = υj, j = 0,1,... . This completes the proof of (ii). □ □ Remark 3.1 The conclusion (ii) of Lemma 3.1 gives the onset of Hopf bifurcation of model (2.7). Here we make the following assumption: (H1) P1Q1 + P2Q2 Q2 1 + Q2 2 > 0, here (H1) P1Q1 + P2Q2 Q2 1 + Q2 2 > 0, where P1 = mu∗ω0 sinω0υ0, P1 = mu∗ω0 sinω0υ0, P2 = mu∗ω0 cosω0υ0, Q1 = qωq–1 0 cos (q – 1) 2 π – mu∗υ cosω0υ0, Q2 = qωq–1 0 sin (q – 1) 2 π + mu∗υ sinω0υ0. 3 Bifurcation analysis of the uncontrolled model P2 = mu∗ω0 cosω0υ0, P2 = mu∗ω0 cosω0υ0, Q1 = qωq–1 0 cos (q – 1) 2 π – mu∗υ cosω0υ0, Q2 = qωq–1 0 sin (q – 1) 2 π + mu∗υ sinω0υ0. Lemma 3.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (3.2). It is easy to see ς(υj) = 0, ω(υj) = ω0, when υ = υj. If (H1) holds, then we have Lemma 3.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (3.2). It is easy to see ς(υj) = 0, ω(υj) = ω0, when υ = υj. If (H1) holds, then we have Lemma 3.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (3.2). It is easy to see ς(υj) = 0, ω(υj) = ω0, when υ = υj. If (H1) holds, then we have Re ds dτ ω=ω0,υ=υ0 > 0. Proof Diﬀerentiating equation (3.2) implicitly with respect to υ, we obtain Proof Diﬀerentiating equation (3.2) implicitly with respect to υ, we obtain ds dυ = smu∗e–sυ qsq–1 – υmu∗e–sυ . (3.7) ds dυ = smu∗e–sυ qsq–1 – υmu∗e–sυ . (3.7) Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 6 of 12 Hence, we deduce that Re ds dυ ω=ω0,υ=υ0 = P1Q1 + P2Q2 Q2 1 + Q2 2 . Hence, we deduce that Re ds dυ ω=ω0,υ=υ0 = P1Q1 + P2Q2 Q2 1 + Q2 2 . Re ds dυ ω=ω0,υ=υ0 = P1Q1 + P2Q2 Q2 1 + Q2 2 . Obviously, hypothesis (H1) means that the transversality condition is satisﬁed. Obviously, hypothesis (H1) means that the transversality condition is satisﬁed. □ Theorem 3.1 For model (2.7), when 0 < m < 1/e, the following results hold: Theorem 3.1 For model (2.7), when 0 < m < 1/e, the following results hold: (i) The equilibrium u∗of model (2.7) is locally asymptotically stable for υ ∈[0,υ0), and unstable when υ > υ0. (ii) Model (2.7) undergoes a Hopf bifurcation at the equilibrium u∗when υ = υ0. Proof Note that the eigenvalue λ = –(m + mu∗) < 0 of the linearized system of (3.1) sat- isﬁes the inequality \|arg(λ)\| > qπ/2 when υ = 0. Therefore, the condition for the Hopf bifurcation is satisﬁed. (i) We can ﬁnd that when υ = 0 the roots of equation (3.2) have negative real parts. 3 Bifurcation analysis of the uncontrolled model In Lemma 3.1, we can see that all the roots of equation (3.2) have negative real parts for υ ∈[0,υ0) by the deﬁnition of υ0. From Lemma 3.2, this implies that equation (3.2) has at least a positive root when υ > υ0. (i) We can ﬁnd that when υ = 0 the roots of equation (3.2) have negative real parts. In Lemma 3.1, we can see that all the roots of equation (3.2) have negative real parts for υ ∈[0,υ0) by the deﬁnition of υ0. From Lemma 3.2, this implies that equation (3.2) has at least a positive root when υ > υ0. (ii) From the above discussion, it is obvious that the occurrence condition of the Hopf bifurcation is satisﬁed for (2.6). Therefore, near the equilibrium u∗, there occurs a Hopf bifurcation when υ = υ0. □ 4 Bifurcation analysis of the controlled model In this part, by choosing the time delay υ as the bifurcation parameter, we are trying to con- trol the Hopf bifurcation of (2.7) based on the fractional-order Proportional and Derivative (PDα) control strategy. For the delayed fractional-order model (2.7), we propose a single input and output PDα controller as follows: ρ(t) = kp u(t) – u∗ + kd dα dtα u(t) – u∗ , (4.1) (4.1) where kp is the proportional control parameter and kd is the derivative control parameter. Hence, the controlled fractional-order survival red blood cells model with time delays becomes where kp is the proportional control parameter and kd is the derivative control parameter. Hence, the controlled fractional-order survival red blood cells model with time delays becomes dαu dtα = –mu(t) + e–u(t–υ) + ρ(t) = –mu(t) + e–u(t–υ) + kp u(t) – u∗ + kd dα dtα u(t) – u∗ . (4.2) dαu dtα = –mu(t) + e–u(t–υ) + ρ(t) dαu dtα = –mu(t) + e–u(t–υ) + ρ(t) = –mu(t) + e–u(t–υ) + kp u(t) – u∗ + kd dα dtα u(t) – u∗ . (4.2) (4.2) Let u(t) – u∗= y(t), The controlled model (4.2) becomes Let u(t) – u∗= y(t), The controlled model (4.2) becomes dαy dtα = 1 1 – kd –my(t) – mu∗y(t – υ) + kpy(t) = 1 1 – kd (kp – m)y(t) – mu∗y(t – υ) , (4.3) (4.3) Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 7 of 12 with the characteristic equation with the characteristic equation with the characteristic equation sq – 1 1 – kd (kp – m) – mu∗e–sυ = 0. (4.4) (4.4) Here, we assume the value range of the control parameters: kp < m, kd < 1. Here, we assume the value range of the control parameters: kp < m, kd < 1. Remark 4.1 Comparing model (4.3) and model (3.1), it is obvious that the controlled fractional-order model and the uncontrolled one have the same equilibrium point. Let s = ω(cos π 2 + isin π 2 ) (ω > 0), then equation (4.4) becomes ωq cos qπ 2 + isin qπ 2 – kp – m 1 – kd + mu∗ 1 – kd (cosωυ – isinωυ) = 0. 4 Bifurcation analysis of the controlled model □ □ eorem 4.1 For model (4.2), when \|a3\| < a4, we get the following results. Theorem 4.1 For model (4.2), when \|a3\| < a4, we get the following results. (i) The equilibrium u∗of model (4.2) is locally asymptotically stable for υ ∈[0,υc 0), and unstable when υ > υc 0. (ii) Model (4.2) undergoes a Hopf bifurcation at the equilibrium u∗when υ = υc 0. (ii) Model (4.2) undergoes a Hopf bifurcation at the equilibrium u∗when υ = υc 0. Proof Note that the eigenvalue λ = 1 1–kd [(kp – m) – mu∗] = a1 – a2 < 0 of the linearized system of (4.3) satisﬁes the inequality \|arg(λ)\| > qπ/2 when υ = 0. Therefore, the condition for the Hopf bifurcation is satisﬁed. Proof Note that the eigenvalue λ = 1 1–kd [(kp – m) – mu∗] = a1 – a2 < 0 of the linearized system of (4.3) satisﬁes the inequality \|arg(λ)\| > qπ/2 when υ = 0. Therefore, the condition for the Hopf bifurcation is satisﬁed. (i) It is easy to see that all the roots of equation (4.4) with υ = 0 have negative real parts. The deﬁnition of υc 0 implies that all the roots of equation (4.4) have negative real parts for υ ∈[0,υc 0). From Lemma 4.2, it indicates that equation (4.4) has at least a root with positive real parts when υ > υc 0. (ii) From the above discussion, it is clear that the occurrence condition of the Hopf bi- furcation is satisﬁed for model (4.2). Therefore, near the equilibrium u∗, there occurs a Hopf bifurcation when υ = υc 0. □ 4 Bifurcation analysis of the controlled model Separating the real and imaginary parts, we get Separating the real and imaginary parts, we get Separating the real and imaginary parts, we get ⎧ ⎨ ⎩ a4 cosωυ = a3 – ωq cos qπ 2 , a4 sinωτ = ωq sin qπ 2 , (4.5) ⎧ ⎨a4 cosωυ = a3 – ωq cos qπ 2 , i q i qπ (4.5) where a3 = kp–m 1–kd , a4 = mu∗ 1–kd . It can be obtained from (4.5) that where a3 = kp–m 1–kd , a4 = mu∗ 1–kd . It can be obtained from (4.5) that ω2q – 2a1ωq cos qπ 2 + a2 3 – a2 4 = 0. (4.6) ω2q – 2a1ωq cos qπ 2 + a2 3 – a2 4 = 0. (4.6) Denote Denote h2(ω) = ω2q – 2a1ωq cos qπ 2 + a2 3 – a2 4. Lemma 4.1 If \|a3\| < a4, equation (4.4) has a pair of purely imaginary roots ±ωc 0 when υ = υc j , j = 0,1,... , where υc j = 1 ωc 0 arccos a3 – ωc 0 cos qπ 2 a4 + 2jπ , j = 0,1,2,..., (4.7) (4.7) where ωc 0 is the unique positive zero of the function h2(ω). where ωc 0 is the unique positive zero of the function h2(ω). Proof By means of \|a3\| < a4, it is easy to see that h2(0) < 0. Combining with q > 0, there exists a unique positive number ωc 0 such that h2(ω) = 0. Then ωc 0 is a root of (4.6). Hence, for υc j as deﬁned in (4.7), (ωc 0,υc j ) is a root of equation (4.5). It can be seen that ±ωc 0 is a pair of purely imaginary roots of equation (4.4), while υ = υc j , j = 0,1,... . □ Remark 4.2 Lemma 4.1 obtains the onset of the delayed fractional-order model’s Hopf bifurcations. We make the following assumption: (H2) M1N1 + M2N2 N2 1 + N2 2 > 0, Page 8 of 12 Sun et al. 4 Bifurcation analysis of the controlled model Advances in Diﬀerence Equations ( 2018) 2018:10 where M1 = ωc 0mu∗ 1 – kd sinωc 0υc 0, M2 = ωc 0mu∗ 1 – kd cosωc 0υc 0, N1 = q ωc 0 q–1 cos (q – 1) 2 π – υmu∗ 1 – kd cosωc 0υc 0, N2 = q ωc 0 q–1 sin (q – 1) 2 π + υmu∗ 1 – kd sinωc 0υc 0. M1 = ωc 0mu∗ 1 – kd sinωc 0υc 0, M2 = ωc 0mu∗ 1 – kd cosωc 0υc 0, N1 = q ωc 0 q–1 cos (q – 1) 2 π – υmu∗ 1 – kd cosωc 0υc 0, N2 = q ωc 0 q–1 sin (q – 1) 2 π + υmu∗ 1 – kd sinωc 0υc 0. Lemma 4.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (4.4). It is easy to see ς(υc j ) = 0, ω(υc j ) = ωc 0 when υ = υc j . If (H2) holds, then we have Lemma 4.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (4.4). It is easy to see ς(υc j ) = 0, ω(υc j ) = ωc 0 when υ = υc j . If (H2) holds, then we have Lemma 4.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (4.4). It is easy to see ς(υc j ) = 0, ω(υc j ) = ωc 0 when υ = υc j . If (H2) holds, then we have Lemma 4.2 Let s(υ) = ς(υ) + iω(υ) be the root of equation (4.4). It is easy to see ς(υc j ) = 0, ω(υc j ) = ωc 0 when υ = υc j . If (H2) holds, then we have Re ds dυ ω=ωc 0,υ=υc 0 > 0. Proof Diﬀerentiating equation (4.4) relative to υ, we get Proof Diﬀerentiating equation (4.4) relative to υ, we get ds dυ = smu∗ 1–kd e–sυ qsq–1 – υmu∗ 1–kd e–sυ . (4.8) ds dυ = smu∗ 1–kd e–sυ qsq–1 – υmu∗ 1–kd e–sυ . (4.8) Hence, we obtain Hence, we obtain Re ds dυ ω=ωc 0,υ=υc 0 = M1N1 + M2N2 N2 1 + N2 2 . □ Re ds dυ ω=ωc 0,υ=υc 0 = M1N1 + M2N2 N2 1 + N2 2 . 5 Numerical simulations In this section, we provide numerical simulations to conﬁrm our theoretical analysis and display the Hopf bifurcation phenomenon of the delayed fractional-order model. For the uncontrolled model (2.7), we take α = 0.92, m = 0.2 used in [7]. Then (2.7) has a positive equilibrium u∗= 1.3267. From (3.6), we can obtain υ0 = 19.6749. The positive Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 9 of 12 Page 9 of 12 Figure 1 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. The equilibrium u∗is asymptotically stable when υ = 18 < υ0 = 19.6749. Figure 2 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 21 > υ0 = 19.6749. Figure 1 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. The equilibrium u∗is asymptotically stable when υ = 18 < υ0 = 19.6749. Figure 1 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. The equilibrium u∗is asymptotically stable when υ = 18 < υ0 = 19.6749. Figure 1 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. The equilibrium u∗is asymptotically stable when υ = 18 < υ0 = 19.6749. Figure 2 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 21 > υ0 = 19.6749. Figure 2 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 21 > υ0 = 19.6749. Figure 2 Waveform plot and phase portrait of the fractional-order model (2.7) with α = 0.92, m = 0.2, and the initial value u0 = 1. 5 Numerical simulations A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 21 > υ0 = 19.6749. Table 1 Bifurcation point υ0 versus fractional order α for model (2.7) Fractional order α Bifurcation point υ0 0.9 21.5435 0.92 19.6749 0.94 17.9960 0.96 16.4929 0.98 15.1312 1.0 13.9045 Table 1 Bifurcation point υ0 versus fractional order α for model (2.7) equilibrium point u∗is asymptotically stable when υ = 18 < υ0 = 19.6749 as illustrated in Figure 1, and when υ = 21 > υ0 = 19.6749, there occurs a Hopf bifurcation at the positive equilibrium point u∗as demonstrated in Figure 2. The eﬀect of the order α from 0.9 to 1 on the values of υ0 for model (2.7) is shown in Table 1. In order to make a comparison with the uncontrolled fractional-order model (2.7), we discuss the controlled model (4.2) with the parameters α = 0.92, m = 0.2. When we choose the control parameters kp = –0.05, kd = 0.5, from (4.7) we can obtain υc 0 = 27.7129. It can be seen that the stable region has been enlarged and the critical value υc 0 has been increased to a larger value than that of the uncontrolled model. This indicates that the Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 10 of 12 Figure 3 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = –0.05, kd = 0.5 and the initial value u0 = 2. The equilibrium u∗is asymptotically stable when υ = 24 < υc 0 = 27.7129. Figure 4 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = –0.05, kd = 0.5 and the initial value u0 = 2. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 30 > υc 0 = 27.7129. Figure 3 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = –0.05, kd = 0.5 and the initial value u0 = 2. The equilibrium u∗is asymptotically stable when υ = 24 < υc 0 = 27.7129. 5 Numerical simulations Figure 3 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = –0.05, kd = 0.5 and the initial value u0 = 2. The equilibrium u∗is asymptotically stable when υ = 24 < υc 0 = 27.7129. Figure 4 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = –0.05, kd = 0.5 and the initial value u0 = 2. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 30 > υc 0 = 27.7129. PDα controller can delay the onset of Hopf bifurcations eﬀectively. The equilibrium u∗ is asymptotically stable when υ = 24 < υc 0 = 27.7129 as illustrated in Figure 3, and when υ = 30 > υc 0 = 27.7129, there occurs a Hopf bifurcation at the positive equilibrium point u∗ as demonstrated in Figure 4. Next, we select other control parameters to validate the eﬀectiveness of our proposed PDα scheme in the bifurcation control. We also take the same parameters α = 0.92, m = 0.2 for the original model (2.7). When kp = 0.05, kd = 0.5, from (4.7) one can ob- tain υc 0 = 6.1794. The critical value υc 0 is smaller than that of the uncontrolled model (2.7), which shows that the PDα controller can advance the onset of the Hopf bifurcation eﬀec- tively. The equilibrium u∗is asymptotically stable when υ = 6 < υc 0 = 6.1794 as illustrated in Figure 5, and when υ = 7 > υc 0 = 6.1794, there occurs a Hopf bifurcation at the positive equilibrium point u∗as demonstrated in Figure 6. Author details 1College of Automation, Nanjing University of Posts and Telecommunications, Nanjing, 210003, China. 2School of Mathematics, Southeast University, Nanjing, 210096, China. 3School of Mathematics and Statistics, Shandong Normal University, Ji’nan, 250014, China. 4College of Mathematics and Statistics, Chongqing Technology and Business University, Chongqing, 400067, China. 5School of Mathematics and Computer Science, Hubei University of Arts and Science, Xiangyang, 441053, China. Xiangyang, 441053, China. 6 Conclusions Therefore, we can choose appropriate values of the Proportional and Derivative parameters to change the charac- teristics of Hopf bifurcation embedding in fractional-order systems with time delays. can successfully delay or advance the onset of Hopf bifurcation. Therefore, we can choose appropriate values of the Proportional and Derivative parameters to change the charac- teristics of Hopf bifurcation embedding in fractional-order systems with time delays. Competing interests Competing interests The authors declare that they have no competing interests. The authors declare that they have no competing interests. Acknowledgements h k l Acknowledgements This work is partly supported by the National Natural Science Foundation of China (Nos. 61573194, 61573096), the Six Talent Peaks High Level Project of Jiangsu Province (No. 2014-ZNDW-004) and the Science Foundation of Nanjing University of Posts and Telecommunications (No. NY213095). Authors’ contributions Authors’ contributions All authors contributed equally to the writing of this paper. All authors read and approved the ﬁnal manuscript. 6 Conclusions In this paper, we have studied the Hopf bifurcation of a fractional-order red blood cells model with time delay and have proposed the conﬁguration for the stable region. In or- der to control the Hopf bifurcation of the delayed fractional-order red blood cells model, we have designed a fractional-order Proportional and Derivative (PDα) controller, which Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 Page 11 of 12 Page 11 of 12 Figure 5 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = 0.05, kd = 0.5 and the initial value u0 = 2. The equilibrium u∗is asymptotically stable when υ = 6 < υc 0 = 6.1794. Figure 6 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = 0.05, kd = 0.5 and the initial value u0 = 2. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 7 > υc 0 = 6.1794. Figure 5 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = 0.05, kd = 0.5 and the initial value u0 = 2. The equilibrium u∗is asymptotically stable when υ = 6 < υc 0 = 6.1794. Figure 5 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = 0.05, kd = 0.5 and the initial value u0 = 2. The equilibrium u∗is asymptotically stable when υ = 6 < υc 0 = 6.1794. Figure 6 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = 0.05, kd = 0.5 and the initial value u0 = 2. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 7 > υc 0 = 6.1794. Figure 6 Waveform plot and phase portrait of the controlled fractional-order model (4.2) with α = 0.92, m = 0.2, kp = 0.05, kd = 0.5 and the initial value u0 = 2. A Hopf bifurcation occurs and the periodic oscillation bifurcates from u∗, where υ = 7 > υc 0 = 6.1794. can successfully delay or advance the onset of Hopf bifurcation. References 1. Xiao, M, Zheng, WX, Jiang, GP, Cao, JD: Undamped oscillations generated by Hopf bifurcations in fractional-o 1. Xiao, M, Zheng, WX, Jiang, GP, Cao, JD: Undamped oscillations generated by Hopf bifurcations in fractional-order recurrent neural networks with Caputo derivative IEEE Trans Neural Netw Learn Syst 26 3201 3214 (2015) 1. Xiao, M, Zheng, WX, Jiang, GP, Cao, JD: Undamped oscillations generated by Hopf bifurcations in fractional-order recurrent neural networks with Caputo derivative. IEEE Trans. Neural Netw. Learn. Syst. 26, 3201-3214 (2015) recurrent neural networks with Caputo derivative. IEEE Trans. Neura 2. Laskin, N: Fractional market dynamics. Physica A 287, 482-492 (2000) 2. Laskin, N: Fractional market dynamics. Physica A 287, 482-492 (2000) y y 3. Djordjevi´c, VD, Jari´c, J, Fabry, B, Fredberg, JJ, Stamenovi´c, D: Fractional derivatives embody essential features of cell 3. Djordjevi´c, VD, Jari´c, J, Fabry, B, Fredberg, JJ, Stamenovi´c, D: Fractional derivatives embody essential features of ce rheological behavior Ann Biomed Eng 31 692 699 (2003) rheological behavior. Ann. Biomed. Eng. 31, 692-699 (2003) g g 4. Arafa, AAM, Rida, SZ, Khalil, M: Fractional modeling dynamics of HIV and CD4+ T-cells during primary infection. N li Bi d Ph 6 1 (2012) g g 4. Arafa, AAM, Rida, SZ, Khalil, M: Fractional modeling dynamics of HIV and CD4+ T-cells during primary infection. Nonlinear Biomed. Phys. 6, 1 (2012) Nonlinear Biomed. Phys. 6, 1 (2012) 5. Rihan, FA: Numerical modeling of fractional-order biological systems. Abstr. Appl. Anal. 2013, 8 (2013) 6. Zhang, CR, Zu, YG, Zheng, BD: Stability and bifurcation of a discrete red blood cell survival model. Chaos Solitons Fractals 28, 386-394 (2006) 7. Song, YL, Wei, JJ, Yuan, Y: Bifurcation analysis on a survival red blood cells model. J. Math. Anal. Appl. 316, 459-471 (2006) 8. Fan, DJ, Wei, JJ: Bifurcation analysis of discrete survival red blood cells model. Commun. Nonlinear Sci. Numer. Simul 14, 3358-3368 (2009) 9. Yin, XW, Thomas, T, Zhang, JF: Multiple red blood cell ﬂows through microvascular bifurcations: cell free layer, cell 9. Yin, XW, Thomas, T, Zhang, JF: Multiple red blood cell ﬂows through microvascular bifurcations: cell free layer, cel trajectory, and hematocrit separation. Microvasc. Res. 89, 47-56 (2013) 9. Yin, XW, Thomas, T, Zhang, JF: Multiple red blood cell ﬂows through microv trajectory, and hematocrit separation. Microvasc. Res. 89, 47-56 (2013) trajectory, and hematocrit separation. Microvasc. Res. 89, 47-56 (2013) 11. Received: 19 October 2017 Accepted: 28 December 2017 Received: 19 October 2017 Accepted: 28 December 2017 References Magin, RL: Fractional calculus models of complex dynamics in biological tissues. Comput. Math. Appl. 59, 1586-1593 (2010) 12. Craiem, D, Magin, RL: Fractional order models of viscoelasticity as an alternative in the analysis of red blood cell (RBC) membrane mechanics. Phys. Biol. 7(1), 13001 (2010) y 13. Podlubny, I: Fractional Diﬀerential Equations. Academic Press, New York (1999) 14. Cheng, ZS, Cao, JD: Hybrid control of Hopf bifurcation in complex networks with delays. Neurocomputing 133, 164-170 (2014) 15. Huang, CD, Cao, JD, Xiao, M: Hybrid control on bifurcation for a delayed fractional gene regulatory network. Chaos Solitons Fractals 87, 19-29 (2016) 16. Yang, XS, Cao, JD, Ho, DWC: Exponential synchronization of discontinuous neural networks with time-varying mixed delays via state feedback and impulsive control Cogn Neurodyn 9 113 128 (2015) 16. Yang, XS, Cao, JD, Ho, DWC: Exponential synchronization of discontinuous neural netw delays via state feedback and impulsive control. Cogn. Neurodyn. 9, 113-128 (2015) g p y delays via state feedback and impulsive control. Cogn. Neurodyn. 9, 113-128 (2015) delays via state feedback and impulsive control. Cogn. Neurodyn. 9, 113-128 (2015) 17. Wang, H, Zhu, QX: Finite-time stabilization of high-order stochastic nonlinear systems in strict-feedback form. Automatica 54, 284-291 (2015) 17. Wang, H, Zhu, QX: Finite-time stabilization of high-order stochastic nonlinear systems in strict-feedback form. Automatica 54, 284-291 (2015) 18. Zhao, HY, Xie, W: Hopf bifurcation for a small-world network model with parameters delay feedback control. Nonlinear Dyn. 63, 345-357 (2011) 19. Fiedler, B, Oliva, SM: Delayed feedback control of a delay equation at Hopf bifurcation. J. Dyn. Diﬀer. Equ. 28, 1357-1391 (2016) 19. Fiedler, B, Oliva, SM: Delayed feedback control of a delay equation at Hopf bifurcation. J. Dyn. Diﬀer. Equ. 28, 1357-1391 (2016) 20. Zhusubaliyev, ZT, Medvedev, A, Silva, MM: Bifurcation analysis of PID-controlled neuromuscular blockade in closed-loop anesthesia. J. Process Control 25, 152-163 (2015) 20. Zhusubaliyev, ZT, Medvedev, A, Silva, MM: Bifurcation analysis of PID-controlled neuromuscular blockade in closed-loop anesthesia. J. Process Control 25, 152-163 (2015) closed-loop anesthesia. J. Process Control 25, 152-163 (2015) 21. Wazewska-Czyzewska, M, Lasota, A: Mathematical problems of the dynamics of the red blood cells system. Mat. Stosow. 6, 23-40 (1976) 21. Wazewska-Czyzewska, M, Lasota, A: Mathematical problems of the dynamics of the red blood cells system. Mat. Stosow. 6, 23-40 (1976) 22. Deng, WH, Li, CP, Lü, JH: Stability analysis of linear fractional diﬀerential system with multiple time delays. Nonlinear Dyn. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional aﬃliations. Page 12 of 12 Sun et al. Advances in Diﬀerence Equations ( 2018) 2018:10 References 48, 409-416 (2007) 23. Matignon, D: Stability results for fractional diﬀerential equations with applications to control processing. In: IMACS-SMC Multi-Conf. Comput. Eng. Syst. Appl., pp. 963-968 (1996)
https://openalex.org/W4200491860	https://www.medrxiv.org/content/medrxiv/early/2021/12/17/2021.12.16.21267904.full.pdf	English	null	Deficits in hospital care among clinically vulnerable children aged 0 to 4 years during the COVID-19 pandemic	medRxiv (Cold Spring Harbor Laboratory)	2,021	cc-by	8,625	Corresponding author: d.etoori@ucl.ac.uk Corresponding author: d.etoori@ucl.ac.uk Title: Deficits in hospital care among clinically vulnerable children aged 0 to 4 years during the COVID-19 pandemic Authors: David Etoori1, Katie Harron2, Louise Mc Grath-Lone1, Maximiliane Verfuerden2, Ruth Gilbert2, Ruth Blackburn1. . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for th this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a p p y Affiliations: 1. University College London Institute of Health Informatics, 222 Euston Road, London, NW1 2DA, UK. 2. University College London Great Ormond Street Institute of Child Health, 30 Guilford St, Holborn, London WC1N 1EH, UK. 2. University College London Great Ormond Street Institute of Child Health, 30 Guilford St, Holborn, London WC1N 1EH, UK. Corresponding author: d.etoori@ucl.ac.uk Study population and data source: Children were included in the cohort if aged 0-4 years and their birth was recorded in hospital administrative data in the English NHS (Hospital Episode Statistics – HES) between January 1, 2010, and March 31, 2021. Births recorded in HES represent 97% of all births in England.(12) Children were followed until their fifth birthday or March 31, 2021, whichever occurred first (Supplementary Figure 1). All contacts with NHS hospitals in England were routinely linked, including admitted patient care (APC) and outpatient attendances. Due to changes in linkage methods used by NHS Digital, accident and emergency (A&E) attendances could not be included.(13) We combined consecutive consultant episodes and hospital transfers to form admissions.(14) Abstract is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for th this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint CC perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in by peer review) The copyright holder for this this version posted December 17, 2021. ; 10.1101/2021.12.16.21267904 . CC-BY 4.0 International license It is made available under a p p y Introduction: Rates of hospital contact among 0-4 year olds increased steadily in England over the past decade and are highest for infants.(1–3) Hospital utilisation patterns differ markedly by age and clinical vulnerability: children born preterm (<37 weeks of gestation), with low birth weight (<2500g) or a congenital anomaly, have substantially more admissions than other children.(4,5) Hospital contacts declined substantially during the COVID lockdown.(6–9) This decline is likely to have impacted most on children with high rates of planned hospital care, such as those born too early or too small or with underlying health conditions.(1,2,10) Planned hospital care that was postponed or cancelled may result in delayed diagnoses or treatments, which could be detrimental to health or development.(11) Fewer unplanned hospital admissions might reflect fewer infections, injuries, or other health problems due to reduced exposure during pandemic restrictions, but could also reflect unmet need for unplanned care. This study aimed to quantify deficits in planned and unplanned hospital care for clinically vulnerable children and their peers during the COVID-19 pandemic. We used national, longitudinal administrative data for England to measure rates of planned hospital contacts (admissions and outpatient attendances) and unplanned admissions among children with chronic health conditions or born too early or too small, and their peers. We quantified the deficit by comparing predicted and observed rates of hospital contacts during the first year of the pandemic. We examined whether rates of contact returned to pre-pandemic levels and described changes in the type of outpatient contact (e.g., face-to-face or virtual). Abstract Objective To quantify deficits in hospital care for clinically vulnerable children during the COVID-19 pandemic. Design Birth cohort in Hospital Episode Statistics (HES). Design Birth cohort in Hospital Episode Statistics (HES). Setting NHS hospitals in England. Setting NHS hospitals in England. Study population All children aged <5 years with a birth recorded in hospital administrative data (January 2010 to March 2021). Main exposure Clinical vulnerability defined by a chronic health condition, preterm birth (<37 weeks gestation) or low birthweight (<2500g). Main outcomes Deficits in care defined by predicted rates for 2020, estimated from 2015- 2019, minus observed rates per 1000 child years during the pandemic (March 2020-2021). Results Of 3,813,465 children, 17.7% (1 in 6) were clinically vulnerable (9.5% born preterm or low birthweight, 10.3% had a chronic condition). Deficits in hospital care during the pandemic were much higher for clinically vulnerable children than peers: respectively, outpatient attendances (314 versus 73 per 1000 child years), planned admissions (55 versus 10), and unplanned admissions (105 versus 79). Clinically vulnerable children accounted for 50.1% of the deficit in outpatient attendances, 55.0% in planned admissions, and 32.8% in unplanned hospital admissions. During the pandemic, weekly rates of planned care returned to pre-pandemic levels for infants with chronic conditions but not older children. Deficits in care differed by ethnic group and level of deprivation. Virtual outpatient attendances increased from 3.2% to 24.8% during the pandemic. sion 1 in 6 clinically vulnerable children accounted for one-third to one half of the onclusion 1 in 6 clinically vulnerable children accounted for one-third to one half of the Conclusion 1 in 6 clinically vulnerable children accounted for one-third to one half of the deficit in hospital care during the pandemic. NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. deficit in hospital care during the pandemic. NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. 1 1 CC-BY 4 0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint CC-BY 4 0 International license It is made available under a perpetuity. Outcome and exposure: Our primary outcome was the deficit in hospital care, defined as the absolute difference between observed and predicted rates (per 1,000 child-years) of hospital contacts 2 2 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for thi this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a p p y (outpatient attendances, planned and unplanned hospital admissions) during the pandemic (March 23, 2020, to March 22, 2021). We also described trends in weekly rates of hospital contacts by year of age (from January 1, 2020 to March 31, 2021, and averaged for 2015- 2019), and uptake (attended, missed, cancelled, postponed) and mode (in person, virtually) of outpatient contacts. (outpatient attendances, planned and unplanned hospital admissions) during the pandemic (March 23, 2020, to March 22, 2021). We also described trends in weekly rates of hospital contacts by year of age (from January 1, 2020 to March 31, 2021, and averaged for 2015- 2019), and uptake (attended, missed, cancelled, postponed) and mode (in person, virtually) of outpatient contacts. Clinical vulnerability was defined by a chronic condition recorded in HES up to age 4, or preterm birth (<37 weeks of gestation) or low birth weight (<2500g) recorded in birth or delivery records, using previously reported methods.(12,15–17) Children with missing gestation and birth weight data who had no chronic conditions were categorised as having no clinical vulnerability. We analysed the following risk factors associated with frequency of hospital contacts: age (0 to 11 months, 1 to 4 years),(3) quintile of deprivation derived from the national distribution of the Index of Multiple Deprivation (IMD) 2004 (an area measure for ~650 households),(18,19) and ethnic group recorded in HES (grouped as White, Black, Asian, or Other, including mixed and Chinese). Hospital contacts pre-pandemic: Hospital contacts were much higher among infants than children aged 1-4 years: 60.1% of infants and 8.2% of 1–4-year-olds had ≥1 outpatient attendance each year (Supplementary Figure 2, Supplementary Table 1). Overall, 31.2% of clinically vulnerable children had ≥1 outpatient attendance compared with 14.1% of those with no known vulnerability (p<0.001). A similar pattern was seen for planned and unplanned hospital admissions. Children with chronic conditions had the highest rates of planned and unplanned admissions across all strata (Supplementary Figures 3-4). In contrast, children born preterm or low birth weight but with no chronic condition had similar rates of admissions to their peers born at term or weighing ≥2500g (Supplementary Figure 2). Population characteristics: Of the 3,813,465 children aged 0-4 years, 394,384 (10.3%) had a record indicating a chronic condition (including congenital anomalies); 363,950 (9.5%) were born preterm or low birthweight, and 83,283 (2.2%) had both vulnerabilities. Overall, 675,051 (17.7%) had one or more of these clinical vulnerabilities (Table 1). Statistical analyses: CC-BY 4.0 International license It is made available under a p p y or attendances by the weekly dynamic denominator population of children within each stratification level (i.e., a child born in week 1 of 2015 would move into the 1-year-old group in week 1 of 2016 and age out of the cohort in week 1 of 2019). We then visually compared weekly plots of rates in 2020 and 2021 with the overall average rates for 2015-2019 for the same week. We also modelled weekly rates between January 1, 2015, and December 31, 2019, using a Poisson model which included a linear effect of time, calendar month to account for seasonality, log of the weekly denominator population as an offset, and second order lagged residuals. A similar approach was then used to estimate weekly deficits during the pandemic. All analyses were performed in Stata 16.(21) or attendances by the weekly dynamic denominator population of children within each stratification level (i.e., a child born in week 1 of 2015 would move into the 1-year-old group in week 1 of 2016 and age out of the cohort in week 1 of 2019). We then visually compared weekly plots of rates in 2020 and 2021 with the overall average rates for 2015-2019 for the same week. We also modelled weekly rates between January 1, 2015, and December 31, 2019, using a Poisson model which included a linear effect of time, calendar month to account for seasonality, log of the weekly denominator population as an offset, and second order lagged residuals. A similar approach was then used to estimate weekly deficits during the pandemic. All analyses were performed in Stata 16.(21) Statistical analyses: We calculated observed rates of hospital contacts per 1,000 child-years in the pre-pandemic period (2015-2019), stratifying by risk factor (clinical vulnerability, age group, area quintile of deprivation, and ethnic group). We calculated child-years at risk by averaging the eligible population of births recorded in HES at the beginning and end of a year, assuming no emigration and ignoring deaths. We used a Poisson model, which included a linear effect of time and log of the mid-year population as an offset, to model rates from January 1, 2015, to December 31, 2019 stratifying by risk factors. Data from January 1 to March 22, 2020, was excluded from the pre-pandemic period as we observed reductions in hospital contact rates before the first lockdown. To calculate the deficit in hospital contacts, we predicted rates for the pandemic period, assuming that the pandemic had not occurred and that previous trends would have continued. The deficit was estimated as the difference between predicted and observed rates. We also calculated deficits within the first national lockdown (March 23 to June 23 2020), easing of restrictions (June 24 to November 4 2020), second national lockdown (November 5 – December 31 2020), and third national lockdown (January 1 – March 22 2021).(20) To understand seasonality and visualise the decline and recovery in hospital contacts during the pandemic, we calculated weekly rates between January 1, 2015, and March 31, 2021 using dynamic denominator study populations. Since we expected the largest differences to be seen when comparing children with and without chronic conditions, we stratified by this risk factor as recorded in any HES admission between birth and the relevant week. Weekly rates of hospital contact were calculated by dividing the total number of weekly admissions 3 3 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . Hospital contacts during the pandemic: There were stark reductions in rates of all types of hospital care during the pandemic (Table 2, Supplementary tables 2-4). Differences between predicted and observed rates of hospital contacts were much larger for children with a chronic condition than those without, and particularly high for children with a chronic condition who were also born preterm or with a low birth weight (Table 2). The 17.7% of children who were categorised as clinically vulnerable accounted for 50.1% of the deficit in outpatient attendances, 55.0% of the deficit in planned hospital admissions, and 32.8% of the deficit in unplanned hospital admissions (Table 2). Deficits were higher for infants than 1-4-year-olds for outpatient attendances and unplanned admissions but not for planned admissions (Figure 1). We found small but 4 4 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for thi this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a p p y statistically significant differences in deficits of planned and unplanned admissions across ethnic groups and in all hospital contacts among children in the most (versus least) deprived quintile (Supplementary tables 3-4). The largest deficits in care was among children with a chronic condition (Supplementary Figures 5-6). Overall, the largest deficits were seen in the first lockdown for outpatient attendances and planned admissions, but for unplanned admissions, the largest deficits occurred in the second lockdown period (Table 3; Supplementary tables 5-7). statistically significant differences in deficits of planned and unplanned admissions across ethnic groups and in all hospital contacts among children in the most (versus least) deprived quintile (Supplementary tables 3-4). The largest deficits in care was among children with a chronic condition (Supplementary Figures 5-6). Overall, the largest deficits were seen in the first lockdown for outpatient attendances and planned admissions, but for unplanned admissions, the largest deficits occurred in the second lockdown period (Table 3; Supplementary tables 5-7). Trends in hospital contacts: Outpatient attendances declined sharply before and during the first national lockdown (Figure 2), among children of all ages with a chronic condition, with less perceptible changes among those without a chronic condition. Outpatient attendances rapidly returned to pre- pandemic rates for infants, but not for older children which remained below 2015-19 averages in the third lockdown. Planned admissions followed a similar pattern, with a return to pre-pandemic rates only for infants (Figure 2). Declines in rates of unplanned admissions during the first lockdown were much greater for children with a chronic condition than those without and remained below pre-pandemic levels for both groups at all ages (Figure 2). In 2020, the autumn-winter peak in unplanned admissions was diminished relative to previous years; however, following the reopening of primary schools at the end of the third lockdown on March 8, 2021, there was an increase in unplanned admission rates for all children, particularly those with a chronic condition (Figure 2, Supplementary Figure 7). Trends did not differ consistently according to deprivation level (Supplementary Figures 8-9) or by ethnic group (Supplementary Figures 10-12). For all age groups, a spike in cancellations and postponement of outpatient appointments preceded the first lockdown by three weeks (Supplementary Figure 13). There was an increase in tele/virtual outpatient attendances during the pandemic and face-to-face visits did not return to pre-pandemic levels in any age group (Supplementary Table 8 and Figure 14). Discussion: This population-based cohort study of all children aged <5 in England found large and disproportionate deficits in planned and unplanned hospital contacts during the COVID-19 pandemic for clinically vulnerable groups and those living in deprived areas. The one in six clinically vulnerable children accounted for over half the deficit in outpatient attendances and planned admissions, and one-third of the deficit in unplanned admissions. We saw some evidence of recovery in the deficit in planned care during the pandemic among infants, but not among older children. This study has several strengths. First, we used a birth cohort of all children born in an NHS hospital in England (97% of all births). Our rates did not account for deaths in the 5 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for thi this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a p p y denominator (0.5%), non-NHS health care, or emigration, but these events are rare. Second, HES captures gestational age and birthweight and we used a clinically developed coding system to define chronic conditions from diagnostic codes recorded in admissions.(15) Third, the large study population made it possible to calculate weekly rates. Fourth, we used data preceding the pandemic to March 2021 to investigate variation across the first year of the pandemic. denominator (0.5%), non-NHS health care, or emigration, but these events are rare. Second, HES captures gestational age and birthweight and we used a clinically developed coding system to define chronic conditions from diagnostic codes recorded in admissions.(15) Third, the large study population made it possible to calculate weekly rates. Fourth, we used data preceding the pandemic to March 2021 to investigate variation across the first year of the pandemic. Limitations include under-ascertainment of chronic conditions for children who could not be admitted to hospital due to the pandemic. Discussion: is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for th this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a p p y access to the internet at home will have been disproportionately impacted by the move to virtual appointments. Deficits in unplanned care may be driven by opposing factors. Previous studies reported reductions in unplanned admissions due to decreases in infection-related hospitalisations, due to reduced social exposure and increased hygiene, with little change in admissions for non-infectious causes such as appendicitis.(25,28,29,32,39) Others have reported reductions in injury.(44) The spike in unplanned admissions after schools reopened in autumn 2020 and in March 2021 when the third lockdown ended likely reflects increased socialisation. Other positive effects could include reduced exposure to triggers for respiratory disease (e.g., air pollution),(45) improved hygiene due to ubiquitous handwashing messaging, and improved medication adherence through increased parental supervision. Negative implications could include reduced extrinsic interventions through education, health, and social care professionals,(7,46) or delaying or avoiding medical care due to restricted access or fears of potential hospital-acquired COVID-19 infection.(36,47– 49) Additionally, these deficits could represent a missed opportunity to catch problems early which could also have consequences in the future.(11,50) This analysis was the first step in quantifying the burden of deferred or foregone hospital care during the pandemic and the potential consequences for young children. Studies using routine administrative data report only the most acute cases are presenting at hospitals,(35,36) and that some of the deficits reflect late or missed diagnoses.(11,50) More research is needed to untangle these different mechanisms. Research is also needed to understand deficits in planned care, the types of care, procedures, or treatments affected and the short- and long-term implications for children with specific conditions. Our findings indicate a need for targeted ‘catch-up’ funding and resources for child health, particularly for vulnerable children who were affected disproportionately.(51) Discussion: These children may have been managed in primary care, or as outpatients, which does not reliably code chronic conditions. Our analyses may underestimate vulnerability for the 10% of children without a record of gestational age or birthweight as we assumed they had non-vulnerable status. Multiple imputation of missing data was not feasible given the study size. Second, we could not quantify the deficit in A&E attendances, as longitudinal linkage is not currently available. However, studies investigating A&E attendances have reported similar deficits to those found in our study.(22,23) Our modelling approach required several assumptions (including continued trends in hospital contacts), and the differences we report are likely conservative estimates of impact. Deficits in hospital care for children during the pandemic have been reported in Europe,(7,24–31) Asia,(32) North,(33–36) and South America.(37) Most studies investigated A&E attendances or unplanned admissions.(23–37) Other studies report a reduction in asthma-related paediatric emergency department attendances,(27) and reduced likelihood of admission, assessment and surgery for children with epilepsy.(38) Furthermore, significant reductions in infection-related hospitalisations have been observed,(28,32,34,39) particularly for children under 5 years.(39) Two studies conducted national level analyses.(28,30) We believe our study is the first to report on deficits in planned care (admissions and outpatient), on disproportionate deficits in hospital care for children with health risk-factors at a population-level, and for a full year of the pandemic. Previous research in adult populations has reported on the disproportionate burden of COVID-19 infection, hospitalisation, and death in minority ethnic groups.(40,41) Our study did not examine COVID-related contacts because children hospitalisation is rare as typically experience mild asymptomatic disease.(42,43) However, we identified small differences in deficits of hospital care for children in the Asian ethnic group, and for children in the most deprived quintile, indicating evidence of structural inequalities in healthcare access. Potential mechanisms underpinning deficits in planned care likely represent restrictions to access, supported by a rise in postponements of outpatient care. Our findings show these restrictions were mitigated in infants, who have a high frequency of hospital care and for whom interventions are likely to be more time-critical than in older children.(3) However, deficits remained large in older children and may reflect unmet need or postponed care that could have longer-term health and care consequences.(11) Furthermore, children without 6 6 . CC-BY 4.0 International license It is made available under a perpetuity. References: 1. Saxena S, Bottle A, Gilbert R, Sharland M. Increasing short-stay unplanned hospital admissions among children in England; time trends analysis ’97-’06. PloS One. 2009 Oct 15;4(10):e7484. 1. Saxena S, Bottle A, Gilbert R, Sharland M. Increasing short-stay unplanned hospital admissions among children in England; time trends analysis ’97-’06. PloS One. 2009 Oct 15;4(10):e7484. 1. Saxena S, Bottle A, Gilbert R, Sharland M. Increasing short-stay unplanned hospital admissions among children in England; time trends analysis ’97-’06. PloS One. 2009 Oct 15;4(10):e7484. 2. Wolfe I, Thompson M, Gill P, Tamburlini G, Blair M, van den Bruel A, et al. Health services for children in western Europe. Lancet Lond Engl. 2013 Apr 6;381(9873):1224– 34. 3. Ruzangi J, Blair M, Cecil E, Greenfield G, Bottle A, Hargreaves DS, et al. Trends in healthcare use in children aged less than 15 years: a population-based cohort study in England from 2007 to 2017. BMJ Open. 2020 May 5;10(5):e033761. 7 7 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint 4. Wijlaars LP, Hardelid P, Woodman J, Allister J, Cheung R, Gilbert R. Contribution of recurrent admissions in children and young people to emergency hospital admissions: retrospective cohort analysis of hospital episode statistics. Arch Dis Child. 2015 Sep;100(9):845–9. 5. Harron K, Gilbert R, Cromwell D, Oddie S, Guttmann A, Meulen J van der. International comparison of emergency hospital use for infants: data linkage cohort study in Canada and England. BMJ Qual Saf. 2018 Jan 1;27(1):31–9. 6. World Health Organisation. COVID-19 significantly impacts health services for noncommunicable diseases [Internet]. [cited 2021 Mar 18]. Available from: https://www.who.int/news/item/01-06-2020-covid-19-significantly-impacts-health- services-for-noncommunicable-diseases 7. Isba R, Edge R, Jenner R, Broughton E, Francis N, Butler J. Where have all the children gone? Decreases in paediatric emergency department attendances at the start of the COVID-19 pandemic of 2020. Arch Dis Child. 2020 Jul 1;105(7):704–704. 8. Public Health England. National COVID-19 surveillance reports [Internet]. References: [cited 2021 Apr 13]. Available from: https://www.gov.uk/government/publications/national-covid- 19-surveillance-reports 9. Bottle A, Aylin P, Warner M, Propper C, Stoye G, Burn S. What happened to English NHS hospital activity during the COVID-19 pandemic? [Internet]. 2021 May [cited 2021 Nov 30]. Available from: https://ifs.org.uk/publications/15432 10. Cecil E, Bottle A, Sharland M, Saxena S. Impact of UK Primary Care Policy Reforms on Short-Stay Unplanned Hospital Admissions for Children With Primary Care-Sensitive Conditions. Ann Fam Med. 2015 Jun;13(3):214–20. 11. Lynn RM, Avis JL, Lenton S, Amin-Chowdhury Z, Ladhani SN. Delayed access to care and late presentations in children during the COVID-19 pandemic: a snapshot survey of 4075 paediatricians in the UK and Ireland. Arch Dis Child. 2021 Feb 1;106(2):e8–e8. 12. Harron K, Gilbert R, Cromwell D, Meulen J van der. Linking Data for Mothers and Babies in De-Identified Electronic Health Data. PLOS ONE. 2016 Oct 20;11(10):e0164667. 13. Hospital Episode Statistics data changes in 2021 [Internet]. NHS Digital. [cited 2021 Nov 30]. Available from: https://digital.nhs.uk/data-and-information/data-tools-and- services/data-services/hospital-episode-statistics/hospital-episode-statistics-data- changes-in-2021 8 8 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint 14. Herbert A, Wijlaars L, Zylbersztejn A, Cromwell D, Hardelid P. Data Resource Profile: Hospital Episode Statistics Admitted Patient Care (HES APC). Int J Epidemiol. 2017 Aug 1;46(4):1093–1093i. 15. Hardelid P, Dattani N, Gilbert R. Estimating the prevalence of chronic conditions in children who die in England, Scotland and Wales: a data linkage cohort study. BMJ Open. 2014 Aug 1;4(8):e005331. 16. Hardelid P, Dattani N, Davey J, Pribramska I, Gilbert R. Overview of child deaths in the four UK countries [Internet]. 2013 [cited 2021 Apr 13]. Available from: https://www.hqip.org.uk/resource/overview-of-child-deaths-in-the-four-uk-countries/ 17. Feudtner C, Christakis DA, Connell FA. Pediatric deaths attributable to complex chronic conditions: a population-based study of Washington State, 1980-1997. Pediatrics. 2000 Jul;106(1 Pt 2):205–9. 18. Office of the Deputy Prime Minister, Neighbourhood Renewal Unit. The English indices of deprivation 2004. [Internet]. References: London: Office of the Deputy Prime Minister; 2004. Available from: https://www.brent.gov.uk/media/323946/SD%2062b%20Indices%20of%20multiple%2 0deprivation%202004.pdf 19. Ward J, Hargreaves D, Rogers M, Firth A, Turner S, Viner R, et al. Recent and forecast post-COVID trends in hospital activity in England amongst 0 to 24 year olds: analyses using routine hospital administrative data [Internet]. 2021 Feb [cited 2021 Nov 30] p. 2021.02.11.21251584. Available from: https://www.medrxiv.org/content/10.1101/2021.02.11.21251584v1 20. Prime Minister’s statement on coronavirus (COVID-19): 23 March 2020 [Internet]. GOV.UK. [cited 2021 Apr 20]. Available from: https://www.gov.uk/government/speeches/pm-address-to-the-nation-on-coronavirus- 23-march-2020 21. StataCorp. Stata Statistical Software. College Station, TX: StataCorp LLC; 2017. 22. Rose KE, Bressan S, Honeyford K, Titomanlio L, Nijman R, Group ESGES. Understanding responses of paediatric emergency departments to the first wave of the Covid-19 pandemic – a pan-European perspective. Arch Dis Child. 2021 Oct 1;106(Suppl 1):A161–2. 23. Nijman R, Honeyford K, Bressan S, Titomanlio L, Study S group E. Epidemiology, severity and outcomes of children presenting to emergency departments across Europe during the SARS-COV-2 pandemic: an observational cohort study. Arch Dis Child. 2021 Oct 1;106(Suppl 1):A469–70. 9 9 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint 24. Scaramuzza A, Tagliaferri F, Bonetti L, Soliani M, Morotti F, Bellone S, et al. Changing admission patterns in paediatric emergency departments during the COVID-19 pandemic. Arch Dis Child. 2020 Jul 1;105(7):704–6. 25. Clavenna A, Nardelli S, Sala D, Fontana M, Biondi A, Bonati M. Impact of COVID-19 on the Pattern of Access to a Pediatric Emergency Department in the Lombardy Region, Italy. Pediatr Emerg Care. 2020 Oct;36(10):e597. 26. Matera L, Nenna R, Ardenti Morini F, Banderali G, Calvani M, Calvi M, et al. Effects of Relaxed Lockdown on Pediatric ER Visits during SARS-CoV-2 Pandemic in Italy. Int J Environ Res Public Health. 2021 Sep 10;18(18):9547. 27. Chavasse R, Almario A, Christopher A, Kappos A, Shankar A. The Indirect Impact of COVID-19 on Children With Asthma. Arch Bronconeumol. References: 2020 Nov;56(11):768–9. 28. Angoulvant F, Ouldali N, Yang DD, Filser M, Gajdos V, Rybak A, et al. Coronavirus Disease 2019 Pandemic: Impact Caused by School Closure and National Lockdown on Pediatric Visits and Admissions for Viral and Nonviral Infections—a Time Series Analysis. Clin Infect Dis. 2021 Jan 15;72(2):319–22. 29. Kruizinga MD, Peeters D, van Veen M, van Houten M, Wieringa J, Noordzij JG, et al. The impact of lockdown on pediatric ED visits and hospital admissions during the COVID19 pandemic: a multicenter analysis and review of the literature. Eur J Pediatr. 2021 Jul 1;180(7):2271–9. 30. Williams TC, MacRae C, Swann OV, Haseeb H, Cunningham S, Davies P, et al. Indirect effects of the COVID-19 pandemic on paediatric healthcare use and severe disease: a retrospective national cohort study. Arch Dis Child. 2021 Sep 1;106(9):911–7. 31. Dann L, Fitzsimons J, Gorman KM, Hourihane J, Okafor I. Disappearing act: COVID-19 and paediatric emergency department attendances. Arch Dis Child. 2020 Aug 1;105(8):810–1. 32. Kishimoto K, Bun S, Shin J, Takada D, Morishita T, Kunisawa S, et al. Early impact of school closure and social distancing for COVID-19 on the number of inpatients with childhood non-COVID-19 acute infections in Japan. Eur J Pediatr. 2021 Sep 1;180(9):2871–8. 33. Hartnett KP. Impact of the COVID-19 Pandemic on Emergency Department Visits — United States, January 1, 2019–May 30, 2020. MMWR Morb Mortal Wkly Rep [Internet]. 2020 [cited 2021 Dec 6];69. Available from: https://www.cdc.gov/mmwr/volumes/69/wr/mm6923e1.htm 10 10 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint 34. Nolen LD, Seeman S, Bruden D, Klejka J, Desnoyers C, Tiesinga J, et al. Impact of Social Distancing and Travel Restrictions on Non–Coronavirus Disease 2019 (Non–COVID-19) Respiratory Hospital Admissions in Young Children in Rural Alaska. Clin Infect Dis. 2021 Jun 15;72(12):2196–8. 35. Isba R, Edge R, Auerbach M, Cicero MX, Jenner R, Setzer E, et al. References: COVID-19: Transatlantic Declines in Pediatric Emergency Admissions. Pediatr Emerg Care. 2020 Nov 1;36(11):551–3. 36. Goldman RD, Grafstein E, Barclay N, Irvine MA, Portales-Casamar E. Paediatric patients seen in 18 emergency departments during the COVID-19 pandemic. Emerg Med J. 2020 Dec 1;37(12):773–7. 37. Ferrero F, Ossorio MF, Torres FA, Debaisi G. Impact of the COVID-19 pandemic in the paediatric emergency department attendances in Argentina. Arch Dis Child. 2021 Feb 1;106(2):e5–e5. 38. Wirrell EC, Grinspan ZM, Knupp KG, Jiang Y, Hammeed B, Mytinger JR, et al. Care Delivery for Children With Epilepsy During the COVID-19 Pandemic: An International Survey of Clinicians. J Child Neurol. 2020 Nov;35(13):924–33. 39. Todd IMF, Miller JE, Rowe SL, Burgner DP, Sullivan SG. Changes in infection-related hospitalizations in children following pandemic restrictions: an interrupted time-series analysis of total population data. Int J Epidemiol. 2021 Nov 10;50(5):1435–43. 40. Mathur R, Rentsch CT, Morton CE, Eggo RM, Bhaskaran K, Tomlinsonn L, et al. Ethnic disparities in COVID-19: increased risk of infection or severe disease? – Authors’ reply. The Lancet. 2021 Jul 31;398(10298):390. 41. Williamson EJ, Walker AJ, Bhaskaran K, Bacon S, Bates C, Morton CE, et al. Factors associated with COVID-19-related death using OpenSAFELY. Nature. 2020 Aug;584(7821):430–6. 42. Ladhani SN, Amin-Chowdhury Z, Davies HG, Aiano F, Hayden I, Lacy J, et al. COVID-19 in children: analysis of the first pandemic peak in England. Arch Dis Child. 2020 Dec;105(12):1180–5. 43. Mehta NS, Mytton OT, Mullins EWS, Fowler TA, Falconer CL, Murphy OB, et al. SARS- CoV-2 (COVID-19): What Do We Know About Children? A Systematic Review. Clin Infect Dis Off Publ Infect Dis Soc Am. 2020 Dec 3;71(9):2469–79. 44. Nuñez JH, Sallent A, Lakhani K, Guerra-Farfan E, Vidal N, Ekhtiari S, et al. Impact of the COVID-19 Pandemic on an Emergency Traumatology Service: Experience at a Tertiary Trauma Centre in Spain. Injury. 2020 Jul 1;51(7):1414–8. 11 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. References: is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for th this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in was not certified by peer review) The copyright holder for this this version posted December 17, 2021. ; https://doi.org/10.1101/2021.12.16.21267904 oi: . CC-BY 4.0 International license It is made available under a p p y We would like to thank Nicolas Libuy, Pia Hardelid, Chloe Parkin and Matthew Lilliman for their contributions to this project. Ethics approval: Approvals for the use of HES data were obtained as part of the standard Hospital Episode Statistics approval process and ethical approval was obtained from London—South East Research Ethics Committee (reference 16/LO/0012). HES records were made available by NHS Digital. Competing interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Authors’ contributions: The study was conceived by RG, KH, and RB. DE conducted all the analyses with input from KH and RG. DE wrote the manuscript with input from all the authors. All the authors approved the final manuscript. Funding: This work is supported by ADR UK (Administrative Data Research UK), an Economic and Social Research Council (part of UK Research and Innovation) programme [grant number ES/V000977/1]. This research was also supported in part by the NIHR Great Ormond Street Hospital Biomedical Research Centre and Health Data Research UK [grant number LOND1], funded by the UK Medical Research Council and eight other funders. This research benefits from and contributes to the NIHR Children and Families Policy Research Unit, but was not commissioned by the National Institute for Health Research (NIHR) Policy Research Programme. R.G. and R.B. are in part supported by the National Institute for Health Research (NIHR) Children and Families Policy Research Unit. References: ; https://doi.org/10.1101/2021.12.16.21267904 doi: medRxiv preprint 45. Krivec U, Seliger AK, Tursic J. COVID-19 lockdown dropped the rate of paediatric asthma admissions. Arch Dis Child. 2020 Aug 1;105(8):809–10. 46. Kubicek K, Liu D, Beaudin C, Supan J, Weiss G, Lu Y, et al. A Profile of Nonurgent Emergency Department Use in an Urban Pediatric Hospital. Pediatr Emerg Care. 2012 Oct;28(10):977–84. 47. Czeisler MÉ, Marynak K, Clarke KEN, Salah Z, Shakya I, Thierry JM, et al. Delay or Avoidance of Medical Care Because of COVID-19–Related Concerns — United States, June 2020. Morb Mortal Wkly Rep. 2020 Sep 11;69(36):1250–7. 48. McDonnell T, Nicholson E, Conlon C, Barrett M, Cummins F, Hensey C, et al. Assessing the Impact of COVID-19 Public Health Stages on Paediatric Emergency Attendance. Int J Environ Res Public Health. 2020 Jan;17(18):6719. 49. Health Foundation. Public perceptions of health and social care in light of COVID-19 (May 2020). [Internet]. 2020. Available from: https://www.health.org.uk/publications/reports/public-perceptions-of-health-and- social-care-in-light-of-covid-19-may-2020 50. Dayal D, Gupta S, Raithatha D, Jayashree M. Missing during COVID-19 lockdown: Children with onset of type 1 diabetes. Acta Paediatr. 2020;109(10):2144–6. 50. Dayal D, Gupta S, Raithatha D, Jayashree M. Missing during COVID-19 lockdown: Children with onset of type 1 diabetes. Acta Paediatr. 2020;109(10):2144–6. 51. NHS England » Core20PLUS5 – An approach to reducing health inequalities [Internet]. [cited 2021 Nov 30]. Available from: https://www.england.nhs.uk/about/equality/equality-hub/core20plus5/ Acknowledgements: The data for this project is part of the ECHILD project. The ECHILD project is in partnership with NHS Digital and the Department for Education (DfE) and we thank the following individuals for their valuable contributions to the project: Garry Coleman, Richard Caulton, Joanna Geisler, Catherine Day (NHS Digital), Chris Douglass and Gary Connell (DfE). We thank all the children, young people, parents and carers who contributed to the ECHILD project, as well as Dr Erin Walker (UCL Partners) who led this involvement. We would particularly like to thank members of the National Children’ Bureau Young Research Advisors, National Children’ Bureau Family Research Advisory Groups, NIHR Great Ormond Street Hospital (GOSH) Biomedical Research Centre (BRC) Parent and Carer Advisory Group, GOSH Young People’s Forum and GOSH Young Persons Advisory Group for their input to this project. We also gratefully acknowledge all children and families whose de-identified data are used in this analysis. 12 . CC-BY 4.0 International license It is made available under a perpetuity. References: The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. R.B. is supported by a UKRI Innovation Fellowship funded by the Medical Research Council [grant number MR/S003797/1]. K.H. is funded by NIHR [grant number 17/99/19]. Data availability: The data used in this analysis is expected to be available to accredited researchers in 2022 (as part of the ECHILD Database) by applying to the data providers (DfE and NHS Digital). ow birthweight; CC – Chronic conditions References: 13 13 Table 1: Demographic characteristics of children born between January 1, 2015, and March 31, 2021, by vulnerability status None (N=3,138,414; 82.3%) LBW/Preterm only (N=280,667; 7.4%) CC only (N=311,101; 8.2%) Both (N=83,283; 2.2%) Total (N=3,813,465) n % n % n % n % n % Sex Male 1,590,308 50.7 134,871 48.1 183,964 59.1 46,304 55.6 1,955,447 51.3 Female 1,547,082 49.3 145,481 51.8 127,006 40.8 36,860 44.3 1,856,429 48.7 Missing 1,024 0.03 315 0.1 131 0.04 119 0.1 1,589 0.04 Age group Infants 480,832 15.3 36445 13.0 27641 8.9 8317 10.0 553,235 14.5 1-4-year-olds 2,657,582 84.7 244,222 87.0 283,460 91.1 74,966 90.0 3,260,230 85.5 Ethnic group White 2,185,934 69.7 184,525 65.8 228,023 73.3 58,366 70.1 2,656,848 69.7 Black/Black British 140,991 4.5 15,477 5.5 14,150 4.6 4,681 5.6 175,299 4.6 Asian/Asian British 332,009 10.6 41,516 14.8 33,610 10.8 10,857 13.0 417,992 11.0 Any other ethnic groups 273,658 8.7 24,085 8.6 25,135 8.1 6,633 8.0 329,511 8.6 Missing 205,822 6.6 15,064 5.4 10,183 3.3 2,746 3.3 233,815 6.1 Deprivation quintile Q1 (most deprived) 739,549 23.6 83,204 29.7 89,425 28.7 27,254 32.7 939,432 24.6 Q2 612,665 19.5 60,337 21.5 67,853 21.8 18,629 22.4 759,484 19.9 Q3 524,863 16.7 45,794 16.3 55,924 18.0 14,117 17.0 640,698 16.8 Q4 461,118 14.7 37,446 13.3 47,670 15.3 11,287 13.6 557,521 14.6 Q5 (least deprived) 436,147 13.9 34,530 12.3 43,052 13.8 10,119 12.2 523,848 13.7 Missing 364,072 11.6 19,356 6.9 7,177 2.3 1,877 2.3 392,482 10.3 Year of birth 2015 512,624 16.3 46,590 16.6 68,814 22.1 16,021 19.2 644,049 16.9 2016 515,510 16.4 49,020 17.5 63,505 20.4 16,042 19.3 644,077 16.9 2017 511,838 16.3 49,358 17.6 55,396 17.8 15,096 18.1 631,688 16.6 2018 500,133 15.9 47,406 16.9 47,628 15.3 13,730 16.5 608,897 16.0 2019 497,682 15.9 41,713 14.9 40,107 12.9 11,646 14.0 591,148 15.5 2020 487,365 15.5 38,128 13.6 30,598 9.8 9,341 11.2 565,432 14.8 2021 113,262 3.6 8,452 3.0 5,053 1.6 1,407 1.7 128,174 3.4 LBW – Low birthweight; CC – Chronic conditions children born between January 1, 2015, and March 31, 2021, by vulnerability status 14 14 Table 2: Difference between predicted and observed rates of hospital contact during the pandemic (March 2020-2021) among children aged 0 to 4 years, by clinical vulnerability group. Table 2: Difference between predicted and observed rates of hospital contact during the pandemic (March 2020-2021) among children aged 0 to 4 years, by clinical vulnerability group. LBW – Low birthweight; CC – Chronic conditions. α Any CC combines CC only and both. * Significantly ombines CC only and both. * Significantly different from children with no known vulnerability (p<0.05). References: Percentage of children seen Number of hospital contacts Rates per 1000 child-years Outpatient attendances n N % Predicted Observed Difference % change Predicted Observed Difference (95% CI) % change Total 370,623 2,765,941 13.4 1,500,674 1,173,091 -327,583 -21.8 543 424 -118 (-115, -122) -21.8 No known vulnerability 232,976 2,243,088 10.4 635,226 471743 -163,483 -25.7 283 210 -73 (-71, -74) -25.7 Any vulnerability 137,647 522853 26.3 865,448 701,348 -164,100 -19.0 1655 1341 -314* (-302, -326) -19.0 • LBW/Preterm only 29,622 226439 13.1 94660 76,039 -18,621 -19.7 418 336 -82* (-77, -88) -19.7 • CC only 86,825 231671 37.5 586,026 475,226 -110,800 -18.9 2530 2051 -478* (-465, -492) -18.9 • Both 21,200 64743 32.7 184,762 150,083 -34,679 -18.8 2854 2318 -536* (-508, -563) -18.8 Any CCα 107,762 296,414 36.4 771,219 625,309 -145,910 -18.9 2602 2110 -492* (-480, -505) -18.9 Planned admissions Total 52,280 2,765,941 1.9 144,980 92,618 -52,362 -36.1 52 33 -19 (-18, -20) -36.1 No known vulnerability 15,287 2,243,088 0.7 40,568 17,022 -23,546 -58.0 18 8 -10 (-10, -11) -58.0 Any vulnerability 36,993 522853 7.1 104,412 75,596 -28,816 -27.6 200 145 -55* (-51, -59) -27.6 • LBW/Preterm only 1,389 226439 0.6 3,675 1,522 -2,153 -58.6 16 7 -10 (-8, -11) -58.6 • CC only 31,038 231671 13.4 85,367 65,506 -19,861 -23.3 368 283 -86* (-81, -91) -23.3 • Both 4,566 64743 7.1 15,370 8,568 -6,802 -44.3 237 132 -105* (-97, -113) -44.3 Any CC 35,603 296,414 12.0 100,929 74,074 -26,855 -26.6 341 250 -91* (-86, -95) -26.6 Unplanned admissions Total 131,134 2,765,941 4.7 442,174 179,366 -262,808 -59.4 160 65 -95 (-93, -97) -59.4 No known vulnerability 66,830 2,243,088 3.0 254263 77713 -176,550 -69.4 113 35 -79 (-78, -80) -69.4 Any vulnerability 64304 522853 12.3 187911 101653 -86,258 -45.9 359 194 -165* (-159, -171) -45.9 • LBW/Preterm only 6046 226439 2.7 25520 7228 -18,292 -71.7 113 32 -81 (-78, -84) -71.7 • CC only 52285 231671 22.6 134055 84126 -49,929 -37.2 579 363 -216* (-209, -222) -37.2 • Both 5973 64743 9.2 28336 10299 -18,037 -63.7 438 159 -279* (-268, -290) -63.7 Any CC 58248 296,414 19.7 162648 94425 -68,223 -41.9 549 319 -230* (-224, -236) -41.9 LBW – Low birthweight; CC – Chronic conditions. α Any CC combines CC only and both. * Significantly different from children with no known vulnerability (p<0.05). References: 15 15 Table 3: Difference in predicted and observed rates of hospital contact per 1,000 child-years among children aged 0 to 4 years during the pandemic (March 2020-2021), by period and clinical vulnerability. References: 1st lockdown March 23 to June 23, 2020 Easing of restrictions June 24 to Nov 4, 2020 2nd lockdown Nov 5 to Dec 31, 2020 3rd lockdown Jan 1 to March 22, 2021 Outpatient attendances Deficit (95% CI) % change Deficit (95% CI) % change Deficit (95% CI) % change Deficit (95% CI) % change Total -167 (-163, -170) -30.7 -88 (-86, -91) -16.3 -96 (-91, -100) -17.9 -128 (-123, -132) -23.5 No known vulnerability -106 (-104, -107) -37.2 -55 (-54, -56) -19.5 -56 (-55, -58) -20.6 -75 (-73, -77) -26.2 Any vulnerability -428* (-417, -439) -25.9 -230* (-221, -240) -13.9 -264* (-249, -278) -16.0 -352* (-337, -367) -21.4 • LBW/Preterm only -121* (-116, -127) -28.7 -57 (-53, -61) -13.8 -73* (-67, -80) -18.0 -84 (-77, -92) -19.9 • CC only -664* (-651, -677) -26.2 -357* (-346, -368) -14.0 -398* (-381, -414) -15.9 -516* (-499, -533) -20.8 • Both -655* (-629, -681) -23.3 -385* (-364, -407) -13.7 -450* (-416, -483) -15.5 -699* (-665, -735) -24.2 Any CC -663* (-652, -675) -25.5 -364* (-354, -374) -14.0 -410* (-395, -425) -15.8 -559* (-544, -574) -21.7 Planned admissions Total -32 (-31, -33) -59.2 -16 (-15, -17) -29.3 -9 (-8, -11) -19.2 -16 (-15, -17) -31.9 No known vulnerability -14 (-14, -15) -79.1 -9 (-9, -9) -48.5 -7 (-6, -7) -38.9 -11 (-11, -12) -62.0 Any vulnerability -105* (-101, -109) -51.6 -46* (-43, -49) -22.0 -22* (-18, -26) -11.6 -36* (-32, -41) -19.4 • LBW/Preterm only -12* (-11, -13) -75.3 -8 (-7, -9) -49.4 -6 (-5, -7) -42.1 -11 (-11, -10) -64.9 • CC only -185* (-180, -190) -49.0 -71* (-67, -75) -18.4 -24* (-18, -30) -6.9 -40* (-34, -46) -11.6 • Both -144* (-137, -152) -60.3 -86* (-80, -93) -35.8 -70* (-61, -79) -31.6 -114* (-105, -125) -48.2 Any CC -177* (-172, -181) -50.8 -75* (-71, -78) -21.2 -35* (-30, -40) -10.7 -57* (-52, -62) -17.7 Unplanned admissions Total -88 (-87, -90) -60.0 -77 (-76, -79) -51.9 -140 (-138, -143) -66.2 -99 (-97, -101) -64.2 No known vulnerability -68 (-67, -69) -66.5 -66 (-66, -67) -63.2 -116 (-115, -118) -75.3 -84 (-83, -85) -76.7 Any vulnerability -176* (-171, -182) -51.6 -124* (-120, -128) -36.9 -245* (-237, -253) -53.2 -161* (-154, -168) -47.1 • LBW/Preterm only -69 (-66, -72) -68.0 -64 (-62, -66) -64.1 -125* (-121, -130) -78.3 -90* (-86, -94) -79.9 • CC only -263* (-257, -270) -47.2 -157* (-152, -162) -28.5 -309* (-300, -318) -43.1 -189* (-181, -197) -34.8 • Both -241* (-231, -251) -59.1 -216* (-208, -225) -54.8 -434* (-419, -450) -73.1 -311* (-298, -325) -72.8 Any CC -259* (-254, -265) -49.3 -171* (-166, -175) -33.0 -337* (-330, -345) -48.8 -217* (-210, -224) -41.8 LBW – Low birthweight; CC – Chronic conditions * Significantly different from children with no known vulnerability (5% level of significance). LBW – Low birthweight; CC – Chronic conditions * Significantly different from children with no known LBW – Low birthweight; CC – Chronic conditions * Significantly different from children with no known vulnerability (5% level of significance). onditions * Significantly different from children with no known vulnerability (5% level of significance). References: 16 Figure 1: Deficit in care during the pandemic (March 2020-2021), estimated from predicted minus observed rates of hospital contacts per 1,000 child-years for children aged 0 to 4 years, by clinical vulnerability status and risk factors. Figure 1: Deficit in care during the pandemic (March 2020-2021), estimated from predicted minus observed rates of hospital contacts per 1,000 child-years for children aged 0 to 4 years, by clinical vulnerability status and risk factors. igure 1: Deficit in care during the pandemic (March 2020-2021), estimated from predicted minus observed rates of hospital contacts per ,000 child-years for children aged 0 to 4 years, by clinical vulnerability status and risk factors. b h h h d 1,000 child-years for children aged 0 to 4 years, by clinical vulnerability status and risk factors. LBW – Low birthweight; CC – Chronic conditions LBW LBW – Low birthweight; CC – Chronic conditions 17 17 Figure 2: Weekly rates of hospital contacts among children aged 0 to 4 years during the pandemic (March 2020-2021) and averaged for 2015-2019, by age and presence of a chronic condition. Figure 2: Weekly rates of hospital contacts among children aged 0 to 4 years during the pandemic (March 2020-2021) and averaged for 2015-2019, by age and presence of a chronic condition. Note: January-March lockdown only affects 2021 data. Note: January-March lockdown only affects 2021 data. Note: January-March lockdown only affects 2021 data. 18
https://openalex.org/W2415671922	https://discovery.dundee.ac.uk/files/9503215/fcell_04_00056.pdf	English	null	Emerging Roles of the Mitogen and Stress Activated Kinases MSK1 and MSK2	Frontiers in cell and developmental biology	2,016	cc-by	8,314	Emerging roles of the mitogen and stress activated kinases MSK1 and MSK2 Reyskens, Kathleen M. S. E.; Arthur, J. Simon C. Published in: Frontiers in Cell and Developmental Biology DOI: 10.3389/fcell.2016.00056 Publication date: 2016 Licence: CC BY Document Version Publisher's PDF, also known as Version of record Link to publication in Discovery Research Portal Citation for published version (APA): Reyskens, K. M. S. E., & Arthur, J. S. C. (2016). Emerging roles of the mitogen and stress activated kinases MSK1 and MSK2. Frontiers in Cell and Developmental Biology, 4, Article 56. https://doi.org/10.3389/fcell.2016.00056 Emerging roles of the mitogen and stress activated kinases MSK1 and MSK2 Reyskens, Kathleen M. S. E.; Arthur, J. Simon C. Reyskens, Kathleen M. S. E.; Arthur, J. Simon C. Link to publication in Discovery Research Portal Citation for published version (APA): Reyskens, K. M. S. E., & Arthur, J. S. C. (2016). Emerging roles of the mitogen and stress activated kinases MSK1 and MSK2. Frontiers in Cell and Developmental Biology, 4, Article 56. https://doi.org/10.3389/fcell.2016.00056 General rights Copyright and moral rights for the publications made accessible in Discovery Research Portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Citation for published version (APA): Reyskens, K. M. S. E., & Arthur, J. S. C. (2016). Emerging roles of the mitogen and stress activated kinases MSK1 and MSK2. Frontiers in Cell and Developmental Biology, 4, Article 56. https://doi.org/10.3389/fcell.2016.00056 Emerging Roles of the Mitogen and Stress Activated Kinases MSK1 and MSK2 Kathleen M. S. E. Reyskens and J. Simon C. Arthur* Division of Cell Signalling and Immunology, School of Life Sciences, University of Dundee, Dundee, UK Mitogen- and stress-activated kinases (MSK) 1 and 2 are nuclear proteins activated downstream of the ERK1/2 or p38 MAPK pathways. MSKs phosphorylate multiple substrates, including CREB and Histone H3, and their major role is the regulation of speciﬁc subsets of Immediate Early genes (IEG). While MSKs are expressed in multiple tissues, their levels are high in immune and neuronal cells and it is in these systems most is known about their function. In immunity, MSKs have predominantly anti-inﬂammatory roles and help regulate production of the anti-inﬂammatory cytokine IL-10. In the CNS they are implicated in neuronal proliferation and synaptic plasticity. In this review we will focus on recent advances in understanding the roles of MSKs in the innate immune system and neuronal function. Edited by: Edited by: Ana Cuenda, Spanish National Research Council, National Centre of Biotechnology, Spain Keywords: MSK1, MSK2, CREB, MAPK, p38, IL-10, innate immunity, synaptic plasticity INTRODUCTION Reviewed by: Piero Crespo, Consejo Superior de Investigaciones Cientíﬁcas - IBBTEC, Spain Cathy Tournier, University of Manchester UK Reviewed by: Piero Crespo, Consejo Superior de Investigaciones Cientíﬁcas - IBBTEC, Spain Cathy Tournier, University of Manchester, UK Correspondence: J. Simon C. Arthur j.s.c.arthur@dundee.ac.uk MSK1 and 2 were ﬁrst identiﬁed in 1998 as proteins that shared homology to the RSK kinase family (Deak et al., 1998; New et al., 1998; Pierrat et al., 1998). MSKs contain 2 kinase domains—an N-terminal kinase domain (NTKD) in the AGC kinase family and a C-terminal kinase domain (CTKD) from the calmodulin kinase family (Caenepeel et al., 2004). MSKs are activated by the ERK1/2 or p38 MAPK pathways. ERK1/2 and/or p38 phosphorylates 3 sites on MSKs, which activates the CTKD. This causes autophosphorylation and activation of the NTKD, which in turn phosphorylates MSK substrates (Deak et al., 1998; McCoy et al., 2005, 2007). MSKs exist in all vertebrate species examined, with the exception of the lamprey. The chordates Ciona intestinalis, Ciona savignyi, and Branchiostoma ﬂoridae possess a single MSK homolog indicating that MSKs diverged from RSK before the onset of vertebrate evolution. MSK1 and 2 probably derive from a duplication event early in vertebrate evolution; orthologs of both MSK1 and 2 exist in the elephant shark, Callorhinchus milii (Venkatesh et al., 2014) and in bony ﬁsh. In cells, MSK1 and MSK2 are functionally redundant (Wiggin et al., 2002). Perhaps because of this, not all vertebrates have retained both genes. In the Ensembl database, MSK2 is absent in 5 bird genomes while 11 out of the 38 mammalian genomes appear to have lost MSK isoforms. Orthologs of MSKs have been identiﬁed in nematodes and insects, although of these only the Drosophila protein Jil-1 has been studied in detail (Jin et al., 1999; Wang et al., 2001). Interestingly the main area of homology for Jil-1 with MSKs is in the NTKD. Unlike the chordate and nematode MSKs, the CTKD of Jil-1 lacks the classical MAPK phosphorylation sites. Correspondence: J. Simon C. Arthur j.s.c.arthur@dundee.ac.uk Specialty section: This article was submitted to Signaling, a section of the journal Frontiers in Cell and Developmental Biology Received: 28 February 2016 Accepted: 30 May 2016 Published: 10 June 2016 Citation: Reyskens KMSE and Arthur JSC (2016) Emerging Roles of the Mitogen and Stress Activated Kinases MSK1 and MSK2. Front. Cell Dev. Biol. 4:56. INTRODUCTION doi: 10.3389/fcell.2016.00056 Specialty section: This article was submitted to Signaling, a section of the journal Frontiers in Cell and Developmental Biology Received: 28 February 2016 Accepted: 30 May 2016 Published: 10 June 2016 Keywords: MSK1, MSK2, CREB, MAPK, p38, IL-10, innate immunity, synaptic plasticity General rights i h d Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Download date: 24. Oct. 2024 Download date: 24. Oct. 2024 Download date: 24. Oct. 2024 Download date: 24. Oct. 2024 MINI REVIEW published: 10 June 2016 doi: 10.3389/fcell.2016.00056 MINI REVIEW Citation: (reviewed in Arthur, 2008). The majority of the genes regulated by MSKs are also targets of CREB. The detailed molecular mechanism by which MSKs regulate CREB is however unclear. MSKs phosphorylate CREB on Ser133, a site that is also targeted by other kinases including protein kinase A (PKA). Phosphorylation of CREB by PKA creates a binding site for the co-activator proteins CBP and p300 (Gonzalez and Montminy, 1989; Chrivia et al., 1993; Cardinaux et al., 2000; Mayr and Montminy, 2001), whose recruitment promotes the eﬃcient transcription of CREB target genes (Yamamoto et al., 1988; Gonzalez and Montminy, 1989). Several studies have demonstrated that CREB phosphorylation downstream of MAPK signaling does not lead to CBP or p300 recruitment (Brindle et al., 1995; Mayr and Montminy, 2001; Mayr et al., 2001; Kasper et al., 2011), leading to the suggestion that MAPKs, and by inference MSKs, do not activate CREB-dependent transcription. Furthermore, PKA- but not MSK-mediated CREB phosphorylation leads to eﬃcient CBP or p300 recruitment to endogenous CREB-dependent promoters as judged by chromatin immunoprecipitation (ChIP) (Naqvi et al., 2014). Despite this, mutation of the Ser133 site to alanine in the endogenous CREB gene actually had a bigger impact on CREB target genes in response to MSK activating stimuli compared to PKA activating stimuli. How MSK-mediated CREB phosphorylation activates CREB is however unresolved (Naqvi et al., 2014). A number of other substrates for MSKs, including NFκB, HMG-14, RAR- related orphan receptor alpha (RORα), KDM3A, Trim7, and Trim28 have been proposed, although their overall importance to MSK function is currently less clear (Soloaga et al., 2003; Vermeulen et al., 2003; Bruck et al., 2009; Cheng et al., 2014; Chakraborty et al., 2015; Singh et al., 2015). While MSKs are expressed in many tissues, their function, as discussed below, has been best studied in the innate immune system and brain. similar to that found in MSK1/2 knockouts with decreased endogenous IL-10. However, additional neutralization of IL-10 in MSK1/2 knockouts did not further aﬀect pro-inﬂammatory cytokine production in isolated macrophages (Ananieva et al., 2008). Interestingly the eﬀect on TNF was partial, suggesting additional IL-10 independent mechanisms of TNF regulation by MSKs. MSKs regulate IL-10 by controlling its transcription in macrophages and dendritic cells (Ananieva et al., 2008; Elcombe et al., 2013). Citation: In addition to IL-10, several other anti-inﬂammatory genes are regulated by MSKs in macrophages, including dual speciﬁcity protein phosphatase 1 (DUSP1), tristetraprolin (TTP) and IL-1 receptor agonist (IL-1ra) (Brook et al., 2006; Ananieva et al., 2008; Darragh et al., 2010) (Figure 1). DUSP1 is a phosphatase that inactivates the p38 and JNK MAPKs, and like MSK1/2 knockout, DUSP1 knockout sensitizes mice to endotoxic shock (reviewed in Lang et al., 2006; Wang and Liu, 2007). TTP is able to bind to AU rich elements in the 3′ UTR of multiple cytokines including TNF, inhibiting translation and promoting mRNA degradation (Brooks and Blackshear, 2013). IL-1ra is a member of the IL-1 family that binds to the IL-1 receptor but cannot activate signaling, thus acting as an inhibitor of IL-1 in vivo (Garlanda et al., 2013). TTP knockout gives rise to multiple inﬂammatory phenotypes leading to early mortality (Taylor et al., 1996). Loss of IL-10 in mice or humans results in the development of colitis (Kühn et al., 1993; Shah et al., 2012) while mutation of IL-1ra in humans gives rise to an early onset auto-inﬂammatory condition, DIRA (Aksentijevich et al., 2009). Given the phenotypes associated with its target genes, it is surprising that MSK knockout does not give a more overt phenotype. This may be because other pathways compensate for MSK in vivo. Related to this, CREB is phosphorylated by other kinases in addition to MSKs including PKA (Johannessen et al., 2004). Prostaglandin E2 (PGE2) activates PKA in macrophages and acts synergistically with TLR agonists to induce IL-10. This eﬀect is dependent on CREB, but does not require PKA-mediated CREB phosphorylation (MacKenzie et al., 2013b). Instead PKA phosphorylates salt-inducible kinase 2 (SIK2) thus inhibiting the ability of SIK2 to phosphorylate CRTC3. As a result CRTC3 becomes dephosphorylated and translocates to the nucleus where it acts as a co-activator for CREB on the IL-10 promoter (Clark et al., 2012; MacKenzie et al., 2013b). Citation: Reyskens KMSE and Arthur JSC (2016) Emerging Roles of the Mitogen and Stress Activated Kinases MSK1 and MSK2. Front. Cell Dev. Biol. 4:56. doi: 10.3389/fcell.2016.00056 y MSKs are predominantly localized to the nucleus and this is reﬂected in their known substrates. The best characterized MSK substrates are Histone H3 and the related transcription factors CREB and ATF1. These have been validated via both pharmacological inhibition and mouse genetics June 2016 \| Volume 4 \| Article 56 Frontiers in Cell and Developmental Biology \| www.frontiersin.org Physiological Roles of MSK1 and 2 Reyskens and Arthur (reviewed in Arthur, 2008). The majority of the genes regulated by MSKs are also targets of CREB. The detailed molecular mechanism by which MSKs regulate CREB is however unclear. MSKs phosphorylate CREB on Ser133, a site that is also targeted by other kinases including protein kinase A (PKA). Phosphorylation of CREB by PKA creates a binding site for the co-activator proteins CBP and p300 (Gonzalez and Montminy, 1989; Chrivia et al., 1993; Cardinaux et al., 2000; Mayr and Montminy, 2001), whose recruitment promotes the eﬃcient transcription of CREB target genes (Yamamoto et al., 1988; Gonzalez and Montminy, 1989). Several studies have demonstrated that CREB phosphorylation downstream of MAPK signaling does not lead to CBP or p300 recruitment (Brindle et al., 1995; Mayr and Montminy, 2001; Mayr et al., 2001; Kasper et al., 2011), leading to the suggestion that MAPKs, and by inference MSKs, do not activate CREB-dependent transcription. Furthermore, PKA- but not MSK-mediated CREB phosphorylation leads to eﬃcient CBP or p300 recruitment to endogenous CREB-dependent promoters as judged by chromatin immunoprecipitation (ChIP) (Naqvi et al., 2014). Despite this, mutation of the Ser133 site to alanine in the endogenous CREB gene actually had a bigger impact on CREB target genes in response to MSK activating stimuli compared to PKA activating stimuli. How MSK-mediated CREB phosphorylation activates CREB is however unresolved (Naqvi et al., 2014). A number of other substrates for MSKs, including NFκB, HMG-14, RAR- related orphan receptor alpha (RORα), KDM3A, Trim7, and Trim28 have been proposed, although their overall importance to MSK function is currently less clear (Soloaga et al., 2003; Vermeulen et al., 2003; Bruck et al., 2009; Cheng et al., 2014; Chakraborty et al., 2015; Singh et al., 2015). While MSKs are expressed in many tissues, their function, as discussed below, has been best studied in the innate immune system and brain. MSKs in Innate Immunity p38α MAPK has been extensively studied in innate immunity and regulates the production of pro-inﬂammatory cytokines in innate immune cells. As a result p38 inhibitors have been developed as potential anti-inﬂammatory drugs, although none have progressed in the clinic. More recently, anti-inﬂammatory roles for p38 have emerged which may contribute to the lack of eﬃcacy of p38 inhibitors in the clinic (reviewed in Arthur and Ley, 2013; Salgado et al., 2014). A key question therefore was whether MSKs regulated pro- or anti- inﬂammatory functions downstream of p38α. MSK1/2 knockout does not result in an overt phenotype in unchallenged mice (Wiggin et al., 2002), however these mice are sensitized to lipopolysaccharide (LPS)-induced endotoxic shock, indicating that MSK activation has an anti-inﬂammatory eﬀect (Ananieva et al., 2008). MSK1/2 knockout also resulted in elevated levels of TNF, IL-6, and IL-12 production downstream of LPS stimulation, but decreased production of the anti- inﬂammatory cytokine IL-10 (Ananieva et al., 2008; Kim et al., 2008). IL-10 is known to repress pro-inﬂammatory cytokine production, and neutralization of IL-10 in wildtype isolated macrophages increases the production of IL-6 and IL-12, MSKs also have complex roles in regulating prostaglandin production. Ptgs2, the rate-limiting enzyme in prostaglandin production, is a CREB regulated gene. Early Ptgs2 transcription downstream of TLR agonists is positively regulated by MSKs, suggesting that MSK inhibition might reduce prostaglandin production. IL-10 can however suppress Ptgs2 mRNA induction. As MSK knockout reduces IL-10 production, at later time points after LPS stimulation MSK1/2 knockout macrophages actually show increased Ptgs2 induction and elevated prostaglandin production (MacKenzie et al., 2013a). In addition to their anti-inﬂammatory roles, various pro- inﬂammatory roles have also been suggested for MSKs. In human neutrophils LPS can induce MSK-dependent CREB phosphorylation. Both CREB phosphorylation and induction of Frontiers in Cell and Developmental Biology \| www.frontiersin.org June 2016 \| Volume 4 \| Article 56 2 Physiological Roles of MSK1 and 2 Reyskens and Arthur FIGURE 1 \| Regulation of innate immune function by MSK1 and 2. In innate immune cells such as macrophages and dendritic cells TLRs, with the exception of TLR3, can activate downstream signaling via Myd88. Upon ligand binding, the TLR recruits Myd88 resulting in the formation of a Myd88osome that also contains IRAK4 and IRAK1 and/or 2. This leads to the recruitment of Traf6 and the formation of K63 and M1 ubiquitin chains that help mediate the activation of Tak1. MSKs in Innate Immunity Tak1 then activates the p38 MAPK cascade and indirectly activates ERK1/2 via signaling to the IKK-mediated activation of Tpl2 through p105. ERK1/2 and p38 are then both able to activate MSK1 and 2, which in turn phosphorylate CREB and Histone H3. This leads to the induction of a number of genes with potential anti-inﬂammatory roles including IL-10, IL-1ra, TTP, DUSP1, and Ptgs2. Ptgs2 is the rate-limiting enzyme in the production of prostaglandins, small lipids that can have both pro- and anti-inﬂammatory roles. In the context of macrophages, prostaglandin E can boost IL-10 production and suppress pro-inﬂammatory cytokines following TLR stimulation. It acts via the G-protein coupled receptors EP2 and EP3 to elevate cAMP levels and activate PKA. PKA phosphorylates SIK2 and inhibits its ability to phosphorylate CRTC3. This allows CRTC3 to be dephosphorylated and translocated to the nucleus where it will act as a co-activator for CREB. FIGURE 1 \| Regulation of innate immune function by MSK1 and 2. In innate immune cells such as macrophages and dendritic cells TLRs, with the exception of TLR3, can activate downstream signaling via Myd88. Upon ligand binding, the TLR recruits Myd88 resulting in the formation of a Myd88osome that also contains IRAK4 and IRAK1 and/or 2. This leads to the recruitment of Traf6 and the formation of K63 and M1 ubiquitin chains that help mediate the activation of Tak1. Tak1 then activates the p38 MAPK cascade and indirectly activates ERK1/2 via signaling to the IKK-mediated activation of Tpl2 through p105. ERK1/2 and p38 are then both able to activate MSK1 and 2, which in turn phosphorylate CREB and Histone H3. This leads to the induction of a number of genes with potential anti-inﬂammatory roles including IL-10, IL-1ra, TTP, DUSP1, and Ptgs2. Ptgs2 is the rate-limiting enzyme in the production of prostaglandins, small lipids that can have both pro- and anti-inﬂammatory roles. In the context of macrophages, prostaglandin E can boost IL-10 production and suppress pro-inﬂammatory cytokines following TLR stimulation. It acts via the G-protein coupled receptors EP2 and EP3 to elevate cAMP levels and activate PKA. PKA phosphorylates SIK2 and inhibits its ability to phosphorylate CRTC3. This allows CRTC3 to be dephosphorylated and translocated to the nucleus where it will act as a co-activator for CREB. TNF, IL-8 (CXCL8), CCL3 and CCL4 were reduced by Ro318220, an inhibitor that targets MSKs amongst other AGC kinases. MSKs in Innate Immunity This may reﬂect a diﬀerence between human and mouse skin; if in human keratinocytes MSK acts predominantly to regulate inﬂammation via IL-8, these diﬀerences may relate to the lack of a direct IL-8 homolog in mice. Herpesviridae are a family of double stranded DNA viruses some of which, including Kaposi’s sarcoma-associated herpes virus (KSHV) and human cytomegalovirus (HCMV), cause human disease. Upon infection these viruses can either enter a lytic replication cycle to produce further virions or enter a latent phase (Roizman and Baines, 1991). Following infection of Human Umbilical Vein Endothelial Cells (HUVEC) with KSHV, MSK is activated and phosphorylates CREB. siRNA- mediated knockdown of either CREB or MSK1 and 2 did not prevent infection, but did reduce the production of infectious virions. This correlated with a drop in the levels of the viral genes involved in lytic replication (Cheng et al., 2015). HCMV can be reactivated in cells with a latent infection via a process stimulated by IL-6 (Hargett and Shenk, 2010; Reeves and Compton, 2011). This process correlates with CREB phosphorylation and furthermore CREB binding can be demonstrated on the viral MIEP during reactivation (Kew et al., 2014). This process was blocked by inhibition of the ERK1/2 pathway, suggesting a role for either MSK or RSK. The RSK inhibitor DI-D1870 did not aﬀect this process while H89, a compound that targets several kinases including MSKs, reduced transcription from the MIEP promoter (Kew et al., 2014). The striatum is also involved in addiction. MSK1 regulates glutamate-stimulated Histone H3 phosphorylation in cultured striatal neurons and cocaine administration activated MSK1 in the striatum in vivo (Brami-Cherrier et al., 2007). MSK1 knockout mice showed increased response to low doses of cocaine in placed preference tests, while following repeated injections of cocaine locomotor sensitization was decreased (Brami-Cherrier et al., 2005). The hippocampus plays important roles in encoding memory and is involved in rodent models of spatial memory and contextual fear conditioning. CREB is implicated in synaptic plasticity and memory (Shaywitz and Greenberg, 1999; Lonze and Ginty, 2002; Carlezon et al., 2005; Benito and Barco, 2010; Sakamoto et al., 2011) while the importance of histone modiﬁcations, including phosphorylation, have also been recognized (Day and Sweatt, 2011; Mifsud et al., 2011; Kandel et al., 2014; Alberini and Kandel, 2015). MSKs in Innate Immunity In support of a role for CREB, transfection of PLB-985 cells, a neutrophil-like cell line, with a dominant negative CREB reduced IL-8 and TNF induction (Mayer et al., 2013). MSK1 has also been linked to IL-8 production in human keratinocytes and airway smooth muscle cells, where siRNA-mediated knockdown of MSK1 reduced IL-8 production (Funding et al., 2006; Rahman et al., 2014). Interestingly both MSK1 and 2 activity are increased in human lesional psoriatic skin (Funding et al., 2006, 2007), June 2016 \| Volume 4 \| Article 56 Frontiers in Cell and Developmental Biology \| www.frontiersin.org 3 Physiological Roles of MSK1 and 2 Reyskens and Arthur Atxn154Q mutation, raising the possibility that MSK inhibitors may be useful for treating this disease (Park et al., 2013). Huntington’s disease results from the expansion of a CAG motif in the Htt gene giving rise to a polyglutamine repeat (Walker, 2007). While it aﬀects many areas of the brain, the striatum is particularly sensitive to damage. Decreased MSK1 expression was observed in the caudate nucleus from the striatum of Huntington’s patients (Roze et al., 2008). In the R6/2 transgenic mouse model of Huntington’s, both MSK1 and Histone H3 Ser10 phosphorylation were decreased (Roze et al., 2008). Furthermore, MSK1 overexpression can promote expression of PGC-1α, a gene that is neuro-protective in Huntington’s and MSK1 knockout mice showed evidence of striatal degeneration upon aging (Martin et al., 2011). Parkinson’s disease is associated with reduced dopamine levels. Dopamine therapy, while beneﬁcial for Parkinson’s can result in Levodopa-induced dyskinesia (LID). In animal models this correlates with increased δFosB expression in the striatum. Two studies showed that MSK knockout decreased δFosB expression in LID, however while in one study LID intensity was attenuated in the other it was not (Brami-Cherrier et al., 2005; Alcacer et al., 2014; Feyder et al., 2016). suggesting a potential involvement of MSKs in the pathology of this disease. Related to this, DMF is used as a treatment for psoriasis and it has been proposed that its eﬃcacy may be in part due to MSK inhibition (Gesser et al., 2007; Peng et al., 2012). In mice however MSK1/2 knockout increases skin inﬂammation following PMA treatment or in oxazolone-induced allergic contact dermatitis (Ananieva et al., 2008; Bertelsen et al., 2011). MSKs in Innate Immunity The MAPK pathway has been also found to be a critical component of consolidation of memory in the hippocampus (Besnard et al., 2014) and ERK1/2-stimulated histone phosphorylation and acetylation in hippocampal CA1 neurons is associated with memory consolidation (Levenson et al., 2004; Chwang et al., 2006). The involvement of both the upstream activators of MSK and its substrates in these processes suggest the involvement of MSKs in memory. Frontiers in Cell and Developmental Biology \| www.frontiersin.org MSKs and Neuronal Function Both MSK1 and MSK2 are expressed in the brain; however MSK1 is the major isoform in most brain structures (Arthur et al., 2004). While MSK1/2 knockout mice do not exhibit gross defects in CNS development, under some circumstances MSKs play roles in neuronal proliferation or survival. Following pilocarpine- induced seizure, proliferation of neuronal progenitors in the subgranular zone (SGZ) of the dentate gyrus of mice was reduced by knockout of MSK1 and 2. In addition reduced neurite arborization was also observed in immature neurons in this region, suggesting that MSKs were helping drive proliferation and the maturation of new neurons (Choi et al., 2012). Similar results were also reported in the SGZ in a model of cerebral ischemia (Karelina et al., 2015). In primary cortical neuronal cultures stimulated with the neurotrophin BDNF, MSK1 was the critical isoform for both CREB phosphorylation and the induction of CREB-dependent IEGs (Arthur et al., 2004). The eﬀect of MSK1 knockout has now been reported in several hippocampal-dependent learning models. In the forced swim test, mice repeatedly placed in a pool with no ability to escape will display a learned helplessness (De Pablo et al., 1989; West, 1990; Korte, 2001). These behavioral changes are accompanied by an increase in phosphorylation and acetylation of Histone H3 in dentate gyrus granule neurons (Bilang-Bleuel et al., 2005), and knockout of MSK1/2 restricted Several reports have also suggested roles for MSKs in neurodegenerative diseases. Spinocerebellar ataxia type 1 (SCA1), a condition resulting from the expansion of a polyglutamine tract in ataxin-1, results in neurodegeneration in the cerebellum and brain stem (Manto, 2005). Ataxin-1 can be phosphorylated by MSKs on Ser766 and lead to stabilization of mutant forms of the protein. Loss of one or more MSK alleles was protective in a mouse model of SCA1 driven by an June 2016 \| Volume 4 \| Article 56 4 Physiological Roles of MSK1 and 2 Reyskens and Arthur this behavioral response and correlated with decreased Histone H3 phosphorylation (Chandramohan et al., 2008). of synaptic transmission in response to TTX (Corrêa et al., 2012). BDNF is also involved in the synaptic scaling and remodeling that occurs in response to environmental enrichment in vivo (Baroncelli et al., 2010; Cowansage et al., 2010). A failure to upregulate synaptic strength during environmental enrichment has been reported in MSK1 kinase dead knockin (Corrêa et al., 2012) and MSK1 knockout mice (Sakamoto et al., 2011). CONCLUSION This review highlights the critical importance of MSKs in limiting inﬂammation in innate immunity and their role in inﬂammatory disease, however their role in the adaptive immune responses remain uncharacterized. MSKs are also implicated in neurodegeneration and synaptic plasticity. Currently, much more work remains to uncover their precise roles in these processes of inﬂammation and neurodegenerative disease. Outside these systems, work has recently suggested further roles for MSKs. For example MSKs may be involved in skin tumor formation (Chang et al., 2011; Dong et al., 2014; Liu et al., 2014) and have been linked to cellular transformation (Raab- Traub, 2002; Pérez-Cadahía et al., 2011; Reyes et al., 2014). Future research will help unravel the exact mechanisms underlying these enzymes in vivo in health and disease. Although the role of MSKs has not been reported in LTP, their role has been investigated in homeostatic scaling. This is a form of non-Hebbian plasticity that allows a neuron to co- ordinately regulate the strengths of all its synaptic inputs in order to maintain its ﬁring rate within physiological boundaries (Davis, 2006), thus allowing protection of neural networks from excessive or inhibitive stimuli. This scaling process has been linked to levels of BDNF (Rutherford et al., 1998; Turrigiano et al., 1998). Homeostatic scaling can be modeled in culture by globally blocking axon potentials using TTX, which blocks voltag- gated sodium channels. Cultured hippocampal neurons from MSK1 kinase-dead mice failed to show homeostatic scaling FUNDING SA is funded by the Medical Research Council and Arthritis Research UK. SA is funded by the Medical Research Council and Arthritis Research UK. AUTHOR CONTRIBUTIONS All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication. MSKs and Neuronal Function Given the role of MSKs in response to environmental enrichment it will be especially interesting to look at the eﬀect of MSK knockout on LTP and memory in animals raised in enriched conditions. In contextual fear conditioning mice learn to associate a speciﬁc context with an aversive stimulus. MSK1 activation has been shown to occur during this process (Sindreu et al., 2007), which is also associated with CREB and Histone H3 phosphorylation (Impey et al., 1998; Taubenfeld et al., 1999). MSK1 knockout resulted in a mild deﬁcit in contextual fear conditioning and decreased CREB and Histone H3 phosphorylation (Chwang et al., 2007). MSK1 knockout also resulted in a mild impairment in the Morris water maze (Chwang et al., 2007), a widely used model for spatial memory (D’Hooge and De Deyn, 2001). Long term potentiation (LTP), which allows the strengthening of speciﬁc synapses in response to high frequency stimulation, is a form of synaptic plasticity considered to provide a molecular model for encoding memory in the hippocampus (Martin et al., 2000). The role of MSKs have not been directly examined in LTP and it would be of interest to address this in the future. Related to this, in response to neurotrophins, MSKs control the production of two closely related miRNAs, miR-132, and miR-212 that have been implicated in synaptic function (Wayman et al., 2008; Remenyi et al., 2010). mir-132/miR-212 knockouts showed normal LTP in the hippocampus but impaired neocortical theta burst-induced LTP (Remenyi et al., 2013). REFERENCES MSK1 regulates the transcription of IL-1ra in response to TLR activation in macrophages. Biochem. J. 425, 595–602. doi: 10.1042/BJ20091062 Brooks, S., and Blackshear, P. (2013). Tristetraprolin (TTP): interactions with mRNA and proteins, and current thoughts on mechanisms of action. Biochim. Biophys. Acta 1829, 666–679. doi: 10.1016/j.bbagrm.2013.02.003 Davis, G. (2006). Homeostatic control of neural activity: from phenomenology to molecular design. Annu. Rev. Neurosci. 29, 307–323. doi: 10.1146/annurev.neuro.28.061604.135751 Bruck, N., Vitoux, D., Ferry, C., Duong, V., Bauer, A., de Thé, H., et al. (2009). A coordinated phosphorylation cascade initiated by p38MAPK/MSK1 directs RARalpha to target promoters. EMBO J. 28, 34–47. doi: 10.1038/emboj.2008.256 Day, J., and Sweatt, J. (2011). Epigenetic mechanisms in cognition. Neuron 70, 813–829. doi: 10.1016/j.neuron.2011.05.019 Caenepeel, S., Charydczak, G., Sudarsanam, S., Hunter, T., and Manning, G. (2004). The mouse kinome: discovery and comparative genomics of all mouse protein kinases. Proc. Natl. Acad. Sci. U.S.A. 101, 11707–11712. doi: 10.1073/pnas.0306880101 Deak, M., Clifton, A. D., Lucocq, J. M., Alessi, D. R., Lucocq, L. M., and Alessi, D. R. (1998). Mitogen- and stress-activated protein kinase-1 (MSK1) is directly activated by MAPK and SAPK2/p38, and may mediate activation of CREB. EMBO J. 17, 4426–4441. doi: 10.1093/emboj/17.15.4426 Cardinaux, J., Notis, J., Zhang, Q., Vo, N., Craig, J., Fass, D., et al. (2000). Recruitment of CREB binding protein is suﬃcient for CREB-mediated gene activation. Mol. Cell. Biol. 20, 1546–1552. doi: 10.1128/MCB.20.5.1546- 1552.2000 De Pablo, J. M., Parra, A., Segovia, S., and Guillamo, A. (1989). Learned immobility explains the behavior of rats in the forced swim test. Physiol. Behav. 46, 229–237. doi: 10.1016/0031-9384(89)90261-8 Dong, Z., Yao, K., Chen, H., Liu, K., Langfald, A., Yang, G., et al. (2014). Kaempferol targets RSK2 and MSK1 to suppress UV radiation-induced skin cancer. Cancer Prev. Res. 7, 958–967. doi: 10.1158/1940-6207.CAPR-14-0126 Carlezon, W. A., Duman, R. S., and Nestler, E. J. (2005). The many faces of CREB. Trends Neurosci. 28, 436–445. doi: 10.1016/j.tins.2005.06.005 Chakraborty, A., Me, D., Mylona, A., Kassel, O., and Behrens, A. (2015). The E3 ubiquitin ligase Trim7 mediates c-Jun / AP- 1 activation by Ras signalling . Nat. Commun. 6, 6782. doi: 10.1038/ncomms7782 Elcombe, S., Naqvi, S., Van Den Bosch, M., MacKenzie, K., Cianfanelli, F., Brown, G., et al. (2013). Dectin-1 regulates IL-10 production via a MSK1/2 and CREB dependent pathway and promotes the induction of regulatory macrophage markers. PLoS ONE 8:e60086. doi: 10.1371/journal.pone.0060086 Chandramohan, Y., Droste, S. K., Arthur, J. S. C., and Reul, J. M. H. M. REFERENCES Arthur, J., Fong, A. L., Dwyer, J. M., Davare, M., Reese, E., Obrietan, K., et al. (2004). Mitogen- and stress-activated protein kinase 1 mediates cAMP response element-binding protein phosphorylation and activation by neurotrophins. J. Neurosci. 24, 4324–4332. doi: 10.1523/JNEUROSCI.5227-03.2004 Aksentijevich, I., Masters, S. L., Ferguson, P. J., Dancey, P., Frenkel, J., van Royen- Kerkhoﬀ, A., et al. (2009). An autoinﬂammatory disease with deﬁciency of the interleukin-1 receptor antagonist. N. Engl. J. Med. 360, 2426–2437. doi: 10.1056/NEJMoa0807865 Arthur, J. S., and Ley, S. C. (2013). Mitogen-activated protein kinases in innate immunity. Nat. Rev. Immunol. 13, 679–692. doi: 10.1038/nri3495 Alberini, C. M., and Kandel, E. R. (2015). The regulation of transcription in memory consolidation. Cold Spring Harb. Perspect. Biol. 7:a021741. doi: 10.1101/cshperspect.a021741 Baroncelli, L., Braschi, C., Spolidoro, M., Begenisic, T., Sale, A., and Maﬀei, L. (2010). Nurturing brain plasticity: impact of environmental enrichment. Cell Death Diﬀer. 7, 1092–1103. doi: 10.1038/cdd.2009.193 Benito, E., and Barco, A. (2010). CREB’s control of intrinsic and synaptic plasticity: implications for CREB-dependent memory models. Trends Neurosci. 33, 230–240. doi: 10.1016/j.tins.2010.02.001 Alcacer, C., Charbonnier-Beaupel, F., Corvol, J., Girault, J., and Hervé, D. (2014). Mitogen- and stress-activated protein kinase 1 is required for speciﬁc signaling responses in dopamine-denervated mouse striatum, but is not necessary for L-DOPA-induced dyskinesia. Neurosci. Lett. 583, 76–80. doi: 10.1016/j.neulet.2014.09.018 Bertelsen, T., Iversen, L., Riis, J. L., Arthur, J. S. C., Bibby, B. M., Kragballe, K., et al. (2011). The role of mitogen- and stress-activated protein kinase 1 and 2 in chronic skin inﬂammation in mice. Exp. Dermatol. 20, 140–145. doi: 10.1111/j.1600-0625.2010.01153.x Ananieva, O., Darragh, J., Johansen, C., Carr, J. M., McIlrath, J., Park, J. M., et al. (2008). The kinases MSK1 and MSK2 act as negative regulators of toll-like receptor signaling. Nat. Immunol. 9, 1028–1036. doi: 10.1038/ni.1644 receptor signaling. Nat. Immunol. 9, 1028–1036. doi: 10.1038/ni.1 Besnard, A., Laroche, S., and Caboche, J. (2014). Comparative dynamics of MAPK/ERK signalling components and immediate early genes in the hippocampus and amygdala following contextual fear conditioning and Arthur, J. (2008). MSK activation and physiological roles. Front. Biosci. 13:5866–5879. doi: 10.2741/3122 June 2016 \| Volume 4 \| Article 56 Frontiers in Cell and Developmental Biology \| www.frontiersin.org 5 Physiological Roles of MSK1 and 2 Reyskens and Arthur retrieval. Brain Struct. Funct. 219, 415–430. doi: 10.1007/s00429-013-0 529-3 Chwang, W. B., Arthur, J. S., Schumacher, A., and Sweatt, J. D. (2007). REFERENCES The nuclear kinase mitogen- and stress-activated protein kinase 1 regulates hippocampal chromatin remodeling in memory formation. J. Neurosci. 27, 12732–12742. doi: 10.1523/JNEUROSCI.2522-07.2007 retrieval. Brain Struct. Funct. 219, 415–430. doi: 10.1007/s00429-013-0 529-3 Bilang-Bleuel, A., Ulbricht, S., Chandramohan, Y., De, C. S., Droste, S. K., and Reul, J. M. H. M. (2005). Psychological stress increases histone H3 phosphorylation in adult dentate gyrus granule neurons: involvement in a glucocorticoid receptor-dependent behavioural response. Eur. J. Neurosci. 22, 1691–1700. doi: 10.1111/j.1460-9568.2005.04358.x Chwang, W., O’Riordan, K., Levenson, J., and Sweatt, J. (2006). ERK/MAPK regulates hippocampal histone phosphorylation following contextual fear conditioning. Learn. Mem. 13, 322–328. doi: 10.1101/lm.152906 Brami-Cherrier, K., Lavaur, J., Pagès, C., Arthur, J. S. C., and Caboche, J. (2007). Glutamate induces histone H3 phosphorylation but not acetylation in striatal neurons: Role of mitogen- and stress-activated kinase-1. J. Neurochem. 101, 697–708. doi: 10.1111/j.1471-4159.2006.04352.x Clark, K., MacKenzie, K. F., Petkevicius, K., Kristariyanto, Y., Zhang, J., Choi, H. G., et al. (2012). Phosphorylation of CRTC3 by the salt- inducible kinases controls the interconversion of classically activated and regulatory macrophages. Proc. Natl. Acad. Sci.U.S.A. 109, 16986–16991. doi: 10.1073/pnas.1215450109 Brami-Cherrier, K., Valjent, E., Hervé, D., Darragh, J., Corvol, J., Pages, C., et al. (2005). Parsing molecular and behavioral eﬀects of cocaine in mitogen- and stress-activated protein kinase-1-deﬁcient mice. J. Neurosci. 25, 11444–11454. doi: 10.1523/JNEUROSCI.1711-05.2005 Corrêa, S. A. L., Hunter, C. J., Palygin, O., Wauters, S. C., Martin, K. J., McKenzie, C., et al. (2012). MSK1 regulates homeostatic and experience-dependent synaptic plasticity. J. Neurosci. 32, 13039–13051. doi: 10.1523/JNEUROSCI.0930-12.2012 Brindle, P., Nakajima, T., and Montminy, M. (1995). Multiple protein kinase A- regulated events are required for transcriptional induction by cAMP. Proc. Natl. Acad. Sci.U.S.A. 92, 10521–10525. doi: 10.1073/pnas.92.23.10521 Cowansage, K., LeDoux, J., and Monﬁls, M. (2010). Brain-derived neurotrophic factor: a dynamic gatekeeper of neural plasticity. Curr. Mol. Pharmacol. 3, 12–29. doi: 10.2174/1874467211003010012 Brook, M., Tchen, C. R., Santalucia, T., Mcilrath, J., Arthur, J. S. C., Saklatvala, J., et al. (2006). Posttranslational regulation of tristetraprolin subcellular localization and protein stability by p38 mitogen-activated protein kinase and extracellular signal-regulated kinase pathways. Mol. Cell. Biol. 26, 2408–2418. doi: 10.1128/MCB.26.6.2408-2418.2006 D’Hooge, R., and De Deyn, P. (2001). Applications of the Morris water maze in the study of learning and memory. Brain Res. Brain Res. Rev. 36, 60–90. doi: 10.1038/nprot.2006.116 Darragh, J., Ananieva, O., Courtney, A., Elcombe, S., and Arthur, J. S. C. (2010). REFERENCES (2008). The forced swimming-induced behavioural immobility response involves histone H3 phospho-acetylation and c-Fos induction in dentate gyrus granule neurons via activation of the N-methyl-D-aspartate/extracellular signal-regulated kinase/mitogen- and stress-a. Eur. J. Neurosci. 27, 2701–2713. doi: 10.1111/j.1460-9568.2008.06230.x Feyder, M., Södersten, E., Santini, E., Vialou, V., LaPlant, Q., Watts, E., et al. (2016). A role for Mitogen- and Stress-Activated Kinase 1 in L-DOPA- induced dyskinesia and δFosB expression. Biol. Psychiatry 79, 362–371. doi: 10.1016/j.biopsych.2014.07.019 Funding, A., Johansen, C., Kragballe, K., and Iversen, L. (2007). Mitogen- and stress-activated protein kinase 2 and cyclic AMP response element binding protein are activated in lesional psoriatic epidermis. J. Invest. Dermatol. 127, 2012–2019. doi: 10.1038/sj.jid.5700821 Chang, S., Iversen, L., Kragballe, K., Arthur, J. S. C., and Johansen, C. (2011). Mice lacking MSK1 and MSK2 show reduced skin tumor development in a two-stage chemical carcinogenesis model. Cancer Invest. 29, 240–245. doi: 10.3109/07357907.2010.550594 Funding, A. T., Johansen, C., Kragballe, K., Otkjaer, K., Jensen, U. B., Madsen, M. W., et al. (2006). Mitogen- and stress-activated protein kinase 1 is activated in lesional psoriatic epidermis and regulates the expression of pro-inﬂammatory cytokines. J. Invest. Dermatol. 126, 1784–1791. doi: 10.1038/sj.jid.57 00252 Cheng, F., Sawant, T. V., Lan, K., Lu, C., Jung, J. U., and Gao, S.-J. (2015). Screening of human kinome identiﬁes MSK1/2-CREB1 as an essential pathway mediating KSHV lytic replication during primary infection. J. Virol. 89, 9262–9280. doi: 10.1128/JVI.01098-15 Garlanda, C., Dinarello, C., and Mantovani, A. (2013). The interleukin- 1 family: back to the future. Immunity 39, 1003–1018. doi: 10.1016/j.immuni.2013.11.010 Cheng, M., Zhang, Y., Cao, C., Zhang, W., Zhang, Y., and Shen, Y. (2014). Speciﬁc phosphorylation of histone demethylase KDM3A determines target gene expression in response to heat shock. PLoS Biol. 12:e1002026. doi: 10.1371/journal.pbio.1002026 Gesser, B., Johansen, C., Rasmussen, M. K., Funding, A. T., Otkjaer, K., Kjellerup, R. B., et al. (2007). Dimethylfumarate speciﬁcally inhibits the mitogen and stress-activated kinases 1 and 2 (MSK1/2): possible role for its anti- psoriatic eﬀect. J. Invest. Dermatol. 127, 2129–2137. doi: 10.1038/sj.jid.57 00859 Choi, Y.-S., Karelina, K., Alzate-Correa, D., Hoyt, K., Impey, S., Arthur, J., et al. (2012). Proliferation and neuron development in the adult dentate gyrus. J. Neurochem. 123, 676–688. doi: 10.1111/jnc.12035 Chrivia, J. C., Kwok, R. P., Lamb, N., Hagiwara, M., Montminy, M. R., and Goodman, R. H. (1993). Phosphorylated CREB binds speciﬁcally to the nuclear protein CBP. Nature 365, 855–859. doi: 10.1038/365 855a0 Gonzalez, G., and Montminy, M. (1989). REFERENCES Double null cells reveal that CBP and p300 are dispensable for p53 targets p21 and Mdm2 but variably required for target genes of other signaling pathways. Cell Cycle 10, 212–221. doi: 10.4161/cc.10.2.14542 Naqvi, S., Martin, K. J., and Arthur, J. S. C. (2014). CREB phosphorylation at Ser133 regulates transcription via distinct mechanisms downstream of cAMP Naqvi, S., Martin, K. J., and Arthur, J. S. C. (2014). CREB phosphorylation at Ser133 regulates transcription via distinct mechanisms downstream of cAMP and MAPK signalling. Biochem. J. 458, 469–479. doi: 10.1042/BJ20131115 Kew, V. G., Yuan, J., Meier, J., and Reeves, M. B. (2014). Mitogen and stress activated kinases act co-operatively with CREB during the induction of human cytomegalovirus immediate-early gene expression from latency. PLoS Pathog. 10:e1004195. doi: 10.1371/journal.ppat.1004195 New, L., Jiang, Y., Zhao, M., Liu, K., Zhu, W., Flood, L. J., et al. (1998). PRAK, a novel protein kinase regulated by the p38 MAP kinase. EMBO J. 17, 3372–3384. doi: 10.1093/emboj/17.12.3372 Park, J., Al-Ramahi, I., Tan, Q., Mollema, N., Diaz-Garcia, J. R., Gallego- Flores, T., et al. (2013). Genetic screens reveal RAS/MAPK/MSK1 modulate ataxin 1 protein levels and toxicity in SCA1. Nature 498, 325–331. doi: 10.1038/nature12204 Kim, C., Sano, Y., Todorova, K., Carlson, B. A., Arpa, L., Celada, A., et al. (2008). The kinase p38 alpha serves cell type-speciﬁc inﬂammatory functions in skin injury and coordinates pro- and anti-inﬂammatory gene expression. Nat. Immunol. 9, 1019–1027. doi: 10.1038/ni.1640 Peng, H., Guerau-de-Arellano, M., Mehta, V. B., Yang, Y., Huss, D. J., Papenfuss, T. L., et al. (2012). Dimethyl fumarate inhibits dendritic cell maturation via nuclear factor κB (NF-κB) and extracellular signal-regulated kinase 1 and 2 (ERK1/2) and mitogen stress-activated kinase 1 (MSK1) signaling. J. Biol. Chem. 287, 28017–28026. doi: 10.1074/jbc.M112.383380 Korte, S. M. (2001). Corticosteroids in relation to fear, anxiety and psychopathology. Neurosci. Biobehav. Rev. 25, 117–142. doi: 10.1016/S0149-7634(01)00002-1 Kühn, R., Löhler, J., Rennick, D., Rajewsky, K., and Müller, W. (1993). Interleukin-10-deﬁcient mice develop chronic enterocolitis. Cell 75, 263–274. doi: 10.1016/0092-8674(93)80068-P Pérez-Cadahía, B., Drobic, B., Espino, P., He, S., Mandal, S., Healy, S., et al. (2011). Role of MSK1 in the malignant phenotype of Ras-transformed mouse ﬁbroblasts. J. Biol. Chem. 286, 42–49. doi: 10.1074/jbc.M110.156687 Lang, R., Hammer, M., and Mages, J. (2006). DUSP meet immunology: dual speciﬁcity MAPK phosphatases in control of the inﬂammatory response. J. Immunol. 177, 7497–7504. doi: 10.4049/jimmunol.177.11.7497 Pierrat, B., Correia, J. S., Mary, J. L., Tomás-Zuber, M., and Lesslauer, W. (1998). REFERENCES RSK-B, a novel ribosomal S6 kinase family member, is a CREB kinase under dominant control of p38alpha mitogen-activated protein kinase (p38alphaMAPK). J. Biol. Chem. 273, 29661–29671. doi: 10.1074/jbc.273.45.29661 Levenson, J., O’Riordan, K., Brown, K., Trinh, M., Molfese, D., and Sweatt, J. (2004). Regulation of histone acetylation during memory formation in the hippocampus. J. Biol. Chem. 279, 40545– 40559. doi: 10.1074/jbc.M402229200 Liu, H., Yang, J., Liu, Q., Jin, C., Wu, S., Lu, X., et al. (2014). Lanthanum chloride impairs spatial memory through ERK/MSK1 signaling pathway of hippocampus in rats. Neurochem. Res. 39, 2479–2491. doi: 10.1007/s11064-014- 1452-6 Raab-Traub, N. (2002). Cooperation between IL-7 and the pre-B cell receptor: a key to B cell selection. Semin. Cancer Biol. 12, 431–441. doi: 10.1016/S1044579X0200086X Rahman, M. M., Alkhouri, H., Tang, F., Che, W., Ge, Q., and Ammit, A. J. (2014). Sphingosine 1-phosphate induces neutrophil chemoattractant IL-8: repression by steroids. PLoS ONE 9:e92466. doi: 10.1371/journal.pone.0092466 Lonze, B. E., and Ginty, D. D. (2002). Function and regulation of CREB family transcription factors in the nervous system. Neuron 35, 605–623. doi: 10.1016/S0896-6273(02)00828-0 Reeves, M. B., and Compton, T. (2011). Inhibition of inﬂammatory interleukin- 6 activity via extracellular signal-regulated kinase-mitogen-activated protein kinase signaling antagonizes human cytomegalovirus reactivation from dendritic cells. J. Virol. 85, 12750–12758. doi: 10.1128/JVI.05878-11 MacKenzie, K. F., Clark, K., Naqvi, S., McGuire, V. A., Nöehren, G., Kristariyanto, Y., et al. (2013b). PGE(2) induces macrophage IL-10 production and a regulatory-like phenotype via a protein kinase A-SIK-CRTC3 pathway. J. Immunol. 190, 565–577. doi: 10.4049/jimmunol.1202462 Remenyi, J., Hunter, C. J., Cole, C., Ando, H., Impey, S., Monk, C. E., et al. (2010). Regulation of the miR-212/132 locus by MSK1 and CREB in response to neurotrophins. Biochem. J. 428, 281–291. doi: 10.1042/BJ20100024 MacKenzie, K. F., Van Den Bosch, M. W. M., Naqvi, S., Elcombe, S. E., McGuire, V. A., Reith, A. D., et al. (2013a). MSK1 and MSK2 inhibit lipopolysaccharide- induced prostaglandin production via an interleukin-10 feedback loop. Mol. Cell. Biol. 33, 1456–1467. doi: 10.1128/MCB.01690-12 Remenyi, J., van den Bosch, M. W., Palygin, O., Mistry, R. B., McKenzie, C., Macdonald, A., et al. (2013). miR-132/212 knockout mice reveal roles for these miRNAs in regulating cortical synaptic transmission and plasticity. PLoS ONE 8:e62509. doi: 10.1371/journal.pone.0062509 Manto, M. (2005). The wide spectrum of spinocerebellar ataxias (SCAs). Cerebellum 4, 2–6. doi: 10.1080/14734220510007914 Reyes, D., Ballaré, C., Castellano, G., Soronellas, D., Bagó, J. R., Blanco, J., et al. (2014). REFERENCES Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at serine 133. Cell 59, 675– 680. doi: 10.1016/0092-8674(89)90013-5 June 2016 \| Volume 4 \| Article 56 Frontiers in Cell and Developmental Biology \| www.frontiersin.org 6 Physiological Roles of MSK1 and 2 Reyskens and Arthur Hargett, D., and Shenk, T. E. (2010). Experimental human cytomegalovirus latency in CD14+ monocytes. Proc. Natl. Acad. Sci. U.S.A. 107, 20039–20044. doi: 10.1073/pnas.1014509107 through CREB and C/EBP factors in human neutrophils. J. Immunol. 191, 4299–4307. doi: 10.4049/jimmunol.1301117 Mayr, B., Canettieri, G., and Montminy, M. (2001). Distinct eﬀects of cAMP and mitogenic signals on CREB-binding protein recruitment impart speciﬁcity to target gene activation via CREB. Proc. Natl. Acad. Sci.U.S.A. 98, 10936–10941. doi: 10.1073/pnas.191152098 Impey, S., Smith, D., Obrietan, K., Donahue, R., Wade, C., and Storm, D. (1998). Stimulation of cAMP response element (CRE)-mediated transcription during contextual learning. Nat. Neurosci. 1, 595–601. doi: 10.1038/2830 Jin, Y., Wang, Y., Walker, D., Dong, H., Conley, C., Johansen, J., et al. (1999). JIL-1: a novel chromosomal tandem kinase implicated in transcriptional regulation in Drosophila. Mol. Cell 4, 129–135. doi: 10.1016/S1097-2765(00)80195-1 Mayr, B., and Montminy, M. (2001). Transcriptional regulation by the phosphorylation-dependent factor CREB. Nat. Rev. Mol. Cell Biol. 2, 599–609. doi: 10.1038/35085068 Johannessen, M., Delghandi, M., and Moens, U. (2004). What turns CREB on? Cell. Signal. 16, 1211–1227. doi: 10.1016/j.cellsig.2004.05.001 McCoy, C. E., Campbell, D. G., Deak, M., Bloomberg, G. B., and Arthur, J. S. C. (2005). MSK1 activity is controlled by multiple phosphorylation sites. Biochem. J. 387, 507–517. doi: 10.1042/BJ20041501 Kandel, E. R., Dudai, Y., and Mayford, M. R. (2014). The molecular and systems biology of memory. Cell 157, 163–186. doi: 10.1016/j.cell.2014.03.001 McCoy, C. E., Macdonald, A., Morrice, N. A., Campbell, D. G., Deak, M., Toth, R., et al. (2007). Identiﬁcation of novel phosphorylation sites in MSK1 by precursor ion scanning MS. Biochem. J. 402, 491–501. doi: 10.1042/BJ20061183 Karelina, K., Liu, Y., Alzate-Correa, D., Wheaton, K. L., Hoyt, K. R., Arthur, J. S. C., et al. (2015). Mitogen and stress-activated kinases 1/2 regulate ischemia-induced hippocampal progenitor cell proliferation and neurogenesis. Neuroscience 285, 292–302. doi: 10.1016/j.neuroscience.2014.10.053 Mifsud, K. R., Gutierrez-Mecinas, M., Trollope, A. F., Collins, A., Saunderson, E. A., and Reul, J. M. H. M. (2011). Epigenetic mechanisms in stress and adaptation. Brain Behav. Immun. 25, 1305–1315. doi: 10.1016/j.bbi.2011.06.005 Kasper, L., Thomas, M., Zambetti, G., and Brindle, P. (2011). REFERENCES Activation of mitogen- and stress-activated kinase 1 is required for proliferation of breast cancer cells in response to estrogens or progestins. Oncogene 33, 1570–1580. doi: 10.1038/onc.2013.95 Martin, E., Betuing, S., Pagès, C., Cambon, K., Auregan, G., Deglon, N., et al. (2011). Mitogen- and stress-activated protein kinase 1-induced neuroprotection in Huntington’s disease: role on chromatin remodeling at the PGC-1-alpha promoter. Hum. Mol. Genet. 20, 2422–2434. doi: 10.1093/hmg/ddr148 Roizman, B., and Baines, J. (1991). The diversity and unity of Herpesviridae. Comp. Immunol. Microbiol. Infect. Dis. 14, 63–79. doi: 10.1016/0147-9571(91)90122-T Martin, S., Grimwood, P., and Morris, R. (2000). Synaptic plasticity and memory: an evaluation of the hypothesis. Annu. Rev. Neurosci. 23, 649–711. doi: 10.1146/annurev.neuro.23.1.649 Roze, E., Betuing, S., Deyts, C., Marcon, E., Brami-Cherrier, K., Pagès, C., et al. (2008). Mitogen- and stress-activated protein kinase-1 deﬁciency is involved in expanded-huntingtin-induced transcriptional dysregulation and striatal death. FASEB J. 22, 1083–1093. doi: 10.1096/fj.07-9814 Mayer, T., Simard, F., Cloutier, A., Vardhan, H., Dubois, C., and McDonald, P. (2013). The p38-MSK1 signaling cascade inﬂuences cytokine production June 2016 \| Volume 4 \| Article 56 Frontiers in Cell and Developmental Biology \| www.frontiersin.org 7 Physiological Roles of MSK1 and 2 Reyskens and Arthur Rutherford, L., Nelson, S., and Turrigiano, G. (1998). BDNF has opposite eﬀects on the quantal amplitude of pyramidal neuron and interneuron excitatory synapses. Neuron 21, 521–530. doi: 10.1016/S0896-6273(00)80563-2 Vermeulen, L., De Wilde, G., Van Damme, P., Vanden Berghe, W., and Haegeman, G. (2003). Transcriptional activation of the NF-kB p65 subunit by mitogen- and stress-activated protein kinase-1 (MSK1). EMBO J. 22, 1313–1324. doi: 10.1093/emboj/cdg139 Sakamoto, K., Karelina, K., and Obrietan, K. (2011). CREB: a multifaceted regulator of neuronal plasticity and protection. J. Neurochem. 116, 1–9. doi: 10.1111/j.1471-4159.2010.07080.x Walker, F. (2007). Huntington’s disease. Lancet 369, 218–228. doi: 10.1016/S0140- 6736(07)60111-1 Salgado, E., Maneiro, J. R., Carmona, L., and Gomez-Reino, J. J. (2014). Safety proﬁle of protein kinase inhibitors in rheumatoid arthritis: systematic review and meta-analysis. Ann. Rheum. Dis. 73, 871–882. doi: 10.1136/annrheumdis- 2012-203116 Wang, X., and Liu, Y. (2007). Regulation of innate immune response by MAP kinase phosphatase-1. Cell. Signal. 19, 1372–1382. doi: 10.1016/j.cellsig.2007.03.013 Wang, Y., Zhang, W., Jin, Y., Johansen, J., and Johansen, K. (2001). The JIL- 1 tandem kinase mediates histone H3 phosphorylation and is required for maintenance of chromatin structure in Drosophila. Cell 105, 433–443. doi: 10.1016/S0092-8674(01)00325-7 Shah, N., Kammermeier, J., Elawad, M., and Glocker, E. (2012). Interleukin-10 and interleukin-10-receptor defects in inﬂammatory bowel disease. Curr. REFERENCES Allergy Asthma Rep. 12, 373–379. doi: 10.1007/s11882-012-0286-z Wayman, G., Davare, M., Ando, H., Fortin, D., Varlamova, O., Cheng, H., et al. (2008). An activity-regulated microRNA controls dendritic plasticity by down-regulating p250GAP. Proc. Natl. Acad. Sci. U.S.A. 105, 9093–9098. doi: 10.1073/pnas.0803072105 Shaywitz, A. J., and Greenberg, M. E. (1999). CREB : a stimulus-induced transcription factor activated by a diverse array of extracellular signals. Ann. Rheum. Dis. 68, 821–861. doi: 10.1146/annurev.biochem.68.1.821 Sindreu, C., Scheiner, Z., and Storm, D. (2007). Ca+2-stimulated adenylyl cyclases regulate ERK-dependent activation of MSK1 during fear conditioning. Neuron 53, 79–89. doi: 10.1016/j.neuron.2006.11.024 West, A. P. (1990). Neurobehavioral studies of forced swimming: the role of learning and memory in the forced swim test. Prog. Neuropsychopharmacol. Biol. Psychiat. 14, 863–877. doi: 10.1016/0278-5846(90)90 073-P Singh, K., Cassano, M., Planet, E., Sebastia, S., Min Jang, S., Sohi, G., et al. (2015). A KAP1 phosphorylation switch controls MyoD function during skeletal muscle diﬀerentiation. Genes Dev. 29, 513–525. doi: 10.1101/gad.254532.114 Wiggin, G. R., Soloaga, A., Foster, J. M., Murray-Tait, V., Cohen, P., and Arthur, J. S. C. (2002). MSK1 and MSK2 are required for the mitogen- and stress- induced phosphorylation of CREB and ATF1 in ﬁbroblasts. Mol. Cell. Biol. 22, 2871–2881. doi: 10.1128/MCB.22.8.2871-2881.2002 Soloaga, A., Thomson, S., Wiggin, G. R., Rampersaud, N., Dyson, M. H., Hazzalin, C. A., et al. (2003). MSK2 and MSK1 mediate the mitogen- and stress-induced phosphorylation of histone H3 and HMG-14. EMBO J. 22, 2788–2797. doi: 10.1093/emboj/cdg273 Yamamoto, K. K., Gonzalez, G. A., Biggs, W. H., and Montminy, M. R. (1988). Phosphorylation-induced binding and transcriptional eﬃcacy of nuclear factor CREB. Nature 334, 494–498. doi: 10.1038/334494a0 Taubenfeld, S., Wiig, K., Bear, M., and Alberini, C. (1999). A molecular correlate of memory and amnesia in the hippocampus. Nat. Neurosci. 2, 309–310. doi: 10.1038/7217 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Taylor, G. A., Carballo, E., Lee, D. M., Lai, W. S., Thompson, M. J., Patel, D. D., et al. (1996). A pathogenetic role for TNFalpha in the syndrome of cachexia, arthritis, and autoimmunity resulting from tristetraprolin (TTP) deﬁciency. Immunity 4, 445–454. Frontiers in Cell and Developmental Biology \| www.frontiersin.org June 2016 \| Volume 4 \| Article 56 REFERENCES doi: 10.1016/S1074-7613(00)80411-2 Copyright © 2016 Reyskens and Arthur. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Turrigiano, G., Leslie, K., Desai, N., Rutherford, L., and Nelson, S. (1998). Activity- dependent scaling of quantal amplitude in neocortical neurons. Nature 391, 892–896. doi: 10.1038/36103 Venkatesh, B., Lee, A., Ravi, V., Maurya, A., Lian, M., Swann, J., et al. (2014). Elephant shark genome provides unique insights into gnathostome evolution. Nature 505, 174–179. doi: 10.1038/nature12826 June 2016 \| Volume 4 \| Article 56 Frontiers in Cell and Developmental Biology \| www.frontiersin.org 8
https://openalex.org/W4307016228	https://zenodo.org/records/7234706/files/ILM.%E2%80%9CIslimiy%20kompozitsiya%20tuzish%E2%80%9D.pdf	Kirghiz, Kyrgyz	null	"Islimiy kompozitsiya tuzish"	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	703	Annotatsiya Ushbu maqolada islimiy naqsh kompozitsiyasini tuzish va chizish haqida, kompozitsiya qonun-qoidalari haqida malumotlar keltirilgan. Kalit sozlar: Kompozitsiya, Islimiy, naqsh chizish, Qadim. Абстрактный В этой статье представлена информация о композиции и рисовании исламских узоров и правилах композиции. Ключевые слова: Композиция, исламский, узорный рисунок, древний. Abstract This article provides information on the composition and drawing of Islamic patterns and the rules of composition. Key words: Composition, Islamic, pattern drawing, Ancient. KONFERENSIYA \| ILM – FAN TA’LIMDA INNOVATSION YONDASHUVLAR, MUAMMOLAR, TAKLIF VA YECHIMLAR Dunyoning oddiy odamlar ko'ra olmaydigan qirralarini ko'rish faqat rassomlarga xosdir! Ular shunday ko'rsatadilar! ular yaratgan asarlardagi narsalar! Bunday go'zallik Qo'qon naqshiga xos bo'lib, uni yaratish orqali vadunyoni ulug'lash! Gozallik bilan juda gozal kompozitsiya inson qo`li bilan yaraladi. Kompozitsiya – lotincha so’z bo’lib, joylashtirish, qurish, tuzish degan ma’noni anglatadi. Bunda tanlangan naqsh namunasiga ko’ra uning elementlari qog’oz yuzasida bir-biriga bog’liq holda joylashib, yaxlit kompozitsiyani hosil qiladi. Istalgan mavzu mazmuniga mos keladigan naqsh kompozitsiyalarini tuzishda avval ko’plab xomaki ishlar, eskizlar ishlanadi va ulardan eng chiroyli namunalar tanlab olinadi. Naqsh kompozitsiyalarini tuzishda elementlar o’rnini almashtirish, qo’shimcha elementlar kiritish orqali uning mazmunini yanada boyitish kabilar ham muhim axamiyatga egadir. Naqsh mazmunini ifodalashda kompozitsiyaning asosiy qonuniyatlariga rioya qilish lozim. Bular kompozitsiyaning bir butunligi, elementlar, shakllar, ranglarning takrorlanishi va ular o’rtasidagi o’zaro bog’liqlik, kompozitsiyaning tarkibiy qismi bo’lgan rang tuslarining ma’lum bir koloritga bo’ysundirishdan iborat. Kompozitsiyada naqsh yaratishning shunday qoidasini simmetriya, assimetriya, ritm, markazni topish, stilizatsiya, muvozanat, elementlarning o’lchovi, shakli, rangi kabilarga rioya qilishni tashkil etadi va bunga qat’iy amal qilinadi. O‘simliksimon naqshlari esa tabiatdagi barg, band, daraxt, buta, g‘uncha va boshqa narsalarni naqqosh tomonidan stillashtirib olingan shaklini ma’lum qonuniyatlar asosida takrorlanishidan hosil qilingan. 2 KONFERENSIYA \| 2 ILM – FAN TA’LIMDA INNOVATSION YONDASHUVLAR, MUAMMOLAR, TAKLIF VA YECHIMLAR Islimiy — o‘simliksimon naqshi amaliy san'atning deyarli barcha turlarida juda keng tarqalgan. Islimiy klassik naqsh turi bo‘lib, barg, gul, kurtak, tanob va boshqalardan tashkil topgan bir–biriga ulanib takrorlanishdan hosil bo‘ladigan naqsh. Islimiyning turlari juda ko‘p, usta kompozitsiyasi qaysi shakl bilan to‘ldirilgan bo‘lsa, shu shaklning nomini qo‘shib aytiladi. Chunonchi, islimi mehrob, islimi bodom, islimi bulbul va boshqalar. O‘rta Osiyo, Hindiston, Yaqin Sharq, Afg‘oniston, Eron ustalari islimiy uslubda buyumlarni badiiy bezashda ajoyib yutuqlarga erishib kelmoqdalar. Bu uslubda Xiva, Buxoro, Samarqand, Toshkent, Marg‘ilon, Qo‘qon, Andijon shaharlarida ustalar qadimdan hozirgacha ishlab kelyaptilar. Demak, naqsh chizishni o‘rganishdan oldin elementlarni chizishni, qanday nomlanishini va qaerlarda ishlatilishini bilish muhim ahamiyatga ega. Barg, gul, novda, marg‘ula, bofta, shukufta (shkuft), tanob, bog‘lam va sirtmoq kabi naqsh elementlari mavjud bo‘lib, ular tuzilishi, xarakteri va xususiyatlaridan kelib chiqib nomlanadi. Qadimda asosan naqsh bezaklari bilan jamoat inshoatlari, boy-badavlat kishilarning uylari jimjilador qilib bezatilgan. Binolarning tashqi va ichki qismlari: shiftlar, ustunlar, devor yuzalari va uy-ro’zg’or buyumlari mohirona bezatilgan. Milliy bino bezakchiligida asosan mehmon xona va ayvonlar nihoyatda bejirim jilolangan. Ularga islimiy, o’simliksimon, xandasiy, geometrik, murakkab (aralash) ko’rinishigagi naqshlar ko’plab ishlangan. Naqsh namunalarining ko’plab turlari mavjud bo’lib ulardan ayrimlari quyidagicha nomlanadi: xoshiya, ruta mehrob, A \| 3 KONFERENSIYA \| 3 ILM – FAN TA’LIMDA INNOVATSION YONDASHUVLAR, MUAMMOLAR, TAKLIF VA YECHIMLAR madohil, zanjira, bandi rumi, mornech, turunj, davra turunj, pargori, daraxti munabbat, namoyon va boshqalar. madohil, zanjira, bandi rumi, mornech, turunj, davra turunj, pargori, daraxti munabbat, namoyon va boshqalar. Adabiyotlar: Ortiqboy, S. S., & Makhsudali, J. (2022). A Desire Sung in a Pattern. CENTRAL ASIAN JOURNAL OF ARTS AND DESIGN, 3(5), 6-9. Prozorova, M. N., Zharkov, A. D., Zharkova, A. A., Khamdamova, V. A., Sattorova, S. O., & Losev, V. S. (2021). University graduate career potential in digital transformation period: Features and problems of formation. Linguistics and Culture Review, 5(S3), 1574-1585. Sattorova, S. O. (2021). PATTERN SUNG ON THE CARPET. Экономика и социум, (5-1), 445-448. Abdurahimovich, M. A., & Qizi, S. S. O. (2020). Master Student Etiquette. www. conferencepublication. com, 22. Khaydarovich, A. A. (2020). Colors in Descriptive Art. International Journal on Economics, Finance and Sustainable Development, 2(12), 20-22. Mamadjananivich, B. B. IT is Necessary to Feed the World with Beauty!. International Journal on Integrated Education, 3(8), 215-217. Bulatov S.S., Dadashev I. Naqsh alifbosi. – T: Cho’lpon, 2012. S.Bulatov “O`zbek xalq amaliy bezak san’ati” T-2010 Ахмедов М.Б. Ёгоч уймакорлиги. -Т.: Адабиёт учкунлари, 2017 KONFERENSIYA \| 4 KONFERENSIYA \| 4 ILM – FAN TA’LIMDA INNOVATSION YONDASHUVLAR, MUAMMOLAR, TAKLIF VA YECHIMLAR Муратов Х.Х., Жаббаров P.P. Амалий ва бадиий безак санъати. -Т.: Ижод- принт, 2020 5 5 KONFERENSIYA \|
https://openalex.org/W2134986484	https://www.scielo.br/j/csp/a/NZcJth9nLcXPHwkmt7MhRpv/?lang=pt&format=pdf	Portuguese	null	Fórum: legitimidade, expansão e sustentabilidade das Ciências Sociais e Humanas em Saúde Coletiva: Posfácio	Cadernos de Saúde Pública	2,012	cc-by	2,945	Ciências Sociais em Saúde: construindo e consolidando um campo O que não pode ocorrer é a vulgarização e ba- nalização do social. Conservar o rigor teórico e conceitual é fundamental. Isso se aplica a todos os níveis de ensino – da graduação à pós-gradu- ação. Maria Cecília de Souza Minayo 2, de for- ma sintética, mas com precisão, aborda o tema e faz interessantes sugestões. Concordo que não se pode isolar o ensino das ciências sociais dos problemas contemporâneos e acrescento, tanto no plano local como no internacional. Concor- do, também, que apesar dos avanços nos cam- pos teórico e metodológico, o ensino necessita ser revisitado. Ou seja, no passado houve muitas tentativas de inovar, por exemplo, a adaptação da história natural da doença para uma história social, ou de acompanhar a “carreira do doen- te/paciente” a fim de reconstituir a sua história pessoal, familiar e social, mais aderentes a uma “sociologia na medicina”, ou com base em as- pectos macroanalíticos (estrutura social, sistema social), mais próximos de uma “sociologia da me- dicina”. Certamente, estamos num momento em que o ensino precisa rever as suas formas e criar estratégias que possibilitem ao estudante incor- porar em sua formação o básico conceitual da sociologia, da antropologia e da ciência política, na graduação e avançar na pós-graduação o bom uso dos “clássicos” e dos “novos clássicos”. Assim, não se trata apenas de estratégias pedagógicas, mas dos conteúdos a serem transmitidos. Isso nos encaminha para a questão da teoria e da me- todologia nas pesquisas em ciências sociais no Num momento de franco desenvolvimento in- ternacional das ciências sociais em saúde, a ela- boração de um fórum nacional que trate do tema é da mais relevante importância. Minhas preo- cupações atuais, pesquisando a construção da identidade desse campo entre nós 1, encontram nos textos propostos ao debate a comprovação de que os caminhos foram abertos, mas neces- sitamos de uma vigilância permanente no ensi- no, na produção científica, na divulgação do co- nhecimento e na relação com as práticas sociais. Estar presente nesse debate é muito honroso, e poder dialogar com diferentes gerações de pes- quisadores traz a sensação da vitalidade de um campo cuja institucionalização está completan- do cinco décadas. 2393 FÓRUM FORUM 2393 FÓRUM FORUM 2393 FÓRUM FORUM Correspondência E. D. Nunes Departamento de Saúde Coletiva, Faculdade de Ciências Médicas, Universidade Estadual de Campinas. Cidade Universitária, Campinas, SP 13081-970, Brasil. evernunes@uol.com.br Fórum: legitimidade, expansão e sustentabilidade das Ciências Sociais e Humanas em Saúde Coletiva. Posfácio Forum: legitimacy, expansion, and sustainability of Social and Human Sciences in Public Health. Postscript Forum: legitimacy, expansion, and sustainability of Social and Human Sciences in Public Health. Postscript Everardo Duarte Nunes 1 Everardo Duarte Nunes 1 Ciências Sociais em Saúde: construindo e consolidando um campo 1 Faculdade de Ciências Médicas, Universidade Estadual de Campinas, Campinas, Brasil Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012 Ciências Sociais em Saúde: construindo e consolidando um campo Tendo começado os comentários por um tra- balho no campo do ensino que se situa como parte da construção da “identidade social” das ciências sócias em saúde, voltamos neste momento a dois outros trabalhos que também concorrem para esta construção. Estamos entendendo que essa identidade “é forjada pelo processo de institucio- nalização através do qual a disciplina procura se estabilizar do ponto de vista de sua organização” 8 (p. 38) e que, em nosso campo, isto tem sido rea- lizado pelas diversas modalidades de reprodução do saber, pelo ensino, publicação e divulgação e exposição oral e visual nos congressos. Sem dúvida, como exposto por Maria Lúcia Magalhães Bosi 9, a questão da avaliação acadê- mica da produção científica em ciências sociais na saúde coletiva é desafiante, especialmente quando referida às pesquisas qualitativas. Ana- lisa que no padrão vigente predomina o que de- nomina “produtivismo”, ou seja, a quantidade de artigos requeridos aos pesquisadores dentro de modelos de avaliação dessa produção científica que marginaliza formas de pesquisa (pesquisas com metodologias quantitativas), estilos de nar- rativas (as discursivas) e formas de divulgação (li- vros). Lembramos que essas análises não se apli- cam somente à produção em ciências sócias em saúde, mas vem sendo objeto de estudos sobre o próprio campo das ciências sociais. O último relatório da Organização das Nações Unidas para a Educação, a Ciência e a Cultura (UNESCO) 10 destaca que num processo de internacionaliza- ção, em um primeiro momento, esse processo “reduz as divisões do conhecimento das ciências sociais entre diferentes regiões do mundo sem des- truir a diversidade” 10 (p. 143), mas quando visto pelos critérios de avaliação da produção (coauto- ria entre pesquisadores de diferentes países) co- mo também onde são produzidos os periódicos e os artigos, a internacionalização tem sido ques- tionada, pela “estrutura altamente assimétrica de troca”, domínio da América do Norte e Europa Ocidental nas duas últimas décadas, quando analisado por meio do número de publicações conjuntas, o predomínio de citações nos artigos produzidos, sendo que estas regiões foram gran- demente favorecidas pela globalização e interna- cionalização da pesquisa. Outro aspecto refere- se ao risco da exclusão de “pesquisadores menos favorecidos” dos países em desenvolvimento, em nome de um discurso que aborde “temas de re- levância global”, assim como a preferência pelo uso do inglês. Ciências Sociais em Saúde: construindo e consolidando um campo Dos desafios dos anos 60 aos dias atuais, as experiências com o ensino de graduação, primei- ramente na medicina, e depois na formação de outros profissionais de saúde, na medida em que trabalhar em equipe tornou-se prática corrente, foram acumuladas e evidenciaram que a biome- dicina não pode sozinha dar conta da complexi- dade dos fenômenos saúde, doença e cuidado. Nesse sentido, o instrumental sociológico tem a possibilidade de ser usado tanto para o ensino na clínica como na comunidade, tanto no hospital como nos serviços públicos de saúde, tanto para a epidemiologia como para a gestão de serviços. Nunes ED 2394 Nunes ED 2394 campo da saúde, objeto do terceiro artigo deste fórum, elaborado por Suely Ferreira Deslandes & Jorge Alberto Bernstein Iriart 3. muitos dos trabalhos elaborados na emergência das ciências sociais em saúde, década de 70, tra- balharam densamente a teoria sociológica, como pode ser visto em Pereira 7, utilizando as formu- lações de Fernandes antes citadas. Essa questão é tema recorrente para os estu- diosos desse campo, sem dúvida, fundamental para a construção da sua “identidade cognitiva”. Como apontado por muitos estudiosos, a relação entre as ciências sociais e saúde e a teoria socioló- gica é crucial para a subdisciplina, inclusive por- que é ela que, usualmente, distingue a pesquisa social da saúde dos “estudos socialmente orien- tados” em campos conexos como saúde pública, saúde coletiva, serviço social, planejamento em saúde. O fato de muitos trabalhos terem cará- ter aplicado levou, no passado, a se considerar o campo das ciências sociais e saúde como ateóri- co, e isto é surpreendente, considerando-se que a sociologia médica emerge nos anos 50, com o trabalho de Parsons, essencialmente teórico 4. Assim, considero que o artigo sobre usos teóri- cos e metodológicos aqui apresentado colabora, de um lado, para o melhor entendimento do su- posto caráter ateórico do campo – dos 266 artigos analisados, 124 (46,6%) definem “claramente sua filiação a um referencial teórico-metodológico”, embora não se saiba de forma mais detalhada como o teórico é trabalhado –, de outro lado, ressaltando as fragilidades e insuficiências en- contradas na utilização do aparato metodológico (categorias metodológicas, técnicas de pesquisa e de análise) das ciências sociais. Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012 Ciências Sociais em Saúde: construindo e consolidando um campo As análises feitas pela autora são pertinentes e concordo que ho- je essa produção, em um movimento pendular, passe novamente a dar atenção à macroanálise e não apenas às microanálises como foi detectado em 68% dos trabalhos empíricos. Certamente, lições antigas como as de Charles Wright Mills 13 ou de Norbert Elias14 , ou mais re- centes como as de Philippe Corcuff 15 não podem ser esquecidas, tanto para pensar o conhecimen- to como para organizá-lo em próximos eventos. O encontro dos clássicos e dos contemporâneos, como lembrado por Minayo em seu texto, assim como o de aprofundar as relações não somente com a biologia, mas com outras áreas do conhe- cimento continuam a ser a pauta dos cientistas sociais em saúde. Deixamos para o final o artigo de Leny Alves Bonfim Trad 12, que também incluo quando se tenta estabelecer a “identidade social” do cam- po e que se refere aos congressos. Sem dúvida, essa forma amplia a participação e visibilidade de qualquer campo de conhecimento nacional e internacionalmente e pode possibilitar uma (re)definição de temáticas e redimensionamen- to do campo. A análise se prende ao V Congresso de Ciências Sociais e Humanas em Saúde, São Paulo, 2011, cujo tema central foi O Lugar das Ciências Sociais na Saúde Coletiva, em especial, sobre a produção científica, veiculada pelos Grupos de Trabalho (GTs). Não se prendendo a todos os aspectos do evento, a autora destaca o que se revela em termos de produção científica e suas tendências quando são analisados 1.482 trabalhos, de 21 GTs. Com exceção de três GTs que optaram por um corte disciplinar (histórico, sociológico) e um sobre aspectos metodológi- cos, os demais são sobre os mais diversos temas, sendo que os cinco primeiros versam sobre: mo- vimentos sociais e educação popular em saú- de, integralidade, mediações sociais e justiça na demanda por cuidado em saúde, humanidades e humanização em saúde, análise de narrativa e experiência de saúde, adoecimento e cuida- do e sociologia do campo da análise de políticas públicas e de saúde no Brasil e racionalidades médicas e práticas em saúde. Embora tenha sido apontada a grande diversidade temática, alguns temas não estiveram presentes, como a saúde da população negra, saúde mental, determinantes sociais em saúde e a pequena presença do tema da violência. Ciências Sociais em Saúde: construindo e consolidando um campo Portanto, são extremamente opor- Em 1990, no seminário que a Universidade do Estado do Rio de Janeiro organizou sobre As Ciências Sociais no Brasil, dentre os vários temas abordados, Castro-Santos 5 apontava que o pa- drão de trabalho dos estudantes de graduação e pós-graduação apresentava uma dupla face: uma que se caracterizava por uma “estreita seletivi- dade dos temas”, outra, em que se encontravam as inúmeras formas de recortá-los, o que levava a uma pulverização e não aprofundamento dos mesmos. No mesmo trabalho, Castro-Santos 5 (p. 260-1), ao “evocar a antiga preocupação de Flores- tan Fernandes 6 sobre os fundamentos empíricos da explicação sociológica”, alertava para as duas situações em oposição – trabalhos de pesquisa com forte conotação empírica, mas “sem que se note neles um balizamento conceitual que oriente estrategicamente as próprias tarefas no ‘campo’ e o próprio esforço de análise, e que permita explicar sociologicamente os fenômenos” e, aqueles traba- lhos que “descrevem uma espécie de loop teórico sem jamais se aproximarem do ‘solo’ empírico”. Essas advertências parecem-nos procedentes e precisam ser levadas em consideração quando enfrentamos um campo específico de pesqui- sas – a produção e construção social da saúde, da doença e do cuidado. Somente lembraria que Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012 FÓRUM: CIÊNCIAS SOCIAIS E HUMANAS EM SAÚDE COLETIVA 2395 tunas as análises feitas por Bosi, quando trata especificamente do campo da saúde, fato que corrobora estudos realizados que salientam, den- tre outros aspectos, a “impertinência da avalia- ção da produção intelectual das Ciências Sociais e Humanas na Saúde Coletiva (CSHSC) segundo o parâmetro do prestígio de periódicos do campo biomédico” 11 (p. 683) e da própria avaliação dos cursos de pós-graduação que com “a hegemonia ou a predominância de critérios, culturas e proce- dimentos das ciências exatas e naturais, as quais migraram para outras áreas e funcionaram como uma camisa de força” 11 (p. 696), fato reconheci- do pela própria Coordenação de Aperfeiçoamen- to de Pessoal de Nível Superior (Capes) em seu plano decenal – 2011-2020 11. lado, permite verificar, comparativamente, com outros congressos a produção científica, não apenas consolidada em publicações, como vi- mos no artigo de Deslandes & Iriart, mas em seu movimento – exposição, análise e interação com outros participantes do grupo. Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012 Referências 1. Nunes ED. A sociologia da saúde no Brasil: a cons- trução de uma identidade. II Seminário Interna- cional de Sociologia da Saúde. Campinas: Faculda- de de Ciências Médicas, Universidade Estadual de Campinas; 2012. 9. Bosi MLM. Produtivismo e avaliação acadêmica na Saúde Coletiva brasileira: desafios para a pes- quisa em Ciências Humanas e Sociais. Cad Saúde Pública 2012; 28:2387-92. 10. United Nations Educational, Scientific and Cul- tural Organization. World social science report: knowledge divides. Paris: United Nations Educa- tional, Scientific and Cultural Organization; 2010. 2. Minayo MCS. Herança e promessas do ensino das Ciências Sociais na área da Saúde. Cad Saúde Pú- blica 2012; 28:2367-72. 3. Deslandes SF, Iriart JAB. Usos teórico-metodoló- gicos das pesquisas na área de Ciências Sociais e Humanas em Saúde. Cad Saúde Pública 2012; 28:2380-6. 11. Costa NR. A avaliação da produção intelectual e o declínio da interdisciplinaridade na Saúde Coleti- va. Physis (Rio J.) 2012; 22:681-9. 12. Trad LAB. Temas e enfoques contemporâneos nas Ciências Sociais e Humanas no Brasil: expressões e tendências refletidas no V congresso da área. Cad Saúde Pública 2012; 28:2373-9. 4. Cockerham WC. Medical sociology and sociologi- cal theory. In: Cockerham WC, editor. The Black- well Companion to Medical Sociology. Oxford: Blackwell Publishers Ltd.; 2001, p. 3-22. 13. Mills CW. A imaginação sociológica. Rio de Janei- ro: Editora Zahar; 1959. 5. Castro-Santos LA. Alguns dilemas da universida- de brasileira e o ensino da sociologia. In: Bomeny H, Birman P, organizadores. As assim chamadas ciências sociais. Rio de Janeiro: Editora Relume- Dumará; 1991. p. 251-64. 14. Elias N. A sociedade dos indivíduos. Rio de Janei- ro: Jorge Zahar Editor; 1994. 15. Corcuff P. As novas sociologias: construções da re- alidade social. São Paulo: Edusc; 2001. 15. Corcuff P. As novas sociologias: construções da re- alidade social. São Paulo: Edusc; 2001. 6. Fernandes F. Fundamentos empíricos da explica- ção sociológica. São Paulo: Nacional; 1967. 6. Fernandes F. Fundamentos empíricos da explica- ção sociológica. São Paulo: Nacional; 1967. 16. Cândido A. formação da literatura brasileira. São Paulo: Itatiaia; 1975. 16. Cândido A. formação da literatura brasileira. São Paulo: Itatiaia; 1975. 7. Pereira JCM. A explicação sociológica na medicina social. São Paulo: Editora Unesp; 2006. 7. Pereira JCM. A explicação sociológica na medicina social. São Paulo: Editora Unesp; 2006. Recebido em 20/Ago/2012 Aprovado em 30/Ago/2012 8. Villas Bôas GK. A tradição renovada. In: Bomeny H, Birman P, organizadores. As assim chamadas ciências sociais. Ciências Sociais em Saúde: construindo e consolidando um campo O texto tem muitos e importantes dados, e o destaque feito é apenas para carac- terizar que essa “cartografia” do campo, não somente informa, mas permite estabelecer, na minha opinião, “trajetos do conhecimento” pro- duzidos e vivenciados pelos autores na medida em que puderam apresentar e discutir os acha- dos de pesquisas empíricas e teóricas. De outro Os quatro trabalhos apresentados mostram a contemporaneidade do campo das ciências so- ciais em saúde e se há entraves, há realizações, se há desafios, há perspectivas. Penso que a sua identidade como “campo disciplinar” vem sendo construída. Particularizo a sociologia da saúde, mas estendo a ideia às ciências sociais em geral. Nessa construção à sua “identidade histórica” que pode ser buscada desde as suas “manifes- tações”, para usar a linguagem de Antônio Cân- dido 16 nos viajantes, naturalistas, folcloristas, médicos-cientistas sociais (Nina Rodrigues, Ar- thur Ramos, Afrânio Peixoto), cientistas sociais (Donald Pierson, Kalervo Oberg, Roger Bastide, Charles Wagley, Oracy Nogueira, Maynard de Araujo, Armando Piovesan, Luis Fontenelle, Ar- thur Rios e outros) que associaram seus nomes aos primeiros estudos sobre comunidades, edu- cação sanitária, ciências sociais e saúde pública, pacientes internados, doença mental, incorpo- ra-se a sua “identidade social”. Inicialmente, no ensino, depois na participação nos movimentos das reformas sanitárias, nos grupos de pesquisa, nos eventos (encontros, seminários, congressos) as ciências sociais foram forjando formas de atu- ação nos domínios acadêmico, político-social e nas instituições de saúde (Serviço Único de Saú- de e centros de saúde). Sem dúvida, a sua visi- bilidade e relação com a comunidade científica se constroem no espaço da produção científica – “identidade cognitiva”. Essa se encontra em franco crescimento, tanto na literatura publicada como nas teses e dissertações. Retomo, para encerrar o que apresentei re- centemente, quando disse: “Certamente, como já apontado, a identidade da Sociologia da Saúde Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012 Nunes ED 2396 Nunes ED 2396 está sendo construída e, passada a ‘fase heróica’, enfrentada pelos primeiros cientistas sociais que se aventuraram nesse caminho na fase inicial de institucionalização desse campo, há a necessidade de que se formalizem os canais para a formação das novas gerações de cientistas sociais em saúde, se desenvolvam meios de divulgação da produção científica e se institucionalizem as relações, de um lado, com as matrizes sociológicas e de outro com o campo da saúde” 1 (p. 21). Referências das novas gerações de cientistas sociais em saúde, se desenvolvam meios de divulgação da produção científica e se institucionalizem as relações, de um lado, com as matrizes sociológicas e de outro com o campo da saúde” 1 (p. 21). Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012 das novas gerações de cientistas sociais em saúde, se desenvolvam meios de divulgação da produção científica e se institucionalizem as relações, de um lado, com as matrizes sociológicas e de outro com o campo da saúde” 1 (p. 21). Recebido em 20/Ago/2012 Aprovado em 30/Ago/2012 Referências Rio de Janeiro: Editora Relume- Dumará; 1991. p. 21-41. Cad. Saúde Pública, Rio de Janeiro, 28(12):2393-2396, dez, 2012
https://openalex.org/W2966317242	https://www.nature.com/articles/s41598-019-47864-4.pdf	English	null	Distribution and ecological risk assessment of PEDCs in the water, sediment and Carex cinerascens of Poyang Lake wetland, China	Scientific reports	2,019	cc-by	7,353	OPEN Received: 12 April 2019 Accepted: 25 July 2019 Published: xx xx xxxx Received: 12 April 2019 Accepted: 25 July 2019 Published: xx xx xxxx Pinghua Yang1,2 Phenolic endocrine disrupting chemicals (PEDCs), such as 4-nonylphenol (NP), 4-t-octylphenol (OP), bisphenol A (BPA), and nonylphenol-di-ethoxylate (NP2EO), can cause feminization and carcinogenesis. This study assessed the distributions of NP, OP, BPA, and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake wetland. The four PEDCs were ubiquitous. The concentrations of NP and OP in the water and sediment of the wetland were significantly lower than those in other regions of China. Average BPA concentrations in the water, sediment, and Carex cinerascens samples were 40.49 ± 18.42 ng/L, 9.840 ± 3.149 ng/g, and 3.25 ± 1.40 ng/g, respectively; the BPA concentration in the water was similar to that of other rivers in China. Average NP2EO concentrations in the wetland were 3125.9 ± 478.1 ng/L, 650.0 ± 209.9 ng/g, and 275.8 ± 59.0 ng/g in the water, sediment, and Carex cinerascens samples, respectively. The predicted no-effect concentrations in sediment for NP, OP, BPA, and NP2EO were estimated to be 75.41, 45.25, 8.22, and 237.5 ng/g, respectively. The risk quotient (RQ) method was used to characterise the ecological risk from these PEDCs. A high ecological risk (RQ ≥ 1) from BPA was observed for 0%, 57.69%, and 5.00% of water, sediment, and C. cinerascens samples, respectively, while a high risk from NP2EO was observed for 71.43%, 96.15%, and 55.00% of samples. Ecological risk varied spatially. The high ecological risk from NP2EO in Poyang Lake wetland may be a result of non-point pollution from rural areas and sewage from Poyang Lake basin. Phenolic endocrine disrupting chemicals (PEDCs) are phenolic substances derived mainly from non-ionic sur- factants that are widely used in industrial and household products such as detergents, emulsifiers, solubilisers, and dispersing agents1–4. PEDCs include 4-nonylphenol (NP), 4-t-octylphenol (OP), bisphenol A (BPA), and nonylphenol-di-ethoxylate (NP2EO). PEDCs interfere with the secretion, synthesis, metabolism, transportation, reactions, combination, and elimination of natural hormones in organisms. Due to their persistence, lipophilicity, toxicity, and endocrine disruption, PEDCs have harmful effects on various organisms, including feminization and carcinogenesis5–10. The wide distribution of PEDCs in the natural environment, and the associated adverse effects on wildlife, necessitate assessment of their distribution in ecosystems.h y In China, there are many ecological and human health risk assessments for PEDCs. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports OPEN These studies include assessments of estrogens in the water, sediment, and biota of northern Lake Taihu11; NP and BPA in the Cape D’Aguilar Marine Reserve, Hong Kong12; APs, BPA, and TBBPA in Lake Taihu13; bisphenol analogues in the water and sediment of Liaohe River Basin and Lake Taihu14; EDCs in the Pearl River15; NP in Jiaozhou Bay in Qingdao16; estrogens and BPA in arid and semiarid areas of northwest China17; bisphenol analogues in the surface water and sediment of shallow freshwater lakes18; NP and OP in the riverine waters and surface sediments of the Pearl River Estuary19; 50 phenolic compounds in the surface water, sediment, and suspended particulate matter of three important rivers in Tianjin20; and eight EDCs of Huai River21. To date, however, there has been no study of the ecological risk from PEDCs in Poyang Lake—the largest freshwater lake in China. g y g g Poyang Lake (115°49′–116°46′E and 28°24′–29°46′N; Fig. 1) is extremely important to migratory birds and regional ecological security. Poyang Lake is a typical shallow lake of eastern China, developed on alluvial plains with high-nutrient sediments. The area of Poyang Lake basin is 162 200 km2, which is 97% of the area of Jiangxi 1College of Chemistry and Environment Engineering, Jiujiang University, Jiujiang, 332005, China. 2Jiangxi Province Engineering Research Center of Ecological Chemical Industry, Jiujiang, 332005, China. Correspondence and requests for materials should be addressed to P.Y. (email: yangph1999@126.com) Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ Figure 1. Sampling sites in Poyang Lake wetland, China. Figure 1. Sampling sites in Poyang Lake wetland, China. Province and its water level fluctuates considerably between the dry (from October to March) and wet seasons (from April to September). Poyang Lake is the only member of the International Living Lakes Network in China and is one of the most important ecological zones in the world (http://www.jxpoyanglake.gov.cn). The wetland is the largest winter migratory bird habitat in East Asia, hosting more than 500,000 migratory birds belonging to 105 species in 2016 and more than 95% and 80% of the global Grus leucogeranus and Ciconia boyciana popula- tions, respectively. Carex cinerascens is the dominant species in the Poyang Lake wetland, with coverage between 63.27–100%. This species provides good wild pasture for cattle, sheep, and geese and its roots provide habitat for vivipara, mussels, and other food favorites of migratory birds. Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 OPEN The lake has unique hydrological characteristics and colloid behavior; most colloids overflow into the Yangtze River, with short hydraulic residence times22,23. There are 18 protected areas in Poyang Lake, with a total area of 2006.18 km2 designated for migratory birds and wetland ecosystem protection. The lake is an important reservoir in the main stream of the Yangtze River, and plays an important role in regulating floods and protecting biodiversity in the Yangtze River Basin.h l The generic framework and guidelines of the US Environmental Protection Agency (EPA) propose the fol- lowing process for ecological risk assessments: (1) derive the predicted no-effect concentration (PNEC), and (2) calculate the risk quotient (RQ), which is the ratio of the predicted environmental concentration (PEC) to the PNEC24. In the technical guidelines for risk assessments formulated by the European Commission, the allocation balance formula can be used to calculate the predicted no-effect concentration in sediment (PNECsediment)25. There are several reports of the PNEC or water quality criteria for PEDCs. Four species sensitivity distribution (SSD) models were used to calculate the PNEC for BPA26. The PNEC values for nonylphenol were also derived using SSD models and these were used to assess its ecological risk in the coastal waters of China27. The PNEC for BPA was derived from both acute and chronic toxicity endpoints using the SSD model and was then used to assess the ecological risk from BPA in surface waters of China28. Both maximum and continuous concentration criteria were derived for nonylphenol and used to assess its ecological risk to aquatic life in Chinese surface freshwater29. This study aimed to investigate the distributions of NP, OP, BPA, and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake and use these to assess the ecological risk. The SSD model was used to estimate This study aimed to investigate the distributions of NP, OP, BPA, and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake and use these to assess the ecological risk. The SSD model was used to estimate the HC5 and PNECwater values for NP, OP, BPA, and NP2EO. OPEN Concentrations of 4-nonylphenol (NP), 4-t-octylphenol (OP), bisphenol A (BPA), and nonylphenol- di-ethoxylate (NP2EO) in surface water, sediment, and Carex cinerascens samples from the Poyang Lake wetland in the dry season (October to February). Line segment (—) indicates that no samples were collected at this site. dw = dry weight. (from the technical guidance document of the EU), the PNECsediment values of NP, OP, BPA, and NP2EO w estimated and used to assess the ecological risk from these four PEDCs. (from the technical guidance document of the EU), the PNECsediment values of NP, OP, BPA, and NP2EO were estimated and used to assess the ecological risk from these four PEDCs. OPEN According to the distribution equilibrium method Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ Sites Water (ng/L) Sediment (ng/g dw) Carex cinerascens (ng/g dw) NP OP BPA NP2EO NP OP BPA NP2EO NP OP BPA NP2EO Y1 1.359 1.819 21.58 3028.2 2.03 0.665 9.058 483.5 0.544 0.036 2.42 233.6 Y2 2.585 2.237 22.21 3818.7 1.534 0.774 6.069 321.3 0.169 0.062 0.4 121.1 Y3 1.327 1.290 23.07 3651.0 2.683 3.578 14.005 422.8 0.607 0.134 3.39 188.6 Y4 7.387 21.241 25.81 3410.2 2.278 1.028 11.101 410 2.19 0.16 3.3 104.4 Y5 0.362 0.480 25.6 2333.4 1.922 4.237 9.288 369.3 0.446 0.072 4.54 241.3 Y6 2.223 1.831 37.03 4411.6 2.325 1.43 5.712 1063.9 0.516 0.066 2.54 206.4 Y7 1.207 0.887 24.97 2653.0 3.535 5.327 2.249 3107.7 0.809 0.115 4.22 273.1 Y8 2.281 1.160 25.74 3434.3 3.579 0.507 4.378 1173.5 0.541 0.085 3.01 244.6 Y9 1.751 1.929 22.98 3617.0 1.551 0.193 2.615 598.5 0.366 1.22 2.09 801.9 Y10 1.692 2.140 21.58 4242.4 0.126 0.504 15.779 127.2 0.229 0.313 4.14 311.5 Y11 1.183 1.587 17.21 3617.0 2.422 0.745 13.75 576.9 0.257 0.106 4.39 332.4 Y12 3.782 5.075 28.98 162.1 2.759 1.631 9.637 563.2 0.22 0.159 4.3 406.4 Y13 1.852 0.875 18.51 2750.2 — — — — — — — — Y14 1.297 1.450 19.97 3048.2 2.081 1.142 21.415 530.9 0.328 0.082 4.09 340.2 Y15 0.672 1.717 34.55 3023.4 — — — — — — — — Y16 1.038 1.618 48.59 1532.5 0.667 1.148 9.5 403.7 — — — — Y17 0.337 1.448 27.85 2374.2 0.717 1.367 3.025 503 0.308 0.304 17.37 353.8 Y18 0.746 4.548 66.86 4473.0 0.216 1.749 1.328 545.6 1.1 0.147 1.02 131.5 Y19 2.002 3.832 45.41 4182.0 0.548 1.051 1.485 306.9 0.424 0.136 0.55 121.2 Y20 0.059 1.093 177.6 850.2 0.724 1.822 1.688 600.7 — — — — Y21 0.442 2.526 21.69 1698.3 0.158 1.106 38.38 436.3 0.421 0.188 0.45 206.1 Y22 0.903 3.534 32.27 3909.0 1.159 1.232 18.69 495 — — — — Y23 0.037 0.143 3.84 421.8 1.115 1.523 11.07 600.1 0.048 0.147 0.38 210.9 Y24 0.944 3.206 242.2 4485.0 0.818 1.326 4.068 574.5 0.323 0.232 1.15 394.9 Y25 0.744 2.644 17.66 2878.2 0.235 1.502 12.43 626.8 — — — — Y26 0.480 5.540 45.87 5574.0 1.707 1.508 9.352 814.4 — — — — Y27 0.534 4.526 9.60 4770.0 3.98 1.226 17.39 515.5 0.234 0.258 1.15 292.9 Y28 1.883 2.706 24.57 3177.0 17.56 3.672 2.385 728.7 — — — — mean± SD 1.468 ± 0.531 2.967 ± 1.409 40.49 ± 18.42 3125.9 ± 478.1 2.247 ± 1.273 1.615 ± 0.471 9.840 ± 3.149 650.0 ± 209.9 0.504 ± 0.178 0.201 ± 0.097 3.25 ± 1.40 275.8 ± 59.0 Table 1. from the technical guidance document of the EU), the PNECsediment values of NP, OP, BPA, and NP2EO were stimated and used to assess the ecological risk from these four PEDCs. Results and Discussion Ecological risk assessment of NP, OP, BPA, and NP2EO in Poyang Lake wetland. The RQ values of NP, OP, BPA, and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake wetland are listed in Supplementary Information Table S11. The RQ values of NP and OP for all sampling sites were significantly lower than 1.0 and lower than those of other regions in China, indicating a low ecological risk from NP and OP13,19,27.h The ecological risk from BPA and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake wet- land is shown in Fig. 3; the horizontal line at RQ = 1.0 indicates high ecological risk. The RQ values of BPA in the water ranged from 0.0076 to 0.4751, with an average of 0.0794, and 10.71% of sites showed medium ecological risk. The RQ values of NP2EO in the water ranged from 0.059 to 2.045, with an average of 1.147. The ecological risk from NP2EO was medium and high at 25.00% and 71.43% of sites, respectively. The RQ values of BPA in the sediment ranged from 0.161 to 4.671, with an average of 1.198, and 42.31% and 57.69% of sites showed medium and high ecological risk, respectively. The RQ values of NP2EO in the sediment showed that 96.15% of sites had high ecological risk; values ranged from 0.536 to 13.09, with an average of 2.737. The RQ values of BPA and NP2EO in Carex cinerascens, showed that 75.00% and 5.00%, and 45.00% and 55.00% of sites showed medium and high ecological risk, respectively. In the water, sediment, and Carex cinerascens of Poyang Lake, there is potential ecological risk from BPA and NP2EO. The ecological risk from BPA in the water and sediment was similar to that in Lake Taihu11,13,28. Spatial ecological risk of Poyang Lake wetland. The spatial RQ values for BPA and NP2EO in the water and sediment of Poyang Lake wetland are shown in Fig. 4. The ecological risk from BPA in Gan River (Y9), Xiu River (Y10), Fu River (Y16), Xin River (Y19), and Rao River (Y24)—the five main tributaries of Poyang Lake—was low in water and high in sediment; the ecological risk of NP2EO was high in both water (except for Fu River) and sediment (except Xiu River). Results and Discussion Results and Discussion Distributions of NP, OP, BPA, and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake wetland. The concentrations of NP, OP, BPA, and NP2EO in water, sediment and Carex cin- erascens samples from twenty-eight sampling sites of the Poyang Lake wetland are shown in Table 1. For com- parison, the range (mean) concentrations of NP, OP, and BPA in the surface waters and sediment of lakes and rivers of China are shown in Supplementary Information Table S9. All four PEDCs were ubiquitous. Average NP concentrations in the water, sediment, and Carex cinerascens samples were 1.468 ± 0.531 ng/L, 2.247 ± 1.273 ng/g, and 0.504 ± 0.178 ng/g, respectively. Average OP concentrations were 2.967 ± 1.409 ng/L, 2.247 ± 1.273 ng/g, and 0.201 ± 0.097 ng/g, respectively and average BPA concentrations were 40.49 ± 18.42 ng/L, 9.840 ± 3.149 ng/g, and 3.25 ± 1.40 ng/g, respectively. The concentrations of NP and OP in the water of the Poyang Lake wetland were significantly lower than those in Lake Taihu, Dianchi Lake, and other rivers in China13,17,30–33. This may be because some pollutants are removed by exchanged water and the water exchange speed is higher in Poyang Lake than in Lake Taihu and Dianchi Lake22,34. The BPA concentration in the water of the Poyang Lake wetland was similar to that of Lake Taihu, Dianchi Lake, and other rivers in China. Notably, the concentrations of NP, OP, and BPA in the waters of the Pearl River Delta region were found to be significantly higher than those in other regions of China, which may be attributable to the Pearl River coast being the most economically developed area in China, with the largest population density and high consumption of detergents15,19,30. The concentrations of NP, OP, and BPA in the sediment of Poyang Lake wetland were significantly lower than those in other regions of China. Average NP2EO concentrations in the wetland were 3125.9 ± 478.1 ng/L, 650.0 ± 209.9 ng/g, and 275.8 ± 59.0 ng/g in the water, sediment, and Carex cinerascens samples, respectively. Due to the high water-exchange speed, the Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ Figure 2. The species sensitivity distribution (SSD) curves for exposure concentrations of 4-nonylphenol (NP), 4-t-octylphenol (OP), bisphenol A (BPA), and nonylphenol-di-ethoxylate (NP2EO) in freshwater. Figure 2. Results and Discussion The species sensitivity distribution (SSD) curves for exposure concentrations of 4-nonylphenol (NP), 4-t-octylphenol (OP), bisphenol A (BPA), and nonylphenol-di-ethoxylate (NP2EO) in freshwater. concentrations of NP2EO, the primary degradation product of detergent, were significantly higher than those of NP, the advanced degradation product. concentrations of NP2EO, the primary degradation product of detergent, were significantly higher than those of NP, the advanced degradation product. SSD curves and PNEC values. Using the chronic toxicity data for NP, OP, BPA, and NP2EO (S1 to S4), SSD models were constructed with a sigmoid distribution and the simulated curves for NP, OP, BPA, and NP2EO for freshwater ecosystems are shown in Fig. 2. The HC5 values of NP, OP, BPA, and NP2EO for freshwater ecosystems were estimated according to the constructed SSD curves (shown in Supplementary Information Table S10). The PNECwater values of NP, OP, BPA, and NP2EO were calculated using the HC5 values, and these are shown in Fig. 2. In this study, the HC5 value of NP in water (1.524 μg/L) was similar to that in the literature (1.43 μg/L)27. This may be due to the adoption of similar SSD and data fitting methods. The PNECwater of NP (0.508 μg/L) was slightly higher than that in the literature (0.33 μg/L)35 and the PNECwater of BPA (0.510 μg/L) was lower than that in the literature (0.86 μg/L)28. These results may be due to differences in the chronic toxicological data used; thirty-five samples were used in this study, compared with ten in previous studies. The PNECwater of NP2EO (2.725 μg/L) was 25 times higher than the 0.11 μg/L estimated from LC50 divided by the extrapolation factor36. The PNECsediment of NP2EO (0.2375 μg/g) was similar to that in the literature (0.25 mg/kg)37.hh ( μg g) ( g g) The SSD methodology requires the least amount of data in order to reliably estimate the PNEC. The OECD recommends a minimum of five no observed effect concentrations (NOECs)38, while the EU’s technical guidance document requires at least ten from species of at least eight taxonomic groups25. The data in this study meet the above requirements and provide a reference for ecological risk assessment. To date, there has been no report on the ecological risk from NP, OP, BPA, and NP2EO in Carex cinerascens. In this study, the HC5 and PNECsediment values were used to assess the ecological risk from NP, OP, BPA, and NP2EO to Carex cinerascens. Results and Discussion The average RQ values were 1.25 for NP2EO in water, 0.74 for BPA in sediment, and 1.764 for NP2EO in sediment, indicating high ecological risk for input water to Poyang Lake. In Poyang Lake Wetland Park (Y25-Y28), the ecological risk from BPA was low in water and high in sed- iment, except for Baishazhou (Y28); the ecological risk from NP2EO was high in both water and sediment. The Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ Figure 3. Risk quotient (RQ) values of bisphenol A (BPA) and nonylphenol-di-ethoxylate (NP2EO) in the water, sediment, and Carex cinerascens of Poyang Lake wetland. Figure 3. Risk quotient (RQ) values of bisphenol A (BPA) and nonylphenol-di-ethoxylate (NP2EO) in the t di t d C i f P L k tl d Figure 3. Risk quotient (RQ) values of bisphenol A (BPA) and nonylphenol-di-ethoxylate (NP2EO) in the water, sediment, and Carex cinerascens of Poyang Lake wetland. Figure 4. Spatial risk quotient (RQ) values of bisphenol A (BPA) and nonylphenol-di-ethoxylate (NP2EO) in the water and sediment of Poyang Lake. Figure 4. Spatial risk quotient (RQ) values of bisphenol A (BPA) and nonylphenol-di-ethoxylate (NP2EO) i the water and sediment of Poyang Lake. average RQ values were 1.504 for NP2EO in water, 1.264 for BPA in sediment, and 2.827 for NP2EO in sediment. Surrounding non-point pollution from rural areas and sewage wastewater are the likely cause for the high eco- logical risk. average RQ values were 1.504 for NP2EO in water, 1.264 for BPA in sediment, and 2.827 for NP2EO in sediment. Surrounding non-point pollution from rural areas and sewage wastewater are the likely cause for the high eco- logical risk. g In Jiangxi Poyang Lake National Nature Reserve (Y4-Y6, Y11), the ecological risk from BPA was low in water and high in sediment, except for in South Lake (Y6); the ecological risk from NP2EO was high in water and sedi- ment, except for in Gongqingcheng (Y5). The average RQ values were 1.264 for NP2EO in water, 1.212 for BPA in sediment, and 2.548 for NP2EO in sediment. The high ecological risk may be due to the location of Jiangxi Poyang Lake National Nature Reserve at the convergence of the main branches of the Gan and Xiu Rivers; the reserve is also surrounded by Nanchang (which had an urban population exceeding 2 million in 2018, http://www.nc.gov. cn/), Anyi, Yongxiu, Gongqingcheng, De’an, and other cities. Results and Discussion In Junshan Lake and Chenjia Lake (Y12-Y15), the ecological risk from BPA was low in water and high in sediment, except for in Junshan Lake (Y13); the ecological risk from NP2EO was high in both water (except in Sanlixiang (Y12)) and sediment. The average RQ values were 0.82 for NP2EO in water, 1.506 for BPA in sediment, and 2.124 for NP2EO in sediment. In the Kangshan area (Y21-Y23), BPA and NP2EO in the sediment were at high ecological risk levels, with averages of 2.764 and 2.149, respectively. Construction of dykes can alter hydrologic factors such that water exchange times are longer and water flow is slower. This may explain the high ecological risk from NP2EO in Zhu Lake, Junshan Lake, Chenjia Lake, and Kangshan, which are isolated from Poyang Lake by dykes. The particularly high ecological risk from NP2EO (RQ = 13.09) at Jishan (Y7) may be related to its proximity to Duchang city. Similarly, Xingzi (Y3) is also located close to Xingzi city, resulting in high ecological risk (except for BPA in water). Although the concentra- tions of NP and OP in water, and NP in Carex cinerascens at Liaonan (Y4) were higher than those at other sites, Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ they were still lower than those of other lakes in China13,32 and the ecological risk was still very low. The concen- trations of NP2EO in the water of Sanlixiang (Y12) and NP in the sediment of Baishazhou (Y28) were similar. The ecological risk from NP2EO in Gan River (Y9) was high, which may be attributable to the high consumption of detergents near the Gan River coast, the most developed area in Jiangxi province. The high ecological risk from BPA in Carex cinerascens from the Fu River Tributary (Y17) may be due to the fact that there are some small plastic factories upstream (http://www.ncx.gov.cn/articles/2019/05/14/577160.shtml), however, the risk was still lower than in other lakes of China13,14,32,39. The high ecological risk from NP2EO may be due to the fact that the Poyang Lake basin (162 200 km2) covers 97% of the area of Jiangxi Province, and sewage produced by the con- sumption of household products by 46.2 million people (2017), converges in Poyang Lake (http://www.jiangxi. gov.cn/). Methods E l i l Ecological risk assessment procedure based on SSDs. The SSD process was carried out according to the work of Gao et al.27 and involved (1) screening of toxicity data; (2) selection of a distribution model and fitting SSD curves; (3) calculating values for HC5 (hazardous concentration for 5% of species) and PNEC; and (4) describing the ecological risks. Chronic toxicity data [no observed effect concentrations (NOEC) or lowest observed effect concentrations (LOEC)] were screened from the US EPA ECOTOX database (http://cfpob.epa. gov/ecotox) following the EU technical guidance document25. Selected toxicity data for NP, OP, BPA, and NP2EO for freshwater are listed in the Supplementary Information Tables S1–S4, respectively. Additional toxicity data published in recent years were used that were not available for previous studies27,30. A sigmoid distribution was selected for constructing the SSD model. The probability distribution function and cumulative distribution func- tion of the sigmoid distribution are shown in Supplementary Information S5. Calculation of HC5, PNECwater, and PNECsediment. HC5 is the concentration expected to be hazardous to 5% of the species in an ecosystem40. According to the assessment factor methods in the EU technical guidance document25, to calculate the values of PNECwater, HC5 is divided by an appropriate assessment factor (AF). The AF is commonly between 1 and 5, reflecting the uncertainty of the data. In this study, the AF value was set as 3, based on the number of species tested, quantity and quality of the toxicity data, and model goodness of fit. y y y gi PNECsediment was calculated from PNECwater according to the distribution equilibrium method in the EU guid- ance document25. The following formula, which is based on equilibrium partitioning theory, was applied: PNEC K RHO PNEC (1) sediment susp water susp water = ⋅ − (1) where Ksusp-water is the partition coefficient of suspended matter water (L/L); RHOsusp is the bulk density of wet suspended matter (g/L); PNECwater is the predicted no effect concentration in water (μg/L); and PNECsediment is the predicted no effect concentration in sediment (μg/g). where Ksusp-water is the partition coefficient of suspended matter water (L/L); RHOsusp is the bulk density of wet suspended matter (g/L); PNECwater is the predicted no effect concentration in water (μg/L); and PNECsediment is the predicted no effect concentration in sediment (μg/g). Characterization of ecological risks. Methods E l i l The risk quotient (RQ) was calculated from the ratio of predicted environmental concentrations (PEC) to PNEC values, and was used to determine the ecological risk41. Due to a shortage of detailed geographic distribution data for the usage and discharge amounts of PEDCs in Poyang Lake wetland, the measured environmental concentrations (MEC) were used to represent the predicted environmental concentrations. Risk levels were low, medium, or high for RQ values of <0.1, 0.1–1.0, and ≥1, respectively. Sampling and analysis of PEDCs. Sample collection locations were identified using a global positioning system (Fig. 1, with detailed coordinates shown in the Supplementary Information Table S6). The sampling sites were located in the Poyang Lake Wetland Park (Y25-Y28), Jiangxi Poyang Lake National Nature Reserve (Y4-Y6, Y11), five main tributaries of Poyang Lake (Y9, Y10, Y16, Y17, Y19, Y24), typical wetland lakes (Y12-Y15, Y18, Y20-Y23), typical wetlands (Y2, Y3, Y7, Y8), and the entrance of Poyang Lake into the Yangtze River (Y1). Surface water (0–1.0 m), sediment samples (0–0.15 m), and Carex cinerascens samples were collected from 28, 26, and 20 sites, respectively, in Poyang Lake in the dry season (October to February), according to sampling specifications. PEDCs were extracted from the water, sediment, and Carex cinerascens samples by ultrasound using dichlo- romethane, derivatised with N, O-bis(trimethylsilyl) trifluoroacetamide. The details of the extraction, clean up, and derivatization of PEDCs in water and Carex cinerascens are shown in Supplementary Information S7. Sediment samples were freeze-dried using a vacuum freeze dryer (FD-1B-50, Shanghai Bilon Instrument Manufacturing Co. Ltd., Shanghai, China), and then homogenised with a stainless steel spoon, placed into pre-cleaned brown glass bottles, and finally stored at 4 °C. Sediment samples (2.0 g) were mixed in a conical bottle with 5 mL of 0.1 mol/L HCl, and pH values < 1.0 were adjusted. Next, 10 mL dichloromethane (DCM) was added, Sampling and analysis of PEDCs. Sample collection locations were identified using a global positioning system (Fig. 1, with detailed coordinates shown in the Supplementary Information Table S6). The sampling sites were located in the Poyang Lake Wetland Park (Y25-Y28), Jiangxi Poyang Lake National Nature Reserve (Y4-Y6, Y11), five main tributaries of Poyang Lake (Y9, Y10, Y16, Y17, Y19, Y24), typical wetland lakes (Y12-Y15, Y18, Y20-Y23), typical wetlands (Y2, Y3, Y7, Y8), and the entrance of Poyang Lake into the Yangtze River (Y1). Results and Discussion In the water, sediment, and Carex cinerascens of Poyang Lake wetland, the high ecological risk from BPA and NP2EO may cause harm to some organisms, especially migratory birds, which warrants further investigation. Conclusionsh The distributions of NP, OP, BPA, and NP2EO in the water, sediment, and Carex cinerascens of Poyang Lake and their ecological risk were investigated. The four PEDCs were ubiquitous. This study adopted the SSD methodology to derive PNECwater and PNECsediment values for NP, OP, BPA, and NP2EO. The RQ method was used to quantify ecological risk. The ecological risk from NP and OP in the water, sediment, and Carex cinerascens from all sampling sites was low. The ecological risk from BPA was low in the water and high in the sediment and Carex cinerascens, while that of NP2EO was high in the water, sediment, and Carex cinerascens. The ecological risks from BPA and NP2EO in the five main tributaries of Poyang Lake were high indicating a high ecological risk in the input waters to Poyang Lake. In Poyang Lake Wetland Park, Jiangxi Poyang Lake National Nature Reserve, and other lakes, the ecological risks from BPA and NP2EO were high, which may cause harm to organisms, especially migratory birds. www.nature.com/scientificreports/ and ultrasound extraction was performed for 15 min using an ultrasonic cleaner (KQ-500E, Shanghai Bilon Instrument Manufacturing Co. Ltd.). The extraction process was repeated twice. The extracted organic phase was concentrated to approximately 3 mL. Three grams of anhydrous sodium sulfate, 4.0 g Florisil, 3.0 g anhydrous sodium sulfate, and 15–20 mL n-hexane were used to wet the pre-clean column, and the extract was poured into the column. When almost all of the extract entered the adsorption layer, it was eluted with 5.0 mL n-hexane/ether (v:v = 9:1) and 3.0 mL n-hexane/acetone (v:v = 7:3) three times. The eluent was collected and was evaporated to near-dryness using nitrogen. Fifty microliters of the derivative reagent BSTFA, 30 μL pyridine, and 20 μL 1.0 mg/L tribromophenol (internal standard, dissolved in n-hexane) were added, and then reacted for 45 min at 65 °C in a water bath. The solution was filtered to the inner column using a disposable needle filter with 0.22 μm organic membrane, to 100 μL for GC-MS analysis. μ y GC-MS analysis was performed with a TRACE™ 1310/TSQ8000Evo (Thermo Fisher Scientific™, Waltham, MA, USA). Separation of the target compounds was achieved using a TG-5SILMS (30 m × 0.25 mm × 0.25 μm) capillary column (Thermo Fisher Scientific™). The carrier gas was helium (99.999%) with a constant flow rate of 1 mL/min. The injector temperature was held at 260 °C, and the injection volume was 1.0 μL in splitless mode. The column temperature was programmed as follows: the initial temperature was 60 °C, increased to 150 °C at 15 °C/min, then to 220 °C at 8 °C/min, held at 220 °C for 1 min, then finally increased to 290 °C at 15 °C/min and then held at 290 °C for 5 min. The MS transfer line temperature was maintained at 280 °C, whereas the ion source temperature was 300 °C. Mass spectra were scanned in full scan mode from 50–650 m/z mass range for qualitative analysis and selected ion monitoring mode (SIM) for quantitative analysis. Electron impact ionization energy was 70 eV. All reagents used were chromatographically pure or were re-purified first. Instrumental calibration curves for NP and OP were established from standard solutions with nine concentrations ranging from 5 to 2000 μg/L. Instrument calibration curves for NP2EO and BPA were established from standard solutions with eight concentrations ranging from 5 to 1000 μg/L. References The trenbolone acetate affects the immune system in rainbow trout, Oncorhynchus mykiss. Aquat. Toxicol. 163 109–120, https://doi.org/10.1016/j.aquatox.2015.04.007. (2015). p g j q 10. Michałowicz, J. Bisphenol A - sources, toxicity and biotransformation. Environ. Toxicol. Pharmacol. 37(2), 738–758, https://doi. org/10.1016/j.etap.2014.02.003. (2014). 11. Wang, Y. H., Wang, Q. Y., Hu, L. F., Lu, G. H. & Li, Y. Occurrence of estrogens in water, sediment and biota and their ecological risk in northern Lake Taihu in China. Environ. Geochem. Health 37(1), 147–156, https://doi.org/10.1007/s10653-014-9637-0 (2015). g 12. Xu, E. G. B., Morton, B., Lee, J. H. W. & Leung, K. M. Y. Environmental fate and ecological risks of nonylphenols and bisphenol A in the Cape D’Aguilar Marine Reserve, Hong Kong. Mar. Pollut. Bull. 91(1), 128–138, https://doi.org/10.1016/j.marpolbul.2014.12.017. (2015). 13. Liu, D. et al. Occurrence, distribution, and risk assessment of alkylphenols, bisphenol A, and tetrabromobisphenol A in surface water, suspended particulate matter, and sediment in Taihu Lake and its tributaries. Mar. Pollut. Bull. 112(1–2), 142–150, https://doi. org/10.1016/j.marpolbul.2016.08.026. (2016). g j p 4. Jin, H. B. & Zhu, L. Y. Occurrence and partitioning of bisphenol analogues in water and sediment from Liaohe River Basin and Taihu Lake, China. Water Res. 103, 343–351, https://doi.org/10.1016/j.watres.2016.07.059. (2016). p g j ( ) 15. Gong, J., Huang, Y. D., Huang, W., Ran, Y. & Chen, D. Y. Multiphase partitioning and risk assessment of endocrine-disrupting chemicals in the Pearl River, China. Environ. Toxicol. Chem. 35(10), 2474–2482, https://doi.org/10.1002/etc.3419. (2016). g j 15. Gong, J., Huang, Y. D., Huang, W., Ran, Y. & Chen, D. Y. Multiphase partitioning and risk asses chemicals in the Pearl River, China. Environ. Toxicol. Chem. 35(10), 2474–2482, https://doi.org/10. 15. Gong, J., Huang, Y. D., Huang, W., Ran, Y. & Chen, D. Y. Multiphase partitioning and risk assessment of endocrine d chemicals in the Pearl River, China. Environ. Toxicol. Chem. 35(10), 2474–2482, https://doi.org/10.1002/etc.3419. (2016). 16. Zhang, P., Wang, C. Z., Liang, S. K., Wang, X. Y. & Wang, X. L. Distribution characteristics and ecological risk assessment of nonylphenol in Jiaozhou Bay in Qingdao, China. Fresenius Environ. Bull. 25(12), 5424–5431 (2016). 7. Liu, X. W. et al. Distributions and ecological risk assessment of estrogens and bisphenol A in an arid and semiarid area in northwes China. Environ. Sci. Pollut. Res. 24(8), 7216–7225, https://doi.org/10.1007/s11356-017-8434-6 (2017). p g 18. Yan, Z. Y. et al. www.nature.com/scientificreports/ All calibration coefficients were greater than 0.99, and the method quantification limits were 0.5 μg/L, 0.4 μg/L, 0.7 μg/L, and 1.2 μg/L for NP, OP, BPA, and NP2EO (RSD <5.8%, n = 5), respectively. Quality assurance/quality control (QA/QC). To further validate the precision and accuracy of the method, blank and matrix spiked experiments were carried out. No PEDCs were detected in the blank samples. Mixed standard solutions of 5.0, 20.0, and 50.0 μg/L PEDCs were added (n = 6). The mean recoveries of NP, OP, BPA, and NP2EO are shown in Supplementary Information Table S8. Data Availability y All data generated or analysed during this study are included in this published article and its Supplementary Information. References 1. Duan, X. Y., Li, Y. X., Li, X. G., Zhang, D. H. & Gao, Y. Alkylphenols in surface sediments of the Yellow Sea and East China Sea inne shelf: occurrence, distribution and fate. Chemosphere 107, 265–273, https://doi.org/10.1016/j.chemosphere.2013.12.054 (2014). p p g j p ( ) 2. Servos, M. R. Review of the aquatic toxicity, estrogenic responses and bioaccumulation of alkylphenols and alkylpheno polyethoxylates. Water Qual. Res. J. Can. 34, 123–177 (1999).f p y y 3. Soares, A., Guieysse, B., Jefferson, B., Cartmell, E. & Lester, J. N. Nonylphenol in the environment: a critical review on occurrence fate, toxicity and treatment in wastewaters. Environ. Int. 34, 1033–1049, https://doi.org/10.1016/j.envint.2008.01.004. (2008). 4 Yi G G Willi B & K k R E i l f f lk l h l d lk l h l h l i E i I t 28 3. Soares, A., Guieysse, B., Jefferson, B., Cartmell, E. & Lester, J. N. Nonylphenol in the environment: a critical review on occurrence, fate, toxicity and treatment in wastewaters. Environ. Int. 34, 1033–1049, https://doi.org/10.1016/j.envint.2008.01.004. (2008). 4. Ying, G. G., Williams, B. & Kookana, R. Environmental fate of alkylphenols and alkylphenol ethoxylates – a review. Environ. Int. 28, 215–226, https://doi.org/10.1016/S0160-4120(02)00017-X. (2002). fate, toxicity and treatment in wastewaters. Environ. Int. 34, 1033–1049, https://doi.org/10.1016/j.envint.2008.01.004. (2008). 4. Ying, G. G., Williams, B. & Kookana, R. Environmental fate of alkylphenols and alkylphenol ethoxylates – a review. Environ. Int. 28, 215–226, https://doi.org/10.1016/S0160-4120(02)00017-X. (2002).f 215–226, https://doi.org/10.1016/S0160-4120(02)00017-X. (2002).f p g ( ) ( ) 5. Capaldo, A. et al. Endocrine-disrupting effects of nonylphenol in the newt, Triturus carnifex (Amphibia, Urodela). Comp. Biochem. Physiol Part C: Toxicol. Pharmacol. 155(2), 352–358, https://doi.org/10.1016/j.cbpc.2011.10.004. (2012). y ( ) p g j p ( ) 6. Chen, B. S. & Yen, J. H. Effect of endocrine disruptor nonylphenol on physiologic features and proteome during gr Arabidopsis thaliana. Chemosphere 91, 468–474, https://doi.org/10.1016/j.chemosphere.2012.11.072 (2013). 6. Chen, B. S. & Yen, J. H. Effect of endocrine disruptor nonylphenol on physiologic features and proteome Arabidopsis thaliana. Chemosphere 91, 468–474, https://doi.org/10.1016/j.chemosphere.2012.11.072 (2013). 7. Chokwe, T. B., Okonkwo, J. O. & Sibali, L. L. Distribution, exposure pathways, sources and toxicity of nonylphenol and nonylphenol ethoxylates in the environment. Water SA. 43, 529–542, https://doi.org/10.4314/wsa.v43i4.01. (2017). 8. Davì, M. L. & Gnudi, F. Phenolic compounds in surface water. Water Res. 33(14), 3213–3219, https://doi.org/10.1016/S0043 1354(99)00027-5 (1999).hf 9. Massart, S. et al. Methods E l i l Surface water (0–1.0 m), sediment samples (0–0.15 m), and Carex cinerascens samples were collected from 28, 26, and 20 sites, respectively, in Poyang Lake in the dry season (October to February), according to sampling specifications. p y y g y y g p g pi PEDCs were extracted from the water, sediment, and Carex cinerascens samples by ultrasound using dichlo- romethane, derivatised with N, O-bis(trimethylsilyl) trifluoroacetamide. The details of the extraction, clean up, and derivatization of PEDCs in water and Carex cinerascens are shown in Supplementary Information S7. Sediment samples were freeze-dried using a vacuum freeze dryer (FD-1B-50, Shanghai Bilon Instrument Manufacturing Co. Ltd., Shanghai, China), and then homogenised with a stainless steel spoon, placed into pre-cleaned brown glass bottles, and finally stored at 4 °C. Sediment samples (2.0 g) were mixed in a conical bottle with 5 mL of 0.1 mol/L HCl, and pH values < 1.0 were adjusted. Next, 10 mL dichloromethane (DCM) was added, Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ References chemicalsafety/testing/34528236.pdf (1992). 39. Wang, Y. H., Wang, Q. Y., Hu, L. F., Lu, G. H. & Li, Y. Occurrence of estrogens in water, sediment and biota and their ecological risk in northern Taihu Lake in China. Environ. Geochem. Health. 37(1), 147–156, https://doi.org/10.1007/s10653-014-9637-0 (2015). y g p 39. Wang, Y. H., Wang, Q. Y., Hu, L. F., Lu, G. H. & Li, Y. Occurrence of estrogens in water, sediment and biota and their ecologica in northern Taihu Lake in China. Environ. Geochem. Health. 37(1), 147–156, https://doi.org/10.1007/s10653-014-9637-0 (201 p g 40. Wheeler, J. R., Grist, E. P. M., Leung, K. M. Y., Morritt, D. & Crane, M. Species sensitivity distributions: data and model choice. Pollut. Bull. 45, 192–202, https://doi.org/10.1016/S0025-326X(01)00327-7. (2002). 41. Cristale, J., Katsoyiannis, A., Sweetman, A. J., Jones, K. C. & Lacorte, S. Occurrence and risk assessment of organophosphorus and brominated flame retardants in the river Aire (UK). Environ. Pollut. 179, 194–200, https://doi.org/10.1016/j.envpol.2013.04.001. (2013). Author Contributions Author Contributions Pinghua Yang has completed all the work of the paper. Author Contributions Pinghua Yang has completed all the work of the paper. References Bisphenol analogues in surface water and sediment from the shallow Chinese freshwater lakes: occurrence, distribution, source apportionment, and ecological and human health risk. Chemosphere. 184, 318–328, https://doi.org/10.1016/j. chemosphere.2017.06.010. (2017). Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4 www.nature.com/scientificreports/ 19. Zhong, M. Q., Yin, P. H. & Zhao, L. Nonylphenol and octylphenol in riverine waters and surface sediments of the Pearl River Estuaries, South China: occurrence, ecological and human health risks. Water Sci. Technol. Water Supply 17(4), 1070–1079, https:// doi.org/10.2166/ws.2017.002. (2017). g ( ) 0. Zhong, W. J., Wang, D. H. & Wang, Z. J. Distribution and potential ecological risk of 50 phenolic compounds in three rivers in Tianjin, China. Environ. Pollut. 235, 121–128, https://doi.org/10.1016/j.envpol.2017.12.037. (2018). Tianjin, China. Environ. Pollut. 235, 121–128, https://doi.org/10.1016/j.envpol.2017.12.037. (2018). 21. Niu, S. P. & Zhang, C. L. Endocrine disrupting compounds from the source water of the Huai River (Huainan City), China. Arch. 1. Niu, S. P. & Zhang, C. L. Endocrine disrupting compounds from the source water of the Huai River (Huainan City), China. Arch Environ. Contam. Toxicol. 74(3), 471–483, https://doi.org/10.1007/s00244-017-0445-2 (2018). p g 2. Wu, P., Qin, B. Q. & Yu, G. Estimates of long-term water total phosphorus (TP) concentrations in three large shallow lakes in th Yangtze River basin, China. Environ. Sci. Pollut. Res. 23(5), 4938–4948, https://doi.org/10.1007/s11356-015-5736-4 (2016). g p g 3. Xu, H. C. et al. Characterization, origin and aggregation behavior of colloids in eutrophic shallow lake. Water Res. 142, 176–186 https://doi.org/10.1016/j.watres.2018.05.059 (2018). p g j 24. U.S. Environmental Protection Agency. Guidelines for ecological risk assessment, https://www.epa.gov/sites/production/ files/2014-11/documents/eco_risk_assessment1998.pdf (1998). i p 5. ECB (European Chemicals Bureau). Technical guidance document on risk assessment, https://ec.europa.eu/jrc/en/publication/eur scientific-and-technical-research-reports/technical-guidance-document-risk-assessment-part-1-part-2 (2003).f i p g p p ( ) 6. Staples, C. A. et al. Comparison of four species sensitivity distribution methods to calculate predicted no effect concentrations fo bisphenol A. Hum. Ecol. Risk Assess. 14(3), 455–478, https://doi.org/10.1080/10807030802074170 (2008). p p g 27. Gao, P., Li, Z. Y., Gibson, M. & Gao, H. W. Ecological risk assessment of nonylphenol in coastal waters of China based on species sensitivity distribution model. Chemosphere 104(3), 113–119, https://doi.org/10.1016/j.chemosphere.2013.10.076. (2014). y p p g j p 28. Guo, L., Li, Z. Y., Gao, P., Hu, H. & Gibson, M. Ecological risk assessment of bisphenol A in surface waters of China based on both traditional and reproductive endpoints. Chemosphere 139, 133–137, https://doi.org/10.1016/j.chemosphere.2015.06.001. (2015). p p p p g j p ( 29. Zhang, L. M., Wei, C. Acknowledgementsh c o edge e ts This work was supported by the National Natural Science Foundation of China (Nos 21567011, 51741805) and the Jiangxi Natural Science Foundation (20151BAB203043). g This work was supported by the National Natural Science Foundation of China (Nos 21567011, 51741805) and the Jiangxi Natural Science Foundation (20151BAB203043). References D., Zhang, H. & Song, M. W. Criteria for assessing the ecological risk of nonylphenol for aquatic life in Ch f f h t Ch ph 184 569 574 htt //d i /10 1016/j h h 2017 06 035 (2017) 9. Zhang, L. M., Wei, C. D., Zhang, H. & Song, M. W. Criteria for assessing the ecological risk of nonylphenol for aquatic life in Chinese surface fresh water. Chemosphere 184, 569–574, https://doi.org/10.1016/j.chemosphere.2017.06.035. (2017). 29. Zhang, L. M., Wei, C. D., Zhang, H. & Song, M. W. Criteria for assessing the ecological risk of nonylphenol for aquatic life in Ch surface fresh water. Chemosphere 184, 569–574, https://doi.org/10.1016/j.chemosphere.2017.06.035. (2017). 30. Chen, R. et al. Spatial-temporal distribution and potential ecological risk assessment of nonylphenol and octylphenol in riverine outlets of Pearl river delta, China. J. Environ. Sci. 26(11), 2340–2347, https://doi.org/10.1016/j.jes.2014.09.019. (2014). 31. Wang, L. et al. Monitoring of selected estrogenic compounds and estrogenic activity in surface water and sediment of the Yellow River in China using combined chemical and biological tools. Environ. Pollut. 165, 241–249, https://doi.org/10.1016/j. envpol.2011.10.005 (2012). p ( ) 32. Yu, F., Pan, X. J. & Wang, B. Determination of four phenolic endocrine disrupting chemicals in Dianchi Lake, China. Int. J. Environ. Anal. Chem. 92(13), 1532–1545, https://doi.org/10.1080/03067319.2010.548604. (2012). 3. Zhang, Y. Z., Tang, C. Y., Song, X. F. & Li, F. D. Behavior and fate of alkylphenols in surface water of the Jialu River, Henan Province China. Chemosphere. 77(4), 559–565, https://doi.org/10.1016/j.chemosphere.2009.07.005 (2009). p p g j p 4. Qin, B. Q., Xu, P. Z., Wu, Q. L., Luo, L. C. & Zhang, Y. L. Environmental issues of Lake Taihu, China. Hydrobiologia. 581(1), 3–14 https://doi.org/10.1007/s10750-006-0521-5 (2007). 35. ECB (European Chemicals Bureau). European Union risk assessment report for 4-Nonylphenol (branched) and Nonylphenol, https://mobil.bfr.bund.de/cm/343/4_nonylphenol_und_nonylphenol.pdf (2001). 36. Fenner, K., Kooijman, C., Scheringer, M. & Hungerbühler, K. Including transformation products into the risk assessment for chemicals: the case of nonylphenol ethoxylate usage in Switzerland. Environ. Sci. Technol. 36(6), 1147–1154, https://doi.org/10.1021/ es015576i. (2002). 37. González, M. M., Martín, J., Santos, J. L., Aparicio, I. & Alonso, E. Occurrence and risk assessment of nonylphenol and nonylphenol ethoxylates in sewage sludge from different conventional treatment processes. Sci. Total Environ. 408(3), 563–570, https://doi. org/10.1016/j.scitotenv.2009.10.027. (2010). 8. OECD (Organization for Economic Cooperation and Development). Report of the OECD workshop on the extrapolation o laboratory aquatic toxicity data to the real environment. OECD Environment Monographs No. 59, http://www.oecd.org chemicalsafety/testing/34528236.pdf (1992). Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-47864-4. Competing Interests: The authors declare no competing interests. Competing Interests: The authors declare no competing interests. Competing Interests: The authors declare no competing interests. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps an institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2019 Scientific Reports \| (2019) 9:11302 \| https://doi.org/10.1038/s41598-019-47864-4
https://openalex.org/W2461690878	https://digital.csic.es/bitstream/10261/197824/1/paclitaxel_resistance.pdf	English	null	Loss of FBXW7 and accumulation of MCL1 and PLK1 promote paclitaxel resistance in breast cancer	Oncotarget	2,016	cc-by	10,329	Loss of FBXW7 and accumulation of MCL1 and PLK1 promote paclitaxel resistance in breast cancer Jessica Gasca1, Maria Luz Flores1, Servando Giráldez4, Manuel Ruiz-Borrego3, María Tortolero4, Francisco Romero4, Miguel A. Japón1,2 and Carmen Sáez1,2 1 Instituto de Biomedicina de Sevilla (IBIS), Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Seville, Spain 2 Department of Pathology, Hospital Universitario Virgen del Rocío, Seville, Spa 2 Department of Pathology, Hospital Universitario Virgen del Rocío, Seville, Spain 3 Oncology Unit, Hospital Universitario Virgen del Rocío, Seville, Spain 3 Oncology Unit, Hospital Universitario Virgen del Rocío, Seville, Spain 4 Department of Microbiology, Faculty of Biology, Universidad de Sevilla, Seville, Spain Correspondence to: Carmen Sáez, email: csaez1@us.es Correspondence to: Miguel A. Japón, email: mjapon@cica.es Keywords: FBXW7, MCL1, PLK1, apoptosis, paclitaxel, Pathology Section Received: January 08, 2016 Accepted: June 21, 2016 Published: July 07, 2016 Published: July 07, 2016 Received: January 08, 2016 Accepted: June 21, 2016 Abstract FBXW7 is a component of SCF (complex of SKP1, CUL1 and F-box-protein)- type ubiquitin ligases that targets several oncoproteins for ubiquitination and degradation by the proteasome. FBXW7 regulates cellular apoptosis by targeting MCL1 for ubiquitination. Recently, we identified PLK1 as a new substrate of FBXW7 modulating the intra-S-phase DNA-damage checkpoint. Taxanes are frequently used in breast cancer treatments, but the acquisition of resistance makes these treatments ineffective. We investigated the role of FBXW7 and their substrates MCL1 and PLK1 in regulating the apoptotic response to paclitaxel treatment in breast cancer cells and their expression in breast cancer tissues. Paclitaxel-sensitive MDA-MB-468 and a paclitaxel-resistant MDA-MB-468R subclone were used to study the role of FBXW7 and substrates in paclitaxel-induced apoptosis. Forced expression of FBXW7 or downregulation of MCL1 or PLK1 restored sensitivity to paclitaxel in MDA-MB-468R cells. By contrary, FBXW7-silenced MDA-MB-468 cells became resistant to paclitaxel. The expression of FBXW7 and substrates were studied in 296 invasive carcinomas by immunohistochemistry and disease-free survival was analyzed in a subset of patients treated with paclitaxel. In breast cancer tissues, loss of FBXW7 correlated with adverse prognosis markers and loss of FBXW7 and MCL1 or PLK1 accumulation were associated with diminished disease-free survival in paclitaxel-treated patients. We conclude that FBXW7 regulates the response to paclitaxel by targeting MCL1 and PLK1 in breast cancer cells and thus targeting these substrates may be a valuable adjunct for paclitaxel treatment. Also, FBXW7, MCL1 and PLK1 may be relevant predictive markers of tumor progression and response to paclitaxel treatment. www.impactjournals.com/oncotarget/ Oncotarget, Vol. 7, No. 33 Loss of FBXW7 and accumulation of MCL1 and PLK1 promote paclitaxel resistance in breast cancer Research Paper: Pathology www.impactjournals.com/oncotarget/ Oncotarget, Vol. 7, No. 33 Loss of FBXW7 and accumulation of MCL1 and PLK1 promote paclitaxel resistance in breast cancer Jessica Gasca1, Maria Luz Flores1, Servando Giráldez4, Manuel Ruiz-Borrego3, María Tortolero4, Francisco Romero4, Miguel A. Japón1,2 and Carmen Sáez1,2 1 Instituto de Biomedicina de Sevilla (IBIS), Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Seville, Spain 2 Research Paper: Pathology www.impactjournals.com/oncotarget/ RESULTS Paclitaxel is a microtubule polymerizing agent that blocks mitotic progression by activating the spindle assembly checkpoint and either causes cell death in mitosis by apoptosis or cells exit mitosis into a tetraploid G1 state, a process known as slippage [15]. Paclitaxel improves overall and disease-free survival in combination with anthracyclines or alkylating agents in metastatic and early stage breast cancer, although after the initial response, resistance to paclitaxel frequently ensues [16]. Several mechanisms for paclitaxel resistance in breast cancer have been reported including alterations in apoptotic pathways [17]. In fact, we reported that resistance to paclitaxel- induced apoptosis in MDA-MB-231breast cancer cells is related to the inability to disrupt the interaction between BCL-xL, an anti-apoptotic BCL-2 family protein, and BAK; and BCL-xL accumulation may be of prognostic value in paclitaxel-treated patients [18]. Defective degradation and accumulation of related proteins may underlie the acquisition of resistance. MCL1, a pro- survival BCL-2 family member that inhibits apoptosis by blocking the cell death mediators BAX and BAK [19], is targeted by FBXW7 for degradation upon phosphorylation by glycogen synthase kinase 3-beta (GSK3β) [20]. During anti-tubulin-induced mitotic arrest, MCL1 levels decline to potentiate cell death. Loss or inactivation of FBXW7 conferred resistance to anti-tubulin agents paclitaxel and vincristine by blocking proteasomal degradation of MCL1 [21]. We also identified this important role of MCL1 in preventing cell death during mitosis in prostate cancer cells after paclitaxel treatment [22]. Polo-like kinase 1 (PLK1) is a major mitotic kinase that plays a key role in the division of eukaryotic cells [23]. PLK1 is recruited in metaphase by several proteins such as the spindle checkpoint component BubR1 [24], which inhibits Cdc20 to prevent the activation of the anaphase promoting complex/cyclosome (APC/C). During late anaphase, PLK1 is degraded by APC/C-proteasome to properly control the exit from mitosis and cytokinesis [25]. PLK1- depleted cells are arrested in prometaphase when the spindle checkpoint is activated, but these cells fail to arrest if MAD2 or BubR1 are co-depleted [26]. PLK1 is also implicated in the replication machinery [27]. In fact, Expression of FBXW7 and its substrates in breast cancer cell lines We selected hormone-dependent MCF7, T47D, and hormone-independent MDA-MB-231, MDA- MB-468 breast cancer cell lines to study the expression levels of FBXW7 and its substrates by western blot in basal conditions (Figure 1A). FBXW7 protein levels in MCF7, and MDA-MB-468 cells were higher than those in T47D, and MDA-MB-231 cells. Accordingly, high levels of FBXW7 correlated with low levels of its substrates. Similar results were also observed in BT474 and SKBR3 cells (Supplementary Figure S1). To confirm these results we silenced the expression of FBXW7 in MCF7 and MDA-MB-468 cells by using siRNA oligonucleotides. AURKA, c-Myc, Cyclin E, PLK1 and MCL1 levels all increased after FBXW7 downregulation in both cell lines (Figure 1B). MCF7 and MDA-MB-468 are sensitive to paclitaxel, so we generated paclitaxel-resistant MDA-MB- 468R cells by selecting survival clones after paclitaxel treatment of the MDA-MB-468 cell line. In order to investigate the potential role of FBXW7 in paclitaxel resistance, we compared FBXW7 protein levels of parental MDA-MB-468 cells (IC50 0.026 ± 0.004 µM) with those of paclitaxel-resistant MDA-MB-468R cells (IC50 0.183 ± 0.002 µM) (Figure 1C). In basal conditions, paclitaxel- resistant MDA-MB-468R cells showed lower levels of FBXW7 and concordantly higher levels of AURKA, c-Myc, Cyclin E, PLK1 and MCL1 than paclitaxel- sensitive MDA-MB-468 cells (Figure 1C), suggesting the involvement of FBXW7 in paclitaxel resistance. Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells INTRODUCTION mammals [2]. Because many of the substrates that FBXW7 targets for degradation are well-known oncoproteins, such as Cyclin E, c-Myc, Aurora kinase A (AURKA) and Notch-1, FBXW7 is generally accepted as a tumor suppressor [3, 4]. FBXW7 is frequently inactivated by deletion, mutation or promoter hypermethylation in multiple neoplasias. Inactivating mutations have been most frequently found in cholangiocarcinoma, T-cell acute lymphocytic leukemia, and carcinomas of colon, FBXW7 is the F-box protein of the SKP1-CUL1- FBXW7 (SCFFBXW7) E3 ubiquitin ligase, responsible for recruiting specific substrates for ubiquitination and degradation by the proteasome [1]. FBXW7 comprises three alternatively spliced isoforms, FBXW7 α, β and γ, of which FBXW7α is the predominant nuclear isoform in www.impactjournals.com/oncotarget Oncotarget 52751 endometrium and stomach [5]. In breast cancer, FBXW7 promoter hypermethylation has been identified in 51% of a series of primary tumors [6]. Many of the primary tumors with FBXW7 mutations exhibit elevated levels of Cyclin E leading to deregulation of the cell cycle [7, 8]. Loss of expression of FBXW7 has been reported in a series of breast carcinomas where Cyclin E was upregulated and associated with poor prognosis [9] but its relationship with other substrates has not been addressed. Other substrates such as c-Myc and AURKA have important roles in tumorigenesis by promoting cancer cell survival and migration [10-13] and AURKA has been found overexpressed in a series of triple negative breast carcinomas [14]. we have recently reported that PLK1 is ubiquitinated and degraded by SCFFBXW7α/proteasome to modulate the intra- S-phase checkpoint, where PLK1 plays an important role in the regulation of DNA replication by contributing to the formation of pre-replicative complexes [28]. Knowledge of paclitaxel resistance mechanisms will allow the identification of predictive tumor markers and therapeutic targets. Herein, we investigate the role of FBXW7 and relevant substrates in regulating apoptosis induced by paclitaxel in breast cancer cells, their expression in breast cancer tissues, and their potential predictive value in paclitaxel-treated patients. Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells Next, we examined paclitaxel-induced apoptosis in MDA-MB-468 and MDA-MB-468R cells by western www.impactjournals.com/oncotarget Oncotarget 52752 Figure 1: Expression of FBXW7 and its substrates in breast cancer cell lines. A. Western blot analyses of FBXW7, AURKA, c-Myc, Cyclin E, PLK1 and MCL1 in MCF7, T47D, MDA-MB-231 and MDA-MB-468 cells in basal conditions are shown, using β-actin as loading control. Quantifications of indicated proteins are shown as histograms and data are presented as mean ± SEM. B. Western blot analyses of MCF7 and MDA-MB-468 cells silenced with FBXW7 siRNA and with a nontargeting control are shown; β-actin was used to ensure equal loading. Histograms show the densitometric analyses of indicated proteins. Data are presented as mean ± SEM comparing iRNA control versus siRNA FBXW7 in MDA-MB-468 cells and siRNA control versus siRNA FBXW7 in MCF7 cells. p < 0.05, p < 0.01 and *p < 0.001 from Student’s t-test. C. FBXW7, AURKA, c-Myc, Cyclin E, PLK1 and MCL1 were detected by Western blot n MDA-MB-468 and MDA-MB-468R cells in basal conditions. β-actin is shown as loading control. Densitometric analyses of indicated proteins are shown as histograms. Data are presented as mean ± SEM comparing MDA-MB-468 and MDA-MB-468R cells in basal conditions. p < 0.05, p < 0.01 and p < 0.001 from Student’s t-test. IC50 curves for paclitaxel in MDA-MB-468 (solid line) and MDA-MB-468R (dotted line) cell lines are shown. Data are presented as mean ± SEM. Figure 1: Expression of FBXW7 and its substrates in breast cancer cell lines. A. Western blot analyses of FBXW7, AURKA, c-Myc, Cyclin E, PLK1 and MCL1 in MCF7, T47D, MDA-MB-231 and MDA-MB-468 cells in basal conditions are shown, using β-actin as loading control. Quantifications of indicated proteins are shown as histograms and data are presented as mean ± SEM. B. Western blot analyses of MCF7 and MDA-MB-468 cells silenced with FBXW7 siRNA and with a nontargeting control are shown; β-actin was used to ensure equal loading. Histograms show the densitometric analyses of indicated proteins. Data are presented as mean ± SEM comparing siRNA control versus siRNA FBXW7 in MDA-MB-468 cells and siRNA control versus siRNA FBXW7 in MCF7 cells. p < 0.05, p < 0.01 and p < 0.001 from Student’s t-test. C. FBXW7, AURKA, c-Myc, Cyclin E, PLK1 and MCL1 were detected by Western blot in MDA-MB-468 and MDA-MB-468R cells in basal conditions. β-actin is shown as loading control. FBXW7 overexpression restores sensitivity to paclitaxel in resistant MDA-MB-468R cells To confirm our analysis, MDA-MB-468R cells were transiently transfected with pCMVHA-FBXW7 or the empty vector and treated with 0.025 µM paclitaxel during 24 h. First, we examined the substrates of FBXW7 after transfection and treatment with paclitaxel (Figure 4A and Supplementary Figure S4A). AURKA, c-Myc, Cyclin E, PLK1, and MCL1 decreased after transfection of FBXW7 in untreated cells. After paclitaxel treatment, the changes were most remarkable in PLK1, which accumulated in mock-transfected cells and was very low in FBXW7- transfected cells, and MCL1, which dramatically reduced in the latter cell type. Next, we analyzed apoptosis and mitotic proteins in transfected cells after paclitaxel treatment (Figure 4B). Apoptosis was efficiently induced by paclitaxel in FBXW7-transfected MDA-MB-468R cells as shown by more cleavage of PARP, caspase-9 and caspase-3. Flow-cytometric sub-G1 analysis also confirmed these results (Figure 4C). Cyclin B1, PTTG1, and BubR1 levels decreased after paclitaxel treatment in FBXW7-transfected cells (Figure 4B and Supplementary Figure S4B) and showed higher ploidy after FISH analysis (Supplementary Figure S4C), indicating that these cells recover the slippage mechanism to become sensitive to paclitaxel, similarly to the parental MDA-MB-468 cells. Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells FISH analysis showed that 60% of paclitaxel- treated MDA-MB-468 cells had higher than basal ploidy demonstrating slippage, whereas the percentage of ploidy remained unchanged in paclitaxel-treated MDA-MB-468R cells reflecting diminished slippage as a mechanism of paclitaxel resistance in these cells (Figure 2C). to conclude they are not cell line specific. Flow-cytometric sub-G1 analysis was performed to confirm the results in MDA-MB-468 cells (Figure 3C). The accumulation of PTTG1, p-His H3 and Cyclin B1 illustrated mitotic arrest in paclitaxel-treated FBXW7-silenced MDA-MB-468, reinforcing the influence in mitotic arrest after FBXW7 downregulation. We also confirmed increased protein levels of both FBXW7 substrates, MCL1 and PLK1. This further supports the importance of FBXW7 in achieving the complete paclitaxel response. Quantifications of western blot bands are shown in Supplementary Figure S3. Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells Densitometric analyses of indicated proteins are shown as histograms. Data are presented as mean ± SEM comparing MDA-MB-468 and MDA-MB-468R cells in basal conditions. p < 0.05, p < 0.01 and p < 0.001 from Student’s t-test. IC50 curves for paclitaxel in MDA-MB-468 (solid line) and MDA-MB-468R (dotted line) cell lines are shown. Data are presented as mean ± SEM. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52753 blot (Figure 2A). Cells were treated with 0.025µM paclitaxel and collected at 24 and 48 h. Cleavage of PARP, caspase-9 and caspase-3 were induced earlier and more potently in parental MDA-MB-468 cells than in MDA-MB-468R cells. Anti-apoptotic protein MCL1 decreased after paclitaxel treatment in MDA-MB-468 cells but not in MDA-MB-468R cells, in which MCL1 remained unchanged. Paclitaxel treatment also induced the increase of Cyclin B1 and PTTG1 at 24 h in parental MDA-MB-468 cells and the later decrease of Cyclin B1, PTTG1, and BubR1 at 48 h to facilitate slippage and post-mitotic cell death. Mitotic exit is also reflected by the phospho-histone H3 (p-His H3) decrease at 48 h. In resistant MDA-MB-468R cells, Cyclin B1, PTTG1, and BubR1 increased after 24 h of paclitaxel treatment and did not fall at 48 h, p-His H3 also remained present, illustrating an association between slippage inhibition and impaired apoptosis in these cells. PLK1 needs to be degraded to allow mitotic progress and post-mitotic cell death. In sensitive MDA-MB-468 cells, PLK1 levels increased after 24 h of paclitaxel treatment and later decreased at 48 h. On the contrary, in paclitaxel treated MDA-MB-468R cells, high levels of PLK1 were still present at 48 h. Quantifications of western blot bands are shown in Supplementary Figure S2. We then analyzed the cell cycle and ploidy of paclitaxel-treated MDA-MB-468 and MDA-MB-468R cells by flow cytometry and FISH. Paclitaxel treatment induced sensitive MDA-MB-468 cells to arrest in G2/M at 24 h and increased the sub-G0/G1 apoptotic population at 48 h (Figure 2B). The cell cycle of resistant MDA-MB-468R cells was not affected after paclitaxel treatment at 24 h, but these cells showed a reduction of the G0/G1 population and fewer sub-G0/G1 apoptotic events than parental cells at 48 h. In resistant cells, FISH analysis showed higher ploidy in DMSO than in MDA-MB-468 sensitive cells. As a result of slippage, cells duplicate their DNA without cytokinesis and increase ploidy. Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells FISH analysis showed that 60% of paclitaxel- treated MDA-MB-468 cells had higher than basal ploidy demonstrating slippage, whereas the percentage of ploidy remained unchanged in paclitaxel-treated MDA-MB-468R cells reflecting diminished slippage as a mechanism of paclitaxel resistance in these cells (Figure 2C). blot (Figure 2A). Cells were treated with 0.025µM paclitaxel and collected at 24 and 48 h. Cleavage of PARP, caspase-9 and caspase-3 were induced earlier and more potently in parental MDA-MB-468 cells than in MDA-MB-468R cells. Anti-apoptotic protein MCL1 decreased after paclitaxel treatment in MDA-MB-468 cells but not in MDA-MB-468R cells, in which MCL1 remained unchanged. Paclitaxel treatment also induced the increase of Cyclin B1 and PTTG1 at 24 h in parental MDA-MB-468 cells and the later decrease of Cyclin B1, PTTG1, and BubR1 at 48 h to facilitate slippage and post-mitotic cell death. Mitotic exit is also reflected by the phospho-histone H3 (p-His H3) decrease at 48 h. In resistant MDA-MB-468R cells, Cyclin B1, PTTG1, and BubR1 increased after 24 h of paclitaxel treatment and did not fall at 48 h, p-His H3 also remained present, illustrating an association between slippage inhibition and impaired apoptosis in these cells. PLK1 needs to be degraded to allow mitotic progress and post-mitotic cell death. In sensitive MDA-MB-468 cells, PLK1 levels increased after 24 h of paclitaxel treatment and later decreased at 48 h. On the contrary, in paclitaxel treated MDA-MB-468R cells, high levels of PLK1 were still present at 48 h. Quantifications of western blot bands are shown in Supplementary Figure S2. We then analyzed the cell cycle and ploidy of paclitaxel-treated MDA-MB-468 and MDA-MB-468R cells by flow cytometry and FISH. Paclitaxel treatment induced sensitive MDA-MB-468 cells to arrest in G2/M at 24 h and increased the sub-G0/G1 apoptotic population at 48 h (Figure 2B). The cell cycle of resistant MDA-MB-468R cells was not affected after paclitaxel treatment at 24 h, but these cells showed a reduction of the G0/G1 population and fewer sub-G0/G1 apoptotic events than parental cells at 48 h. In resistant cells, FISH analysis showed higher ploidy in DMSO than in MDA-MB-468 sensitive cells. As a result of slippage, cells duplicate their DNA without cytokinesis and increase ploidy. FBXW7 downregulation induces mitotic arrest and resistance to paclitaxel in MDA-MB-468 cells We used siRNA oligonucleotides to study whether PLK1 or MCL1 downregulation may have major impact on the response to paclitaxel of resistant MDA-MB-468R cells. As expected, MCL1 downregulation increased apoptosis after 24 h of paclitaxel treatment in these cells as shown by increased levels of cleaved PARP, caspase-9 and caspase-3, therefore reverting their resistant phenotype (Figure 5A). Similarly, PLK1 downregulation was also able to promote paclitaxel-induced apoptosis efficiently in MDA-MB-468R cells (Figure 5B). Thus, we may conclude that MCL1 and PLK1 are contributors to the development of paclitaxel resistance in breast cancer cells. To gain insight into the implication of FBXW7 in paclitaxel resistance, we investigated whether the downregulation of FBXW7 influenced the paclitaxel response of sensitive MDA-MB-468 cells. As shown in Figure 3A, interfering FBXW7 diminished the sensitivity to paclitaxel treatment as demonstrated by decreased levels of cleaved PARP, caspase-9 and caspase-3. Same results were also observed in MCF7 cells (Figure 3B), so www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52754 Figure 2: Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells. Cells wer treated with 0.025 µM paclitaxel during 24 and 48 h. A. Intact and cleaved PARP, cleaved caspase-9 (CASP-9), cleaved caspase-3 (CASP 3), MCL1, Cyclin B1, PTTG1, phospho-histone H3 (p-His H3), PLK1 and BubR1 proteins were detected by western blot. β-actin is show as loading control. Densitometric analysis of cleaved PARP, cleaved CASP-9 and cleaved CASP-3 are shown as histograms. Data from triplicate experiments are presented as mean ± SEM comparing paclitaxel-treated MDA-MB-468 versus paclitaxel-treated MDA-MB-468R at 24 and 48 h. p < 0.05 and **p < 0.001 from Student’s t-test. B. Cell cycle analysis of propidium iodide-stained cells by flow cytometry Quantification of each phase and sub-G1 cells are shown as histograms. C. Ploidy analysis by FISH. The number of signals per cells wa determined for chromosomes 1, 11 and 17 in at least 100 cells. Representative photographs are shown. Histograms represent the percentag of cells with normal or higher ploidy for each condition. Figure 2: Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells. Cells were treated with 0.025 µM paclitaxel during 24 and 48 h. A. Intact and cleaved PARP, cleaved caspase-9 (CASP-9), cleaved caspase-3 (CASP- 3), MCL1, Cyclin B1, PTTG1, phospho-histone H3 (p-His H3), PLK1 and BubR1 proteins were detected by western blot. β-actin is shown as loading control. Densitometric analysis of cleaved PARP, cleaved CASP-9 and cleaved CASP-3 are shown as histograms. Data from triplicate experiments are presented as mean ± SEM comparing paclitaxel-treated MDA-MB-468 versus paclitaxel-treated MDA-MB-468R at 24 and 48 h. p < 0.05 and **p < 0.001 from Student’s t-test. B. Cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms. C. Ploidy analysis by FISH. The number of signals per cells was determined for chromosomes 1, 11 and 17 in at least 100 cells. Representative photographs are shown. Histograms represent the percentage of cells with normal or higher ploidy for each condition. Figure 2: Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells. Cells were treated with 0.025 µM paclitaxel during 24 and 48 h. A. Intact and cleaved PARP, cleaved caspase-9 (CASP-9), cleaved caspase-3 (CASP- 3), MCL1, Cyclin B1, PTTG1, phospho-histone H3 (p-His H3), PLK1 and BubR1 proteins were detected by western blot. www.impactjournals.com/oncotarget β-actin is shown as loading control. Densitometric analysis of cleaved PARP, cleaved CASP-9 and cleaved CASP-3 are shown as histograms. Data from triplicate experiments are presented as mean ± SEM comparing paclitaxel-treated MDA-MB-468 versus paclitaxel-treated MDA-MB-468R at 24 and 48 h. p < 0.05 and **p < 0.001 from Student’s t-test. B. Cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms. C. Ploidy analysis by FISH. The number of signals per cells was determined for chromosomes 1, 11 and 17 in at least 100 cells. Representative photographs are shown. Histograms represent the percentage of cells with normal or higher ploidy for each condition. Figure 2: Paclitaxel-induced apoptosis and cell cycle analysis of MDA-MB-468 and MDA-MB-468R cells. Cells were treated with 0.025 µM paclitaxel during 24 and 48 h. A. Intact and cleaved PARP, cleaved caspase-9 (CASP-9), cleaved caspase-3 (CASP- 3), MCL1, Cyclin B1, PTTG1, phospho-histone H3 (p-His H3), PLK1 and BubR1 proteins were detected by western blot. β-actin is shown as loading control. Densitometric analysis of cleaved PARP, cleaved CASP-9 and cleaved CASP-3 are shown as histograms. Data from triplicate experiments are presented as mean ± SEM comparing paclitaxel-treated MDA-MB-468 versus paclitaxel-treated MDA-MB-468R at 24 and 48 h. p < 0.05 and **p < 0.001 from Student’s t-test. B. Cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms. C. Ploidy analysis by FISH. The number of signals per cells was determined for chromosomes 1, 11 and 17 in at least 100 cells. Representative photographs are shown. Histograms represent the percentage of cells with normal or higher ploidy for each condition. www.impactjournals.com/oncotarget Quantification of each phase and sub-G1 cells are shown as histograms and data from triplicates are presented as mean ± SEM. Figure 3: FBXW7 downregulation induces mitotic arrest and resistance to paclitaxel in MDA-MB-468 and MCF7 cells. MDA-MB-468 A. and MCF7 B. cells were silenced with FBXW7 siRNA or with nontargeting control and treated with 0.025 µM paclitaxel during 24 h. Expression levels of the indicated proteins were assessed by western blot, using β-actin as loading control. Densitometric analyses of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) are shown as histograms. Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated siRNA control versus paclitaxel-treated siRNA FBXW7. p < 0.05 from Student’s t-test. C. MDA-MB-468 cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms and data from triplicates are presented as mean ± SEM. wnregulation induces mitotic arrest and resistance to paclitaxel in MDA-MB-468 and MCF www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52755 paclitaxel-treated MCL1-silenced cells (Figure 5A-5B and Supplementary Figure S5). These results may indicate that PLK1 is more relevant for the regulation of slippage than MCL1 after paclitaxel treatment. To analyze the relevance of each protein in slippage, we have analyzed Cyclin B1 and p-His H3 in PLK1- silenced or MCL1-silenced MDA-MB-468R cells after treatment with DMSO or paclitaxel for 24h. In PLK1- silenced cells paclitaxel treatment decreased CyclinB1 and p-His H3 after 24h while no differences were observed in Oncota get 52756 www impactjournals com/oncotarget we have analyzed Cyclin B1 and p-His H3 in PLK1- silenced or MCL1-silenced MDA-MB-468R cells after treatment with DMSO or paclitaxel for 24h. In PLK1- silenced cells paclitaxel treatment decreased CyclinB1 and p-His H3 after 24h while no differences were observed in Supplementary Figure S5). These results may indicate that PLK1 is more relevant for the regulation of slippage than MCL1 after paclitaxel treatment. Figure 3: FBXW7 downregulation induces mitotic arrest and resistance to paclitaxel in MDA-MB-468 and MCF7 cells. MDA-MB-468 A. and MCF7 B. cells were silenced with FBXW7 siRNA or with nontargeting control and treated with 0.025 µM paclitaxel during 24 h. Expression levels of the indicated proteins were assessed by western blot, using β-actin as loading control. Densitometric analyses of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) are shown as histograms. Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated siRNA control versus paclitaxel-treated siRNA FBXW7. p < 0.05 from Student’s t-test. C. MDA-MB-468 cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms and data from triplicates are presented as mean ± SEM. Figure 3: FBXW7 downregulation induces mitotic arrest and resistance to paclitaxel in MDA-MB-468 and MCF7 cells. MDA-MB-468 A. and MCF7 B. cells were silenced with FBXW7 siRNA or with nontargeting control and treated with 0.025 µM paclitaxel during 24 h. Expression levels of the indicated proteins were assessed by western blot, using β-actin as loading control. Densitometric analyses of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) are shown as histograms. Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated siRNA control versus paclitaxel-treated siRNA FBXW7. p < 0.05 from Student’s t-test. C. MDA-MB-468 cell cycle analysis of propidium iodide-stained cells by flow cytometry. Loss of FBXW7 correlates with substrate accumulation and adverse prognosis markers in breast cancer tissues characterizing all samples for their FBXW7 levels, we analyzed the correlation with clinicopathological variables (Table 1). Our data show that the loss of FBXW7 expression correlated with adverse prognostic markers, e.g. high tumor grade, lack of hormone receptors or HER2 positivity. To further investigate whether FBXW7 expression correlated with its substrates in breast cancer tissues, we decided to study the protein levels of Cyclin E, MCL1, AURKA and PLK1 in these samples (Figure 6). AURKA and MCL1 were located in the cytoplasm, Cyclin E showed a nuclear pattern and PLK1 cytoplasmic and nuclear location. FBXW7 showed a statistically significant correlation with the expression of Cyclin E, MCL1 and To analyze FBXW7 expression and its relation to the relative abundance of substrates, and with prognostic variables in breast cancer, we performed immunohistochemistry for FBXW7, AURKA, Cyclin E, MCL1 and PLK1 in tissue microarrays containing up to 296 samples of breast carcinomas. FBXW7 was present in nuclear location of 149 tumors (50.3%) and was diminished or absent in 147 tumors (49.7%). After Oncotarget 52757 www impactjournals com/oncotarget Figure 4: FBXW7 overexpression restores sensitivity to paclitaxel in resistant MDA-MB-468R cells. Cells were transiently transfected with pCMVHA-FBXW7 or with empty vector and treated with 0.025 µM paclitaxel during 24 h. A. FBXW7, HA, AURKA, c-Myc, Cyclin E, PLK1 and MCL1 proteins were detected by western blot. β-actin is shown as loading control. B. Western blot analysis of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) is shown; β-actin was used to ensure equal loading. Densitometric analyses of cleaved PARP, cleaved CASP-9 and cleaved CASP-3 are shown as histograms. Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated empty vector versus paclitaxel-treated HA-FBXW7. p < 0.05 from Student’s t-test. C. Cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms. Figure 4: FBXW7 overexpression restores sensitivity to paclitaxel in resistant MDA-MB-468R cells. Cells were transiently transfected with pCMVHA-FBXW7 or with empty vector and treated with 0.025 µM paclitaxel during 24 h. A. FBXW7, HA, AURKA, c-Myc, Cyclin E, PLK1 and MCL1 proteins were detected by western blot. β-actin is shown as loading control. B. Western blot analysis of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) is shown; β-actin was used to ensure equal loading. Densitometric analyses of cleaved PARP, cleaved CASP-9 and cleaved CASP-3 are shown as histograms. www.impactjournals.com/oncotarget Oncotarget 52756 Loss of FBXW7 correlates with substrate accumulation and adverse prognosis markers in breast cancer tissues Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated empty vector versus paclitaxel-treated HA-FBXW7. p < 0.05 from Student’s t-test. C. Cell cycle analysis of propidium iodide-stained cells by flow cytometry. Quantification of each phase and sub-G1 cells are shown as histograms. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52757 Figure 5: Silencing PLK1 or MCL1 restores sensitivity to paclitaxel in resistant MDA-MB-468R cells. Cells were silenced A. with siRNA MCL1 or B. with siRNA PLK1 or with nontargeting controls and treated with 0.025 µM paclitaxel during 24 h. Apoptosis induction was measured by appearance of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) by western blot. MCL1, PLK1, Cyclin B1 and p-His H3 were also assessed by western blot, using β-actin as a loading control. Densitometric analysis of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) are shown as histograms. Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated siRNA control versus paclitaxel-treated siRNA MCL1 or siRNA PLK1. p < 0.05 and p < 0.01 from Student’s t-test. K1 or MCL1 restores sensitivity to paclitaxel in resistant MDA-MB-468R cells. Cells were silence Figure 5: Silencing PLK1 or MCL1 restores sensitivity to paclitaxel in resistant MDA-MB-468R cells. Cells were silenced A. with siRNA MCL1 or B. with siRNA PLK1 or with nontargeting controls and treated with 0.025 µM paclitaxel during 24 h. Apoptosis induction was measured by appearance of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) by western blot. ure 5: Silencing PLK1 or MCL1 restores sensitivity to paclitaxel in resistant MDA-MB-468R ce Figure 5: Silencing PLK1 or MCL1 restores sensitivity to paclitaxel in resistant MDA-MB-468R cells. Cells were silenced A. with siRNA MCL1 or B. with siRNA PLK1 or with nontargeting controls and treated with 0.025 µM paclitaxel during 24 h. Apoptosis induction was measured by appearance of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) by western blot. MCL1, PLK1, Cyclin B1 and p-His H3 were also assessed by western blot, using β-actin as a loading control. Densitometric analysis of cleaved PARP, cleaved caspase-9 (CASP-9) and cleaved caspase-3 (CASP-3) are shown as histograms. Data from triplicates are presented as mean ± SEM comparing paclitaxel-treated siRNA control versus paclitaxel-treated siRNA MCL1 or siRNA PLK1. p < 0.05 and **p < 0.01 from Student’s t-test. Figure 6: Immunohistochemical expression of FBXW7 and substrates in breast carcinomas. Representative micrographs of FBXW7, Cyclin E, AURKA, MCL1 and PLK1 in breast carcinomas, tumor grade 1 to 3 (top to bottom), are shown. Scale bars, 100 µm. Figure 6: Immunohistochemical expression of FBXW7 and substrates in breast carcinomas. www.impactjournals.com/oncotarget Representative micrographs of FBXW7, Cyclin E, AURKA, MCL1 and PLK1 in breast carcinomas, tumor grade 1 to 3 (top to bottom), are shown. Scale bars, 100 µm. www.impactjournals.com/oncotarget Oncotarget 52758 Table 1: FBXW7 expression in breast carcinomas and its relationship with clinicopathological variables and substrates. FBXW7 Low High p-valuea Number (%) Number (%) Tumor <2 cm ≥2 cm 119 28 (81.0) (19.0) 133 16 (89.3) (10.7) 0.044 Node Negative 77 (52.4) 93 (62.4) 0.080 Positive 70 (47.6) 56 (37.6) Grade 1 + 2 3 49 98 (33.3) (66.7) 107 42 (71.8) (28.2) <0.0001 ER Negative Positive 59 88 (40.1) (59.9) 26 123 (17.4) (82.6) <0.0001 PR Negative Positive 75 72 (51.0) (49.0) 39 110 (26.2) (73.8) <0.0001 HER2 Negative Positive 114 32 (78.1) (21.9) 139 10 (93.3) (6.7) 0.0002 Ki67 Negative Positive 33 114 (22.4) (77.6) 52 97 (34.9) (65.1) 0.017 PLK1 Low High 28 119 (19.0) (81.0) 103 45 (69.6) (30.4) < 0.0001 MCL1 Low High 44 103 (29.9) (70.1) 124 24 (83.8) (16.2) < 0.0001 Cyclin E Low High 40 105 (27.6) (72.4) 85 64 (57.0) (43.0) < 0.0001 AURKA Low High 63 84 (42.9) (57.1) 69 78 (46.9) (53.1) 0.481 ER, estrogen receptor; PR, progesterone receptor ap-values were obtained from χ2 test Table 1: FBXW7 expression in breast carcinomas and its relationship with clinicopathological variables and substrates. PLK1, and the expression levels of these three proteins increased concordantly in high grade tumors. PLK1, and the expression levels of these three proteins increased concordantly in high grade tumors. FEC (5-Fluorouracil, epirubicin, cyclophosphamide; n = 41) or FEC with paclitaxel (FECP; n = 44). Diagnostic biopsies of breast cancer patients were analyzed by immunohistochemistry for the expression of FBXW7, MCL1 and PLK1. We performed Kaplan-Meier analyses and the disease free survival was evaluated in months as the time variable. With a mean follow up time of 113.9 months after FEC treatment, disease progression was demonstrated in 30.0%, 25.0% and 20.7% patients with low FBXW7, MCL1 or PLK1 expression, respectively; and in 19%, 23.8% and 33.3% patients with high FBXW7, MCL1 or PLK1 expression. The mean time for achieving disease progression was 97.6 months, 118.9 months Immunohistochemical expression levels of FBXW7, MCL1 and PLK1 in cancer patients are associated with paclitaxel therapy response To assess the relationship between FBXW7, MCL1 and PLK1 expression and the response to chemotherapy in breast cancer, we performed an analysis of FBXW7, MCL1 and PLK1 expression in two groups of node-negative breast cancer patients treated with www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52759 and 119.3 months in patients with low FBXW7, MCL1 or PLK1-expressing tumors; and 122.9 months, 97.6 months and 91.0 months in patients with high FBXW7, MCL1 or PLK1-expressing tumors. Kaplan-Meier curves were not significantly different (p = 0.241, p = 0.670 and p = 0.192 from Mantel and Cox, log-rank test; χ2 = 1.377, χ2 = 0.182, χ2 = 1.7; 1 degree of freedom) between both groups for FBXW7, MCL1 and PLK1 expression, respectively (Figure 7). In contrast, at the mean follow-up of 120.6 months post FECP treatment, disease progression was demonstrated in 55.5% patients with low FBXW7 expression, 6.25% patients with low MCL1 or PLK1 expression; and in 2.8% patients with high FBXW7 expression and in 33.3% patients with high MCL1 or PLK1 expression. The mean time for achieving disease progression was 82.5 months in patients with low FBXW7-expressing tumors, and 129.5 months in patients with high FBXW7-expressing tumors. Kaplan-Meier curves were significantly different (p <0.001 from Mantel and Cox, log-rank test; χ2 = 20.972; 1 degree of freedom) between both groups. The mean time for achieving disease and 119.3 months in patients with low FBXW7, MCL1 or PLK1-expressing tumors; and 122.9 months, 97.6 months and 91.0 months in patients with high FBXW7, MCL1 or PLK1-expressing tumors. Kaplan-Meier curves were not significantly different (p = 0.241, p = 0.670 and p = 0.192 from Mantel and Cox, log-rank test; χ2 = 1.377, χ2 = 0.182, χ2 = 1.7; 1 degree of freedom) between both groups for FBXW7, MCL1 and PLK1 expression, respectively (Figure 7). In contrast, at the mean follow-up of 120.6 months post FECP treatment, disease progression was demonstrated in 55.5% patients with low FBXW7 expression, 6.25% patients with low MCL1 or PLK1 expression; and in 2.8% patients with high FBXW7 expression and in 33.3% patients with high MCL1 or PLK1 expression. The mean time for achieving disease progression was 82.5 months in patients with low FBXW7-expressing tumors, and 129.5 months in patients with high FBXW7-expressing tumors. Kaplan-Meier curves were significantly different (p <0.001 from Mantel and Cox, log-rank test; χ2 = 20.972; 1 degree of freedom) between both groups. The mean time for achieving disease progression was 127.3 months in patients with low MCL1 or PLK1-expressing tumors and 101.2 months in patients with high MCL1 or PLK1-expressing tumors. www.impactjournals.com/oncotarget Kaplan- Meier curves were significantly different (p = 0.012 from Mantel and Cox, log-rank test; χ2 = 6.35; 1 degree of freedom) between both groups (Figure 7). DISCUSSION FBXW7 acts as a tumor suppressor involved in the degradation of substrates with oncogenic activity frequently overexpressed in breast cancer such as Cyclin E, c-Myc and AURKA [14, 29-31]. It also controls degradation of MCL1 [20, 21], which has an important role in the regulation of apoptosis, and PLK1 [28], a protein kinase involved in essential events of the cell cycle [25, 32]. Treatment with taxanes improves survival of node-negative breast cancer patients [33], but acquisition of resistance can cause this treatment to lose effectivity [15, 17]. We sought to investigate the role of FBXW7 in the response to paclitaxel in breast cancer cells and Figure 7: Expression levels of FBXW7, MCL1 and PLK1 and association with disease-free survival (DFS) in paclitaxel-treated patients. Kaplan-Meier analysis of disease-free survival in patients treated with FEC (5-Fluorouracil, epirubicin and cyclophosphamide; left) and FECP (5-Fluorouracil, epirubicin, cyclophosphamide and paclitaxel; right) showing either high (dotted line) or low (solid line) expression of FBXW7, MCL1 and PLK1. Tick marks represent censored patients. The p values were obtained from the long-rank test of Mantel and Cox. 1 and association with disease-free survival (DFS) in i l i ti t t t d ith FEC (5 Fl il i bi i d Figure 7: Expression levels of FBXW7, MCL1 and PLK1 and association with disease-free survival (DFS) in paclitaxel-treated patients. Kaplan-Meier analysis of disease-free survival in patients treated with FEC (5-Fluorouracil, epirubicin and cyclophosphamide; left) and FECP (5-Fluorouracil, epirubicin, cyclophosphamide and paclitaxel; right) showing either high (dotted line) or low (solid line) expression of FBXW7, MCL1 and PLK1. Tick marks represent censored patients. The p values were obtained from the long-rank test of Mantel and Cox. www.impactjournals.com/oncotarget This also occurred when FBXW7 mRNA is interfered in paclitaxel-sensitive MDA-MB-468 cells. Sensitivity to paclitaxel-induced apoptosis can be restored in resistant cells by either reintroducing FBXW7 expression or interfering MCL1. Furthermore, we demonstrated that PLK1 is also involved in paclitaxel resistance because decreasing the levels of PLK1 by siRNA sensitized resistant breast cancer cells to treatment with paclitaxel. In this sense, PLK1 levels have been related to improved sensitivity to paclitaxel and trastuzumab when combined with siRNA targeting PLK1 in breast cancer cell lines in a synergistic way [38]. We noticed that the paclitaxel-resistant subclone had aneuploid DNA content. This genetic instability could be related to FBXW7 loss as it contributes to the accumulation of Cyclin E, AURKA, BubR1 and PLK1. These findings are consistent with those reported by Finkin et al., where FBXW7-deficient cells showed aneuploidy associated with increased Cyclin E and AURKA, and that AURKA could mediate the induction of aneuploidy by the p53-LATS2 pathway [39]. Moreover, PLK1 contributes to SAC regulation [42], and components of the SAC, such as BubR1, are associated with chromosomal instability [43, 44]. Dysregulated activity of PLK1 results in abnormalities in microtubules and increased centrosome size and number, and both phenomena have been related to aneuploidy [45]. Therefore, FBXW7 could be contributing to increased genetic instability through the regulation of PLK1 since FBXW7-deficient cells have higher levels of PLK1. We show that paclitaxel resistant cells are aneuploid and accumulate MCL1 and PLK1 as well as in sensitive cells silenced for FBXW7. In cells silenced for FBXW7, paclitaxel treatment arrested cells more efficiently in mitosis, thus preventing the degradation of both substrates. These results point to the involvement of FBXW7 in the control of the degradation of these substrates in breast cancer cells and in response to apoptosis in cells arrested in mitosis. PLK1 needs to be degraded by the APC/CCDH1- proteasome in late anaphase to adequately control the exit from mitosis and cytokinesis so that cells can enter in G1 phase [25], but as we see in Figure 3, BubR1 is increased in cells silenced for FBXW7 so the SAC remains active. This could contribute to the inactivation of APC/Cdc20, thus preventing the degradation of PTTG1, Cyclin B1 and PLK1. Elowe et al. demonstrated that diminished BubR1 contributes to cells escaping by slippage [24]. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52760 in patients with node-negative disease. Commonly used breast cancer cell lines have different responses to paclitaxel [18]. We chose basal-like phenotype, paclitaxel- sensitive MDA-MB-468 cells that express abundant FBXW7 to study the role of this protein in mitotic arrest and apoptosis induced by paclitaxel, with special focus on the regulation of MCL1 and PLK1. We observed that paclitaxel-resistant subclonal MDA-MB-468R cells had lower levels of FBXW7 and accumulated PLK1 and MCL1 to impair mitosis progress and apoptosis after paclitaxel treatment. The silencing of FBXW7 in sensitive MDA-MB-468 and MCF7 cells induced the accumulation of both substrates and disabled apoptosis in response to paclitaxel. This prompted us to consider FBXW7 as an important mediator of the paclitaxel response through the control of MCL1 and PLK1 degradation in breast cancer. mitotic arrest with drugs that inhibit mitotic slippage can lead to increased cell survival and proliferation, whereas inducing mitotic slippage in cells treated with antimitotic drugs seems to lead to cell death [40]. We previously observed that slippage-prone PC3 cells suffered less efficient apoptosis when they were arrested in mitosis and MCL1 was not properly degraded [22]. Therefore, MCL1 degradation has great importance in the balance between mitosis and apoptosis [41]. In this sense, the control of MCL1 degradation by FBXW7 seems determinant and we observed that FBXW7 transfection in paclitaxel resistant cells decreased the levels of MCL1 and restored sensitivity to apoptosis. Also, paclitaxel-resistant cells transfected with FBXW7 were prone to mitotic slippage as noted by the downregulation of Cyclin B1, PTTG1 and BubR1, and thus they recapitulated the behavior of paclitaxel-sensitive cells. MCL1 is an anti-apoptotic protein involved in resistance to cancer therapies such as cisplatin in lung cancer [34] or rituximab in lymphomas [35, 36]. MCL1 is also relevant in the response to anti-mitotic agents because cell death in mitosis may depend on MCL1 phosphorylation by CDK1-Cyclin B1 complex and its subsequent degradation [37]. Wertz et al. showed that MCL1 was stabilized in FBXW7-deficient cells arrested in mitosis and that JNK/p38/CKII kinases were regulators of the MCL1 degradation. These authors also showed that sensitivity to anti-mitotic agents was regulated by FBXW7 through degradation of MCL1 in colon and ovarian cancer cells. We observed that resistant MDA- MB-468R cells, which are FBXW7-deficient, arrest in mitosis after paclitaxel treatment but MCL1 levels remain high and apoptosis diminishes. www.impactjournals.com/oncotarget We observed a similar response in the paclitaxel sensitive cells, where BubR1 and PLK1 levels are low, and taxol treatment induced the cells to exit mitosis by slippage with subsequent apoptosis. Consistent with these After anti-mitotic treatments, cells may die in mitosis, or may either undergo slippage and then die, stay in G0, or enter a new cell cycle with increased ploidy [15]. We observed that paclitaxel-sensitive MDA-MB-468 cells die after slippage while paclitaxel-resistant MDA- MB-468R cells survived, probably after slippage, and continued cycling into an aneuploid state. These results are in concordance with those of Finkin et al., who reported that FBXW7-deficient cells are more likely to escape mitosis after anti-mitotic treatments and accumulate genetic errors [39]. Whether mitotic slippage constitutes a cell survival strategy or a cell death mechanism is not yet well understood. Riffell et al. showed that reinforcing www.impactjournals.com/oncotarget Oncotarget 52761 serum, 50 U/mL penicillin, 50 µg/mL streptomycin, 10 mmol/L HEPES buffer and 1 mM glutamine in a 37ºC humidified incubator under 5% CO2. Subconfluent cell cultures were harvested by trypsinization. All the experiments were performed using cells that had not exceeded the first ten passages after receipt of the initial vial. Stock solution of paclitaxel (Calbiochem) was prepared at 10 mM in dimethyl sulfoxide (DMSO, Sigma) and stored frozen. In all the experiments, DMSO was added to untreated cells. The paclitaxel-resistant MDA- MB-468R cell line was generated in our laboratory as follows. Parental MDA-MB-468 cells were treated with 4 nM paclitaxel for 2 months; then, the concentration of the drug in the medium was reduced by half. The surviving cells were maintained with 1 nM paclitaxel for 1 month and finally, the drug was withdrawn from the medium. The surviving cells were again subjected to the same treatment and the resulting paclitaxel-resistant cells were used for the assays. data, van Vugt et al. [26] established the role of PLK1 in the control of mitosis by regulation of the spindle checkpoint. They showed that PLK1 depleted cells arrest in prometaphase with an activated spindle checkpoint, but if these cells are BubR1 co-depleted, they fail to arrest in mitosis. Accordingly, resistant cells and cells silenced for FBXW7 accumulate both proteins, particularly PLK1, an accumulation that remains after paclitaxel treatment and is accompanied by efficient mitotic arrest with less death by apoptosis. www.impactjournals.com/oncotarget y p p We also looked at the relative expression of FBXW7 and its substrates in a series of breast carcinoma tissues to establish whether these proteins could be valuable prognostic markers and predict patient response to paclitaxel treatment. We observed that expression levels of FBXW7 correlated inversely with the levels of substrates in a statistically significant way with Cyclin E, PLK1 and MCL1. Ibusuki et al. previously reported that a decreased expression of FBXW7 was associated with increased levels of Cyclin E and c-Myc in a group of patients with breast carcinoma of a more aggressive tumor stage [9]. We see that loss of nuclear FBXW7 associates significantly with higher tumor size and grade and with HER2-positive and triple negative phenotypes. Our study also shows that loss of FBXW7 and increased levels of MCL1 and PLK1 are associated with lower disease-free survival in patients whose treatment includes paclitaxel. We conclude that FBXW7 can be a marker of prognostic value and that the patient’s response to paclitaxel is most effective in slowing disease progression when FBXW7 is present and MCL1 and PLK1 do not accumulate. Small interfering RNA (siRNA) and plasmid transfections Validated pools of FBXW7α-, MCL1-, PLK1- siRNA and non-targeting siRNA as negative control were obtained from GE Dharmacon (ON-TARGETplus SMART pools L-004264, L-004501, L-003290, and D-001810). Transfections were carried out using DharmaFECT 2 reagent (GE Dharmacon) according to manufacturer’s instructions. All siRNA pools were used at 50 nM. Cells were subjected to different treatments 24 h after silencing. Transient transfections of pCMVHA and pCMVHA-FBXW7 plasmids [28] were carried out using FuGENE reagent (Promega) according to manufacturer’s instructions. Cells were subjected to different treatments 24 h after transfection. In summary, we show that FBXW7 plays a prominent role in paclitaxel-induced apoptosis by regulating MCL1 and PLK1 levels in breast cancer, as resistant cells loose FBXW7 and accumulate these substrates. Loss of FBXW7 is associated with adverse prognostic factors such as tumor grade or non-luminal phenotypes and FBXW7, MCL1 and PLK1 levels may have predictive value in patients treated with paclitaxel. Reintroducing FBXW7 activity or interfering MCL1 and PLK1 restores paclitaxel sensitivity in resistant cells and thus, targeting these substrates may be a valuable adjunct to paclitaxel for the treatment of breast cancer. Notably, for the first time, we observed a direct association between FBXW7 and PLK1 expression in breast cancer tissues. MATERIALs AND METHODS MATERIALs AND METHODS Cell culture MCF7, T47D, MDA-MB-231, MDA-MB-468, SKBR3 and BT474 breast cancer cell lines from the Interlab Cell Line Collection (Genoa, Italy) were routinely grown in RPMI 1640 supplemented with 10% fetal bovine Antibodies The antibodies and dilutions used for western blots were as follows: rabbit polyclonal anti-FBXW7 (1:500) and anti-AURKA (1:2,000) from Novus; rabbit polyclonal anti-MCL1 (S-19, 1:1,000), anti-Cyclin B1 (1:1,000), anti-phospho-histone H3 (Ser10, 1:500) and mouse monoclonal anti-PTTG1 (1:1,000) from Santa Cruz; mouse monoclonal anti-β-actin (Ac15, 1:20,000) and rabbit polyclonal anti-c-Myc (1:1,000) from Sigma; rabbit polyclonal anti-cleaved caspase-9 (Asp 315, 1:500), anti-cleaved caspase-3 (Asp 175, 1:500) from Cell Signaling; rabbit polyclonal anti-BubR1 (1:3,000) from Bethyl; mouse monoclonal anti-PARP (1:500) from BD Bioscience; mouse monoclonal anti-Cyclin E (1:500) from Monosan; mouse monoclonal anti-PLK1 (1:10,000) from Millipore, and rat monoclonal anti-HA-peroxidase (50 mU/mL) from Roche. For immunohistochemical www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 52762 analyses, anti- FBXW7 (1:2,000), anti- Cyclin E (1:100), anti- MCL1 (1:1,250), anti-AURKA (1:700), and anti- PLK1 (1:1,000) were used. for 5 min and hybridized at 37ºC overnight in a humid chamber (Thermomixer, Eppendorf). After hybridization, slides were washed in 2x SSC/0.3% NP40 solution at 72ºC for 5 min, counterstained, mounted with DAPI/antifade solution (Vysis), and visualized under fluorescence microscopy equipped with appropriate filter sets and a digital camera (Leica). At least one hundred cells were counted to calculate the percentage of cells with normal ploidy and higher ploidy in each condition. Western blotting Cells were lysed in Nonidet P-40 (NP40) lysis buffer [10 mM Tris-HCl (pH 7.5), 150 mM NaCl, 10% glycerol, and 1% NP40]. Twenty micrograms of total protein, as determined by the BCA protein assay kit (Pierce), were separated by SDS-PAGE on 8% polyacrylamide gels and electroblotted onto nitrocellulose membranes (GE Healthcare). Membranes were stained with Ponceau S to ensure that protein amounts were comparable. For immunodetection, blots were blocked in 1% blocking reagent (Roche) in 0.05% Tween 20-phosphate buffered saline (PBS) for 1 hour and incubated with primary antibody diluted in blocking buffer overnight at 4ºC. Blots were then washed in 0.05% Tween 20-PBS and incubated with either goat anti-mouse (1:20,000; GE Healthcare) or goat anti-rabbit (1:20,000; GE Healthcare) peroxidase-labeled antibodies in blocking buffer for 1 hour. The enhanced chemoluminescent system was applied according to the manufacturer’s protocol (GE Healthcare). The experiments were performed at least three times and densitometric analysis was performed using ImageJ software (http://imagej.nih.gov/ij/). Arbitrary densitometric units of the proteins of interest were corrected for those of β-actin. Data comparing differences between the two conditions were statistically analyzed, when indicated, using paired Student’s t-test. Statistical analyses were performed with Prism 4.0 software (GraphPad). Differences were considered as significant when p < 0.05. Cytotoxicity assay 6 x 103 cells/well in exponential cell growth were plated in 96-well plates (Nunc). Following cell adherence (growth for 24 h), experimental medium containing paclitaxel was added to triplicate wells, and serial dilutions were performed to span the dose range suitable for the analysis. Cells were exposed to the drug for 72 h and a cell viability assay was performed using alamarBlue Cell Viability Assay Reagent (Thermo Scientific) according to manufacturer’s instructions. Fluorescence was measured with excitation wavelength at 545 nm and emission wavelength at 590 nm with FLx800 Microplate Fluorescence Reader (Bio-Tek). Data were normalized to vehicle treatment, and the half inhibitory concentrations (IC50) were calculated using OriginPro 8 and Prism 4.0 software (GraphPad). The experiments were performed at least three times. Immunohistochemical staining and scoring Formalin-fixed, paraffin-embedded tissue blocks from 296 patients with invasive breast carcinoma were selected to build up tissue microarrays with 1 mm cores. The study was approved by the Ethical Committee of the hospital. Five µm tissue sections were dewaxed, rehydrated, and immersed in 3% H2O2 aqueous solution for 30 minutes to exhaust endogenous peroxidase. Heat-induced epitope retrieval was performed with 1 mM EDTA (pH 9.0) in a microwave oven. Sections were incubated overnight at 4ºC with the primary antibodies. Peroxidase-labeled secondary antibodies and 3,3’-diaminobenzidine were applied according to manufacturer’s protocol (EnVision, Dako). Slides were then counterstained with hematoxylin and mounted. Sections where the primary antibody was omitted were used as negative controls. Immunostains were scored as low (<25%) or high expression (≥25%) according to the extent of positive cells. Correlations between proteins and clinicopathological variables were analysed by χ2 test using Prism 4.0 software (GraphPad). Differences were considered as significant when p < 0.05. The association of FBXW7, MCL1 and PLK1with paclitaxel therapy response was analyzed by the method of Kaplan-Meier. Flow cytometric analysis of cell cycle Cells were trypsinized and fixed in 70% ethanol. Propidium iodide staining of nuclei was performed with a CycleTest Plus DNA reagent kit (BD Biosciences), and the DNA content was measured with a FACScan instrument (BD Biosciences). Data were acquired with CellQuest Pro software (BD Biosciences). Fluorescence in situ hybridization (FISH) Cultured cells were imprinted onto silanized slides, fixed in ice-cold methanol/glacial acetic acid (3:1) for 10 min and air-dried. Then, slides were immersed in a 2x SSC/0.3% NP40 solution at 37ºC during 30 min, dehydrated in an increasing series of graded ethanols and air-dried. Cellular DNA and centromeric probes for chromosomes 8 (Spectrum red), 11 (Spectrum green) and 17 (Spectrum aqua) from Vysis were co-denatured at 72ºC www.impactjournals.com/oncotarget GRANT SUPPORT This work was supported by research grants from the Instituto de Salud Carlos III (FIS PI13/02282), Ministerio de Economía y Competitividad (SAF2014- 53799-C2-1/2-R), Consejería de Salud (AI-0025-2015) and Consejería de Innovación, Ciencia y Empresa (P10- CTS-6243), Junta de Andalucía. ML Flores was supported by a pre-doctoral grant from the Consejería de Innovación, Ciencia y Empresa, Junta de Andalucía (CTS-6243). C Sáez was supported by a contract from Nicolás Monardes Program, Consejería de Salud, Junta de Andalucía. 11. Meyer N, Penn LZ. Reflecting on 25 years with MYC. Nat Rev Cancer. 2008; 8:976–990. 12. Yao J, Yan M, Guan Z, Pan C-B, Xia L-P, Li C-X, Wang L-H, Long Z-J, Zhao Y, Li M-W, Zheng F-M, Xu J, Lin D-J, Liu Q. Aurora-A down-regulates IkappaBα via Akt activation and interacts with insulin-like growth factor-1 induced phosphatidylinositol 3-kinase pathway for cancer cell survival. Mol Cancer 2009; 8:95. 13. Do T, Xiao F, Bickel L, Klein-Szanto A. Aurora kinase A mediates epithelial ovarian cancer cell migration and adhesion. Oncogene. 2014; 33:539–549. www.impactjournals.com/oncotarget Oncotarget 52763 No potential conflicts of interest were disclosed. 10. Dang CV, O’Donnell K A, Zeller KI, Nguyen T, Osthus RC, Li F. The c-Myc target gene network. Semin Cancer Biol. 2006; 16:253–264. ACKNOWLEDGMENTS We thank B. Marquez-Garrido for her support. 9. Ibusuki M, Yamamoto Y, Shinriki S, Ando Y, Iwase H. Reduced expression of ubiquitin ligase FBXW7 mRNA is associated with poor prognosis in breast cancer patients. Cancer Sci. 2011; 102:439–445. Abbreviations hypermethylation is associated with favorable prognosis in primary breast cancer. Breast Cancer Res. 2010; 12:R105. 7. 7. Spruck CH, Strohmaier H, Sangfelt O, Müller HM, Hubalek M, Müller-Holzner E, Marth C, Widschwendter M, Reed SI. hCDC4 Gene Mutations in Endometrial Cancer. Cancer Res. 2002; 62:4535–4539. DAPI, 4´,6-Diamidino-2-phenyl-indole; DMSO, Dimethyl sulfoxide; NP40, Nonidet-P40; PARP, Poly (ADP-ribose) polymerase. 8. Ekholm-Reed S, Spruck CH, Sangfelt O, van Drogen F, Müeller-Holzner E, Widschwendter M, Zetterberg A, Reed SI. Mutation of hCDC4 Leads to Cell Cycle Deregulation of Cyclin E in Cancer. Cancer Res. 2004; 64:795–800. Conflicts of Interest No potential conflicts of interest were disclosed. REFERENCES 14. Xu J, Wu X, Zhou W, Liu A, Wu J, Deng J, Yue C, Yang S, Wang J, Yuan Z, Liu Q. Aurora-A identifies early recurrence and poor prognosis and promises a potential therapeutic target in triple negative breast cancer. PLoS One. 2013; 8:e56919. 1. Welcker M, Clurman BE. FBW7 ubiquitin ligase: a tumour suppressor at the crossroads of cell division, growth and differentiation. Nat Rev Cancer. 2008; 8:83–93. 2. Grim JE, Gustafson MP, Hirata RK, Hagar AC, Swanger J, Welcker M, Hwang HC, Ericsson J, Russell DW, Clurman BE. Isoform- and cell cycle-dependent substrate degradation by the Fbw7 ubiquitin ligase. J Cell Biol. 2008; 181:913–920. 15. Gascoigne KE, Taylor SS. How do anti-mitotic drugs kill cancer cells? J Cell Sci. 2009; 122:2579–2585. 16. Paridaens R, Biganzoli L, Bruning P, Klijn JG, Gamucci T, Houston S, Coleman R, Schachter J, Van Vreckem A, Sylvester R, Awada A, Wildiers J, Piccart M. Paclitaxel Versus Doxorubicin as First-Line Single-Agent Chemotherapy for Metastatic Breast Cancer: A European Organization for Research and Treatment of Cancer Randomized Study With Cross-Over. J Clin Oncol. 2000; 18:724–733. 3. Tan Y, Sangfelt O, Spruck C. The Fbxw7/hCdc4 tumor suppressor in human cancer. Cancer Lett. 2008; 271:1–12. 4. Crusio KM, King B, Reavie LB, Aifantis I. The ubiquitous nature of cancer: the role of the SCF(Fbw7) complex in development and transformation. Oncogene. 2010; 29:4865–4873. 17. McGrogan BT, Gilmartin B, Carney DN, McCann A. Taxanes, microtubules and chemoresistant breast cancer. Biochim Biophys Acta. 2008; 1785:96–132. 5. Akhoondi S, Sun D, Von Der Lehr N, Apostolidou S, Klotz K, Maljukova A, Cepeda D, Fiegl H, Dofou D, Marth C, Mueller-Holzner E, Corcoran M, Dagnell M, Nejad SZ, Nayer BN, Zali MR, Hansson J, Egyhazi S, Petersson F, Sangfelt P, Nordgren H, Grander D, Reed SI, Widschwendter M, Sangfelt O, Spruck C. FBXW7/hCDC4 is a general tumor suppressor in human cancer. Cancer Res. 2007; 67:9006–9012. 18. Flores ML, Castilla C, Ávila R, Ruiz-Borrego M, Sáez C, Japón MA. Paclitaxel sensitivity of breast cancer cells requires efficient mitotic arrest and disruption of Bcl-xL/ Bak interaction. Breast Cancer Res Treat. 2012; 133:917– 928. 19. Huang DC, Strasser A. BH3-Only proteins-essential initiators of apoptotic cell death. Cell. 2000; 103:839–842. 6. Akhoondi S, Lindström L, Widschwendter M, Corcoran M, Bergh J, Spruck C, Grandér D, Sangfelt O. Inactivation of FBXW7/hCDC4-β expression by promoter 20. REFERENCES Inuzuka H, Shaik S, Onoyama I, Gao D, Tseng A, Maser 20. Inuzuka H, Shaik S, Onoyama I, Gao D, Tseng A, Maser 20. Inuzuka H, Shaik S, Onoyama I, Gao D, Tseng A, Maser www.impactjournals.com/oncotarget Oncotarget 52764 S, Calvo L, Margdí Vila M, Anton A, Rodríguez-Lescure A, Seguí-Palmer MA, Muñoz-Mateu M, Dorea Ribugent J, López-Vega JM, Jara C et al. Fluorouracil, Doxorubicin, and Cyclophosphamide (FAC) Versus FAC Followed by Weekly Paclitaxel As Adjuvant Therapy for High- Risk, Node-Negative Breast Cancer: Results From the GEICAM/2003-02 Study. J Clin Oncol. 2013; 31:2593– 2599. S, Calvo L, Margdí Vila M, Anton A, Rodríguez-Lescure A, Seguí-Palmer MA, Muñoz-Mateu M, Dorea Ribugent J, López-Vega JM, Jara C et al. Fluorouracil, Doxorubicin, and Cyclophosphamide (FAC) Versus FAC Followed by Weekly Paclitaxel As Adjuvant Therapy for High- Risk, Node-Negative Breast Cancer: Results From the GEICAM/2003-02 Study. J Clin Oncol. 2013; 31:2593– 2599. RS, Zhai B, Wan L, Gutierrez A, Lau AW, Christie AL, Aster J, Settleman J et al. SCF(Fbw7) regulates cellular apoptosis by targiting MCL1 for ubiquitination and destruction. Nature. 2011; 471:104–109. RS, Zhai B, Wan L, Gutierrez A, Lau AW, Christie AL, Aster J, Settleman J et al. SCF(Fbw7) regulates cellular apoptosis by targiting MCL1 for ubiquitination and destruction. Nature. 2011; 471:104–109. 21. Wertz IE, Kusam S, Lam C, Okamoto T, Sandoval W, Anderson DJ, Helgason E, Ernst JA, Eby M, Liu J, Belmont LD, Kaminker JS, O’Rourke KM et al. Sensitivity to antitubulin chemotherapeutics is regulated by MCL1 and FBW7. Nature. 2011; 471:110–114. 34. Michels J, Obrist F, Vitale I, Lissa D, Garcia P, Behnam- Motlagh P, Kohno K, Wu GS, Brenner C, Castedo M, Kroemer G: MCL-1 dependency of cisplatin-resistant cancer cells. Biochem Pharmacol 2014, 92:55–61. 22. Castilla C, Flores ML, Medina R, Pérez-Valderrama B, Romero F, Tortolero M, Japón MA, Saéz C. Prostate cancer cell response to paclitaxel is affected by abnormally expressed securin PTTG1. Mol Cancer Ther. 2014; 13:2372–2383. 35. Bannerji R, Kitada S, Flinn IW, Pearson M, Young D, Reed JC, Byrd JC. Apoptotic-Regulatory and Complement- Protecting Protein Expression in Chronic Lymphocytic Leukemia: Relationship to In Vivo Rituximab Resistance. J Clin Oncol. 2003; 21:1466–1471. 23. Barr FA, Sillje HHW, Nigg EA. Polo-like kinases and the orchestration of cell division. Nat Rev Mol Cell Biol. 2004; 5:429–441. 36. Hussain S-RA, Cheney CM, Johnson AJ, Lin TS, Grever MR, Caligiuri MA, Lucas DM, Byrd JC. REFERENCES Mcl-1 Is a Relevant Therapeutic Target in Acute and Chronic Lymphoid Malignancies: Down-Regulation Enhances Rituximab-Mediated Apoptosis and Complement- Dependent Cytotoxicity. Clin Cancer Res. 2007; 13:2144– 2150. 24. Elowe S, Hümmer S, Uldschmid A, Li X, Nigg EA. Tension-sensitive Plk1 phosphorylation on BubR1 regulates the stability of kinetochore microtubule interactions. Genes Dev. 2007; 21:2205–2219. 25. Lindon C, Pines J. Ordered proteolysis in anaphase inactivates Plk1 to contribute to proper mitotic exit in human cells. J Cell Biol. 2004; 164:233–241. 37. Harley ME, Allan LA, Sanderson HS, Clarke PR. Phosphorylation of Mcl-1 by CDK1–cyclin B1 initiates its Cdc20-dependent destruction during mitotic arrest. EMBO J. 2010; 29:2407–2420. 26. van Vugt MA, van de Weerdt BCM, Vader G, Janssen H, Calafat J, Klompmaker R, Wolthuis RMF, Medema RH. Polo-like kinase-1 is required for bipolar spindle formation but is dispensable for anaphase promoting complex/Cdc20 activation and initiation of cytokinesis. J Biol Chem. 2004; 279:36841–36854. 38. Spänkuch B, Kurunci-Csacsko E, Kaufmann M, Strebhardt K. Rational combinations of siRNAs targeting Plk1 with breast cancer drugs. Oncogene. 2007; 26:5793–5807. 27. Archambault V, Lépine G, Kachaner D. Understanding the Polo Kinase machine. Oncogene. 2015; 34:4799–4807. 39. Finkin S, Aylon Y, Anzi S, Oren M, Shaulian E. Fbw7 regulates the activity of endoreduplication mediators and the p53 pathway to prevent drug-induced polyploidy. Oncogene. 2008; 27:4411–4421. 28. Giráldez S, Herrero-Ruiz J, Mora-Santos M, Japón MA, Tortolero M, Romero F. SCF(FBXW7α) modulates the intra-S-phase DNA-damage checkpoint by regulating Polo like kinase-1 stability. Oncotarget. 2014; 5:4370–4383. doi: 10.18632/oncotarget.2021. 40. Riffell JL, Zimmerman C, Khong A, McHardy LM, Roberge M. Effects of chemical manipulation of mitotic arrest and slippage on cancer cell survival and proliferation. Cell Cycle. 2009; 8:3025–3038. 29. Shaye A, Sahin A, Hao Q, Hunt K, Keyomarsi K, Bedrosian I. Cyclin E deregulation is an early event in the development of breast cancer. Breast Cancer Res Treat. 2009; 115:651– 659. 41. Topham CH, Taylor SS. Mitosis and apoptosis: How is the balance set? Curr Opin Cell Biol. 2013; 25:780–785. 42. Strebhardt K, Ullrich A. Targeting polo-like kinase 1 for cancer therapy. Nat Rev Cancer. 2006; 6:321–330. 30. Xu J, Chen Y, Olopade OI. MYC and Breast Cancer. Genes Cancer. 2010; 1:629–640. doi: 10.1177/1947601910378691. 43. Shichiri M, Yoshinaga K, Hisatomi H, Sugihara K, Hirata Y. Genetic and epigenetic inactivation of mitotic checkpoint genes hBUB1 and hBUBR1 and their relationship to survival. Cancer Res. 2002; 62:13–7. 31. REFERENCES Singhi AD, Cimino-Mathews A, Jenkins RB, Lan F, Fink SR, Nassar H, Vang R, Fetting JH, Hicks J, Sukumar S, De Marzo AM, Argani P. MYC gene amplification is often acquired in lethal distant breast cancer metastases of unamplified primary tumors. Mod Pathol. 2012; 25:378-387 44. Kops GJ, Weaver BA, Cleveland DW. On the road to cancer: aneuploidy and the mitotic checkpoint. Nat Rev Cancer. 2005; 5:773-785. 32. Ahonen LJ, Kallio MJ, Daum JR, Bolton M, Manke I a, Yaffe MB, Stukenberg PT, Gorbsky GJ. Polo-like kinase 1 creates the tension-sensing 3F3/2 phosphoepitope and modulates the association of spindle-checkpoint proteins at kinetochores. Curr Biol. 2005; 15:1078–1089. 45. Takai N, Hamanaka R, Yoshimatsu J, Miyakawa I. Polo- like kinases (Plks) and cancer. Oncogene. 2005; 24:287– 291. 33. Martín M, Ruiz A, Ruiz Borrego M, Barnadas A, González www.impactjournals.com/oncotarget Oncotarget 52765
https://openalex.org/W4214918998	http://publica.fraunhofer.de/bitstreams/2f949be3-124a-4d6a-85ff-1ccc27e08d61/download	English	null	Leaching of Titanium Dioxide Nanomaterials from Agricultural Soil Amended with Sewage Sludge Incineration Ash: Comparison of a Pilot Scale Simulation with Standard Laboratory Column Elution Experiments	Materials	2,022	cc-by	7,030	Citation: Meisterjahn, B.; Schröder, N.; Oischinger, J.; Hennecke, D.; Weinfurtner, K.; Hund-Rinke, K. Leaching of Titanium Dioxide Nanomaterials from Agricultural Soil Amended with Sewage Sludge Incineration Ash: Comparison of a Pilot Scale Simulation with Standard Laboratory Column Elution Experiments. Materials 2022, 15, 1853. https://doi.org/10.3390/ ma15051853 Keywords: nano titanium dioxide (nTiO2); engineered nanomaterial (ENM); sewage sludge incineration (SSI); ENM containing sewage sludge ash (SSA); leaching; column elution; agricultural use Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. materials materials Boris Meisterjahn 1 , Nicola Schröder 1, Jürgen Oischinger 2 , Dieter Hennecke 1,* , Karlheinz Weinfurtner 1 and Kerstin Hund-Rinke 1 Boris Meisterjahn 1 , Nicola Schröder 1, Jürgen Oischinger 2 , Dieter Hennecke 1,* , Karlheinz Weinfurtner 1 and Kerstin Hund-Rinke 1 1 Fraunhofer Institute for Molecular Biology and Applied Ecology IME, Auf dem Aberg 1, 57392 Schmallenberg, Germany; boris.meisterjahn@ime.fraunhofer.de (B.M.); nicola.schroeder@ime.fraunhofer.de (N.S.); karlheinz.weinfurtner@ime.fraunhofer.de (K.W.); kerstin.hund-rinke@ime.fraunhofer.de (K.H.-R.) 2 Fraunhofer Institute for Environmental, Safety and Energy Technology UMSICHT, An der Maxhütte 1, 92237 Sulzbach-Rosenberg, Germany; juergen.oischinger@umsicht.fraunhofer.de * Correspondence: dieter.hennecke@ime.fraunhofer.de; Tel.: +49-2972-302-209 2 Fraunhofer Institute for Environmental, Safety and Energy Technology UMSICHT, An der Maxhütte 1, 92237 Sulzbach-Rosenberg, Germany; juergen.oischinger@umsicht.fraunhofer.de * Correspondence: dieter.hennecke@ime.fraunhofer.de; Tel.: +49-2972-302-209 Fraunhofer Institute for Environmental, Safety and Energy Technology UMSICHT, An der Maxhütte 1, 92237 Sulzbach-Rosenberg, Germany; juergen.oischinger@umsicht.fraunhofer.de * Correspondence: dieter.hennecke@ime.fraunhofer.de; Tel.: +49-2972-302-209 92237 Sulzbach Rosenberg, Germany; juergen.oischinger@umsicht.fraunhofer.de * Correspondence: dieter.hennecke@ime.fraunhofer.de; Tel.: +49-2972-302-209 g y j g g * Correspondence: dieter.hennecke@ime.fraunhofer.de; Tel.: +49-2972-302-209 Abstract: Nanoscale titanium dioxide (nTiO2 (Hombikat UV 100 WP)) was applied to sewage sludge that was incinerated in a large-scale waste treatment plant. The incineration ash produced was applied to soil as fertilizer at a realistic rate of 5% and investigated in pilot plant simulations regarding its leaching behavior for nTiO2. In parallel, the applied soil material was subject to standard column leaching (DIN 19528) in order to test the suitability of the standard to predict the leaching of nanoscale contaminants from treated soil material. Relative to the reference material (similar composition but without nTiO2 application before incineration) the test material had a total TiO2 concentration, increased by a factor of two or 3.8 g/kg, respectively. In contrast, the TiO2 concentration in the respective leachates of the simulation experiment differed by a factor of around 25 (maximum 91.24 mg), indicating that the added nTiO2 might be signiﬁcantly mobilisable. Nanoparticle speciﬁc analysis of the leachates (spICP-MS) conﬁrmed this ﬁnding. In the standard column elution experiment the released amount of TiO2 in the percolates between test and reference material differed by a factor of 4 to 6. This was also conﬁrmed for the nTiO2 concentrations in the percolates. Results demonstrate that the standard column leaching, developed and validated for leaching prediction of dissolved contaminants, might be also capable to indicate increased mobility of nTiO2 in soil materials. However, experiments with further soils are needed to verify those ﬁndings. Leaching of Titanium Dioxide Nanomaterials from Agricultural Soil Amended with Sewage Sludge Incineration Ash: Comparison of a Pilot Scale Simulation with Standard Laboratory Column Elution Experiments 1 , Nicola Schröder 1, Jürgen Oischinger 2 , Dieter Hennecke 1,* , Karlheinz Weinfurtner 1 Rinke 1 1. Introduction Engineered nanomaterials (ENMs) applied, e.g., in consumer products, can be released to the environment during their use (e.g., release of silver nanoparticles (AgNPs) from facade painting [1] or TiO2-nanoparticles (TiO2-NPs) from sunscreens [2]), while, after use, a major fraction of the ENMs is supposed to be released to wastewater streams [3] and, furthermore, becomes attached to sewage sludge during wastewater treatment [4,5]. The majority of sewage sludge is incinerated and ends up in landﬁlls [6]. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Sewage sludge incineration ash (SSA) can be used for the production of phosphorous fertilizer (e.g., according German fertilizer ordinance [7]). This recycling route will be increased in the future because the recovery of phosphorus from sewage sludge and sewage sludge ash is required by law in Germany and other countries of the EU. https://www.mdpi.com/journal/materials Materials 2022, 15, 1853. https://doi.org/10.3390/ma15051853 Materials 2022, 15, 1853 2 of 12 The use of materials in soils always requires a speciﬁc assessment to prevent soil con- tamination by re-use of waste materials. Integral part of the assessment is the determination of the source strength from those materials for leaching of contaminants. This is performed, e.g., by a standard column elution according to DIN 19528 for dissolved heavy metals and a set of organic substances. g So far, very little has been published regarding the release of nanoparticles from SSA. In a very recent study, Wielinski et al. [8] investigated the release of different nanomaterials from SSA in column experiments, but the column experiments described therein are very different to the standard column elution used for evaluation of source strength in the scope of regulation. Further, laboratory column experiments need references from the “real world” to assess the robustness of the laboratory results. However, we could not ﬁnd any study regarding the release of nanomaterials from SSA after use in agricultural soils as fertilizer. Respective studies mostly focus on heavy metal contaminations (not in the form of NPs) or availability of phosphorous from SSA. 1. Introduction Thus, there were two knowledge gaps identiﬁed for our study: (i) the question of leaching of NPs that might enter agricultural soils by fertilization with NP-contaminated SSA under realistic and environmentally relevant conditions and (ii) the predictive power of a column elution procedure actually used in the German soil protection ordinance [9] for NP leaching. p g These gaps were addressed by the present study. In order to investigate the potential leaching of a representative ENM nano titanium dioxide (nTiO2) from SSA applied to soils, combined elution experiments in pilot and laboratory scale were performed. nTiO2 was selected as test material, as it is among the most produced nanomaterials [10,11] and was used, e.g., in suncreens [2], as photocatalyst [12] or food additives [13]. Ash material was produced in a large-scale waste incinerator from incineration of sewage sludge amended with nTiO2. The resulting SSA was mixed with a reference soil at a rate of about 5%. This is about a factor of 10 greater than the maximum allowed application rate [14] and was considered to be a worst case scenario. The SSA amended soil was used for a pilot scale bioreactor trial in order to investigate nTiO2 leaching from the material under controlled realistic conditions. In a timelapse experiment, three annual summer/winter cycles were simulated within 250 days to verify the inﬂuence of seasons on the leaching behavior. Those pilot scale simulations were accompanied by standard laboratory column elution experiments. Based on the results obtained in both tests, the suitability of the standard column elution with regard to a prediction of nTiO2 leaching from soil treated with SSA should be determined. In case of a signiﬁcant correlation, the standard laboratory column elution could be extended for risk assessment of nanomaterial leaching from SSA amended soils. So far, this has not been considered in soil protection, since transport of nanomaterials in soil cannot be described by common leaching models, as the sorption theory used for dissolved chemicals does not apply for nanomaterials [15]. 2.2. Simulation Experiments 2.2. Simulation Experiments 2.1. Preparation of ENM Containing Sewage Sludge Incineration Ash (SSA) For the production of the ENM-containing sewage sludge ash (SSA), the product Hombikat UV 100 WP from Co. VENATOR was used. The aqueous Hombikat UV 100 WP dispersion consists of 42% (w/w) nTiO2 and about 6.5% (w/w) of poly acrylate acting as stabilizer. As a basis for the formulation of Hombikat UV 100 WP, nTiO2, in the form of the product Hombikat UV 100 with a primary particles size of <10 nm, was employed. Further information can be found in Börner et al. [16] and Oischinger et al. [17]. The application of Hombikat UV 100 WP lies mainly in the photocatalytic area, as the nTiO2 is present in the anatase modiﬁcation. The sewage sludge incineration was performed at the Sewage Sludge Incineration Plant (SSIP) of the waste water treatment plant ZVK Steinhäule at Neu-Ulm. Annually a wastewater amount of 440,000 population equivalents is cleaned in the plant, resulting in 10,000 t of sewage sludge (dry matter (DM)) and 2500 t SSA [18]. The SSIP consists of a centrifuge, a dryer, a ﬂuidised-bed incinerator with selective non-catalytic NOx reduction, Materials 2022, 15, 1853 3 of 12 an electrostatic precipitator, a 2-stage scrubber and an activated-carbon reactor with fabric ﬁlter. A schematic diagram of the plant is depicted in [16]. an electrostatic precipitator, a 2-stage scrubber and an activated-carbon reactor with fabric ﬁlter. A schematic diagram of the plant is depicted in [16]. For the production of the reference ash material on the 1st day, the average amount of sewage sludge mounted up to 2171 kg h−1 (DM) and, for the day with nTiO2 injection, on the 2nd day, up to 2120 kg h−1 (DM). During the measurement with nTiO2 injection, 630 kg of nTiO2 dispersion was added to the sewage sludge with a peristaltic pump over 6 h. On both days, the measured background concentration of titanium was <0.1 wt%. Hence, an augmentation of titanium in the sewage sludge of about 1.25 wt% was achieved by the injection of the dispersion. The SSA was sampled on the basis of the recommendations of LAGA PN98 [19] for moving waste, as far as they were applicable to the large scale plant [20]. 2.2.1. Leaching in Pilot Scale Simulation Reactors For investigation of the release of nanoparticles from SSA applied in soil-related applications, experiments in pilot-scale simulation reactors were conducted. For this purpose, the incineration residue (with and without treatment with nTiO2) was mixed with an agricultural soil (refesol 04-A, details of the soil characteristics see Table 1). Table 1. Refesol 04-A soil characteristics (loamy sand). Soil Type Soil Texture (DIN) Soil Texture Corg (%) Ntotal (g/kg) pHCaCl2 CECeff (mol/kg) WHCmax (g/kg) Sand (%) Silt (%) Clay (%) Refesol 04-A 79.7 14.9 5.4 loamy sand 3.04 1.76 5.11 0.0412 346 Table 1. Refesol 04-A soil characteristics (loamy sand). Soil Type Soil Texture (DIN) Soil Texture Corg (%) Ntotal (g/kg) pHCaCl2 CECeff (mol/kg) WHCmax (g/kg) Sand (%) Silt (%) Clay (%) Refesol 04-A 79.7 14.9 5.4 loamy sand 3.04 1.76 5.11 0.0412 346 The sandy soil with higher organic carbon content was selected in order to simulate worst case conditions regarding potential nanoparticle release and mobility [21,22]. Per reactor, 700 kg of soil material (dry weight basis) were mixed with an amount of SSA, representing 5% with regard to the dry weight of the soil material, corresponding to approximately 190 t sludge ash per ha, taking into account an assumed plough share depth of 25 cm. At mixing, the soil was adjusted to a water content representing 50% of its water holding capacity (WHC), being optimal for microbial activity. The mixture was then subjected to an experiment simulating at least three annual seasonal cycles during a total incubation time of 250 days. Thus, the soil/ash mixture was incubated at 20–22 ◦C simulating a summer phase, followed by freezing of the mixture to −10 to −15 ◦C ﬁnal temperature. The soil in the reactor was kept at this temperature for 14 days (simulated winter period). After each winter period and thawing of the soil/ash mixture, the soil was dug up and samples were taken for the column elution experiments (seasonal cycles and sampling dates see Figure 1). The soil was irrigated over a period of 12 days with 10 L portions until approximately 5 L of seepage/leachate water was collected. The leachate was analysed for total titanium concentration by ICP-OES after microwave assisted digestion and, in addition, was subjected to nanoparticle speciﬁc analysis by single particle (sp)ICP-MS. 2.2.1. Leaching in Pilot Scale Simulation Reactors For the next cycle, the soil was again dried to a water content of 50% WHC by digging the soil up several times and leaving the reactor surface open while incubating at 20 ± 2 ◦C. After reaching the desired water content, the reactor was closed again and the next cycle started. Table 1. Refesol 04-A soil characteristics (loamy sand). The sandy soil with higher organic carbon content was selected in order to simulate worst case conditions regarding potential nanoparticle release and mobility [21,22]. Per reactor, 700 kg of soil material (dry weight basis) were mixed with an amount of SSA, representing 5% with regard to the dry weight of the soil material, corresponding to approximately 190 t sludge ash per ha, taking into account an assumed plough share depth of 25 cm. At mixing, the soil was adjusted to a water content representing 50% of its water holding capacity (WHC), being optimal for microbial activity. The mixture was then subjected to an experiment simulating at least three annual seasonal cycles during a total incubation time of 250 days. Thus, the soil/ash mixture was incubated at 20–22 ◦C simulating a summer phase, followed by freezing of the mixture to −10 to −15 ◦C ﬁnal temperature. The soil in the reactor was kept at this temperature for 14 days (simulated winter period). After each winter period and thawing of the soil/ash mixture, the soil was dug up and samples were taken for the column elution experiments (seasonal cycles and sampling dates see Figure 1). The soil was irrigated over a period of 12 days with 10 L portions until approximately 5 L of seepage/leachate water was collected. The leachate was analysed for total titanium concentration by ICP-OES after microwave assisted digestion and, in addition, was subjected to nanoparticle speciﬁc analysis by single particle (sp)ICP-MS. For the next cycle, the soil was again dried to a water content of 50% WHC by digging the soil up several times and leaving the reactor surface open while incubating at 20 ± 2 ◦C. After reaching the desired water content, the reactor was closed again and the next cycle started. Materials 2022, 15, 1853 4 of 12 Figure 1. Temperature curve over the entire test period with three annual cycles, measured at 20, 40 and 60 cm distance from the edge of the reactor with treated soil ash/mixture. 2.2.1. Leaching in Pilot Scale Simulation Reactors Dotted line marks the three sampling points after the winter phases (23 January, 24 April and 27 June). 2 2 2 C l El i T Figure 1. Temperature curve over the entire test period with three annual cycles, measured at 20, 40 and 60 cm distance from the edge of the reactor with treated soil ash/mixture. Dotted line marks the three sampling points after the winter phases (23 January, 24 April and 27 June). 2.2.2. Column Elution Tests At three time points, samples of the soil/ash mixture in the pilot-scale reactors were sampled and ﬁlled into glass columns (diameter 5.5 cm) according to the instructions given by the standard DIN 19528 [23] for soil column leaching. At both ends of the glass columns, quartz sand was used as a ﬁltration layer. The column was then subjected to an elution according to DIN 19528 [23]. The column was saturated with deionized water from the bottom to the top layer. After saturation, the column was eluted with deionized water and eluate samples taken at ﬁve different solid/water ratios (0.3, 1, 2, 4 and 10 L/kg). The titanium concentration in the eluate was determined as described in Section 2.3.1 without any previous eluate ﬁltration or centrifugation. The amount of soil per soil column and the percolation rates for saturation and percolation are presented in Table 2. The percolation rates were calculated according to DIN 19528: q = l × π × r2 × n t × 60 (1) (1) with q = percolation rate (mL min−1); l = length of soil column; r = inner radius of column; n = porosity (0.43); t = time (2 h for saturation, 5 h for percolation). with q = percolation rate (mL min−1); l = length of soil column; r = inner radius of column; n = porosity (0.43); t = time (2 h for saturation, 5 h for percolation). Table 2. Mass of soil per soil column and percolation rates for saturation and percolation. Table 2. Mass of soil per soil column and percolation rates for saturation and percolation. Soil Mass per Column Percolation Rate (mL min−1) (g) Saturation Percolation RefeSol 04-A 1100 2.64 1.06 2.3. Chemical Analysis y 2.3.1. TiO2 Concentrations in Reactor Leachates and Column Eluates For determination of the total concentration of titanium in the leachates of the reac- tors, as well as for the column eluates, 10 mL of the respective sample was ﬁlled into a Materials 2022, 15, 1853 5 of 12 5 of 12 TeﬂonTM digestion vessel (MLS, Leutkirch, Germany) and evaporated at 105 ◦C to dryness. The residue was then mixed with 4.8 mL of HNO3 (69%, supra pure, Roth) and 0.2 mL hydroﬂuoric acid (HF, 40%, supra pure, Roth) and digested in a microwave (Ultraclave II, MLS, Leutkirch, Germany; digestion parameters: 220 ◦C, 30 min, 100 bar). 2.3.3. Single Particle (sp)ICP-MS Analysis of Reactor Leachates and Column Eluates In addition to the determination of the total titanium concentration, number-based particle size distributions were determined by single particle inductively coupled plasma mass spectrometry (spICP-MS) [24,25]. The aqueous samples were directly measured without any further sample preparation, except for dilution with ultrapure water. The analyses were performed using a triple-quadrupole ICP-MS instrument (ICP-QQQ-MS, Agilent 8900, Agilent Technologies, Waldbronn, Germany). The dwell time in the single particle measurement mode of the ICP-MS was set to 100 µs and time-resolved signals were recorded on the selected m/z for 60 s. Peak detection and integration was conducted automatically by the Agilent MassHuntersoftware. Conversion of signal heights of particle spikes into particle sizes were performed by application of a calibration with a dissolved Ti standard. To apply a dissolved calibration for size calculation of nanoparticles from their signal spikes, the nebulization or transport efﬁciency in the interface was determined by analyzing a gold nanoparticle standard of known concentration and size [24]. Dispersions of 60 nm gold nanoparticles (AuNPs 60 nm, BBI solutions, Kent, UK) were used for the determination of the nebulization efﬁciency and were prepared freshly on the day of measurement. The samples were diluted in ultrapure water by a factor of 102–105 for measurement in order to reach a particle concentration of 200–2000 particle events per minute. Due to possible interferences for the most abundant titanium isotope 48Ti caused by the calcium isotope 48Ca present in the leachate and eluate matrices, titanium was measured in the MS/MS mode with ammonia as reaction gas (10% NH3 + 1 mL/min He). Thus, titanium was measured as [48TiNH]+ with a m/z of 63. The threshold between background and particle signals was deﬁned based on visual inspection of the measured signal distributions. The conversion of signal distributions into number-based size distributions, as well as particle number concentrations, was performed by the Agilent MassHunter software. 3. Results 3.1. Characterisation of the Produced SSA 2.3.2. TiO2 Concentrations in Soil/Ash-Mixed Samples For determination of total titanium concentrations in mixtures of soil and sewage, sludge ash was applied for simulation in the pilot-scale reactors; in addition, column elution experiments, 5 aliquots of 5 g (fresh weight) each, were taken and mixed to obtain a representative sample. The samples were dried at 105 ◦C until constant weight. Aliquots of approximately 200 mg of the dried residues were then weighted in TeﬂonTM digestion vessels and 1 mL HF (40% supra pure, Roth) and 4 mL HNO3 (69%, supra pure, Roth) added. The samples were subjected to microwave-assisted digestion (Ultraclave II, MLS, digestion parameters: 220 ◦C, 30 min, 100 bar). After digestion, 5 mL of boric acid was added to complex remaining HF and the digestate ﬁlled up to a ﬁnal volume of 15 mL with ultrapure water. The solution was analysed for titanium by ICP-OES at a wavelength of 334.941 nm. 2.2.2. Column Elution Tests After digestion, 1 mL boric acid was added in order to complex the remaining HF; the mixture ﬁlled up to a ﬁnal volume of 15 mL with ultrapure water. The concentration of Ti in the solution was then determined by ICP-OES (wavelength 334.941 nm, Instrument: Agilent 5110, Agilent technologies). The obtained results were converted into TiO2 concentrations under the assumption that no other titanium-containing phase was present in the samples. 3.1. Characterisation of the Produced SSA The two SSA produced differed in their titanium content as expected. For the SSA, from the reference day, an average titanium concentration of about 0.3 wt% (dry mass Materials 2022, 15, 1853 6 of 12 (DM)) with a standard deviation (SD) of 0.04 was determined, whereas the SSA treated with nTiO2 had a mean value of 2.9 wt% (DM) with an SD of 0.19 [26]. Hence, the titanium content in the nTiO2-SSA was increased compared to the reference measurement by a factor of about 9.7 [26]. Further information on characterisation of the used SSA can be found elsewhere [26]. 3.2. Leaching from Pilot Scale Simulation Reactors 3.2. Leaching from Pilot Scale Simulation Reactors 3.2. Leaching from Pilot Scale Simulation Reactors 3.2.1. Determination of Total Titanium Content in Mixtures of Soil/SSA and in Leachates 3.2.1. Determination of Total Titanium Content in Mixtures of Soil/SSA and in Leachates Figure 2 shows the determined total titanium concentrations in both mixtures of soil with SSA with additional nTiO2 or without nTiO2 treatment (reference samples) which were ﬁlled into the pilot scale reactors. The SSA/soil mixture containing nTiO2-treated ash shows a titanium concentration elevated by a factor of approximately two compared to the reference sample. Thus, at least this ratio should also be found in the leachates. Figure 2. Total TiO2 concentrations (converted Ti-concentrations) in sludge ash/soil mixtures. left column: Concentration in samples using sludge ash without treatment with nTiO2, right column: Concentration in samples using SSA treated with nTiO2. Error bars refer to standard deviation of four subsamples that were digested and analysed. Figure 2. Total TiO2 concentrations (converted Ti-concentrations) in sludge ash/soil mixtures. left column: Concentration in samples using sludge ash without treatment with nTiO2, right column: Concentration in samples using SSA treated with nTiO2. Error bars refer to standard deviation of four subsamples that were digested and analysed. Figure 2. Total TiO2 concentrations (converted Ti-concentrations) in sludge ash/soil mixtures. left column: Concentration in samples using sludge ash without treatment with nTiO2, right column: Concentration in samples using SSA treated with nTiO2. Error bars refer to standard deviation of four subsamples that were digested and analysed. However, as shown in Figure 3, the titanium/converted TiO2 content in the leachates of treatment and reference reactors collected at the end of the three simulated annual seasonal cycles was found to be higher for the treated samples compared to the references by a factor of ~25. Titanium-containing particles are released from the treatment reactor in a signiﬁcantly higher amount than expected from the total titanium content of the respective mixtures, thus, indicating that the added nTiO2 (or titanium containing particles) might be more mobilisable than the titanium-bearing particles (most likely also being TiO2) already present in the soil and sludge ash. 7 of 12 Materials 2022, 15, 1853 Figure 3. Total TiO2 content (derived from Ti-concentrations) determined in leachates from pilot scale simulation reactors collected after watering of the soil at end of the respective preceding winter phase. Figure 3. 3.2. Leaching from Pilot Scale Simulation Reactors Total TiO2 content (derived from Ti-concentrations) determined in leachates from pilot scale simulation reactors collected after watering of the soil at end of the respective preceding winter phase. 3.2.2. SpICP-MS/MS Analyses of Leachates The signiﬁcantly higher release of titanium in the leachates from the treatment reactor is also visible in the nanoparticle speciﬁc analysis of the leachates by spICP-MS, as shown in Figure 4. The higher TiO2 particle discharge in leachates after the ﬁrst winter period (23 January) is 3.7 times higher in the treated reactor and rises to 640 times after the second simulated winter phase (17 April) and drops to 8.2 times after the third winter phase (27 June) (Figure 4). Again, the differences in discharge are much higher than could be expected from the total Ti-content determined for the soil/ash mixtures applied to the two reactors. Figure 4. TiO2 particle concentrations (Particles/L) determined in leachates from pilot scale simula- tion reactors collected after watering of the soil at end of the respective preceding winter phase. Figure 4. TiO2 particle concentrations (Particles/L) determined in leachates from pilot scale simula- tion reactors collected after watering of the soil at end of the respective preceding winter phase. Materials 2022, 15, 1853 8 of 12 3.3. Column Elution Experiments 3.3. Column Elution Experiments The percolate samples obtained from standard column eluate experiments were anal- ysed analogously to the leachates from the pilot scale experiments for their total TiO2 content (ICP-OES analysis after HF digestion) and particle size distribution (spICP-MS). Figure 5 shows the TiO2 content in the percolates of soil/ash-material taken at end of the three simulated annual seasonal cycles from the treatment and reference reactor to investigate the behaviour during soil-related use. In the percolates, an increased TiO2 load was detected for the treated SSA/soil mixture compared to the reference mixture. In contrast to the analyses of the reactor leachates, the ratio between TiO2 released from nTiO2-treated material and the reference material was found to be more or less stable during all three sampling times. Already, percolates from the ﬁrst sampling (17 January) showed increased TiO2 contents in the nTiO2 treated SSA/soil mixture. This was most likely due to the very high water:solid ratios used in the column elution compared to the pilot scale experiment. However, the maximum TiO2 concentrations determined are in the same range as the maximum concentration found in the pilot scale experiment. Figure 5. Total TiO2 content (µg) in column percolates from reference and nTiO2 treated SSA/soil mixtures out of pilot scale simulation reactors collected at three different sampling times. Figure 5. Total TiO2 content (µg) in column percolates from reference and nTiO2 treated SSA/soil mixtures out of pilot scale simulation reactors collected at three different sampling times. The spICP-MS/MS analyses of the column percolates (reference and nTiO2 treatment) show a higher nTiO2 particle discharge from the nTiO2 treated SSA/soil mixture (Figure 6). The particle number in the percolates of the reference and TiO2 treated SSA/soil mixtures increases in both cases with the water/solid ratio (W/S). In the TiO2-treated percolates, the particle concentration at a W/S of 10 L/kg at the ﬁrst sampling (17 January) extend concentration in the reference percolates by a factor 10. During the following samplings (17 April and 21 June), particle counts in the nTiO2 percolates extend the reference percolates by a factor of 4. In general, the nTiO2 analysis conﬁrms the ﬁndings of the total TiO2 analysis, indicating that most of the TiO2 determined was due to leached nTiO2. Materials 2022, 15, 1853 9 of 12 Figure 6. 3.3. Column Elution Experiments nTiO2 particle concentrations (Particles/L) determined in column percolates from reference and nTiO2 treated SSA/soil mixtures out of pilot scale simulation reactors collected at three different sampling times. Figure 6. nTiO2 particle concentrations (Particles/L) determined in column percolates from reference and nTiO2 treated SSA/soil mixtures out of pilot scale simulation reactors collected at three different sampling times. 4. Discussion However, in principle, the spICP-MS analyses give similar results to the determination of the total concentrations and can possibly be a fast alternative analytical method that, additionally, provides also information on particle sizes. Looking further and in more detail to the obtained data, some differences can be noticed not only between the elution behavior of the test materials but also between the elution procedures itself. The elution procedure in the column and the simulation experiment were very different. In the column elution, within 7 days, a water to solid ratio of 10 L per kg soil material was applied. In the simulation, the maximum water to solid ratio used was 0.5 L per kg of soil material over a time of 250 days. This results in very different ﬂow velocities in both approaches; it is not surprising that these huge differences in the elution conditions give different results in eluate analysis. This signiﬁcant inﬂuence of different elution conditions is strikingly visible if the released TiO2 amounts are related to the total amount of soil/ash mixture in the respective experiments. For instance, for the last sampling date, after three seasonal cycles, around 130 µg/kg (released amount (Figure 3) referred to 700 kg of soil/ash (dry weight) material per reactor) was released from the reactor, while, for the column elution experiments, 5172 µg/kg (TiO2 content in percolate referred to 1.1 kg soil/ash per column) was released. Thus, the column leaching experiments are overestimating the leaching for this kind of particle in comparison to the more realistic simulation experiments. p p p The higher relative release of TiO2 from reference samples in the column experiments indicates that the natural “background” TiO2 might be attached differently to the soil matrix, as it is more easily mobilized under the experimental conditions compared to the simulation experiments. Therefore, the main ﬁnding of the study is that the standard column experiment according to DIN 19528 might be suitable to show increased mobility and emissions of nTiO2 during soil-related recycling. This ﬁnding is conﬁrmed by simulation experiments under more realistic conditions. It proves the general suitability of the standardized column elution method for predicting an increased release of nTiO2 from a soil-ash mixture. 4. Discussion In this study, nTiO2 was used as an exemplary ENM for the preparation of ENM- containing SSA produced for the soil simulation experiments. In the simulation of agricul- tural use of this nTiO2 amended SSA, a clearly increased release of nTiO2 was found in the leachates of the pilot scale simulation reactors after a certain time compared to a reference ash from non-amended sewage sludge. The amount of release was signiﬁcantly higher than could be expected due to the differences in the total TiO2 contents in the soil-ash mixture relative to the reference ash. A cause for the comparatively high releases could not be determined within the project. After three freezing and thawing cycles, the amount of TiO2 released from the reference was increased signiﬁcantly, indicating mobilization by changes of the soil structure due to frost wedging. This is also observed in increasing amounts of TiO2 in the column experiments throughout the three samplings. A difference between the leaching from the simulation reactors and standard column elution is the signiﬁcantly higher ratio between TiO2 released from reference and treated SSA in the reactors compared to the ratios observed for the column experiments, which were nearer to the ratio of total Ti contents of both materials (Figure 2) at the end of the study. For instance, while, for the reactors, the release of TiO2, expressed in terms of total concentrations, from treated samples compared to the reference, was elevated by a factor of approx. 25, the analogous ratio for the standard column experiments was found to be approx. 3 (Figures 3 and 5). The differences are not that pronounced if the particle counts from spICP-MS analyses are assessed with factors of approx. 10 (leachates of treated-against-reference reactors) and 4 (column percolates of treated-against-reference soil/ash-mixtures). Thus, the particle counts agree with the total titanium on a relative scale for the column experiments, whereas, there seems to be an underestimation for the reactor leachates. This might be explained by losses of, e.g., bigger particles, which would signiﬁcantly contribute to the total titanium mass in the sample, due to sedimentation prior to measurement or the underestimation Materials 2022, 15, 1853 10 of 12 10 of 12 of the number of bigger particles due to the applied analysis time of 60 s. In contrast, the total Ti determination after digestion captures the complete amount of Ti in the sample. 4. Discussion However, due to the limitations of a pilot scale experiment, which does not allow to test a reasonable number of different soils that would be needed for a more general statement, results cannot be generalized at this point. However, they indicate that increased leaching of ENM measured in a standard soil column approach relative to a reference material might serve as reasonable prediction of increased leaching of ENM under realistic environmental conditions. As a next step, a larger data set should be determined on both SSAs using a set of different soils for standard column elution to obtain the inﬂuence of different soil matrices. 5. Conclusions Acknowledgments: The authors acknowledge the plant operator at ZVK Steinhäule for his support during the experiments. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. 5. Conclusions The results of this study demonstrate that a standard soil column experiment might be suitable to reliably indicate an increased mobility of nTiO2 from SSA mixed with agricultural soil relative to a respective reference material containing only “background” TiO2. However, due to the limited data set of so far just one soil general conclusions cannot be made at this point. Still, the result is important for future exposure assessment, as it might offer an experimental laboratory tool to determine the risk of leaching from a recycling material. The use of an accepted standard (DIN 19528) for the determination of the mobility offers the chance to get a new standard for testing ENM in soil material just by extension of the scope of an existing standard. In order to establish effective thresholds and to make them available for regulatory practice, respective standards for laboratory use are required. However, further research is necessary to verify current results and to base ﬁndings on a broader database before standardization processes might start. Laboratory elution experiments could give a useful contribution to the classiﬁcation and safer use of incineration ashes, which contain not only potential pollutants but also many components Materials 2022, 15, 1853 11 of 12 11 of 12 that are too valuable for landﬁll disposal. Even if no direct transfer from the laboratory elution tests on the simulation experiments can be derived, the general concept is conﬁrmed. Author Contributions: Conceptualization, B.M., D.H. and K.H.-R.; methodology, B.M., D.H., K.H.-R.; validation, B.M., N.S. and D.H.; formal analysis, N.S. and B.M.; investigation, N.S. and K.W; resources, K.W., J.O.; data curation, B.M. and N.S.; writing—original draft preparation, B.M. and N.S.; writing— review and editing, B.M., D.H. and K.W.; visualization, N.S.; supervision, D.H. and B.M.; project administration, D.H. and B.M.; funding acquisition, D.H., J.O., K.H.-R. and B.M. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the German Federal Environment Agency (UBA) and was part of the project “Investigations on the possible release of nanoparticles in the deposition and soil-related application of mineral wastes” (FKZ 3712 33 327). Data Availability Statement: Restrictions apply to the availability of these data. Data was obtained from the project funded by the German Environment Agency (FKZ 3712 33 327) and are available from the corresponding authors with the permission of the German Environment Agency. References 1. Kaegi, R.; Sinnet, B.; Zuleeg, S.; Hagendorfer, H.; Mueller, E.; Vonbank, R.; Boller, M.; Burkhardt, M. Release of silver nanoparticles from outdoor facades. Environ. Pollut. 2010, 158, 2900–2905. [CrossRef] [PubMed] 1. Kaegi, R.; Sinnet, B.; Zuleeg, S.; Hagendorfer, H.; Mueller, E.; Vonbank, R.; Boller, M.; Burkhardt, M. Release of silver nanoparticles from outdoor facades. Environ. Pollut. 2010, 158, 2900–2905. [CrossRef] [PubMed] 2. Gondikas, A.P.; von der Kammer, F.; Reed, R.B.; Wagner, S.; Ranville, J.F.; Hofmann, T. Release of sunscreens into surface waters: A one-year survey at the old Danube recreational Lake. Environ. Sci. Tec 2. Gondikas, A.P.; von der Kammer, F.; Reed, R.B.; Wagner, S.; Ranville, J.F.; Hofmann, T. Release of TiO2 nanoparticles from sunscreens into surface waters: A one-year survey at the old Danube recreational Lake. Environ. Sci. Technol. 2014, 48, 5415–5422. [CrossRef] [PubMed] 2. Gondikas, A.P.; von der Kammer, F.; Reed, R.B.; Wagner, S.; Ranville, J.F.; Hofmann, T. Release of TiO2 nanoparticles from sunscreens into surface waters: A one-year survey at the old Danube recreational Lake. Environ. Sci. Technol. 2014, 48, 5415–5422. [CrossRef] [PubMed] 3. Keller, A.A.; McFerran, S.; Lazareva, A.; Suh, S. Global life cycle releases of engineered nanomaterials. J. Nanopart. Res. 2013, 15, 1692–1709. [CrossRef] 3. Keller, A.A.; McFerran, S.; Lazareva, A.; Suh, S. Global life cycle releases of engineered nanomaterials. J. Nanopart. Res. 2013, 15, 1692–1709. [CrossRef] ] G.; Hristovski, K.; Kiser, M.A. Occurrence and removal of titanium at full scale wastewater treatment plants 2 nanomaterials. J. Environ. Monit. 2011, 13, 1195–1203. [CrossRef] [PubMed] 4. Westerhoff, P.; Song, G.; Hristovski, K.; Kiser, M.A. Occurrence and removal of titanium at full scale Implications for TiO2 nanomaterials. J. Environ. Monit. 2011, 13, 1195–1203. [CrossRef] [PubMed] 5. Kaegi, R.; Voegelin, A.; Sinnet, B.; Zuleeg, S.; Hagendorfer, H.; Burkhardt, M.; Siegrist, H. Behavior of metallic silver nanoparticles in a pilot wastewater treatment plant. Environ. Sci. Technol. 2011, 45, 3902–3908. [CrossRef] [PubMed] 5. Kaegi, R.; Voegelin, A.; Sinnet, B.; Zuleeg, S.; Hagendorfer, H.; Burkhardt, M.; Siegrist, H. Behavior of metallic silver nanoparticles in a pilot wastewater treatment plant. Environ. Sci. Technol. 2011, 45, 3902–3908. [CrossRef] [PubMed] 6. Wiechmann, B.; Dienemann, C.; Kabbe, C.; Brandt, S.; Vogel, I.; Roskosch, A. Sewage Sludge Treatment in Germany; Umwelt- bundesamt, Dessau-Roßlau. 2015. Available online: https://www.umweltbundesamt.de/sites/default/ﬁles/medien/378/ publikationen/sewage_sludge_management_in_germany.pdf (accessed on 10 December 2021). 6. Wiechmann, B.; Dienemann, C.; Kabbe, C.; Brandt, S.; Vogel, I.; Roskosch, A. Sewage Sludge Treatment in Germany; Umwelt- bundesamt, Dessau-Roßlau. 2015. References Available online: https://www.umweltbundesamt.de/sites/default/ﬁles/medien/378/ publikationen/sewage_sludge_management_in_germany.pdf (accessed on 10 December 2021). 7. DüMV, Verordnung über das Inverkehrbringen von Düngemitteln, Bodenhilfsstoffen, Kultursubstraten und Pﬂanzenhilfsmitteln (Düngemittelverordnung–DüMV) (German Fertilizer Ordinance); Bundesgesetzblatt (Federal Law Gazette) Jahrgang 2012 Teil I Nr. 58. 2012. Available online: https://www.gesetze-im-internet.de/d_mv_2012/ (accessed on 10 December 2021). 8. Wielinski, J.; Gogos, A.; Voegelin, A.; Müller, C.R.; Morgenroth, E.; Kaegi, R. Release of gold (Au), silver (Ag) and cerium dioxide (CeO2) nanoparticles from sewage sludge incineration ash. Environ. Sci. Nano 2021, 8, 3220–3232. [CrossRef] [PubMed] 9. BBodSchV, Bundes-Bodenschutz- und Altlastenverordnung vom 12. Juli 1999 (BGBl. I S. 1554), die Zuletzt durch Artikel 126 der Verordnung vom 19. Juni 2020 (BGBl. I S. 1328) Geändert Worden ist (German Soil Protection Ordinance); Bundesgesetzblatt (Federal Law Gazette) Jahrgang 1999 Teil I, Nr. 36. 1999. Available online: https://www.gesetze-im-internet.de/bbodschv/ (accessed on 10 December 2021). 10. Piccinno, F.; Gottschalk, F.; Seeger, S.; Nowack, B. Industrial production quantities and uses of ten engineered nanomaterials in Europe and the world. J. Nanopart. Res. 2012, 14, 1109. [CrossRef] 10. Piccinno, F.; Gottschalk, F.; Seeger, S.; Nowack, B. Industrial production quantities and uses of ten engineered nanomaterials in Europe and the world. J. Nanopart. Res. 2012, 14, 1109. [CrossRef] 11. Foss Hansen, S.; Heggelund, L.R.; Revilla Besora, P.; Mackevica, A.; Boldrin, A.; Baun, A. Nanoproducts–what is actually available to European consumers? Environ. Sci. Nano 2016, 3, 169–180. [CrossRef] p 12. Landi, S.; Carneiro, J.; Soares, O.S.G.P.; Pereira, M.F.R.; Gomes, A.C.; Ribeiro, A.; Fonseca, A.M.; Parpot, P.; Neves, I.C. Photocatalytic performance of N-doped TiO2nano-SiO2-HY nanocomposites immobilized over cotton fabrics. J. Mater. Res. Technol. 2019, 8, 1933–1943. [CrossRef] 13. Noireaux, J.; López-Sanz, S.; Vidmar, J.; Correia, M.; Devoille, L.; Fisicaro, P.; Loeschner, K. Titanium dioxide nanoparticles in food: Comparison of detection by triple-quadrupole and high-resolution ICP-MS in single-particle mode. J. Nanopart. Res. 2021, 23, 4. [CrossRef] 14. BioAbfV, Bioabfallverordnung in der Fassung der Bekanntmachung vom 4. April 2013 (BGBl. I S. 658), die Zuletzt durch Artikel 3 Absatz 2 der Verordnung vom 27. September 2017 (BGBl. I S. 3465) Geändert Worden ist. Bundesgesetzblatt (Federal Law Gazette) Jahrgang 1998 Teil I, S. 2955. 1998. Available online: https://www.gesetze-im-internet.de/bioabfv/BJNR295500998.html (accessed on 10 December 2021). Materials 2022, 15, 1853 12 of 12 12 of 12 15. Praetorius, A.; Tufenkji, N.; Goss, K.-U.; Scheringer, M.; von der Kammer, F.; Elimelech, M. The road to nowhere: Equilibrium partition coefﬁcients for nanoparticles. Environ. Sci. Nano 2014, 1, 317–323. [CrossRef] p p 16. References Börner, R.; Meiller, M.; Oischinger, J.; Daschner, R. Untersuchung Möglicher Umweltauswirkungen bei der Entsorgung Nanomaterial- haltiger Abfälle in Abfallbehandlungsanlagen; Umweltbundesamt: Dessau-Roßlau, Germany, 2016. 17. Oischinger, J.; Meiller, M.; Daschner, R.; Hornung, A.; Warnecke, R. Fate of nano titanium dioxide during combustion of engineered nanomaterial-containing waste in a municipal solid waste incineration plant. Waste Manag. Res. 2019, 37, 1033–1042. [CrossRef] [PubMed] 18. Schäfer, E. Beispiel einer zukunftsorientierten kommunalen Abwasserreinigung. In Verwertung von Klärschlamm; Holm, O.E., Thomé-Kozmiensky, E., Quicker, P., Kopp-Assenmacher, S., Eds.; Thomé-Kozmiensky Verlag GmbH: Neuruppin, Germany, 2018; pp. 121–130. 19. Ländergemscinaft Abfall, LAGA PN98 Richtlinie für das Vorgehen bei Physikalischen, Chemischen und Biologischen Unter- suchungen im Zusammenhang mit der Verwertung/Beseitigung von Abfällen. Grundregeln für die Entnahme von Proben aus Festen und Stichfesten Abfällen Sowie Abgelagerten Materialien; Ländergemeinschaft Abfall. 2001. Available online: https://www.laga-online.de/documents/m32_laga_pn98_1503993280.pdf (accessed on 10 December 2021). p g g p p ( ) 20. Länderarbeitsgemeinschaft Abfall, LAGA-Mitteilung 19, Merkblatt für die Entsorgung von Abfällen aus Verbrennungsanlagen für Sied- lungsabfälle; ESV: Berlin, Germany, 1994. 21. Chen, G.; Liu, X.; Su, C. Distinct effects of humic acid on transport and retention of TiO2 rutile nanoparticles in saturated sand columns. Environ. Sci. Technol. 2012, 46, 7142–7150. [CrossRef] [PubMed] 22. Wang, Y.; Gao, B.; Morales, V.L.; Tian, Y.; Wu, L.; Gao, J.; Bai, W.Y.L. Transport of titanium dioxide nanoparticles in saturated porous media under various solution chemistry conditions. J. Nanopart. Res. 2012, 14, 1095–1104. [CrossRef] p y p 23. DIN 19528:2009-01; Leaching of Solid Materials-Percolation Method for the Joint Examination of th Inorganic and Organic Substances. Beuth: Berlin, Germany, 2009. g g y 24. Montano, M.D.; Olesik, J.W.; Barber, A.G.; Challis, K.; Ranville, J.F. Single Particle ICP-MS: Advances toward routine analysis of nanomaterials. Anal. Bioanal. Chem. 2016, 408, 5053–5074. [CrossRef] [PubMed] [ ] [ ] 25. Meermann, B.; Nischwitz, V. ICP-MS for the analysis at the nanoscale–A tutorial review. J. Anal. At. Spectrom. 2018, 33, 1432–1468. [CrossRef] 26. Oischinger, J.; Meiller, M.; Daschner, R.; Hennecke, D.; Hund-Rinke, K.; Meisterjahn, B.; Schröder, N. Untersuchungen zur Möglichen Freisetzung von Nanopartikeln bei der Ablagerung und Bodenbezogenen Anwendung von Mineralischen Abfällen-Abschlussbericht; Umweltbundesamt: Dessau-Roßlau, Germany, 2020.
https://openalex.org/W3027404967	http://www.scielo.br/pdf/neco/v29nspe/1980-5381-neco-29-spe-1277.pdf	Portuguese	null	Música, redes e tecnologia na periferia: impactos tecnológicos na produção de rap na zona sul de São Paulo	Nova Economia	2,019	cc-by	11,183	Música, redes e tecnologia na periferia: impactos tecnológicos na produção de rap na zona sul de São Paulo Rodrigo Cavalcante Michel (1) Ana Flávia Machado (2) Natália Guimarães Duarte Sátyro (3) (1) Universidade Federal de Minas Gerais (2) Universidade Federal de Minas Gerais (3) Universidade Federal de Minas Gerais Rodrigo Cavalcante Michel (1) Ana Flávia Machado (2) Natália Guimarães Duarte Sátyro (3) (1) Universidade Federal de Minas Gerais (2) Universidade Federal de Minas Gerais (3) Universidade Federal de Minas Gerais Resumo The ICT technological paradigm brought signiﬁ cant changes in the economics of music, reducing costs and expanding possi- bilities production access. The objective of the present work is to verify if the current technological impacts on the music produc- tion are also perceived in urban peripheral zones, starting from the case study of the production of rap music in Grajaú, district of São Paulo, Brazil. Thus, the hypothesis is considered: the technological impacts re- duced production costs and increased access to music production. In addition, new orga- nizational production forms emerged, based on networks. The results are summarized as follows: a) the impacts of ICTs on music production are veriﬁ ed in the speciﬁ c case study, suggesting that innovations also reach the periphery; b) the local music production occurs in organizations based on networks, incorporating monetary and community re- lations. It is noted that the rap music produc- tion in Grajaú presents community and col- laborative aspects associated with economic and monetary. O paradigma tecnológico das TICS trouxe mu- danças signiﬁ cativas na economia da música, reduzindo custos e ampliando possibilidades de acesso à produção. O objetivo do presente artigo é veriﬁ car se os atuais impactos tecnológicos na produção de música também são percebidos em zonas periféricas urbanas, a partir do estudo de caso da produção de música rap no Grajaú. As- sim, considera-se a hipótese: os impactos tecnoló- gicos possibilitaram redução nos custos de produ- ção e ampliação do acesso à produção musical. Ademais, novas formas organizacionais de pro- dução emergem baseadas em redes. Os resultados encontrados resumem-se em: a) os impactos das TICs na produção musical veriﬁ cam-se no estu- do de caso especíﬁ co, sugerindo que as inovações atingem também a periferia; b) a produção de música local ocorre em organizações baseadas em redes, incorporando relações monetárias e comu- nitárias. Nota-se que a produção de música rap no Grajaú apresenta aspectos comunitários e co- laborativos associados a econômicos e monetários. DOI: http://dx.doi.org/10.1590/0103-6351/5797 DOI: http://dx.doi.org/10.1590/0103-6351/5797 1 No artigo são apresentadas referências que servem como principal arcabouço teórico e conceitual para o problema em questão. Os trabalhos de Wikstrom (2009), Dolata (2011) e Leyshon (2014) constatam o impacto das TICs no aumento da participação de entrantes no mercado musical. Palavras-chave Palavras-chave economia criativa, economia da música, redes, inovações tecnológicas. Palavras-chave economia criativa, economia da música, redes, inovações tecnológicas. creative economy, economics of music, so- cial networks, technological innovations. Códigos JEL Z11, L82. Códigos JEL Z11, L82. JEL Codes Z11, L82. JEL Codes Z11, L82. JEL Codes Z11, L82. 1277 v.29 n.especial p.1277-1303 2019 Nova Economia 1277 v.29 n.especial p.1277-1303 2019 Nova Economia Michel, Machado & Sátyro 2 Streaming são plataformas que transmitem música e/ou vídeo sem a necessidade de down- load pelo usuário, como YouTube, Spotify, Tidal, Vimeo. 1 Introdução O setor de produção musical passa recorrentemente por transformações tecnológicas que modiﬁcam a forma produtiva, a estrutura de mercado, a atuação dos agentes, o formato do produto e, consequentemente, a forma do consumo. Atualmente, o setor se encontra na onda tecnológica das Tec- nologias de Informação e Comunicação (doravante, TICs), com intenso uso de fatores digitais e cibernéticos. A digitalização da música nos anos 2000 trouxe mudanças no que tange ao processo de produção, distribuição e consumo. Algumas teorias e análises acadêmicas aﬁrmam que o atual cenário tecnológico e estrutural do setor favorece a presença de agentes de menor porte no mercado, bem como redução das barreiras à entrada.1 Na esteira dessa revolução tecnológica, a partir de meados da década de 1990, os processos de produção, comercialização e consumo de música so- freram modiﬁcações, pois a produção (gravação, mixagens e sonorização) foi alvo de alterações nos processos. Essas modiﬁcações se deram, tam- bém, na estrutura de custos e investimentos na produção de música, pois possibilitaram a utilização de computadores pessoais e uso de softwares livres ou facilmente adquiridos na internet. Ademais, as TICs ocasionaram ampliação de mercado e formação de novos nichos por meio da ampla difusão do uso da internet e de redes de compartilhamento de arquivos, vídeos e redes sociais (Nakano et al., 2010; Wikstrom, 2009; Dolata, 2011; Leyshon, 2014;). Nesse período, a música passou a se consolidar no for- mato digital, com o surgimento de novas possibilidades de consumo, bem como “trocas” e compras de música através das plataformas peer-to-peer e streaming. Além disso, as plataformas streaming2 permitem a fruição de música pelo usuário sem a necessidade de “baixar” ou comprar as músi- cas. Assim, empresários nessa indústria se deparam com a necessidade de uma reorganização produtiva e mercadológica (Dias, 2000; Herschmann; Kischinhevsky, 2005). Nesse novo cenário, criam-se oportunidades para maior divulgação de músicas realizadas fora das maiores empresas, e fora dos grandes centros 1278 Nova Economia v.29 n.especial 2019 1278 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia urbanos como, por exemplo, as músicas produzidas nas periferias urbanas. O horizonte de atuação também se amplia, uma vez que se torna mais fácil divulgar e consumir músicas de diversas partes do país. 3 Compreende-se como majors empresas gravadoras e distribuidoras com ampla atuação no mercado, que contam em seu leque uma variedade de produtos e artistas cosolidados e com apelo de público grande. Alguns exemplos podem ser citados como, Warner, Sony, EMI etc. 4 Trata-se de um distrito bastante povoado, com crescimento populacional nos últimos anos acima da média paulistana, porém com densidade populacional baixa. A renda dos habitan- tes se situa em patamares baixos quando comparada com outros distritos do município. Por outro lado, o Grajaú se apresenta como um território de forte presença do rap, representando, em conjunto com outros territórios da zona sul de São Paulo, um epicentro de criatividade desse movimento. Nomes como Racionais MCs, Sabotage e Criolo são grandes expoentes do rap nacional, sendo todos originários da zona sul de São Paulo, e o último, especificamente do Grajaú. 5 Entende-se por indústrias criativas o conjunto de atividades econômicas que apresentem, em seu processo criativo, o uso da criatividade como insumo básico ou prioritário. indústrias criativas. Bendassolli et al. (2009) listam quatro principais características das indústrias criati- vas: (1) o uso da criatividade como aspecto central das atividades, sendo entendida nas con- cepções de Hesmondhalgh (2002) como a capacidade que os indivíduos possuem de decifrar e manipular símbolos para gerar algo inovador; (2) a valorização da arte pela arte, uma vez que os fatores culturais e artísticos têm forte relevância no processo produtivo; (3) O intenso uso de tecnologias de informação, que possibilitam a descentralização das atividades; e (4) o uso de “equipes polivalentes” e diversas nas criações. 1 Introdução As TICs, ao desonerarem o processo produtivo e ampliarem o acesso à produção, re- velam estilos musicais que antes eram localmente concentrados, como é o caso do tecnobrega, mangobeat, rap, entre outros. E o alargamento do mercado destes estilos se dá pela atuação de organizações de menor porte. Ao contrário das majors,3 empresas que, inicialmente, concentravam e defi- niam o repertório musical, as empresas indies, de produção independente, tornam-se importantes canais de produção e distribuição musical. A partir disso, analisamos de que forma o uso da tecnologia possibilitou (e tem possibilitado) a inserção de novos atores no mercado de música, ou ainda, tem oportunizado outra forma de atuação de músicos já inseridos no mercado, através de uma análise do arranjo das relações entre os agen- tes nessas produções específicas. A investigação do fenômeno apresentado é realizada através da análise de um caso específico, a produção de rap na zona sul de São Paulo/SP, especificamente no distrito do Grajaú, um dos 96 distritos do município paulistano.4 Dessa forma, o presente trabalho visa contribuir com a discussão dos impactos tecnológicos na economia criativa ao discorrer acerca desses efeitos em um espaço urbano específico e periférico. Grande parte da literatura especializada advém de autores de países desenvolvidos como Estados Unidos, Austrália e Reino Unido e, por muitas vezes, não analisa disparidades regionais, sociais e econô- micas que permeiam os espaços urbanos de países menos desenvolvidos. A contribuição inova também com o desenho de pesquisa e a técnica aqui empregados – análise de redes sociais – pouco utilizados nas Ciências Eco- nômicas, mas de relevância para estudos em Economia Criativa (Ormerod et al., 2008). A pertinência da metodologia empregada começa pela junção v.29 n.especial 2019 Nova Economia 1279 Michel, Machado & Sátyro da técnica com uma abordagem teórica que pressupõe laços entre os atores envolvidos, e isso se reﬂete nos resultados alcançados, podendo assim se tornar base para trabalhos vindouros no tema. O objetivo principal da pesquisa é, então, veriﬁcar se as mudanças tecnológicas proporcionadas pelas TICs no processo produtivo de músi- ca ocasionaram maior participação e dinamismo de agentes no mercado localizados em periferias urbanas. Além disso, propõe-se inferir sobre a organização da produção de música nesses espaços, buscando medir a natureza das relações estabelecidas entre os agentes inseridos na rede de produção analisada. Tendo-se tal objeto em tela, organizamos este artigo em cinco seções, incluindo esta introdução. 1 Introdução Na próxima seção, tratamos do papel de redes no contexto da digitalização da música. Em seguida, na terceira seção, apresentamos a estrutura metodológica desse estudo de caso. Na quarta seção, descrevemos os resultados encontrados e, por ﬁm, apresentamos as considerações ﬁnais. 2 Redes de produção periférica na economia criativa e impactos das inovações tecnológicas recentes na música As indústrias culturais e criativas5 utilizam componentes culturais em seus processos, dependentes da história e cultura local, o que acarreta heteroge- neidade nos processos e produtos decorrentes das diferenças locais. Soma- -se a isso, a importância de relações entre diferentes setores das indústrias criativas (a relação entre música e cinema, por exemplo; ou ainda publici- dade e audiovisual), que também contribui para a constituição dessas di- ferenças. Diante disso, podemos compreender as indústrias criativas como sendo sistemas abertos na concepção de Beinhoker (2006) e Comunian (2011), uma vez que as produções nas indústrias criativas envolvem intera- 1280 Nova Economia v.29 n.especial 2019 1280 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia ções entre diversos agentes pertencentes ao mesmo setor ou a setores dis- tintos. Trata-se de interações não lineares, que podem ser permanentes ou temporárias e, por vezes, geram consequências em todo o sistema criativo. Tais interações são características de processos não estáticos, dinâmicos e não determinísticos. Devido à presença de fatores culturais, as relações existentes entre diferentes agentes podem envolver influências que mo- dificam o bem ou serviço de maneira significativa, gerando assim, uma característica de sistema aberto, fora do equilíbrio, baseado em interações inter e intrassistemas. Tornando-se um sistema aberto com relações entre os agentes, pode- mos dizer que as indústrias culturais e criativas podem ser compreendidas como redes, a partir do conceito de Beinhoker (2006) e em consonância com Potts et al. (2008). Os padrões relacionais são específicos em cada se- tor das indústrias criativas. A quantidade de atores, o número de relações e a distribuição do poder, também, podem ser distintos. O conceito de rede é amplamente utilizado no presente trabalho. Defendemos que as indústrias criativas (e em específico a produção de música) sejam analisadas a partir da noção de redes, quando se busca compreender os mecanismos explicativos do funcionamento geral da produção, distribuição e consumo das mesmas. Nos setores criativos, a diversidade de agentes que são importantes ao processo produtivo pode extrapolar os limites conceituais do setor. Por exemplo, a produção musical pode sofrer grande influência de agentes de fora do setor, como profissionais do audiovisual e comunicação. A interconexão entre as in- dústrias criativas é premissa presente, e uma análise em rede permite captar esse fato. 6 A esse grupo de gravadoras menores que apresentam um “produto diferenciado” dire- cionado a públicos específicos e menores convencionou-se chamar de ‘gravadoras indies’. Apresentam um leque reduzido de artistas e opções (em relação às majors), geralmente são empresas focadas em um ou poucos estilos musicais. 2 Redes de produção periférica na economia criativa e impactos das inovações tecnológicas recentes na música As TICs foram protagonistas desse processo, uma vez que possibilita- ram redução nos custos de produção democratizando o acesso a softwares para a produção de CDs e música digital. Com essa possibilidade, estúdios de gravação musical de menor porte surgiram, inclusive alguns home stu- dios. Essa reconfiguração do setor levou a novas possibilidades de produ- ção musical para a pequena produtora (em gravadoras indies6), assegurando novos meios de divulgação, consumo e uma nova janela de oportunidade 128 v.29 n.especial 2019 Nova Economia 1281 Michel, Machado & Sátyro para competitividade com as grandes empresas (Wikstrom, 2009; Dolata, 2011; Watson, 2016; Tschmuck, 2016). Analisando algumas consequências da digitalização da música, Wiks- tröm (2009) aﬁrma que houve redução do controle do ﬂuxo de informa- ção pelos maiores players do mercado. Antes da digitalização, a infor- mação apresentava um alto ﬂuxo das empresas em direção à audiência, com conexões fracas no sentido inverso, o que ocasionava um poder às empresas em relação ao processo de informação e comunicação. Cabia a essas empresas a escolha dos meios de difusão, quantidade e público- -alvo. Esse cenário mudou. Devido ao maior ﬂuxo de informações, re- dução de custos e maior acessibilidade aos meios de gravação Wikström (2009) pontua ainda que outro impacto decorrente da digitalização da música (e da conectividade que a acompanha) é a maior presença de mú- sicos e artistas “amadores” em relação aos “proﬁssionais” no mercado. A conectividade permite que consumidores de música passem a produzir conteúdos musicais e carregar esses arquivos nas plataformas como You- Tube, Soundclound, Myspace etc. O fenômeno da digitalização na música é analisado também por Wat- son (2016), que aborda o impacto do que se denomina digital disruption no setor de produção/gravação de música. A partir da ascensão das tec- nologias domésticas (e maior participação de agentes médios e pequenos no setor) e das modiﬁcações ocorridas no modo de produção e consumo, o autor aﬁrma que as empresas de pequeno e médio porte passam a ser grandes expoentes do setor. Tschmuck (2016) também aborda o fenôme- no e chega a conclusões similares. Aﬁrma que as relações no mercado sofrem alterações, pois o peso das grandes empresas é diminuído, en- quanto artistas e pequenos agentes são deslocados para os centros de decisões e participações. Com base nisso, podemos dizer que, mais uma vez, as novas tecno- logias no setor musical diminuem as barreiras à entrada no mercado. 1282 Nova Economia v.29 n.especial 2019 2 Redes de produção periférica na economia criativa e impactos das inovações tecnológicas recentes na música A possibilidade de participação no mercado é ampliada a músicos não ex- perientes e independentes. Janelas de distribuição como o YouTube e as plataformas de streaming funcionam como uma entrada ao segmento de música de nicho, possibilitando maior visibilidade e possibilidade de con- tato com gravadoras e empresários do setor. O aumento da participação de “amadores” no mercado ocorre possibilitado pela tecnologia. Os agentes que estão espacialmente distantes de empresas gravadoras e dos grandes Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia centros dispõem agora de ferramentas que possibilitam sua participação. Assim, presumimos que uma das consequências das novas tecnologias na produção musical é diversificar os locais de produção, alcançando outras regiões do território nacional e outros públicos. Tais conclusões de Wikström (2009) também apresentam diálogo com a hipótese do trabalho. O autor afirma a maior participação de artistas amadores e de menor porte no mercado concorrendo com grandes artis- tas e players em geral. Esse fato gera, de acordo com ele, maiores possibi- lidades de inserção no mercado de trabalho. No entanto, devemos refletir sobre a realidade dessa suposta redução de barreiras à entrada. Talvez seja uma visão otimista e superficial falarmos em redução de barreiras à entrada, quando a própria tecnologia que causa a suposta redução não é ubíqua. As tecnologias de informação e comunicação, embora apresen- tem aumento em número de usuários, ainda são exclusivas de alguns grupos na sociedade global. Assim, buscamos compreender o fenômeno do pequeno, local e inde- pendente. A música produzida em periferias urbanas tem a característica de ser independente em diversos aspectos: independência em relação às grandes gravadoras localizadas nos centros urbanos, à grande mídia e veí- culos de comunicação, e, por vezes, às políticas públicas culturais. Esse caso pode ser observado em algumas produções, especialmente o funk carioca, do tecnobrega paraense, da música rap produzida nas periferias de São Paulo, e também em outras cidades brasileiras como Brasília (Dayrell, 2002; Camargos, 2015). A música é um dos produtos culturais mais consumidos e produzidos nas periferias urbanas. E esse produto é dividido em diferentes estilos mu- sicais que formam grupos com intensos laços de amizade e profissionais, trocas de experiência e criação coletiva. 1283 v.29 n.especial 2019 Nova Economia 2 Redes de produção periférica na economia criativa e impactos das inovações tecnológicas recentes na música Assim, a experiência de consumo e de criação musical nesses espaços gera uma rede de trocas sociais e, tam- bém, econômicas por esses agentes com uma identidade comum e, assim, o sentimento de pertencimento se constrói. Essa rede possui, então, carac- terísticas conceituais próximas à sociologia econômica, como também ca- racterísticas próprias por se tratar de uma rede de criação cultural e criativa em zona periférica que as diferencia da forma como a produção cultural se dá nos lugares que representam o centro da produção. Porém, as redes criativas diferem, em algum sentido, dos conceitos de redes da sociologia econômica, porque apresentam em sua composição 1283 v.29 n.especial 2019 Nova Economia 1283 Michel, Machado & Sátyro fatores especíﬁcos de produção cultural que trazem singularidades ao seu funcionamento. As características de redes apresentadas por Powell (2003) como a comunicação baseada nas relações, as aptidões complementares balizadoras de normas, a reputação dos agentes como resolutórias de con- ﬂitos, incentivo para aprendizado, troca de experiências e conhecimen- to tácito existem nas redes criativas. Entretanto, fatores como a troca de serviços artísticos por outros serviços artísticos e a função colaborativa – que não envolvem trocas monetárias nas produções – também se fazem presentes. A produção e atuação dos agentes no sentido da arte pela arte diferencia essa rede criativa (e cultural) das demais redes econômicas deﬁ- nidas na literatura. Assim, nas redes musicais discutidas por Leyshon (2014), podemos en- contrar possibilidades de diálogo com os conceitos de Powell (2003) e ou- tros autores da sociologia econômica. Entretanto, existem características próprias da produção musical que diferenciam as redes musicais das outras organizações em redes no sistema econômico. Os agentes se organizam através de relações de pré-produção e pós-produção para a construção de produtos e serviços musicais diversos para a comercialização e distribui- ção no mercado. Embora o processo técnico de produção de música seja semelhante, quando analisamos diferentes espaços e territórios, as relações e, por consequência, a estrutura e o signiﬁcado da rede são distintos. Como dito, a música em periferias urbanas é um instrumento de construção de identidade social, e esse fato acarreta em predicados especíﬁcos e únicos das redes de produção musical periféricas. 1284 Nova Economia v.29 n.especial 2019 7 Por se tratar de análise de redes sociais, utilizando a técnica de amostragem de snowball, pode ser que algum dos entrevistados leve a agentes e atores que estejam fora do específico território que, são, posteriormente, incluídos na pesquisa. O limite espacial se dá em relação à produção de música e de estilo. 2 Redes de produção periférica na economia criativa e impactos das inovações tecnológicas recentes na música Além do objetivo físico e ma- terial ou digital, ou seja, a produção do objeto cultural CD ou música di- gital, existe o objetivo da construção identitária e ampliação da atividade musical do território para a comunidade, dilatando assim o sentimento e a possibilidade de pertença dos moradores e indivíduos que frequentam o local. Os agentes envolvidos nessas redes musicais periféricas atuam, então, com intuito econômico, realizando trocas de bens e serviços por unidades monetárias, mas também com objetivo comunitário, realizan- do os mesmos serviços por valores menores ou ainda por trocas de favo- res alternativos. Assim, na próxima seção, descrevemos os instrumentos utilizados para delineamento de uma rede de produção de rap no Grajáu, segundo arca- bouço metodológico da Análise de Redes Sociais. 1284 Nova Economia v.29 n.especial 2019 Música, redes e tecnologia na periferia 3 Aspectos metodológicos da construção da rede A análise de redes, metodologia deste estudo, se baseou em pesquisa de campo com questionário semiestruturado onde se questionam informa- ções concernentes ao uso e impacto de tecnologias digitais, bem como às relações interpessoais. Para esse último aspecto, utilizamos de elementos da Análise de Redes Sociais (doravante ARS). Os dados são construídos a partir de coleta primária de informações sobre o rap na zona sul de São Paulo, especificamente o distrito de Grajaú. Dessa forma, pode-se enten- der que o trabalho está baseado na verificação das hipóteses levantadas a partir da análise do estudo de caso. Ou seja, o trabalho realiza a pesquisa de campo no intuito de responder às questões sobre o uso e impacto das tecnologias, bem como a forma estrutural da produção de rap na região do Grajaú em São Paulo. O primeiro passo foi delimitar os horizontes de alcance da pesquisa de campo para a aplicação dos questionário e constru- ção das redes sociais, bem como determinar as unidades analíticas que são buscadas em campo. Em se tratando de entrevistas iniciais da pesquisa de campo, questiona-se sobre as relações que esses atores realizam. The unit of observation is the entity on which measurements are taken. Most often social network data are collected by observing, interviewing, or questioning individual actors about the ties from these actors to other actors in the set. Thus, the unit of observation is an actor, from whom we elicit information about ties (Wasserman; Faust, 1994, p.43) O primeiro limite (boundarie) ocorre em relação à delimitação urbana es- pecífica da atividade de produção da música rap no território analisado. A proposta analítica aqui é de compreender como se dá a relação de produ- ção e o uso de tecnologia nas periferias urbanas, tendo como caso a zona sul de São Paulo. São, assim, incluídas apenas as unidades analíticas que es- tejam inseridas nesse território,7 a região analisada é a zona sul do municí- pio de São Paulo (especificamente a produção de rap no distrito do Grajaú). 8 A discussão acerca de espaço e território é muito mais ampla e complexa do que as ideias brevemente apresentadas. Aqui a preocupação reside em apenas delimitar os limites de atua- ção da pesquisa de campo a partir do conceito de território. Para maiores detalhes das deﬁni- ções, recomenda-se a leitura das obras referenciadas e também de trabalhos de Edward Soja, Henri Lefébvre, além de outras obras de Milton Santos e Rogério Haesbaert. 9 Os questionários podem ser socilitados através do e-mail rod.cmichel@gmail.com. 3 Aspectos metodológicos da construção da rede Na busca de compreender esse limite, recorremos à obra de Rogério Haesbaert, que apresenta um enfoque do território baseado em três as- pectos: a) jurídico-político: um espaço que recebe influência direta de um determinado poder (estatal); b) cultural: em um território existe uma 1285 v.29 n.especial 2019 Nova Economia 1285 Michel, Machado & Sátyro priorização de aspectos simbólicos e subjetivos; o território pode ser com- preendido como um produto da apropriação do imaginário ou da identida- de sobre o espaço; e c) econômica: analisado em sua perspectiva material, de embate de classes e relações de capital-trabalho (Haesbaert, 2002).8 Baseando nessas deﬁnições de território, delimitamos as áreas analisa- das traduzidas no aspecto cultural do território. Embora tenhamos sele- cionado um (ou mais) bairro(s) de análise dos objetos de estudo, ressalta- mos que os limites são mais ﬂexíveis do que os “bairros administrativos”. Assim, na pesquisa de campo, entendemos que o bairro vizinho (não an- tes selecionado) pode se tornar parte desse território. O bairro é um epi- centro analítico, sabendo, entretanto, que o horizonte de território pode ser mais amplo. O segundo limite da pesquisa de campo diz respeito ao tipo de música realizada nesse território, sendo a música rap o alvo de pesquisa. Unidades analíticas que estejam no mesmo espaço urbano e produzam música em outros estilos não são considerados na amostra. Por ﬁm, o terceiro limite se estendeu à função exercida pela unidade analíti- ca na produção musical. O foco nessa etapa encontrou-se na produção e difusão de música nas periferias selecionadas. Assim, num primeiro mo- mento, identiﬁcaram-se quais os agentes estão envolvidos nessa produ- ção. O RAP apresenta uma estrutura onde diversos agentes são envolvi- dos como artistas, MCs, estúdios de gravação, organizadores de batalhas de rimas, casas de shows e produtores musicais, porém outras funções são envolvidas no processo como artistas visuais para as artes de CDs e shows e outras funções podem ser encontradas em campo. Uma vez deﬁnidas as unidades analíticas, realizou-se a coleta de da- dos em campo a partir da aplicação de questionários com os agentes especíﬁcos.9 Os instrumentos de coleta foram construídos com o intuito de obter as informações de todas as unidades possíveis de análise (pes- soas físicas e jurídicas), com variáveis de atributos e relacionais. Para ﬁns de descrição dos atributos pessoais e socioeconômicos, incluíram-se variáveis como sexo, idade, escolaridade, tamanho da empresa, tempo de atuação, renda etc. 3 Aspectos metodológicos da construção da rede 1286 Nova Economia v.29 n.especial 2019 1286 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia Dois questionários foram elaborados: um voltado para músicos/artistas, e o outro, para agentes diversos, em especial empresários de empresas gra- vadoras, distribuidoras etc. No primeiro foram descritos atributos pessoais e socioeconômicos, bem como as questões relacionais. Os questionários foram divididos em blocos de questões para possibilitar melhor organiza- ção das variáveis. Cada bloco associa-se a uma problemática específica, porém o conjunto dos dados possibilitou uma análise mais detalhada do objeto. Estamos lidando, grosso modo, com o uso de tecnologias e a pro- dução musical em um espaço periférico do município de São Paulo. Para compreender o fenômeno em completude, fez-se necessário o questiona- mento acerca de informações que possibilitem situar cada entrevistado (unidade de análise) em relação à condição social e econômica, à atuação musical e artística, ao uso tecnológico e, por fim, ao posicionamento e ao papel na rede social de produção de música rap. De posse das informações descritas, a próxima etapa compreendeu a pesquisa de campo através da coleta e sistematização de informações da produção de música rap e o uso da tecnologia nas produções realizadas em zonas periféricas urbanas na zona sul de São Paulo. A seleção de entrevistados foi baseada a partir de pesquisa com especialistas nos as- suntos, como músicos consolidados pelo público, e gravadoras de música rap de São Paulo. Como salientado por Lazega e Higgins (2014), existe dificuldade de deli- mitação amostral da coleta de dados das variáveis relacionais e de atributos de uma rede. Os métodos de amostragem, se utilizados para a delimitação das observações a serem analisadas em uma ARS, pressupõem que, a partir de uma rede, seja possível extrair uma amostra que descreva a realidade da rede como um todo. Contudo, utilizar técnicas amostrais onde todos os indivíduos apresentem a mesma probabilidade de serem incluídos na amostra pode ser um problema, pois nem todos os atores apresentam a mesma função, grau relacional e outras características. Assim, no processo amostral pode ser excluído um agente crucial na rede, que liga grupos de atores, por exemplo. Como uma forma de superar esses obstáculos os autores sinalizam para métodos amostrais que se va- lem de técnicas não probabilísticas, como o snowball sampling. v.29 n.especial 2019 Nova Economia 3 Aspectos metodológicos da construção da rede Destarte, a amostra foi iniciada com rappers, MCs e grupos de rap de várias trajetórias na região, tanto aqueles já consolidados no mercado nacional, bem como expoentes na nova geração. Além disso, existem agentes de produção tam- v.29 n.especial 2019 Nova Economia 1287 Michel, Machado & Sátyro bém antigos e novos no cenário do Grajaú. A heterogeneidade de atores escolhidos como seeds garantiu que as entrevistas fossem levadas a outros perﬁs de agentes, partindo do pressuposto de que os indivíduos tendem a ter relações mais próximas com agentes de capital social, cultural e eco- nômico correlatos. O trabalho em campo contou, então, com a deﬁnição dos seeds para a amostra, considerando os instrumentos possíveis (pesqui- sas anteriores, websites, facebook) e buscando a heterogeneidade na seleção para minimizar vieses de seleção. Partindo desses agentes, pressupõe-se que a rede coletada apresentaria também uma diversidade de atores de- correntes dessa delimitação inicial, com o intuito de realizar inferências válidas acerca da produção musical nos territórios. Em ARS, existem diferentes modelos analíticos, tais como redes socio- métricas totais, redes egocentradas, redes de dois modos (ou ﬁliação, ou two-mode), redes multiníveis e outros desenvolvimentos metodológicos. No presente estudo, optamos pelas redes sociométricas totais, pois buscamos compreender as relações estabelecidas entre agentes diversos como um todo. As redes sociométricas totais são redes sem especiﬁcidades de agentes (tipos) e relações.10 Mostram todas as relações entre os agentes, apenas diferenciando-os por seus atributos, e buscam atingir a totalidade de agen- tes em uma rede. Assim, para o presente caso, os músicos, artistas, em- presas e demais agentes são compreendidos como agentes. Então, nessa etapa da metodologia, buscamos responder como é a forma do mercado de produção do rap na região analisada. Quais são as etapas produtivas e as relações estabelecidas pelos músicos para a produção de um álbum. E ainda, analisamos se existem relações únicas dentro da rede, ou seja, se a região é caracterizada por um polo de produção local onde todos os agen- tes (ou a maioria) se relacionam entre si, ou se existem relações estabele- cidas em grupos especíﬁcos (clusters). Na sequência, dividimos essa rede geral em duas redes correlatas, sendo (A’) a rede resultante das relações efetuadas que envolvam trocas monetárias e (A’’) a rede de “colaboração”, apenas com as trocas não monetárias. 10 Para maiores detalhes, ver Lazega e Higgs (2014) e Wasserman e Faust (1994). 1288 Nova Economia v.29 n.especial 2019 3 Aspectos metodológicos da construção da rede O objetivo dessa rede foi respon- der à indagação de que a forma produtiva que emerge na região após os adventos das TICs apresenta um aspecto colaborativo intenso. Pudemos aqui estabelecer as diferenças entre as duas redes e os agentes destacados em cada uma, assim como realizado na rede geral. Ou seja, o desenho de 1288 Nova Economia v.29 n.especial 2019 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia pesquisa e a técnica aqui empregados incorporam elementos essenciais da abordagem teórica ora utilizada e permite estabelecer as conexões entre os agentes e atores envolvidos, de forma a tratar de forma adequada a hipótese trazida. pesquisa e a técnica aqui empregados incorporam elementos essenciais da abordagem teórica ora utilizada e permite estabelecer as conexões entre os agentes e atores envolvidos, de forma a tratar de forma adequada a hipótese trazida. 4 Redes de produção musical periféricas e os impac- tos da tecnologia nos agentes e em suas relações: resultados e discussões A amostra foi composta por 26 músicos e 4 empresários que atuam na pro- dução musical no território do Grajaú. Esse número é justificado, pois de acordo com a técnica amostral apresentada não partimos de um número fixo nem de metas para o cálculo da amostra. Notamos o ponto de satura- ção a partir da 26ª entrevistas, pois as informações coletadas não apresen- tavam mais ineditismo e, assim, optamos por esse número. As entrevistas foram realizadas em dois períodos, o primeiro realizado entre os meses de abril e maio de 2016, e outubro e novembro de 2017. A idade média é de 29,8 anos, sendo o mais jovem entrevistado de 19 anos, e o mais velho 41, indicando boa variabilidade. Todos os entrevis- tados são do sexo masculino. Além disso, verifica-se maior participação de indivíduos que se declaram como negro e pardo, com apenas 12% da amostra declarada como branco. Em relação à escolaridade, observou-se que que a maior parte da amostra é composta por indivíduos que concluí- ram o ensino médio (64% das observações), e 16% fizeram graduação No que tange à educação musical, cerca de 54% dos entrevistados afirmaram ter realizado algum tipo de curso. O tempo de atuação no RAP enquanto agente ligado à produção de música é variável na amostra, nota-se que 56% das pessoas estão há mais de 10 anos no setor, enquanto 24% afirmam atuar no RAP desde um a cinco anos. Esses dados sugerem que a amostra conseguiu captar a heterogeneidade de agentes já estabelecidos e novatos no setor. Mais de 90% dos indivíduos afirmaram que a inserção no RAP se deu no bairro de moradia, evidenciando a existência de um cenário dinâmico da produção musical no local. No que tange à inserção do trabalho artístico na cidade, 43% afirmaram que a atuação até hoje é no bairro onde habitam no distrito de Grajaú; 1289 1289 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro outros 40% registraram que sua atuação é dividida entre as periferias da cidade de São Paulo, como também no centro da cidade; os demais atuam com maior ênfase fora da cidade, na região metropolitana e outros estados. Em se tratando dessa atuação proﬁssional das pessoas, veriﬁcamos que a maior parte dos indivíduos entrevistados considera a atividade no RAP como emprego secundário. CD hoje cada vez tá ﬁcando mais difícil vender CD. A gente chegou até num consenso de quando a gente faz show faz uma premiação, no palco ali distribui. Porque acaba que você ﬁca com material parado, e não compensa. Então se você tá em outra área, cê cativa, cê fala: “dá uma escutada lá, depois você comenta, entra na página”. Daí fala de página, das redes sociais. Então é mais um cartão de visita. A gente não espera mais grana do CD. A gente já insistiu nisso, numa época anterior (Entrevistado A8). 1290 Nova Economia v.29 n.especial 2019 4 Redes de produção musical periféricas e os impac- tos da tecnologia nos agentes e em suas relações: resultados e discussões Cerca de 80% possuem outro emprego, sendo que, desse montante, 25% realizam atividades ligadas às indústrias criati- vas como, por exemplo, produtor cultural, professor de hip-hop, tatuador e administração de galeria de arte. A remuneração por shows é a fonte de renda da atividade musical mais enunciada pelos entrevistados como principal, seguida por venda de produtos relacionados, como artigos de vestuário e adereços. Tanto a venda de CDs físicos como a remuneração através das plataformas de broadcasting e streaming foram citadas por ape- nas um entrevistado. Dessa forma, no sentido de remuneração pelo traba- lho musical dos artistas analisados, a lógica segue a tendência da indústria fonográﬁca, em proporções menores. Assim como atualmente os grandes artistas mundiais auferem lucros através de shows, os artistas situados na periferia do Grajaú também encontram nas contratações de shows fonte de renda para a atividade musical. CD hoje cada vez tá ﬁcando mais difícil vender CD. A gente chegou até num consenso de quando a gente faz show faz uma premiação, no palco ali distribui. Porque acaba que você ﬁca com material parado, e não compensa. Então se você tá em outra área, cê cativa, cê fala: “dá uma escutada lá, depois você comenta, entra na página”. Daí fala de página, das redes sociais. Então é mais um cartão de visita. A gente não espera mais grana do CD. A gente já insistiu nisso, numa época anterior (Entrevistado A8). Notamos, então, que a venda de unidades físicas de CDs e EPs repre- senta uma oportunidade de maior visibilidade e divulgação do trabalho. O recurso de plataformas de música digital também se comporta da mesma forma, uma vez que poucos entrevistados aﬁrmam conseguir recursos com essas ferramentas. Em suma, estamos diante de uma amostra que apre- senta as seguintes características, em geral: o trabalho artístico-musical é a segunda ocupação com uma geração de incremento de renda pequena. A pesquisa conseguiu captar indivíduos novatos e veteranos na atuação no rap e assim, permite que possamos analisar informações de agentes que sentiram as mudanças das novas tecnologias durante sua atuação, bem como de agentes que iniciaram suas carreiras no novo cenário tecnoló- gico. Dessa forma, as inovações tecnológicas, a digitalização da música, Notamos, então, que a venda de unidades físicas de CDs e EPs repre- senta uma oportunidade de maior visibilidade e divulgação do trabalho. 4 Redes de produção musical periféricas e os impac- tos da tecnologia nos agentes e em suas relações: resultados e discussões Notamos, então, que a venda de unidades físicas de CDs e EPs repre- senta uma oportunidade de maior visibilidade e divulgação do trabalho. 1290 Nova Economia v.29 n.especial 2019 Música, redes e tecnologia na periferia afetaram e têm afetado os entrevistados ao possibilitar uma nova forma de produção mais acessível em termos espaciais e econômicos (através da redução de custos). 1 v.29 n.especial 2019 Nova Economia 4.1 A rede de produção de rap no Grajaú: centralidades reforçadas pela tecnologia Primeiramente, apresentamos os resultados da rede geral, ou seja, aquela onde todas as relações são medidas e mostradas, não fazendo diferença entre relações que envolvem trocas monetárias ou não. A partir da com- preensão dessa rede, pudemos identificar o tamanho da rede de produção do rap, baseada nos entrevistados, horizonte de alcance das atividades em relação a funções e territórios alcançados e estrutura relacional. Ficou clara a importância que os entrevistados percebem na dinami- zação de um mercado de produção local, como forma de atuação no mercado musical e, além disso, notamos outras informações importantes como a criação e consolidação de uma rede de produção própria, que é capaz de produzir música, sentido identitário, sentido de coletividade e autonomia do Grajaú e da periferia. A ideia de movimentação e valori- zação dos trabalhos artísticos locais foi percebida nas entrevistas reali- zadas. Alguns entrevistados afirmaram a preferência em trabalhar com indivíduos do território, pois existe uma identificação e a realização em movimentar a cena local. A Figura 1 apresenta a rede geral de produção do RAP. As cores repre- sentam a função exercida na rede. Indivíduos em vermelho representam rappers e músicos em geral, sendo que o vermelho mais escuro representa indivíduos que foram entrevistados diretamente na pesquisa de campo, enquanto vermelho mais claro são agentes citados por esses entrevista- dos. Observações em verde indicam agentes responsáveis por gravação, masterização, produção, fabricação de beats e DJs; indivíduos em azul rea- lizam funções de design, audiovisual e artes gráficas, enquanto agentes em amarelo foram classificados em outras funções. As formas geométricas indicam a localização dos agentes, onde círculos representam o Grajaú, quadrados indicam agentes localizados em outros espaços do município de São Paulo, triângulos representam outras cidades, estados e países e, por fim, losangos indicam agentes com localização indefinida. O tama- 1291 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro nho das formas geométricas está relacionado ao grau de intermediação dos agentes, obtida através das estatísticas de betweeness. 1292 Nova Economia v.29 n.especial 2019 Figura 1 Rede geral de produção do rap no Grajaú Fonte: Elaboração própria. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráﬁcas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por ﬁm, losangos: agentes com localização indeﬁnida. Fonte: Elaboração própria. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráﬁcas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por ﬁm, losangos: agentes com localização indeﬁnida. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráﬁcas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por ﬁm, losangos: agentes com localização indeﬁnida. A entrevista nos levou a uma rede composta por 230 atores. Cerca de 50% desses residem no distrito do Grajaú, 18% em outras regiões da capital paulista, 7% outras cidades, estados e municípios, e cerca de 25% com localização indeﬁnida. Esse fato evidencia que, embora se tratando de uma periferia urbana, a produção de música encontra diversos agentes locais no próprio distrito do Grajaú, porém estabelece diálogo com outras regiões do município bem como com outras cidades. Desses agentes, 22% foram classiﬁcados como rappers e músicos, 38% agentes produtores e gravado- res, 17% trabalham com artes gráﬁcas e/ou audiovisual, e 14% atuam em funções classiﬁcadas como “outros”. O restante da amostra corresponde aos agentes entrevistados em campo (rappers, DJs e produtores) A densidade da rede geral da rede é de 0,010, indicando uma rede pouco densa, onde apenas 1% de todos os laços possíveis entre os agentes são estabelecidos. Esse baixo valor não surpreende, posto a metodologia de co- leta dos dados relacionais. A pesquisa de campo captou informações acerca das relações exercidas pelos agentes entrevistados, não sendo possível es- Música, redes e tecnologia na periferia tabelecer as relações entre os agentes citados por estes. Logo, não podemos levantar informações e conclusões acerca da densidade total da rede. 1293 v.29 n.especial 2019 Nova Economia Fonte: Elaboração própria. Fonte: Elaboração própria. Figura 1 Rede geral de produção do rap no Grajaú A distância média entre os agentes da rede é de 4,3 contatos, indicando que, perante a estrutura relacional da rede, para que um agente qualquer estabeleça relações com qualquer outro agente é preciso passar por 4,3 indivíduos. Essa estatística apresentou-se baixa, indicando facilidade de contatos na rede. Assim, as relações, indicações e reputações dos agentes nessa rede são importantes, pois, como os contatos são de fácil acesso, a troca de informações entre os agentes nela envolvidos é alta, facilitando o conhecimento da atuação e história pessoal entre eles. Tabela 1 Estatísticas de centralidade na rede geral Agente Intermediação Intermediação normalizada Grau A7 18.654.938 37.346 15.000 A10 14.913.010 29.855 39.000 A3 12.890.900 25.807 10.000 A18 11.817.079 23.657 31.000 A6 10.340.000 20.700 16.000 A13 10.202.692 20.425 28.000 B149 10.192.000 20.404 2.000 A5 9.852.552 19.724 28.000 A19 9.781.632 19.582 11.000 B199 6.402.000 12.816 2.000 A12 6.214.000 12.440 15.000 A1 4.752.000 9.513 13.000 B117 4.409.452 8.827 6.000 A15 4.383.987 8.776 14.000 A9 4.330.000 8.668 11.000 A2 3.796.558 7.600 12.000 A16 3.531.901 7.071 11.000 A8 3.397.931 6.802 7.000 A17 3.052.000 6.110 9.000 A14 2.761.643 5.529 12.000 A4 2.095.491 4.195 7.000 A11 1.756.000 3.515 5.000 1293 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro Michel, Machado & Sátyro Em relação à centralidade e importância dos agentes na rede geral, apre- sentamos as estatísticas de grau e intermediação dos principais agentes na Tabela 1. Os nomes dos agentes foram construídos da seguinte forma A(X) indica os indivíduos entrevistados diretamente na pesquisa de campo, en- quanto B(X) são agentes que foram citados pelos entrevistados. O agente com maior grau de intermediação (A7), embora tenha trabalhos enquanto rapper e música, exerce primordialmente a função de produtor musical com um estúdio montado em sua residência. Esse agente foi citado por 15 indivíduos e respondeu estabelecer contatos com 8 indivíduos. O segundo agente com maior intermediação (A10) atua como rapper solo e apresenta a maior rede pessoal entre os entrevistados, com 39 contatos. Desse fato, pudemos observar que a possibilidade de criação de um home studio leva a atuações importantes na rede de produção. O agente A6 apresenta a quinta intermediação mais elevada e também trabalha como produtor em um estúdio montado em sua residência. Aqui nós temos vários proﬁssionais, várias ramiﬁcações. Basta a gente se olhar, se valorizar, né? E fazendo essa troca, essa parceria, que eu acho maneiro. Envolve gra- na assim, que a gente tá se envolvendo, apostando num trabalho. Se isso repercutir, se isso tiver uma repercussão, eu acho justo partilhar, eu acho justo. O cara se dispôs a fazer o trampo, eu acho isso justo, tá ligado? Mas não é na base da troca, cê me dá isso eu te dou aquilo. Mas se acontecer, não tem porque não, tá ligado. Eu acho justo com o cara. Não é um proﬁssional? Assim como eu quero ser entendido como um pro- ﬁssional? Eu tenho que respeitar o trabalho do cara, mano. Aﬁnal, de contas, senão eu vou tá pensando só em mim, e como ﬁcam as outras pessoas? (Entrevistado A5). Figura 1 Rede geral de produção do rap no Grajaú Assim, a possibili- dade que as TICs trouxeram de construção dessas empresas de gravação no domicílio leva esses agentes a realizarem produções de artistas locais e a se caracterizarem como um ponto intermediador na rede, que conecta agentes de diversas funções. 1294 Nova Economia v.29 n.especial 2019 4.2 As trocas monetárias na produção de rap: a centralidade dos agentes de produção nas relações A produção de música rap no Grajaú envolve diversas relações, como vis- to na rede geral. O número de laços efetuados pelos entrevistados é va- riável, sendo que alguns entrevistados realizam seu trabalho com poucos parceiros e alguns criam redes de produção e criação mais abrangentes. Apresentamos nesta seção a rede de relações que envolvem apenas trocas monetárias entre os agentes. Aqui nós temos vários proﬁssionais, várias ramiﬁcações. Basta a gente se olhar, se valorizar, né? E fazendo essa troca, essa parceria, que eu acho maneiro. Envolve gra- na assim, que a gente tá se envolvendo, apostando num trabalho. Se isso repercutir, se isso tiver uma repercussão, eu acho justo partilhar, eu acho justo. O cara se dispôs a fazer o trampo, eu acho isso justo, tá ligado? Mas não é na base da troca, cê me dá isso eu te dou aquilo. Mas se acontecer, não tem porque não, tá ligado. Eu acho justo com o cara. Não é um proﬁssional? Assim como eu quero ser entendido como um pro- ﬁssional? Eu tenho que respeitar o trabalho do cara, mano. Aﬁnal, de contas, senão eu vou tá pensando só em mim, e como ﬁcam as outras pessoas? (Entrevistado A5). 1294 Nova Economia v.29 n.especial 2019 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia As relações estabelecidas que envolvem trocas monetárias por bens e ser- viços para a produção de música no local podem ser observadas na Figura 2, que apresenta apenas as relações que compõem o que denominamos de rede monetária. Essa rede é composta por 106 agentes. Do total, cerca de 57% residem no Grajaú, 18% em outras regiões da cidade de São Paulo, 7% em outras cidades, e 18% apresentam localidade indefinida. Da variedade dos agentes envolvidos, cerca de 42% dos atores desem- penham função de produção e/ou gravação, representando a maior par- te das relações monetizadas na rede de produção de música RAP; 15% atuam como atividades de artes gráficas e audiovisual; 5% são rappers e músicos (não entrevistados na pesquisa de campo) e participam como contribuições artísticas; 20% atuam em funções diversas que classifi- camos como “outros”; o restante da amostra é composto pelos agentes entrevistados em campo. 4.2 As trocas monetárias na produção de rap: a centralidade dos agentes de produção nas relações Aqui se percebe uma diferença em relação à rede geral: no que condiz à distribuição das funções, a maior parte dos indivíduos que estabelecem relações monetárias são os produtores e gra- vadores, similar com a rede anterior, porém a participação de rappers e músicos nas relações monetárias é de apenas 5%. Portanto, os indivíduos que possuem a tecnologia para a produção e gravação de músicas são os que mais detêm relações monetárias na rede de produção do Grajaú. Figura 2 Rede monetária Figura 2 Rede monetária Figura 2 Rede monetária Fonte: Elaboração própria. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráficas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por fim, losangos: Figura 2 Rede monetária Figura 2 Rede monetária Fonte: Elaboração própria. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráficas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por fim, losangos: ç p p Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráficas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por fim, losangos: 1295 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro Michel, Machado & Sátyro agentes com localização indeﬁnida. A densidade da rede geral da rede é de 0,021, indicando, também, uma rede de baixa densidade, com apenas 2,1% de todos os laços possíveis de fato estabelecidos. Entretanto, devemos lembrar a observação sobre essa esta- tística quando analisamos a rede geral. A distância média entre os agentes da rede é de 2,19 contatos. Sendo menor do que a rede geral, o que é espe- rado, pois o número de agentes também é menor. Mais uma vez, notamos que as relações e reputações dos agentes são importantes, dada a possibi- lidade de ﬂuxo de informações e facilidade de contatos entre os agentes. A densidade da rede geral da rede é de 0,021, indicando, também, uma rede de baixa densidade, com apenas 2,1% de todos os laços possíveis de fato estabelecidos. Entretanto, devemos lembrar a observação sobre essa esta- tística quando analisamos a rede geral. A distância média entre os agentes da rede é de 2,19 contatos. Sendo menor do que a rede geral, o que é espe- rado, pois o número de agentes também é menor. Mais uma vez, notamos que as relações e reputações dos agentes são importantes, dada a possibi- lidade de ﬂuxo de informações e facilidade de contatos entre os agentes. Tabela 2 Estatísticas de centralidade na rede monetária Agente Intermediação Intermediação normalizada Grau A7 3.197.036 29.277 13.000 A5 3.137.430 28.731 21.000 A18 1.998.352 18.300 8.000 A13 1.512.594 13.852 10.000 B117 1.454.788 13.322 5.000 A10 1.364.121 12.492 9.000 A8 1.129.067 10.339 5.000 A3 1.120.000 10.256 8.000 A15 1.038.600 9.511 9.000 A16 1.004.267 9.197 8.000 A14 810.000 7.418 7.000 A11 652.000 5.971 5.000 A4 559.636 5.125 4.000 A2 426.000 3.901 6.000 B159 246.000 2.253 1.000 A1 166.000 1.520 3.000 B116 63.255 0.579 2.000 A9 20.000 0.183 5.000 A6 12.000 0.11 4.000 A12 6.000 0.055 3.000 A17 6.000 0.055 3.000 B92 0.2 0.002 1.000 B175 0.2 0.002 1.000 F t El b ã ó i Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia A Tabela 2 mostra as estatísticas de Grau e Intermediação dos principais agentes na rede monetária. O agente com maior grau de intermediação (A7) é o mesmo agente da rede geral. Esse agente foi citado treze vezes e respondeu estabelecer contatos com oito indivíduos. O segundo agente com maior intermediação na rede monetária (A5) é diferente da rede geral. agentes com localização indeﬁnida. Trata-se de um agente que atua como produtor e DJ na região, trabalhando com diversos agentes entrevistados em campo. O quinto também é um produtor musical (B117). Notamos, portanto, que as tran- sações monetárias levam os agentes produtores a se inserirem com maior poder na rede de relações. 1298 Nova Economia v.29 n.especial 2019 4.3 Relações colaborativas e as criações coletivas de músicos e artistas do rap no Grajaú A última rede a ser apresentada mostra as relações que não envolvem trocas monetárias entre os agentes. Trata-se de serviços artísticos ou não que envolveram a criação, produção, gravação de músicas, vídeos e outras funções ligadas à produção musical de rap no Grajaú. Embora essas relações não envolvam trocas monetárias, compreendemos que elas são baseadas na confiança mútua e na reputação criada na trajetória do indivíduo. Na verdade, são outras coisas que fazem girar assim. É que cê fica com uma dí- vida de gratidão que ela fica uma coisa natural. Cê não precisa falar – oh, mano, cê tá me devendo depois cê paga – Num tem, quando cê tiver cê paga. Isso é uma coisa que já tá predefinida assim, transcendental. Ninguém precisa explicar isso, a gente sabe que as pessoas que ajudam as coisas acontecem (Entrevistado A1). A rede colaborativa é composta por 157 agentes e pode ser observada na Figura 3. Cerca de 52% dos indivíduos residem no Grajaú, 15% em outras regiões da cidade, 6% em outras localidades, e o restante com localiza- ção indefinida. Quase 30% dos agentes são músicos e rappers; 34% de- sempenham função de produção, gravação e afins; 14% funções de artes gráficas e visuais, 10% em outras, e o restante da rede é composta pelos agentes entrevistados em campos. Observamos aqui a maior participação de músicos e rappers quando comparamos com a rede monetária. Assim, concluímos que as relações estabelecidas entre os músicos se dão no sen- tido de colaboração na criação de músicas e de participações em álbuns e faixas, não envolvendo trocas monetárias. Embora a participação seja de 1297 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro gravadores e produtores, é perceptível que, quando artistas e músicos se relacionam entre si, a probabilidade de essa relação não ser monetária é maior, sugerindo a colaboração artística. Figura 3 Rede colaborativa Figura 3 Rede colaborativa Fonte: Elaboração própria. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráﬁcas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por ﬁm, losangos: agentes com localização indeﬁnida. Figura 3 Rede colaborativa Figura 3 Rede colaborativa Fonte: Elaboração própria. Legenda: Preto: rappers e músicos em geral; cinza-escuro: responsáveis por gravação, masterização, produção, fabricação de beats e DJs; cinza-claro: funções de design, audiovisual e artes gráﬁcas, bran- co: outras funções. As formas geométricas indicam a localidade, círculos: Grajaú, quadrados: outros espaços do município de São Paulo, triângulos: outras cidades, estados e países e, por ﬁm, losangos: agentes com localização indeﬁnida. Nota-se, assim, que a rede não monetária é maior do que as relações mo- netárias na produção de música RAP, indicando forte presença de aspectos colaborativos e comunitários no local. A quantidade de relações entre in- divíduos residentes do Grajaú é similar nas redes econômicas e comuni- tárias, indicando que o mercado de produção local se sustenta econômica e comunitariamente. Assim, o mercado que se originou e se desenvolve a partir da tecnologia garante um ﬂuxo econômico entre os agentes, mas também o fortalecimento das relações comunitárias, contribuindo para uma identidade musical local e coletiva. A densidade da rede geral da rede é de 0,014, o que caracteriza uma rede de baixa densidade com apenas 1,4% de todos os laços possíveis estabele- cidos de fato. A distância média entre os agentes da rede é de 2,14 conta- tos, sendo muito similar à rede monetária. A Tabela 3 traz as estatísticas de grau e intermediação dos principais agentes na rede colaborativa. O agente com maior grau de intermediação (A10) é um rapper bastante conhecido e Música, redes e tecnologia na periferia citado por boa parte dos agentes entrevistados na rede comunitária, bem como estabelece diversas relações. O segundo agente com maior interme- diação na rede colaborativa (A7) corresponde ao agente com maior grau de intermediação nas redes geral e econômica, mantendo também sua impor- tância na rede colaborativa. Figura 3 Rede colaborativa Tabela 3 Estatísticas de centralidade na rede colaborativa Agente Intermediação Intermediação normalizada Grau A10 3.221.650 26.647 32.000 A7 2.981.567 24.661 13.000 A18 2.823.917 23.357 28.000 A13 2.091.167 17.297 21.000 A19 2.064.683 17.078 8.000 B199 1.495.000 12.366 2.000 A12 1.458.000 12.060 13.000 A1 1.225.000 10.132 11.000 A5 1.025.067 8.479 11.000 A15 625.000 5.170 6.000 A17 502.000 4.152 6.000 A2 398.250 3.294 7.000 A4 378.000 3.127 4.000 A14 328.300 2.715 7.000 B91 204.400 1.691 4.000 A3 127.000 1.050 2.000 A8 127.000 1.050 2.000 A6 115.000 0.951 11.000 A9 81.000 0.67 7.000 B126 44.000 0.364 2.000 Fonte: Elaboração própria Tabela 3 Estatísticas de centralidade na rede colaborativa Em resumo, podemos notar que o maior número de relações entres agen- tes entrevistados corresponde àquelas voltadas para as funções de grava- ção e produção de música nas três diferentes redes. Através das análises das entrevistas, percebemos que tais funções foram possibilitadas no ter- ritório através das TICs (reduzindo custos e permitindo a produção “do- méstica”) e com o advento da internet. Tanto na rede geral como na rede 1299 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro monetária, o agente com maior intermediação corresponde a um produtor que possui um estúdio caseiro que aﬁrmou ter adquirido o conhecimento de gravação através da internet. Na rede colaborativa, esse mesmo agente aparece como o segundo maior intermediador. Os rappers e músicos em geral apresentam baixa participação na rede monetária e alta participação na rede colaborativa, indicando que a troca monetária entre os agentes do Grajaú não se dá entre músicos, e sim em relações músicos-produtores; músicos-gravadores; artistas-artistas visuais. Os resultados do estudo mostram que, embora o setor seja recorrente- mente alvo de inovações tecnológicas desde seu princípio, as TICs trouxe- ram nova dinâmica ao setor, com alterações que, quando comparadas às tecnologias anteriores, proporcionam nova forma de atuação dos agentes, sejam estes músicos ou produtores. A redução de custos de produção e distribuição de música possibilitou ampliação do acesso às tecnologias de produção. Os entrevistados discorreram sobre tal fato a partir de suas ex- periências próprias, bem como da observação do comportamento de seus pares, evidenciando a importância da internet como forma de obtenção dos instrumentos produtivos (softwares e plataformas) e também como for- ma de obtenção de conhecimento técnico para essa produção. 1300 Nova Economia v.29 n.especial 2019 Figura 3 Rede colaborativa O mesmo ocorre com a distribuição, uma vez que músicos e produtores atualmente dispõem da possibilidade de divulgação de seus trabalhos nas redes so- ciais, plataformas de streaming e broadcasting, bem como sites pessoais e outros. Assim, podemos dizer que, para a amostra analisada, as TICs trou- xeram maior dinamização na produção de música a partir de maiores fa- cilidades no acesso à produção. Nesse sentido, é possível creditar às novas tecnologias a inserção de novos atores no mercado da música, inclusive aqueles situados em zonas periféricas, além de ofertar mais facilidade de manutenção daqueles já consolidados no mercado. A maior proximidade com os processos produtivos e tecnológicos trou- xe aos artistas situados em zonas periféricas um modo de produção de música baseado nas relações locais. As redes construídas no presente tra- balho mostram a preponderância de agentes locais (residentes no Grajaú e região) na produção e distribuição de música. O resultado são músicas, vídeos, e produtos derivados tipicamente do território em questão. A mo- vimentação das relações segue uma tendência de privilegiar os laços co- munitários. Essa tendência se sustenta tanto pelo respeito e pelas caracte- rísticas de identidade como, também, por questões econômicas, pois, em 1300 Nova Economia v.29 n.especial 2019 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia algumas entrevistas, foi dito que os preços praticados pelos serviços de produção para os residentes são mais baixos. Assim, caracterizamos a estrutura de produção de rap no Grajaú como intensa em aspectos colaborativos entre os músicos, com altas participa- ções e reciprocidades entre eles, ao mesmo tempo que as relações mone- tárias são efetuadas com produtores e gravadores, beneficiados pelas TICs nas mudanças de processos produtivos. 1 v.29 n.especial 2019 Nova Economia 5 Considerações finais As inovações tecnológicas ocorrem no cerne do setor de produção musi- cal desde sua gênese e geram impactos decorrentes em plurais aspectos: a) no formato do produto e, consequentemente, no modo de consumo; b) na estrutura de mercado, através de fusões e aquisições de empresas, surgimento de novas empresas e expulsão de outras, e; c) na forma de produção musical, com novas tecnologias e novos processos tecnológicos. Especificamente em relação a este último aspecto, novos processos tec- nológicos levam a novas formas organizacionais e produtivas no interior do setor. Esses impactos se refletem, também, na redução de custos para a produção de música, bem como na descentralização da produção, uma vez que a produção doméstica é facilitada através dos home studios. O uso de análise de redes permite também observar como se dão os laços e como isso ajuda a criar esses impactos, entretanto, o desenho aqui utilizado nos leva a crer que, como tais laços são fortemente comunitários, eles sugerem uma restrição ao alcance de atores e instituições para além do Grajaú. No entanto, essa é uma hipótese a ser investigada em agenda de pesquisa fu- tura. O que é certo são os impactos gerados pelas TICs no sentido de for- mato, processos produtivos, formas organizacionais e acesso à tecnologia. No caso analisado, percebemos que esses impactos chegam aos músicos e produtores situados na periferia em questão, alterando a forma de pro- dução, distribuição e consumo. Permite-nos, também, observar como os artistas podem ser levados a se tornarem produtores e a se ligarem com consumidores que antes não alcançariam. Para além dos resultados alcançados neste trabalho, sugerimos que, em potenciais trabalhos sobre segmentos da economia criativa, recorra-se a métodos como o de Análises de Redes Sociais por possibilitarem mapear 1301 v.29 n.especial 2019 Nova Economia Michel, Machado & Sátyro as relações socioeconômicas em estruturas de produção e de distribuição de menor escala e de maior diferenciação de produto como o reportado neste estudo. as relações socioeconômicas em estruturas de produção e de distribuição de menor escala e de maior diferenciação de produto como o reportado neste estudo. Referências Edward Elgar Publishing, 2016. WASSERMAN, S.; FAUST, K. Social network analysis: Methods and applications. [S.l.]: Cam- bridge University Press, 1994. WATSON, A. Digital disruption and recording studio diversiﬁcation: Changing business models for the digital age. In: ____. Business innovation and disruption in the music industry. Edward Elgar Publishing, 2016. WIKSTRÖM, P. The Music Industry: digital media and society series. [S.l.]: Cambridge: Polity Press, 2009. Sobre os autores Rodrigo Cavalcante Michel – rod.cmichel@gmail.com Rodrigo Cavalcante Michel – rod.cmichel@gmail.com Doutor em Economia pelo Centro de Desenvolvimento e Planejamento Regional (Cedeplar) da Universidade Federal de Minas Gerais (UFMG). Belo Horizonte, Minas Gerais, Brasil. ORCID: https://orcid.org/0000-0002-8962-9715. Doutor em Economia pelo Centro de Desenvolvimento e Planejamento Regional (Cedeplar) da Universidade Federal de Minas Gerais (UFMG). Belo Horizonte, Minas Gerais, Brasil. Ana Flávia Machado – anaﬂaviatavaresmachado@gmail.com Professora Associada da Faculdade do Centro de Desenvolvimento e Planejamento Regional (Cedeplar) da Universidade Federal de Minas Gerais (UFMG). Belo Horizonte, Minas Gerais, Brasil. ORCID: https://orcid.org/0000-0001-8573-7906. Natália Guimarães Duarte Sátyro – nsatyro@gmail.com Professora Associada do Departamento e do Programa de Pós-Graduação de Ciência Política da Universidade Federal de Minas Gerais (UFMG), Belo Horizonte, Minas Gerais, Brasil. ORCID: https://orcid org/0000-0001-7298-342X Ana Flávia Machado – anaﬂaviatavaresmachado@gmail.com Professora Associada da Faculdade do Centro de Desenvolvimento e Planejamento Regional (Cedeplar) da Universidade Federal de Minas Gerais (UFMG). Belo Horizonte, Minas Gerais, Brasil. ORCID: https://orcid.org/0000-0001-8573-7906. Professora Associada da Faculdade do Centro de Desenvolvimento e Planejamento Regional (Cedeplar) da Universidade Federal de Minas Gerais (UFMG). Belo Horizonte, Minas Gerais, Brasil. p g Natália Guimarães Duarte Sátyro – nsatyro@gmail.com Professora Associada do Departamento e do Programa de Pós-Graduação de Ciência Política da Universidade Federal de Minas Gerais (UFMG), Belo Horizonte, Minas Gerais, Brasil. ORCID h // id /0000 0001 7298 342X Natália Guimarães Duarte Sátyro – nsatyro@gmail.com Professora Associada do Departamento e do Programa de Pós-Graduação de Ciência Política da Universidade Federal de Minas Gerais (UFMG), Belo Horizonte, Minas Gerais, Brasil. ORCID: https://orcid.org/0000-0001-7298-342X. Professora Associada do Departamento e do Programa de Pós-Graduação de Ciência Política da Universidade Federal de Minas Gerais (UFMG), Belo Horizonte, Minas Gerais, Brasil. Referências BEINHOCKER, E. D. The origin of wealth: Evolution, complexity, and the radical remaking of economics. [S.l.]: Harvard Business Press, 2006. BENDASSOLLI, P. F. et al. Indústrias criativas: deﬁnição, limites e possibilidades. RAE-Revista de Administração de Empresas, Fundação Getulio Vargas, v. 49, n. 1, 2009. CAMARGOS, R. Rap e política: percepções da vida social brasileira. São Paulo: Boitempo Editorial, 2015. COMUNIAN, R. Exploring the role of networks in the creative economy of North East Eng- land: economic and cultural dynamics. In: ___. Encounters and engagements between economic and cultural geography. Springer, Dordrecht, 2011. p. 143-157. DAYRELL, J. O rap e o funk na socialização da juventude. Educação e pesquisa, v. 28, n. 1, p. 117-136, 2002. DIAS, M. T. Os donos da voz: indústria fonográﬁca brasileira e mundialização da cultura. [S.l.]: Boitempo Editorial, 2000. DOLATA, U. The music industry and the internet: A decade of disruptive and uncontrolled sectoral change. SOI Discussion Paper 2011-02, 2011. HAESBAERT, R. Territórios alternativos. São Paulo: Contexto, 2002. HESMONDHALGH, D. The cultural industries. London: Sage, 2002. HERSCHMANN, M.; KISCHINHEVSKY, M. Indústria da música – uma crise anunciada. Tra- balho apresentado ao NP, 2005. LAZEGA, E.; HIGGINS, S. S. Redes sociais e estruturas relacionais. Belo Horizonte: Fino Traço Editora, 2014. LEYSHON, A. Reformatted: Code, networks, and the transformation of the music industry. Oxford: Oxford University Press, USA, 2014. NAKANO, D. et al. A produção independente e a desverticalização da cadeia produtiva da música. Gestão & Produção, Universidade Federal de São Carlos, v. 17, n. 3, p. 627-638, 2010. ORMEROD, P; POTTS, J; CUNNINGHAM, S; HARTLEY, J. Social network markets: A new deﬁnition of the creative industries. Journal of Cultural Economics, 32. 167-185. 10.1007/ s10824-008-9066-y, 2008 POTTS, Jason et al. Social network markets: a new deﬁnition of the creative industries. Journal of cultural economics, v. 32, n. 3, p. 167-185, 2008. POWELL, W. Neither market nor hierarchy. The sociology of organizations: classic, contemporary, and critical readings, v. 315, p. 104-117, 2003. 1302 Nova Economia v.29 n.especial 2019 Música, redes e tecnologia na periferia Música, redes e tecnologia na periferia TSCHMUCK, P. From record selling to cultural entrepreneurship: The music economy in the digital paradigm shift. In: ___ Business innovation and disruption in the music industry. Edward Elgar Publishing, 2016. TSCHMUCK, P. From record selling to cultural entrepreneurship: The music economy in the digital paradigm shift. In: ___ Business innovation and disruption in the music industry. This article is licensed under a Creative Commons Attribution 4.0 International License. 1303 v.29 n.especial 2019 Nova Economia Sobre o artigo Recebido em 16 de outubro de 2019. Aprovado em 18 de fevereiro de 2020.
https://openalex.org/W4246932797	https://www.qeios.com/read/HH3WQL/pdf	English	null	Aldesleukin	Definitions	2,020	cc-by	168	Qeios · Definition, February 2, 2020 Open Peer Review on Qeios Open Peer Review on Qeios Aldesleukin National Cancer Institute Qeios ID: HH3WQL · https://doi.org/10.32388/HH3WQL Open Peer Review on Qeios Source National Cancer Institute. Aldesleukin. NCI Thesaurus. Code C1498. A recombinant analog of the endogenous cytokine interleukin-2 (IL-2) with immunoregulatory and antineoplastic activities. Aldesleukin binds to and activates the IL- 2 receptor, followed by heterodimerization of the cytoplasmic domains of the IL-2R beta and gamma(c) chains; activation of the tyrosine kinase Jak3; and phosphorylation of tyrosine residues on the IL-2R beta chain, resulting in an activated receptor complex. Various cytoplasmic signaling molecules are recruited to the activated receptor complex and become substrates for regulatory enzymes that are associated with the receptor complex. This agent enhances lymphocyte mitogenesis; stimulates long-term growth of human IL-2 dependent cell lines; enhances lymphocyte cytotoxicity; induces lymphokine- activated killer (LAK) cell and natural killer (NK) cell activities; and induces expression of interferon-gamma. Aldesleukin may induce T cell-mediated tumor regression in some tumor types. 1/1
https://openalex.org/W4390235107	https://iieta.org/download/file/fid/115564	English	null	Global Trends in Poliomyelitis Research: A Bibliometric Analysis from 1857–2019	International journal of design & nature and ecodynamics	2,023	cc-by	10,654	ERROR: type should be string, got "https://doi.org/10.18280/ijdne.180606 \nABSTRACT https://doi.org/10.18280/ijdne.180606 \nABSTRACT \nReceived: 23 October 2023 \nRevised: 9 November 2023 \nAccepted: 15 December 2023 \nAvailable online: 26 December 2023 \nThe evolution o\naffecting childre\nevaluated throu\nthousand one hu\nof Science and S The evolution of poliomyelitis (polio) research, an acute viral infection predominantly \naffecting children and ranging from mild illness to disabling paralysis, was systematically \nevaluated through a bibliometric analysis of publications from 1857 to 2019. Six \nthousand one hundred thirty-nine polio-related publications were extracted from the Web \nof Science and Scopus databases, employing 'Polio' as the keyword. Publications were \npredominantly in article format (71.16%), with English-language documents constituting \n30% of the dataset. The analysis, executed using EndNote, Microsoft Excel, R \n(Biblioshiny), and VOSviewer, identified 2014 as the peak year of polio literature output, \nfeaturing 340 publications. A shift from single to collaborative authorship was observed, \nwith multiple authorship patterns emerging more prominently. This trend highlights a \ntransition towards interdisciplinary collaboration in the field. Despite the dominance of \narticles, a diverse array of sixteen document types was catalogued. The journal 'Vaccine' \nwas the pivotal publication venue, representing a nexus for polio-related discourse. The \ninstitutional analysis placed the World Health Organization at the epicenter of polio \nresearch, with universities and private research entities following suit. The geographical \nspread of contributions was vast, covering 114 regions. Nonetheless, a significant \nproportion of the literature (33.5%) remained uncited, suggesting an untapped potential \nfor research impact. Citation analysis yielded an average citation rate of 9.52%, and co-\ncitation analysis unveiled a network of 4,758 references despite incomplete citation data \nin several records. Visualization maps elucidated recurrent thematic terms such as 'polio', \n'children', 'eradication', 'vaccine', and 'virus', indicating the focal points of scholarly \ncommunication. This study not only delineates the historiography of polio research but \nalso underscores the necessity for enhanced dissemination and integration of findings to \nfortify the global health response to poliomyelitis. Keywords: \nvirus, poliomyelitis, citation analysis, research \ntrends, public health, scientometrics, Web of \nScience, Scopus International Journal of Design & Nature and Ecodynamics \nVol. 18, No. 6, December, 2023, pp. 1333-1346 \nJournal homepage: http://iieta.org/journals/ijdne International Journal of Design & Nature and Ecodynamics \nVol. 18, No. 6, December, 2023, pp. 1333-1346 \nJournal homepage: http://iieta.org/journals/ijdne lobal Trends in Poliomyelitis Research: A Bibliometric Analysis from 1857–2019 Shiza Shamin1\n, Malik Muhammad Saad Missen1\n, Ali Mustafa Qamar2*\n, Amnah Firdous3\n, Qurat Ul Ain1\n, \nV. B. Surya Prasath4,5 1 Department of Computer Science and IT, The Islamia University, Bahawalpur 63100, Pakistan \n2 Department of Computer Science, College of Computer, Qassim University, Buraydah 51452, Saudi Arabia \n3 Department of Computer Science and IT, Government Sadiq College Women University, Bahawalpur 63100, Pakistan \n4 Division of Biomedical Informatics, Cincinnati Children’s Hospital Medical Center 3333 Burnet Ave, Cincinnati 45229, USA \n5 Departments of Biomedical Informatics, Electrical Engineering and Computer Science University of Cincinnati, Cincinnati \n45221, USA 4 Division of Biomedical Informatics, Cincinnati Children’s Hospital Medical Center 3333 Burnet Ave, Cincinnati 45229, USA \n5 Departments of Biomedical Informatics, Electrical Engineering and Computer Science University of Cincinnati, Cincinnati \n45221, USA Corresponding Author Email: al.khan@qu.edu.sa Corresponding Author Email: al.khan@qu.edu.sa Copyright: ©2023 IIETA. This article is published by IIETA and is licensed under the CC BY 4.0 license \n(http://creativecommons.org/licenses/by/4.0/). 2. LITERATURE REVIEW A lot of bibliometric analysis research was done in the field \nof medical sciences. Here, we review some related literature \nrelated to medical sciences, such as the analysis of critical \nissues in healthcare integrity. Pandey et al. [1] analyze \nessential healthcare integrity matters. They report that \nmanaging data integrity is a difficult task for the researcher. Therefore, this study makes a scientometrics analysis of the \nresearch publications in healthcare data integrity. This paper \nhighlights the data integrity issues in healthcare and provides \na literature review of the previous findings. Moreover, they \ndiscuss the methodology and issues and give a road map to \nfuture researchers by giving them various ideas. The \nconcluding part of the paper presents an objective and \nsensitivity analysis for finding the difficulties in the studies \nwhile informing about the feasible solutions. Hossain [4] discussed in his work that the coronavirus \n(COVID-19) disease is the primary health concern due to its \ninfective and global distribution. He performed the \nbibliometric analysis to evaluate the current scientific \nliterature to find the development of knowledge on \ncoronavirus. A total of 422 records were extracted, with 1,652 \nauthors. One thousand five hundred eighty-one documents had \nmultiple-authored documents. Moreover, 2.47 citations were \nreceived on average for each document. The author, E. Mahase, had the highest number of publications (n=13). It also \nincludes information about the top 10 journals which publish \narticles on COVID-19. British Medical Journal was the top \njournal with 47 articles. We believe the data could be more \nextensive according to the state-of-the-art bibliometric \nanalysis of COVID-19. Even the number of publications \nconsidered in our research is more than ten times. g\nEsmaeilpour-Bandboni et al. [2] conducted a bibliometric \nanalysis aimed at performing analysis of Radiology, Nuclear \nMedicine, and Medical Imaging articles written by Iranian \nresearchers, published in journals and extracted from WoS and \nScopus databases from 2001 to 2016. Three thousand three \nhundred thirty-five documents were retrieved. HistCite, \nMicrosoft Excel, and VOSviewer were used to analyze the \ndata. The study analyzes that Iran’s contribution was meagre \nin both databases, i.e., WoS (0.32%) and Scopus (0.52%). The \nTehran University of Medical Sciences was the database's \nmost contributing institution. According to these findings, the \nstudy concludes that Iran may contribute with the USA, Japan, \nand Germany to increase the publication count. 1. INTRODUCTION potentially leading to paralysis. Globally, two types of vaccines are instrumental in the fight \nagainst polio. Both are designed to confer immunity and \ninterrupt person-to-person transmission of the virus, thereby \nprotecting individuals and the wider community. The \npioneering vaccine, featuring an attenuated strain of the virus, \nwas developed by microbiologist Hilary Koprowski and first \nadministered to a child in 1950. Subsequently, Jonas Salk \nintroduced an inactivated poliovirus vaccine at the University \nof Pittsburgh in 1952, followed by Albert Sabin's development \nof a live oral vaccine. After three doses, over 95% of recipients \nachieve protection from all three poliovirus types. Notably, \npictorial evidence of polio dates back to prehistoric times. Highly contagious, poliomyelitis (polio) is transmitted \npredominantly via the fecal-oral route, primarily in areas \nafflicted by epidemics, where polioviruses may infect entire \nhuman populations. The disease's latency—from initial \nexposure to symptom onset—usually spans six to twenty days, \nwith a broader range of three to thirty-five days possible. Following infection, viral particles are shed in feces for several \nweeks. Transmission occurs primarily through ingesting \ncontaminated food or water, with secondary transmission via \nthe oral-oral route. The infectious period extends from seven \nto ten days post-symptom onset, but the virus can be excreted \nand potentially transmitted as long as it remains in the feces. Factors such as immune deficiency and malnutrition increase \nsusceptibility to infection and can also exacerbate the severity, Paralytic polio outbreaks, unknown before the 20th century, \nbegan to surface in Europe and the United States around 1900. The first documented outbreak occurred in Louisiana in 1841, 1333 publications on other viral diseases. We discuss the literature \nreview on some diseases and viruses such as COVID-19, \nCoronavirus, Celiac disease, Poliovirus, Echinococcosis \ndisease, \nAlzheimer’s \ndisease, \nLeishmaniasis \ndisease, \nRotavirus, cancer disease and Zika virus. Colavizza et al. [3] \nanalyzed the COVID-19 Open Research Dataset (CORD-19) \nof publications on COVID-19 and coronavirus research. They \ncompare the CORD-19 account with the Web of Science \ndatabase. On one hand, the indication is that CORD-19 \nextends beyond COVID-19 and coronavirus studies. On the \nother hand, it might not encompass a significant portion of the \nliterature solely dedicated to COVID-19 and coronaviruses, \nsuggesting a potential omission in the dataset. They \nthoroughly analyzed CORD-19 in different ways. They \ngenerated a map of interconnected terms sourced from \nCOVID-19 publications. A topic modeling analysis indicated \nthat CORD-19 publications connect to medical research \nencompassing various viruses, including COVID-19 and \ncoronaviruses. 1. INTRODUCTION Predominant themes within CORD-19 include \nstudies on public health, epidemics, molecular biology, and a \nspectrum of viruses like coronaviruses and influenza. Using \nAltmetric data, 46,996 publications were explored on CORD-\n19. They found that 4,096 publications belong to the year \n2020. They selected social media engagements concerning \nCORD-19 publications, focusing on platforms like Altmetric, \nincluding blogs, Twitter, news media, policy documents, and \ncitations in Wikipedia. Notably, Twitter dominated the \ninteractions, comprising over 95% of all social media \ninteractions. News mentions accounted for 3.6% of the \nengagement, followed by blog citations at 0.6%, highlighting \ntheir relative importance in this context. Our research also \nfocuses on a virus. However, the number of publications is \nmuch lesser. with a fifty-year hiatus until a subsequent report of 26 cases in \nBoston in 1893. The widespread adoption of polio vaccines in \nthe 1950s precipitated a marked decrease in incidence across \nmany industrialized nations. In 1988, the World Health \nOrganization (WHO), UNICEF, and The Rotary Foundation \nlaunched a concerted global effort to eradicate polio, which \nachieved a 99% reduction in cases—from an estimated \n350,000 in 1988 to 483 by 2001. Currently, the incidence \nfluctuates between 1,000 to 2,000 cases annually. As of the \nWHO’s 2015 Annual Report, polio remains endemic in \nPakistan and Afghanistan, with sporadic re-emergence in other \nnations, including developed ones. The present study employs a bibliometric approach to \nanalyze a corpus of 6,139 polio-related publications from 1857 \nto 2019, extracted from the Scopus and Web of Science \ndatabases using 'Polio' as the search term. Employing \nanalytical tools such as EndNote, Microsoft Excel, R \n(Biblioshiny), and VOSviewer, the research delineates the \ntrajectory of polio literature, examining document types, \nlanguage distribution, author collaboration patterns, journal \nactivity, and contributions by country and institution, along \nwith citation and co-citation analyses to identify prevailing \nterms in the discourse. This paper is structured as follows: Section 2 presents a \ndetailed literature review, while the methodology is delineated \nin Section 3. Section 4 discusses the experiments and findings, \nand the paper concludes with Section 5, which also provides \nfuture research directions. 2. LITERATURE REVIEW Indian Institute of Chemical \nBiology, Kolkata, has 566 articles published on Leishmaniasis \nand remains at the top. The most active journal in this research \nis PLOS Neglected Tropical Diseases, with 86 publications \nand 2.54% of the share in total papers published. Regarding \ninternational collaboration, India has the most collaborated \nresearch with the USA (339 articles; 10% share). There were \nsix articles which received more than 500 citations. Koster et al. [10] reported that Rotavirus is the leading cause \nof diarrheal disease. Every child in the world is affected by this \ndisease at least once by the age of five. A scientific analysis \nwas made to identify the quantity and quality of research \narticles on Rotavirus. From 1900 to 2013, 5,906 publications \nwere published in 138 countries. The USA authored 2,037 \narticles, accounting for 34.5% of all publications. They also \nmade a gender analysis of authors, finding that most of the \nauthors are women and from Brazil. They also made the \ncooperation and subject analysis. Singh [11] studied the Zika virus, which is transmitted \nthrough the bite of mosquitoes and was initially found in \nmonkeys. Later 1952, it was found in humans in Uganda, \nAmerica, and Africa. It mainly affects pregnant women, \nleading to severe congenital disabilities like microcephaly. After applying a filter to data, 567 documents remained. Lotka’s law was used to analyze the authors’ productivity, and \nBroadford’s law was used to find the distribution pattern of \narticles. The authorship distribution did not align with Lotka’s \nlaw, indicated by a value of n = 2, suggesting an irregular \npattern. However, the distribution of articles across journals \nclosely adhered to Bradford’s law of scattering. This confirms \nthe presence of a select few core journals that significantly \ncontribute to Zika virus research. Ma et al. [7] stated that both humans and animals are \naffected by Echinococcosis disease caused by the larval stage \nof Echinococcus. Scientists tried to minimize this viral \ndisease, but many cases have been reported in the last twenty \nyears. Echinococcus, which is a pandemic, is found in sheep, \ncattle, and dogs. The humans were affected directly or \nindirectly by eating foodstuff or drinking water contaminated \nwith Echinococcosis. This virus causes cysts in organs such as \nthe liver and lungs. The relevant data were retrieved from the \nWeb of Science database, and hot topics and trends about this \ndisease were evaluated. The data were visualized using \nciteSpace software. 2. LITERATURE REVIEW A total of 7,775 articles were downloaded. The period between 2008 and 2017 was found to be a fast \ndevelopment period. France was considered the most \nrepresented country in this research area (n=582/7,688). The \nUniversity of Zurich was at the top. More than 22,445 authors \ncontributed to this research, with P. S. Craig appearing in 171 \npublications. They also analyzed the citations and co-citations. The top cited author was J. Eckert (1,508 citations). Basu et al. [12] discussed the cancer disease and conducted \na scientometric analysis on herbal medicine for cancer \ntreatment. Cancer is increasing health problems worldwide, \nespecially with an increase in city life, population, and changes \nin lifestyle and environment. According to WHO, it is the \nsecond major cause of death worldwide. The data were \ndownloaded from Medline, Scopus, and Web of Science \ndatabases, while most of the data belonged to WoS and was \nconsequently used for analysis. Seventeen thousand authors \nbelonged to 135 countries and 3,900 different institutions. The \ndata were published in 1,290 different journals and got 82,879 \ncitations. The data were divided into periods, followed by the \nanalysis of period-wise publication growth. Dong et al. [8] made a bibliometric analysis of Alzheimer’s \ndisease worldwide, particularly in China. The WoS and \nPubMed databases were selected for search with data from \n1988 to 2017. One hundred eighty-one thousand one hundred \nsixteen articles were downloaded and analyzed. According to \nthis research, 30.93% were global publications, and 95.31% \nwere from China. A total of 8,935 journals participated in the \npublication of this disease. The Neurobiology of Aging journal \npublished the most articles, with 5,206 publications. Many \norganizations worked together and sponsored 11,809 articles. Missen et al. [13] analyzed Pakistan’s social science and \nscience disciplines. Two thousand publications were \nexamined, which belong to 50 authors of different disciplines. This analysis has three levels: Research, domain, and field \nlevel. They discussed single and multiple authors of articles, \nreadability scoring, title format, publication rate, gender-wise \ncontribution of researchers, and contribution of PhD students \nin research publications. This analysis found that more \nresearch is being conducted on the science discipline than the \nsocial science discipline. It was also discovered that the \nreadability scores of authors changed with time, while the \nreadability score of science articles was higher than the social \nscience ones. Ram [9] contributed to the bibliometric analysis of \nLeishmaniasis disease. 2. LITERATURE REVIEW On the other \nhand, our research is not limited to a particular country and \ncovers more than 150 years of Polio research. Demir and Comba [5] explain in their work that Celiac \ndisease (CD) is an immune disease in which people cannot eat \ngluten because it can destroy the small intestine. They analyze \nthe publications on celiac disease and find the related trends \nabout this disease, such as top articles, journals, and countries. The data were downloaded from the Web of Science database \nbetween 1980 and 2018. It was found that 13,306 articles \nrelated to celiac disease were downloaded. The most popular \nresearch area on CD was gastroenterology hepatology (2,703, \n41.3%). They also analyze the publications during the next ten \nyears, from 2019 to 2028. The Journal of Pediatric \nGastroenterology and Nutrition was regarded as the most \nactive journal. They also made the co-citation analysis and \nterm analysis just like our research. In recent years, bibliometric analysis has been widely \nadministered to analyze and evaluate scientific research The aim of the paper by Ravi and Jeyshankaris [6] is to 1334 make a scientometrics analysis of polio research papers in \nIndia. The data were downloaded from the WoS database for \nthis purpose. The study period was quite limited (1994–2018). They found that India published 488 records on Polio; all \npublications are in English, and publishing was done in \ndifferent journals. The study tells us that India published most \nof its publications (44) in 2017. The document type was also \nanalyzed. It was found that from 488 records, 317 are articles, \nwhereas letters, proceeding papers, reviews, and meeting \nabstracts are also present. While considering the global \ncitations, it was found that 20% of records did not receive any \nglobal citations. In local citations, about half, i.e., 49.5%, did \nnot get any single local citation, while only one record had \nmore than 100 citations. Of 488 papers, 76 have single-\nauthored documents, and 73 have joint authors. The most \nproductive author is T. J. John, who has 41 publications. The \njournals contributing to this Polio research are 19. By contrast, \nour research is not limited to a particular country and contains \nmuch more papers. Furthermore, we used WoS and Scopus to \nretrieve the publications. the highest h-index (211) and citation rate (3,17,646 citations). The citation rate per paper is 1.75. 2. LITERATURE REVIEW According to WHO, Leishmaniasis is \none of the neglected diseases in the world, which has become \nmore challenging for the medical sciences. The disease has \nspread to 98 countries, which is alarming. It is caused by the \nbite of phlebotomine sandflies on humans, dogs, foxes, and \nrats. For this analysis, 39,302 articles on Leishmaniasis are \nfetched from the Scopus database. The time duration is \nselected from 1968 to 2017 (50 years). Comparing the global \npublications on Leishmaniasis with India, 2008–2017 seems \nto be the most productive years for both. In the case of India, \nit is 60.34% and globally, it is 44.64%. The USA is the most \nproductive country for Leishmaniasis research, with 7,609 \npapers (19.36%). Similarly, India was ranked fifth with 3,395 \npapers. Hirsch Index (h-index) is also calculated. The USA has Fuzzy logic resolved many problems in the research areas \nand various applications, like mobile computing, for data \nanalysis and cloud computing [14]. Researchers have tried to \nmake many changes in fuzzy logic within the last few decades. For analyzing fuzzy logic, 410 research publications were 1335 were collected between 1991 to 2017. downloaded from the Web of Science (WoS) database. This \nresearch answers various questions related to essential \nconcepts of hesitant fuzzy sets. Furthermore, it determines co-\nauthorship patterns, research areas of hesitant fuzzy sets, \ncontributions of countries with various research contributions, \nco-citation patterns, and the bibliographic coupling for the \ninstitution in this field. Fang et al. [21] studied climate change and tourism \ninteractions. This has been one of the most critical research \nareas in tourism during the last few years. For this analysis, \n976 documents were extracted between 1990 to 2015. Later, \nthey identified and visualized the collaboration network of the \nauthors of these publications to find the emerging trends. They \nanalyzed from their findings that publications in this area are \nconstantly increasing. The most productive authors belong to \nthe USA, Canada, and Europe. They find out the most trending \ntopics about climate, such as tourist behavior and response \ntoward the environment. Yang et al. [15] performed an analysis on hepatocellular \ncarcinoma (HCC) using magnetic resonance imaging (MRI). With the development of MRI, many articles have been \npublished. A total of 835 articles on MRI were published from \n2008 to 2017. Journal of Magnetic Resonance Imaging has the \nmost publications (79 or 9.46%). 2. LITERATURE REVIEW South Korea is the most \nactive contributor to MRI publications, with 199 publications \n(21.82%). In the top 10 co-cited authors, J. Bruix is considered \nfirst in rank with 398 citations. Forty-seven of the most \ncommon keywords were found, with microvascular invasion \nbeing the most popular during the last three years. Therefore, \npublications on HCC MRI continuously increased from 2008 \nto 2017. The extensive analysis of diverse fields showcases the \nbreadth of bibliometric research conducted across numerous \nscientific disciplines. Various health-related studies demonstrated extensive \nbibliometric analysis: Esmaeilpour-Bandboni et al. [2] \nevaluated Iranian research contributions in Radiology, while \nHossain [4] reviewed the current scientific literature on \ncoronavirus. The research on specific diseases, such as \nAlzheimer's [8] and Leishmaniasis [9], highlighted global \ntrends, top contributing countries, institutions, and prolific \nauthors. Basu et al. [12] addressed cancer's impact and the \nscientometric evaluation of herbal medicine in its treatment, \nunderscoring the increasing challenge of cancer as a global \nhealth issue. Senel and Demir [16] made a bibliometric analysis of the \narticles about religion and health between 1975 and 2016. Fields like religion and fitness have become growing areas of \nstudy, but only a few bibliometric studies have been made in \nthese fields. The WoS database found two thousand six \nhundred eighty-three research articles. The original articles \nwere 1,655, making 62.1% of the total, whereas the remaining \nare letters, books, and proceeding papers. With 1,665 \n(62.45%) publications, the USA has the highest chunk of \npublications. Cornell University was found at the top with 73 \npapers. Two thousand nine hundred seventy-three keywords \nwere considered for use (n = 253, 250, 97, 71, and 41 times). Other analyses concentrated on neglected areas, like \nColavizza et al. [3], who explored the COVID-19 Open \nResearch Dataset (CORD-19), unraveling its broader scope \nbeyond \nCOVID-19 \nstudies. Similarly, \nChandra \n[17] \nscrutinized \nthe \nevolution \nof \nentrepreneurship \nusing \nscientometric techniques. Demir and Comba [5], also shed \nlight on Celiac disease, emphasizing research trends, top \narticles, journals, countries, and co-citation analyses. Koster et \nal. [10] focused on Rotavirus, elucidating its significant impact \non diarrheal diseases. Chandra [17] explained in his article the scientometrics \ntechniques used to study the evolution of entrepreneurship \nbetween 1990 and 2013. He used a combination of topic \nmapping, author citation analysis, and co-citation analysis. Moreover, he also made a visualization graph to show the hot \nand trending topics. 2. LITERATURE REVIEW Prakash and Arumugam [18] analyze the \ngrowth of literature related to biotechnology in India between \n2002 and 2016. The Scopus database is used to analyze Indian \ncontributions to biotechnology literature. The study analyzes \nthe performance based on many parameters such as literature \npublishing trends, authorship and co-authorship patterns, \ncollaboration patterns, top journals by the scientists and \naccording to the number of citations. The number of articles \npublished from 2002–2016 is 5,514. The trend is going upward \naccording to the number of published papers. Multiple studies employed scientometric tools and \ndatabases, such as Web of Science, Scopus, and PubMed, to \nclarify publication patterns, collaborations, and citation \nanalyses. These varied analyses collectively contribute to the \nexpanding knowledge and comprehension of diverse scientific \ndisciplines, guiding future research directions and informing \ndecision-making processes across different fields of study. 3. METHODOLOGY The present study uses a bibliometric method to analyze the \nliterature on Polio. The idea is to analyze and measure the \nscientific literature related to Polio. Quantitative, qualitative, \nand computational methods are used by bibliometrics. The \nprimary purpose of this study is to map institutional \nproductivity and the institutional ranking by research area and \njournal rank while identifying the countries and authors who \ncooperated the most. This method is beneficial for visualizing \nconsequential patterns and trends in a vast body of literature \nand scientific data. It allows the researchers to make literature-\nrelated inventions that are challenging through any other \nanalysis. The research methodology comprised data collection, \ntools and techniques, and bibliometric techniques. Kullenberg and Kasperowski [19] stated that citizen science \ninvolves the public in research to improve scientific \nknowledge. This study obtained two data sets (N=1,935, \nN=633) from the WoS database. The results reveal that there \nwere three main points. First, the citizens used different \nmethodologies for collecting and classifying data. Second, the \ncitizens are involved in collecting the geographical data. Third \nincludes public participation related to environmental and \nhealth issues. In another paper, Moraes et al. [20] showed that Mercury \n(Hg) has a hazardous effect on humans and wildlife. Because \nof its high toxic capacity in food, mercury is classified as a \npersistent toxic substance (PTS). It is obtained from natural \nsources and enters the environment from anthropogenic \nsources. Amazon rain forests are relevant in this regard on a \nglobal scale. Scientometrics analysis was applied to find out \nabout the latest trends of publications on mercury. The data The selection of WoS and Scopus databases as search \nengines was based on their widespread recognition and \nestablished prominence in the analysis of scientific literature \n[22]. Furthermore, they are well-known index databases for 1336 publications remained. This data contains articles, book \nchapters, proceeding papers, letters, and notes. It is observed \nthat these articles were written in many different languages, \nsuch as English, French, Spanish, German, Portuguese, \nJapanese, and Chinese (Complete details about document \ntypes and languages of articles are described in the \nExperiments section). Table 1 shows a detailed summary of \nthe general results. Involvement from 114 regions, 30 \nlanguages, and contributions from 4,306 institutions \ndemonstrate a widespread and diverse global engagement in \nPolio research, underscoring its international significance. This research uses quantitative analysis, graphs, and network \nmodels to visualize our findings efficiently. 3.1 Tools and techniques In total, 13390 publications on Polio were obtained, 3,635 \nfrom the Web of Science database and 9,755 from the Scopus \ndatabase. The data of these two databases was merged, \nresulting in 2,988 duplicates. After removing all the duplicates, \n10,402 publications remained. A revisit of the title and abstract \nwas done in order to ensure that the data must contain the word \nPolio. However, after revision, we found that much data was \nunrelated to our research work, and data related to different \ndiseases and viruses other than Polio was also present. Several bibliometric analysis tools exist, each with strengths \nand abilities [23]. Therefore, after thoroughly examining each \ntool, we find that the appropriate use of various tools for \nvarious types of analysis is essential. Therefore, we take an \noverview of tools like VOSviewer, Gephi, HistCite, Citespace, \nBibexcel, and Rtools. However, we find VOSviewer and \nRtools more flexible and accurate than others. 3. METHODOLOGY The flow chart of \nthis study is shown in Figure 2. The selected research papers \nare analyzed using the current acclaimed bibliometric mapping \ntools, such as VOSviewer and R-studio. the scientific literature. However, the reason for selecting \nthese databases for data extraction is the large dataset involved, \nand many previous analyses were also based on these \ndatabases. Following strict selection criteria, we collect the \nworld’s most prominent scientific and technical journals from \nboth databases. This ensures a high quality of retrieved papers \non Polio. The data were downloaded from Scopus and WoS \ndatabases using the keywords ‘Polio’ or ‘Poliomyelitis’. The keywords ‘Polio’ or ‘Poliomyelitis’ are prevalent in this \nresearch field, ensuring comprehensive coverage of related \nareas. Employing these keywords in the search process across \nthe title, abstract, and keywords sections of documents was \ncrucial. The literature search spanned from 2007 to June 2019 \nin the WoS database and from 1857 to 2019 in the Scopus \ndatabase to capture all relevant articles for the study. Figure 1. Selection process of the studies for exclusion in the \nreview Table 1. Summary of general results \n \nCriteria \nQuantity \nTime Span \n1857–2019 \nDocument type \n16 \nLanguages \n30 \nDocuments \n6,139 \nAuthors \n4,572 \nSource Journals \n1,725 \nRegions \n114 \nInstitutions \n4,306 \nCited References \n58,438 \nTerm Analysis \n10,416 \nDocuments per Author \n0.506 \n \n \nFigure 2. Research process of Polio research Table 1. Summary of general results Figure 1. Selection process of the studies for exclusion in the \nreview The year 2020 was excluded to improve the results' \naccuracy. If the most recent (incomplete) year had been \nincluded, getting the same number of documents might not be \npossible if the analysis is repeated. The reason is that several \njournal issues are released and uploaded to databases after \nseveral months of publication to be online. The search results \nare saved in BibTeX format, and the cited reference \ninformation is included. The selection criteria for inclusion are \nshown in Figure 1. Figure 2. Research process of Polio research 3.3 SCImago Journal Rank (SJR) SJR measures the scientific impact of scholarly journals that \ncount both the importance or prestige of the journals, from \nwhere the citations come from and the citations received by a \njournal. A journal’s SCImago Journal Rank shows a numeric \nvalue that indicates the average number of weighted citations \nreceived during a selected year, during the previous three years \nper document published in the journal. The higher the value of \nSJR, the greater the prestige of the journal [29]. SJR does not \nonly rank journals but countries as well. It can tell the citations \nof that journal, H-index, place of publication, editor name, \nsubject category to which certain journal belongs, scope, \ncitations per document, and documents cited. All of the \naforementioned fields are shown in graph format as well. (2) \nOverlay visualization: This visualization is the same \nas network visualization, except the items’ colors differ. In the \noverlay visualization, the items can be colored differently. In \nthe first case, the color of an item is determined by the score \nwhen the items are scored. Therefore, when the score is low, \nthe color by default is blue, and when the score is higher, it \nranges from green to yellow. If the items have no scores, then \noverlay visualization is not shown. In this research, we only \nmake one overly visualization graph to find the new and old \nterms within the given period. (2) \nOverlay visualization: This visualization is the same \nas network visualization, except the items’ colors differ. In the \noverlay visualization, the items can be colored differently. In \nthe first case, the color of an item is determined by the score \nwhen the items are scored. Therefore, when the score is low, \nthe color by default is blue, and when the score is higher, it \nranges from green to yellow. If the items have no scores, then \noverlay visualization is not shown. In this research, we only \nmake one overly visualization graph to find the new and old \nterms within the given period. (3) \nDensity visualization: In this type of visualization \ngraph [24], the items are also represented similarly to the \nnetwork and overlay visualization. Each point of the item in \ndensity visualization is represented by a color that indicates \nthe density. By default, it ranges from blue to green and green \nto yellow. 3.1.2 R (Biblioshiny) R-Studio is an integrated development environment (IDE) \ntailored for the R language, specifically designed for statistical \ncomputation and graphics [26]. R-studio provides several \npackages such as Shiny, RMarkdown, flex-dashboard, \nTidymodels, Sparklyr, Stringr, Reticulate, and Plumber. However, the Bibliometrix package of R-Studio was selected \nfor bibliometric analysis in related research. Biblioshiny is the \ninterface of bibliometrix [27]. 𝐼𝐹2018 = \n𝐶𝑖𝑡𝑎𝑡𝑖𝑜𝑛𝑠(2017) + 𝐶𝑖𝑡𝑎𝑡𝑖𝑜𝑛𝑠(2016)\n𝑃𝑢𝑏𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠(2017) + 𝑃𝑢𝑏𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠(2016) \n(1) 𝐼𝐹2018 = \n𝐶𝑖𝑡𝑎𝑡𝑖𝑜𝑛𝑠(2017) + 𝐶𝑖𝑡𝑎𝑡𝑖𝑜𝑛𝑠(2016)\n𝑃𝑢𝑏𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠(2017) + 𝑃𝑢𝑏𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠(2016) \n(1) (1) It has many features that are very helpful to carry out deep \nbibliometric analysis. The interface of biblioshiny is well \nmanaged and divided according to the workflow [28]. A main \nmenu presents all data summaries, such as year span, total \nauthors, total citations, total references, and single-author and \nmulti-author documents. The analysis options are divided into \nseven categories: Overview, Sources, Authors, Documents, \nConceptual structures, Intellectual structure, and social \nstructure. The graphs that have been generated can be exported These metrics filter out top journals in a particular domain \nwherever they are needed. 3.3 SCImago Journal Rank (SJR) In this study, the analysis of the top active journals is \nconducted along with finding the SJR of 2019 and the subject \ncategory related to those journals. This analysis helps us to \nfind out which journal belongs to which specific areas of \nresearch [30] (“SJR: Scimago Journal and Country Rank”, \n2020). The weights of the neighboring items are higher when the \nitems of the neighborhood are higher in point and the color of \nthe point is yellow. On the other hand, the weights of the \nneighboring items are smaller when the neighbourhood items \nare lower in point, and the point color is blue. An academic journal's Impact Factor (IF) is a bibliometric \nindex representing the average citations received annually. Its \ncalculation involves aggregating the total citations garnered by \narticles over the last two years and dividing this by the total \nnumber of articles published within that period or a specific \ntimeframe [31]. Across academic disciplines, the IF helps in \nassessing the significance of publications or academic \ncontributions, typically considering journals with higher IF as \nmore influential. The formula to compute the IF for the year \n2018 is expressed in Eq. (1). 3.1.1 VOSviewer Therefore, we applied a filter on the title and extracted only \nPolio-related data, or applied a filter on the title and extracted \ndata that must contain the word Polio or Poliomyelitis. After \nfiltering the data, it was discovered that 4,263 publications are \nirrelevant. After excluding all the irrelevant data, 6,139 VOSviewer is a software that offers the main functionality \nrequired for visualizing, producing, and exploring bibliometric \nnetworks [24]. This idea of visualizing bibliometric networks \nreceived much attention during the early days of bibliometric \nresearch. Bibliometric network visualization has become a 1337 to several file formats; maps can be exported to Pajek and then \nfurther analyzed in VOSviewer. powerful approach to analyzing bibliometric networks in a \nlarge variety, which can depict the relationship between \nauthors and co-authors, find a relationship between countries \nand institutions or citation and co-citation networks, and also \nfind the subject category and term analysis. With time, \nresearchers have started to analyze large networks, thus \nrequiring more advanced visualization tools [25]. The \nVOSviewer (version 1.6.15) was used in this research and is \nfreely available. All the tabs that were used are described next: Similarly, the tables can be copied, downloaded, or saved as \nExcel, CSV, PDF or printed. Later, the file of Scopus and Web \nof Science literature in BibTeX format was uploaded on the \nBiblioshiny interface. As per the study's objectives, the Excel \nfiles were downloaded and used for further data analysis and \nfor making tables and graphs, as described in the next section. 3.2 Linear forecasting VOSviewer displays graphs in three ways, but we use only \nnetwork visualization and overlay visualization graphs in this \nresearch work. Forecasting is also performed in this research work. Forecasting is the technique that helps us predict the future. Linear forecasting is applied to the yearly publication growth \nof polio articles, and we forecast the number of publications \nduring the next four years. Linear forecasting is a type of time \nseries forecasting. The linear forecasting function is \ncategorized and calculated under the statistical function of \nMicrosoft Excel. It predicts a future value by using linear \nregression. (1) \nNetwork visualization: The network visualization \ngraph represents the items by labels and circles by default. The \nitem's weight determines the label's size and the item's circle. If the item's weight is more significant, its label and circle size \nare also larger. Some items whose labels are not displayed are \nalso present; this is done to avoid overlapping of labels. Their \ncluster represents the color of an item. The lines between items \ndemonstrate the links. By default, one thousand lines are \ndrawn, representing the most vital links between different \nitems. In this work, we make three network visualization \ngraphs and find the cooperation of the authors, countries, and \ninstitutions. We also perform the term analysis. (1) \nNetwork visualization: The network visualization \ngraph represents the items by labels and circles by default. The \nitem's weight determines the label's size and the item's circle. If the item's weight is more significant, its label and circle size \nare also larger. Some items whose labels are not displayed are \nalso present; this is done to avoid overlapping of labels. Their \ncluster represents the color of an item. The lines between items \ndemonstrate the links. By default, one thousand lines are \ndrawn, representing the most vital links between different \nitems. In this work, we make three network visualization \ngraphs and find the cooperation of the authors, countries, and \ninstitutions. We also perform the term analysis. 4. EXPERIMENTS AND RESULTS A comprehensive dataset spanning 162 years (1857 to 2019) \nregarding polio research publications was collected from the \nWeb of Science and Scopus databases, comprising 6,139 1338 records. This extensive dataset was analyzed using tools such \nas R-studio, Microsoft Excel, and VOSviewer to gain insights \nand visualize trends within the data. Table 3 indicates the language-wise literature output on \nPolio. Among the 6,139 papers, 85.77% were published in \nEnglish, showcasing the language's dominance in conveying \nPolio-related research. The total number of documents \npublished in languages other than English is 873, around 14.23% \nof the total publications. 4.2 Publication output trends: Analyzing growth over time Table 2 shows the document types of world contributions to \npolio research. Most of the research papers on Polio are \narticles (4,369, 71.16%), followed by reviews (411), letters \n(344) and editorials (179). Therefore, journal articles stand out \nas researchers' most favored mode of research communication. The scientific documents underwent analysis by periods, as \ndepicted in Figure 3. Publications showed an upward trend \nbeginning in 1963. Over the past three decades, there was \nconsistent growth in the average number of publications per \nperiod: 1990–1999 (95), 2000–2010 (147), and 2011–2017 \n(254). However, there was a decline in publications in 2018 \nand 2019 (n=191). Notably, the pinnacle of publications was \nobserved in 2014, reaching its peak at 340. The annual growth \nrate is 5.34%. The prediction of publication volume was \naccomplished through a statistical function utilizing linear \nregression to forecast future values. The forecast for the \nupcoming four years (including 2020) stands at 148, 149, 151, \nand 153 publications respectively. This cumulative growth \nunderscores the escalating significance of subject-related \npublications, showing an exponential increase over the years. Table 2. Distribution of document types in Polio research D\nt T\nN\nf R\nd\nP\nt Document Types \nNo. of Records \nPercentage \nArticle \n4,369 \n71.16 \nArticle in press \n18 \n0.29 \nArticle, proceedings paper \n3 \n0.04 \nBook \n87 \n1.41 \nConference paper \n87 \n1.41 \nCorrection \n12 \n0.19 \nEditorial \n179 \n2.91 \nErratum \n27 \n0.43 \nLetter \n344 \n5.60 \nMeeting abstract \n117 \n1.90 \nNews item \n79 \n1.28 \nNote \n312 \n5.08 \nPoetry \n2 \n0.03 \nReprint \n2 \n0.03 \nReview \n411 \n6.69 \nShort survey \n90 \n1.46 \nTable 3. Language-wise distribution in Polio research Figure 3. Number and accumulation of publications per year \nin Polio research Table 3. Language-wise distribution in Polio research Language \nNo. of records \nPercentage \nEnglish \n5,266 \n85.77 \nGerman \n224 \n3.64 \nSpanish \n125 \n2.03 \nFrench \n91 \n1.48 \nChinese \n63 \n1.02 \nItalian \n61 \n0.99 \nJapanese \n57 \n0.92 \nSwedish \n39 \n0.63 \nPolish \n34 \n0.55 \nDutch \n31 \n0.50 \nPortuguese \n24 \n0.39 \nRussian \n24 \n0.39 \nDanish \n21 \n0.34 \nRomanian \n18 \n0.29 \nCzech \n10 \n0.16 \nNorwegian \n10 \n0.16 \nTurkish \n8 \n0.13 \nFinnish \n5 \n0.08 \nHungarian \n5 \n0.08 \nHebrew \n4 \n0.06 \nPolyglot \n3 \n0.04 \nSlovak \n3 \n0.04 \nBulgarian \n2 \n0.03 \nCroatian \n2 \n0.03 \nGreek \n2 \n0.03 \nKorean \n2 \n0.03 \nSerbian \n2 \n0.03 \nArabic \n1 \n0.01 \nBosnian \n1 \n0.01 \nThai \n1 \n0.01 Figure 3. Number and accumulation of publications per year \nin Polio research 4.3 Authors and their cooperation Twelve thousand one hundred twenty-four different authors \nwrote the 5,781 publications. Three hundred fifty-eight \npublications are without an author name. Most authors (71.3%; \nn=8,290) are associated with a single publication, while 14.6% \n(n=1,775) are credited in two, 5.2% in three, 2.5% in four, and \n6.2% in five or more. Table 4 highlights the top ten most \nprolific authors based on their total number of publications. The analysis of key authors in Polio research reveals a blend \nof significant contributors from various countries, notably \nSwitzerland and the USA. Notably, R. Sutter emerges as both \nthe most productive and cited author. The USA maintains a \nnotable presence among top authors, with individuals like M. Pallansch, K. Thompson, and S. Cochi contributing \nsignificantly in volume and average citations per publication. 1339 Table 4. Top 10 most-productive authors in Polio research Table 4. Top 10 most-productive authors in Polio research Authors \nCountry \nPublications \nCitations \nAvg. Citations \nPub. as the 1st Author \nSutter, R. Switzerland \n109 \n4042 \n37 \n11 \nPallansch, M. USA \n93 \n3975 \n42.7 \n2 \nThompson, K. USA \n72 \n1854 \n25.7 \n27 \nJohn, T. India \n61 \n792 \n12.9 \n36 \nOberste, M. USA \n60 \n1254 \n20.9 \n2 \nWassilak, S. Europe \n59 \n1571 \n26.6 \n5 \nKew, O. USA \n54 \n3446 \n63.8 \n9 \nCochi, S. USA \n52 \n1655 \n31.8 \n9 \nAylward, R. Switzerland \n50 \n1511 \n30.2 \n10 \nDuintjer, T.R. USA \n47 \n1147 \n24.4 \n25 Table 5. Top 10 most-active journals in Polio research Table 5. Top 10 most-active journals in Polio research \n \nJournal Title \nPubl. 4.3 Authors and their cooperation SJR \n2019 \nImpact \nFactor \n2019 \nSubject Category (SJR) \nVaccine \n245 \n1.683 \n4.76 \nBiochemistry: Genetics and Molecular Biology, \nMolecular Medicine, \nImmunology and Microbiology: Immunology and Microbiology (miscellaneous), \nMedicine: Infectious Diseases, Public Health, Environmental and Occupational \nHealth, Veterinary: Veterinary (misc.) \nJournal of Infectious \nDiseases \n188 \n2.946 \n4.71 \nMedicine: Immunology and \nAllergy Infectious Diseases \nIndian Pediatrics \n114 \n0.285 \n1.163 \nMedicine: Pediatrics, Perinatology and Child Health \nScience \n89 \n13.11 \n41.037 \nArts and Humanities: History and Philosophy of Science, \nMultidisciplinary: Multidisciplinary \nBulletin of the \nWorld \nHealth Organization \n72 \n2.502 \n6.818 \nMedicine: Public Health, Environmental and \nOccupational Health \nArchives of \nPhysical Medicine \nand Rehabilitation \n70 \n1.114 \n2.697 \nHealth Professions: Physical Therapy, Sports Therapy \nand Rehabilitation, \nSports Science Medicine: Rehabilitation \nThe Lancet \n62 \n14.554 \n59.102 \nMedicine: Medicine (miscellaneous) \nDeutsche \nMedizinische \nWochenschrift \n61 \n0.176 \n0.15 \nMedicine: Medicine (miscellaneous) \nBritish Medical \nJournal \n59 \n2.049 \n27.604 \nMedicine: Medicine (miscellaneous) \nNature \n59 \n14.047 \n43.07 \nMultidisciplinary: Multidisciplinary The average number of publications per author is 0.506. The \nauthor's name is missing in 5.8% (n=358/6,139) of the \npublications. Sixty-three publications do not have a specific \nauthorship. 29.3% (n=1,801) of the publications have single \nauthorship. Out of the publications analyzed, 14.2% (n=874) \ninvolved two authors, 12.2% (n=749) had three authors, 9.2% \n(n=565) had four authors, 6.9% (n=423) had five authors, and \n21.3% (n=1,306) had six or more authors, with a maximum of \n15 authors per publication. The authors' collaborative patterns \nin publishing were examined using VOSviewer, as shown in \nFigure 4. The size of the circles indicates the publication \nvolume, while the curved connections represent collaborative \nties. The colors denote distinct collaboration clusters within \nthe network, identifying 27 author clusters. (5.5%) contributed significantly by publishing 11 or more \narticles, with publication counts ranging from 11 to 245. Table \n5 outlines details regarding the top 10 most active journals. These ten journals accounted for 16.5% of all publications \n(n=1019 out of 6,139). According to the Scimago Journal \nRanking (SJR), the primary sub-area category for these most \nrecurring journals falls under Medicine: Infectious Diseases. Journals like \"Vaccine\" and \"Journal of Infectious Diseases\" \ndemonstrate strong relevance to Polio research, focusing on \nImmunology, Infectious Diseases, and Pediatrics. Other \ninfluential journals such as \"Science,\" \"The Lancet,\" and \n\"Nature\" cover broader multidisciplinary areas, indicating \ntheir significant influence beyond specific medical domains. These results reflect a comprehensive landscape, from \nspecialized \nmedical \npublications \nto \nmultidisciplinary \nplatforms, emphasizing the diverse avenues for Polio-related \nresearch dissemination and impact. 4.4 Journals publishing on Polio research Out of the 6,139 published articles, they were distributed \nacross 1,725 journals. Among these journals, 1,011 (58.6%) \nwere responsible for publishing only one article each. 283 \n(16.4%) journals published two articles. One hundred nine \n(6.3%) published three articles. Similarly, 227 (13.1%) \njournals published four to ten articles. Among the journals, 95 Figure 5 demonstrates the publication frequency for the top \nfive most influential journals over the years. There has been a \nnoticeable rise in publications during the past decade. The \nsubstantial volume of publications in both ‘Vaccine’ and \n‘Journal of Infectious Diseases’ can be attributed to including \nconference proceedings and editorials. 1340 Figure 4. Author cooperation network in Polio research the USA emerged as the leading producer of publications in \nthis domain. Figure 6. Distribution of publications across countries or \nterritories in Polio research Figure 4. Author cooperation network in Polio research Figure 6. Distribution of publications across countries or \nterritories in Polio research Figure 5. Frequency of publications across years in the top \nfive most influential journals Figure 5. Frequency of publications across years in the top \nfive most influential journals Figure 7 showcases the top 10 countries and territories with \nthe most publications. There was no publication in the journal \nfrom 1857 to 2017. Therefore, these years were skipped from \nthe graph. Economically developed countries significantly \ncontribute to scientific and academic advancements. The top \n10 most productive countries in polio publications include the \nUSA, India, United Kingdom, Sweden, Germany, France, \nNetherlands, Switzerland, Italy, and Pakistan. Information \nfrom countries and territories' affiliations of the authors is \npresent in 6,417 publications. Figure 7. Countries and territories leading in publication \nproductivity: Top 10 Rankings Figure 5. Frequency of publications across years in the top \nfive most influential journals 4.5 Geographical and institutional distribution, along with \ncollaborative partnerships of \nPublications \nWorld Health Organization \nSwitzerland \n371 \nCenters for Disease Control \nand Prevention \nUSA \n202 \nBill and Melinda Gates \nFoundation \nUSA \n64 \nNational Center for \nImmunization and Respiratory \nDiseases \nUSA \n64 \nCenter for Global Health \nUSA \n52 \nStatens Serum Institute \nDenmark \n50 \nInstitute Pasteur \nFrance \n48 \nLund University \nSweden \n46 \nUniversity of Amsterdam \nAmsterdam \n41 \nLondon School of Hygiene and \nTropical Medicine \nLondon \n36 \n \n \n \nFigure 9. Collaboration network among institutions and \nregions Table 6. Top 10 institutions with the highest productivity to the Netherlands and Bangladesh (yellow group). The \nseventh cluster relates to India (purple group). The eighth \ncluster relates to Sweden (sky-blue group). The ninth cluster \nrelates to Pakistan (orange group). Table 6. Top 10 institutions with the highest productivity \n \nInstitution \nCountry \nNo. of \nPublications \nWorld Health Organization \nSwitzerland \n371 \nCenters for Disease Control \nand Prevention \nUSA \n202 \nBill and Melinda Gates \nFoundation \nUSA \n64 \nNational Center for \nImmunization and Respiratory \nDiseases \nUSA \n64 \nCenter for Global Health \nUSA \n52 \nStatens Serum Institute \nDenmark \n50 \nInstitute Pasteur \nFrance \n48 \nLund University \nSweden \n46 \nUniversity of Amsterdam \nAmsterdam \n41 \nLondon School of Hygiene and \nTropical Medicine \nLondon \n36 Figure 8. Cooperation network between countries and \nregions Figure 8. Cooperation network between countries and \nregions The tenth cluster relates to Turkey (pink group). The \neleventh cluster relates to Italy (grey group). The twelfth \ncluster relates to Switzerland (light purple group). In our \nanalysis of country cooperation in publications, we found that \nthe UK collaborates significantly with Western European \ncountries. Similarly, Germany's collaborations extend notably \nto African nations. Notably, Nigeria is the sole African country \ndisplaying strong connections with the USA and the UK. Pakistan has connections with the USA, Saudi Arabia, \nAustralia, Germany, and Nigeria. The United States is a \ncentral hub in these relationships, while Japan is associated \nwith neighboring Asian countries. Figure 9. Collaboration network among institutions and \nregions Four thousand three hundred six research institutions \ncontributed to 4,617 publications for which affiliation data was \navailable (with the possibility of multiple authors per \npublication or authors affiliated with multiple institutions). To \nensure consistency, the names reported in the databases were \nstandardized across different records in Scopus and WoS. 4.5 Geographical and institutional distribution, along with \ncollaborative partnerships Each publication underwent assignment to a country based \non the available information within the author affiliations and \naddresses documented in WoS and Scopus databases. A total \nof 1,522 publications lacked any identifiable affiliation. Approximately 75.2% of the publications were associated with \na specific region or country. Publications potentially linked to \nmultiple countries or institutions were assigned multiple \naffiliations or authorships, considering all affiliations based on \ncountry or institution. Figure 7. Countries and territories leading in publication \nproductivity: Top 10 Rankings Figure 8 depicts the collaboration network among countries \nand territories in Polio research created using VOSviewer. The \nnetwork explicitly highlights countries with five or more \npublications, excluding those not interconnected. Circle sizes \ncorrespond to publication volumes, while connections signify \ncollaborative efforts between countries. The analysis identifies \nthirteen primary clusters, with the USA forming the central \ngroup, followed by the UK (peach-colored cluster) and a third \ncluster involving Ethiopia and Cuba (red-colored cluster). The \nfourth cluster relates to Japan and Brazil (green). The fifth \ncluster relates to France (blue group). The sixth cluster relates The body of publications regarding Polio originated from \n114 distinct regions worldwide. Among these, 31 are in \nEurope, 32 are in Asia, 33 are in Africa, and 15 are in the \nAmericas. Figure 6 provides a visual representation of the \nglobal distribution showcasing the contribution of these \ncountries and territories. A breakdown reveals that 67 countries or territories (58.8%) \ncontributed fewer than ten publications, while 28 (24.6%) \nproduced between 11 and 50 publications. Furthermore, 11 \n(9.6%) contributed between 51 and 150 publications. Notably, 1341 Table 6. Top 10 institutions with the highest productivity \n \nInstitution \nCountry \nNo. of \nPublications \nWorld Health Organization \nSwitzerland \n371 \nCenters for Disease Control \nand Prevention \nUSA \n202 \nBill and Melinda Gates \nFoundation \nUSA \n64 \nNational Center for \nImmunization and Respiratory \nDiseases \nUSA \n64 \nCenter for Global Health \nUSA \n52 \nStatens Serum Institute \nDenmark \n50 \nInstitute Pasteur \nFrance \n48 \nLund University \nSweden \n46 \nUniversity of Amsterdam \nAmsterdam \n41 \nLondon School of Hygiene and \nTropical Medicine \nLondon \n36 \n \n \n \nFigure 9. Collaboration network among institutions and \nregions Table 6. Top 10 institutions with the highest productivity \n \nInstitution \nCountry \nNo. 4.6 Citation analysis This section delves into how these documents are \nreferenced in other publications. The citation analysis helps \nidentify how often a publication has been referenced. Out of \nthe 6,139 publications, they were cited 58,438 times in other \npublications up to the data extraction point. This accounts for \nan average of 9.519 citations per publication. Of all the institutions involved, 71% (n=3,058) participated \nin just one publication, while 13.5% (n=580) were involved in \ntwo publications, and 13% (n=564) contributed to three \npublications. A smaller percentage, 1.6% (n=68), participated \nin 11 to 20 publications, and a mere 0.8% of institutions (36) \nwere responsible for 20 or more publications. Table 6 outlines \ndetails regarding the top 10 most productive institutions. Most \ninstitutions belong to universities, private organizations, \nresearch centers, and institutes. Institutions like WHO and the \nCenters for Disease Control and Prevention (CDC) in the USA \nemerge as essential contributors, highlighting their noteworthy \nroles in global health and disease control. g\np\np\n33.5% (n=2,057) of the articles had not been cited. 11% \n(n=678) were cited only once. 6.9% (n=424) were cited twice. 24.7% (n=1,519) were cited between 3 to 10 times. 10.7% \n(n=655) were cited between 11 to 20 times. 9.6% (n=588) of \npublications had 21 to 50 citations. 2.6% (n=162) were cited \nbetween 51 to 100 times. 0.9% (n=56) were cited up to 454 \ntimes. The average citation of the top 10 articles was 285. Table 7 enumerates the top five frequently referenced \npublications. The primary authorship of nine publications \naligns with the USA, followed by Europe. Notably, O. Kew is \nthe sole author featured twice on this list. Regarding the \nprimary topics categorized by Scimago Journal-SJR, four \njournals revolve around Medicine, while the others focus on \nMicrobiology. Among the 6,139 journals studied (1,106 of \nwhich received more than one citation), 35.9% (n=619 out of \n1,752) had no citations. Additionally, 9.9% (n=171) received \na single citation, while 26.1% (n=451) fell within the range of Additionally, academic institutions such as Lund University \nand the University of Amsterdam display active engagement \nin Polio research, highlighting the academic community's \ndedication to understanding and fighting this disease. These \ndiverse \ninstitutions \nunderscore \na \ncooperative, \nmultidisciplinary approach towards addressing Polio globally. Figure 9 shows the cooperation network between institutions \nand territories in Polio research. 1342 two to 10 citations. Furthermore, 18.2% (n=314) accrued \ncitations ranging between 11 and 50. 4.6 Citation analysis 6.2% of the journals have \naccumulated citations falling within 51 to 200. 3% (n=52) \nhave between 200 and 1,000 citations, and 0.6% (n=10) have \ncitations ranging from 1,001 to 3,415. Vaccine is the journal \nwith the highest citations, with 5.9% (n=3,415/58,437) of the \ntotal citations. The Journal of Infectious Diseases and \nArchives of Physical Medicine and Rehabilitation collectively \ngarnered 9% (n=5,301) of the total citations. This pattern \nsuggests a strong correlation between the overall publication \ncount and the most frequently referenced journals. Figure 10 \ndepicts the top five journals referenced by other articles. citation analysis serves to quantify interactions among \npublications, aiding in the analysis of their relationships. Specifically, we focus on the top 50 highly referenced articles. Figure 11 portrays the co-citation analysis created using \nVOSviewer. The nodes represent publications, and the edges \nrepresent co-citation relationships. Circle size indicates the \nextent of citation on the subject, while distance signifies the \nlevel of relationship. One can observe that most of the articles \nwere written after 1990. The visual network reveals four \ndistinct clusters, indicating topics or research areas. Notably, \nthe red cluster pertains to Polio vaccination, while the green \ngroup is associated with conditions following polio. The \nyellow and blue groups are associated with vaccination, \neradication, and symptoms, and O. Kew (2005) is highly co-\ncited. Figure 10. Journals with the highest references from other \narticles: Top five rankings \n \nIn total 4 758 references\nere sed for co citation anal sis Figure 10. Journals with the highest references from other \narticles: Top five rankings Figure 10. Journals with the highest references from other \narticles: Top five rankings \n \nIn total, 4,758 references were used for co-citation analysis. Many articles do not contain their cited references. The co-\neradication, and symptoms, and O. Kew (2005) is highly co\ncited. Figure 11. Analysis of co-citations among extensively \nreferenced sources \n \nTable 7. Most cited publications: Top 10 rankings Figure 11. Analysis of co-citations among extensively \nreferenced sources Figure 10. Journals with the highest references from other \narticles: Top five rankings Figure 10. Journals with the highest references from other \narticles: Top five rankings In total, 4,758 references were used for co-citation analysis. Many articles do not contain their cited references. The co- In total, 4,758 references were used for co-citation analysis. Many articles do not contain their cited references. The co- Figure 11. Analysis of co-citations among extensively \nreferenced sources Table 7. 5. CONCLUSION The present study examined the polio research output \npublished from 1857 to 2019, merging Web of Science and \nScopus databases. It is observed that the publications have had \na steady growth in this field till 2014, which was also observed \nas a peak year for publications. However, after 2014, the \nnumber of publications decreased. In forecasting, it is also \nnoticed that the number of publications declined in polio \nresearch. The primary reason for this decline is that according \nto WHO, the entire world except Pakistan and Afghanistan is \npolio-free. It was analyzed that 6,139 documents were \npublished in Polio during 162 years. Twelve thousand one \nhundred twenty-four authors participated in publications in \nthirty languages, 1,725 journals, and sixteen document types. The most contributed country in polio publications is the USA. One hundred fourteen regions participated with 4,306 \ninstitutions. For the 6,139 publications, a total of 58,438 \ncitations were received. The author of the most cited \npublication was O. Kew in 2005, and the most cited journal is \n‘Vaccine’. A total of 4,758 co-cited references were also found. In term analysis, the most repeated words are ‘polio’ and \n‘Poliomyelitis’. Figure 12. Terms analysis Figure 12. Terms analysis In conclusion, these results provide an idea for future \nresearchers of Polio to analyze the current scientific literature \nand fill the gaps in the field for further research. Furthermore, \nresearchers in developing countries like Pakistan need to \nundertake research in polio and its related fields to make such \ndeveloping countries polio-free. The current research \nunderscores the importance of future investigations focusing \non unexplored aspects, including the long-term effects of polio, \naddressing vaccine hesitancy, and exploring socio-economic \nimpacts. Identifying and bridging gaps in cultural implications, \ndemographic impacts, and post-eradication challenges remain \ncritical. Encouraging global collaboration, particularly from \nregions that have eradicated polio, is vital for sustaining \nprogress and redirecting research efforts toward public health \ninitiatives. This study advocates for ongoing research in \ndeveloping countries to sustain polio-free status. It suggests Figure 13 depicts the temporal relationship among terms, \nshowcasing the average publication period represented by the \ncolor of the terms. This average is computed from the \ncollective publication years of all associated publications. The \nspectrum spans from 1990 to 2019, with yellow denoting \nnewer terms and blue indicating older ones. The prominence \nof terms, delineated by the color range and circle size, \nhighlights the prevalence of terms predominantly between \n2000 and 2010. 4.6 Citation analysis Most cited publications: Top 10 rankings Table 7. Most cited publications: Top 10 rankings Title \nReferences \nCountry \nInstitution \nJournal-IF \nSJR (2018-2019) \nTimes \nCited \nAvg. Cit. per \nYear \nMain Topic \n(Category SJR) \nVaccine derived \npolioviruses and the \nendgame strategy for \nglobal polio eradication \nKEW O \n2005 \nUSA \nNational Center \nfor Infectious \nDiseases \nAnnual Review of \nMicrobiology \n(5.62-6.517) \n454 \n28.38 \nMicrobiology \nOutbreak of poliomyelitis \nin Hispaniola associated \nwith circulating type 1 \nvaccine derived \npoliovirus \nKEW O \n2002 \nUSA \nNational Center \nfor Infectious \nDiseases \nScience \n(41.037-13.11) \n423 \n22.26 \nHistory and \nPhilosophy of \nScience \nA long-term follow-up \nstudy of patients with \npost poliomyelitis \nneuromuscular symptoms \nDALAKAS \nMC \n1986 \nUSA \nNational Institute \nof Neurological \nNew England \nJournal \nof Medicine \n(70.670-18.291) \n291 \n8.31 \nMedicine \n(miscellaneous) \nLymphoblastic leukemias \nevidence for the \ninvolvement \nof the polio virus receptor \ncd 155 and nectin2 cd 112 \nPENDE D \n2005 \nEurope \nIstituto \nNazionale \nper la Ricerca sul \nCancro \nBlood \n(16.601-5.416) \n268 \n16.75 \nBiochemistry, Cell \nBiology, \nImmunology, \nMedicine \nA protein covalently \nlinked to poliovirus \ngenome RNA \nUSA \nLEE YF \n1977 \nState University \nof New York \nPNAS \n(9.58-5.165) \n254 \n5.77 \nMultidisciplinary 1343 Lund University and the University of Amsterdam collectively \nspearhead \nPolio \nresearch, \nshowcasing \na \nunited, \ninterdisciplinary effort toward eradicating global disease. 4.7 Analysis of terms The examination of terms present in the publications offers \ninsights into primary subjects and evolving research directions. The terms frequently appearing together in different document \nparts are often considered related. This analysis measures the \nstrength of the association using the frequency of the co-\noccurrence in titles, keywords, and abstracts. Terms occurring \nin at least ten publications were considered, resulting in the \nselection of 662 terms for inclusion in the network. The \noutcomes of this term analysis are depicted in Figure 12, where \ncircles denote term occurrences, and the distance between \nthem signifies their relation. The frequency of their co-\noccurrence indicates relationships among terms. The analysis \nidentified six distinct clusters, notably highlighting a red \ncluster associated with Immunization (disease eradication, \nprevention, developing countries). Countries affected by Polio \nare also indicated (Afghanistan, Africa, Australia, Europe, \nIndia, and Pakistan). The green cluster is related to humans, \npost poliomyelitis syndrome, symptoms, females, children, \nand age groups. The blue cluster represents the poliomyelitis \nvirus, polio virus, and virus antibody. The yellow cluster \nrelates to the poliomyelitis vaccine, oral vaccination, and \nvarious circumstances. The purple cluster represents all the \nterms related to polio. The most frequent words are Polio, \npoliomyelitis, poliomyelitis vaccination, children, eradication, \nclinical condition, paralysis, and developing countries. Figure 13. Analysis of Terms with Temporal Data Figure 13. Analysis of Terms with Temporal Data The WHO and CDC, USA, and academic powerhouses like REFERENCES [15] Yang, D.W., Wang, X.P., Wang, Z.C., Yang, Z.H., Bian, \nX.F. (2019). A scientometric analysis on hepatocellular \ncarcinoma magnetic resonance imaging research from \n2008 to 2017. Quantitative Imaging in Medicine and \nSurgery, \n9(3): \n465-476. https://doi.org/10.21037/qims.2019.02.10 [1] Pandey, A.K., Khan, A.I., Abushark, Y.B., Alam, M.M., \nAgrawal, A., Kumar, R., Khan, R.A. (2020). Key issues \nin \nhealthcare \ndata \nintegrity: \nAnalysis \nand \nrecommendations. IEEE Access, 8: 40612-40628. https://doi.org/10.1109/ACCESS.2020.2976687 [2] Esmaeilpour-Bandboni, M., Alizadeh, I., YektaKooshali, \nM.H., Ramezani, A. (2018). Scientometric analysis of \nradiology, nuclear medicine and medical imaging. Journal \nof \nParamedical \nSciences, \n9(1): \n21-28. https://doi.org/10.22037/jps.v9i1.13591 [16] Senel, \nE., \nDemir, \nE. (2018). Bibliometric \nand \nscientometric analysis of the articles published in the \njournal of religion and health between 1975 and 2016. J \nRelig \nHealth, \n57(4): \n1473-1482. https://doi.org/10.1007/s10943-017-0539-1 [3] Colavizza, G., Costas, R., Traag, V.A., van Eck, N.J., van \nLeeuwen, T., Waltman, L. (2020). A scientometric \noverview of CORD-19. PLOS ONE, 16(1): e0244839. https://doi.org/10.1371/journal.pone.0244839 [17] Chandra, Y. (2018). Mapping the evolution of \nentrepreneurship as a field of research (1990-2013): A \nscientometric analysis. PLOS ONE, 13(1): e0190228. https://doi.org/10.1371/journal.pone.0190228 [4] Hossain, M.M. (2020). Current status of global research \non \nnovel \ncoronavirus \ndisease \n(COVID-19): \nA \nbibliometric \nanalysis \nand \nknowledge \nmapping. F1000Research, \n9: \n374. https://doi.org/10.12688/f1000research.23690.1 [18] Prakash, M., Arumugam, J. (2017). Indian contribution to \nbiotechnology research: Scientometric analysis. Journal \nof Advances in Library and Information Science, 6(1): 36- \n40. [19] Kullenberg, C., Kasperowski, D. (2016). What is citizen \nscience?–A scientometric meta-analysis. PLOS ONE, \n11(1): \ne0147152. https://doi.org/10.1371/journal.pone.0147152 [5] Demir, E., Comba, A. (2020). The evolution of celiac \ndisease \npublications: \nA \nholistic \napproach \nwith \nbibliometric analysis. Irish Journal of Medical Science, \n189: \n267-276. https://doi.org/10.1007/s11845-019-\n02080-x [20] Moraes, P.L.C., Dorea, J.G., Bernardi, J.V.E., Gomes, L.F. (2019). Mapping the evolution of mercury (Hg) research \nin the amazon (1991-2017): A scientometric analysis. International Journal of Environmental Research and \nPublic \nHealth, \n16(7): \n1111. https://doi.org/10.3390/ijerph16071111 [6] Ravi, C., Jeyshankar, R. (2019). Scientometric analysis of \nIndian research output on polio literature during 1994-\n2018. Journal of Advances in Library and Information \nScience, 8(2): 77-85. [7] Ma, X., Zhang, L., Wang, J., Luo, Y. (2019). Knowledge \ndomain and emerging trends on echinococcosis research: \nA scientometric analysis. International Journal of \nEnvironmental Research and Public Health, 16(5): 842. https://doi.org/10.3390/ijerph16050842 [21] Fang, Y., Yin, J., Wu, B. (2017). Climate change and \ntourism: A scientometric analysis using CiteSpace. Journal of Sustainable Tourism, 26(1): 108-126. https://doi.org/10.1080/09669582.2017.1329310 [8] Dong, R., Wang, H., Ye, J., Wang, M., Bi, Y. (2019). 5. CONCLUSION The clusters with more contemporary terms \nrelate to Polio and the physical condition after Polio. The \ncluster associated with medical aspects pertains to the older \nterms. One can easily observe through the yellow cluster that \nthe world is moving towards the completely eradicating the \ndisease. The WHO and CDC, USA, and academic powerhouses like 1344 importance of anticancer research using herbal medicine: \nA scientometric analysis. Scientometrics, 110(3): 1375-\n1396. https://doi.org/10.1007/s11192-016-2223-8 adapting research strategies to support broader vaccination \nefforts or study other vaccine-preventable diseases, utilizing \nexisting expertise and resources. [13] Missen, M.M.S., Qureshi, S., Salamat, N., Akhtar, N., \nAsmat, H., Coustaty, M., Prasath, V.B.S. (2020). Scientometric analysis of social science and science \ndisciplines in a developing nation: A case study of \nPakistan in the last decade. Scientometrics, 123(1): 113-\n142. https://doi.org/10.1007/s11192-020-03379-8 ACKNOWLEDGEMENT The researchers would like to thank the Deanship of \nScientific \nResearch, \nQassim \nUniversity \nfor \nfunding \npublication of this project. [14] Vij, S., Jain, A., Tayal, D., Castillo, O., Kahraman, C. (2020). Scientometric inspection of research progression \nin hesitant fuzzy sets. Journal of Intelligent & Fuzzy \nSystems, 38(1): 619-626. https://doi.org/10.3233/jifs-\n179435 REFERENCES Publication trends for Alzheimer’s disease worldwide and \nin China: A 30-year bibliometric analysis. Frontiers in \nHuman \nNeuroscience, \n13: \n259. https://doi.org/10.3389/fnhum.2019.00259 [22] Yang, L., Chen, Z., Liu, T., Gong, Z., Yu, Y., Wang, J. (2013). Global trends of solid waste research from 1997 \nto 2011 by using bibliometric analysis. Scientometrics, 96: \n133-146. https://doi.org/10.1007/s11192-012-0911-6 [23] Darko, A., Chan, A.P.C., Adabre, M.A., Edwards, D.J., \nHosseini, M.R., Ameyaw, E.E. (2020). Artificial \nintelligence in the AEC industry: Scientometric analysis \nand visualization of research activities. Automation in \nConstruction, \n112: \n103081. https://doi.org/10.1016/j.autcon.2020.103081 [9] Ram, S. (2018). India’s contribution and research impact \nin leishmaniasis research: A bibliometric analysis. Journal \nof \nScientometric, \n7(1): \n38-44. https://doi.org/10.5530/jscires.7.1.5 [10] Koster, C., Klingelhofer, D., Groneberg, D.A., Schwarzer, \nM. (2016). Rotavirus - Global research density equalizing \nmapping and gender analysis. Vaccine, 34(1): 90-100. https://doi.org/10.1016/j.vaccine.2015.11.002 [24] van Eck, N.J. Waltman, L. (2010). Software survey: \nVOSviewer, a computer program for bibliometric \nmapping. Scientometrics, \n84(2): \n523-538. https://doi.org/10.1007/s11192-009-0146-3 [11] Singh, N. (2016). Scientometric analysis of research on \nZika \nvirus. Virus \ndisease, \n27(3): \n303-306. https://doi.org/10.1007/s13337-016-0339-3 [25] van Eck, N.J., Waltman, L. (2014). Visualizing \nbibliometric networks. Measuring Scholarly Impact, 285-\n320. https://doi.org/10.1007/978-3-319-10377-8_13 p\ng\n[12] Basu, T., Mallik, A., Mandal, N. (2017). Evolving 1345 29(1). [26] RStudio. https://en.wikipedia.org/wiki/RStudio, accessed\non Nov. 22, 2023. Professional, \n29(1). https://doi.org/10.3145/epi.2020.ene.03 Professional, \n29(1). https://doi.org/10.3145/epi.2020.ene.03 [27] Shonhe, L. (2020). Continuous Professional Development\n(CPD) of librarians: A bibliometric analysis of research\nproductivity viewed through WoS. The Journal of\nAcademic \nLibrarianship, \n46(2): \n102106. https://doi.org/10.1016/j.acalib.2019.102106 [29] Rodriguez, M.J. (2015). A critical review of SCImago\nJournal & Country Rank. Research Evaluation, 24(4):\n343-354. https://doi.org/10.1093/reseval/rvu008 [30] SJR: Scimago Journal & Country Rank. (2020). Retrieved\nfrom https://www.scimagojr.com/. [28] Moral-Munoz, J.A., Herrera-Viedma, E., Santisteban-\nEspejo, A., Cobo, M.J. (2020). Software tools for\nconducting bibliometric analysis in science: An up-to-\ndate review. Profesional De La información Information [31] Kumar, A. (2018). Is “Impact” the “Factor” that matters...? (Part I). Journal of Indian Society of Periodontology,\n22(2): 95-96. https://doi.org/10.4103/jisp.jisp_195_18 1346"
https://openalex.org/W2973465475	https://www.jstage.jst.go.jp/article/jea/30/10/30_JE20190065/_pdf	English	null	Epidemiologic Features of Kawasaki Disease in Shanghai From 2013 Through 2017	Journal of epidemiology	2,020	cc-by	6,422	Li-ping Xie1, Wei-li Yan2, Min Huang3, Mei-rong Huang4, Sun Chen5, Guo-ying Huang1, and Fang Liu1, on Behalf of the Shanghai Kawasaki Disease Research Group 1Heart Center, Children’s Hospital of Fudan University, Shanghai, China 5Department of Pediatric Cardiology, Xinhua Hospital, Aﬃliated to Shanghai Jiao Tong Un Received April 16, 2019; accepted August 12, 2019; released online September 21, 2019 Received April 16, 2019; accepted August 12, 2019; released online September 21, 2019 ABSTRACT Of 4,325 patients (97.0%) receiving intravenous immunoglobulin (IVIG), 362 (8.4%) were resistant to initial IVIG. CAL occurred in 406 (9.1%) patients, including 118 (2.7%) with medium aneurysms and 31 (0.7%) with giant aneurysms. Recurrent cases were 60 (1.3%). No death was found in this survey. Higher platelet levels, lower albumin levels, male sex, incomplete KD, IVIG resistance, and receiving initial IVIG ≤4 days or >10 days, were independently associated with CAL. Results: A total of 4,452 cases were enrolled. Male-to-female ratio was 1.7:1. The incidence of KD was 68.8 to 107.3 per 100,000 children aged <5 years from 2013 to 2017. Age at onset ranged from 15 days to 14.0 years (median: 1.8 years). KD occurred more frequently in spring and summer. Of 4,325 patients (97.0%) receiving intravenous immunoglobulin (IVIG), 362 (8.4%) were resistant to initial IVIG. CAL occurred in 406 (9.1%) patients, including 118 (2.7%) with medium aneurysms and 31 (0.7%) with giant aneurysms. Recurrent cases were 60 (1.3%). No death was found in this survey. Higher platelet levels, lower albumin levels, male sex, incomplete KD, IVIG resistance, and receiving initial IVIG ≤4 days or >10 days, were independently associated with CAL. p y Conclusions: The incidence of KD in Shanghai had substantially increased while the proportion of CAL had substantially decreased as compared with our previous surveys. Higher platelet levels, lower albumin levels, male sex, incomplete KD, IVIG resistance, and receiving initial IVIG ≤4 days or >10 days, were risk factors for CAL. Key words: Kawasaki disease; epidemiology; coronary artery lesion; Shanghai Key words: Kawasaki disease; epidemiology; coronary artery lesion; Shanghai Copyright © 2019 Li-ping Xie et al. This is an open access article distributed under the terms of Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. a decreasing trend with the increasing administration of large dose of intravenous immunoglobulin (IVIG).3,4,8,9 In Shanghai, the proportion of CAL decreased from 19.8% to 15.9% from 1998 through 2012.6,11 Address for correspondence. Fang Liu, Heart Center, Children’s Hospital of Fudan University, Shanghai 201102, China (e-mail: liufang@fudan.edu.cn); Guoying Huang, Heart Center, Children’s Hospital of Fudan University, Shanghai 201102, China (e-mail: gyhuang@shmu.edu.cn). DOI https://doi.org/10.2188/jea.JE20190065 HOMEPAGE http://jeaweb.jp/english/journal/index.html Journal of Epidemiology Original Article ABSTRACT Background: We sought to investigate epidemiologic features of Kawasaki disease (KD) in Shanghai from 2013 through 2017 and identify risk factors for coronary artery lesions (CAL). Background: We sought to investigate epidemiologic features of Kawasaki disease (KD) in Shanghai from 2013 through 2017 and identify risk factors for coronary artery lesions (CAL). Methods: As in our previous three surveys, a set of questionnaires and diagnostic guidelines for KD were sent to 50 hospitals providing pediatric medical care in Shanghai. Medical records of KD patients diagnosed from January 2013 through December 2017 were retrospectively analyzed. Multivariate logistic regression analysis was performed to identify risk factors for CAL. Methods: As in our previous three surveys, a set of questionnaires and diagnostic guidelines for KD were sent to 50 hospitals providing pediatric medical care in Shanghai. Medical records of KD patients diagnosed from January 2013 through December 2017 were retrospectively analyzed. Multivariate logistic regression analysis was performed to identify risk factors for CAL. Methods: As in our previous three surveys, a set of questionnaires and diagnostic guidelines for KD were sent to 50 hospitals providing pediatric medical care in Shanghai. Medical records of KD patients diagnosed from January 2013 through December 2017 were retrospectively analyzed. Multivariate logistic regression analysis was performed to identify risk factors for CAL. Results: A total of 4,452 cases were enrolled. Male-to-female ratio was 1.7:1. The incidence of KD was 68.8 to 107.3 per 100,000 children aged <5 years from 2013 to 2017. Age at onset ranged from 15 days to 14.0 years (median: 1.8 years). KD occurred more frequently in spring and summer. Of 4,325 patients (97.0%) receiving intravenous immunoglobulin (IVIG), 362 (8.4%) were resistant to initial IVIG. CAL occurred in 406 (9.1%) patients, including 118 (2.7%) with medium aneurysms and 31 (0.7%) with giant aneurysms. Recurrent cases were 60 (1.3%). No death was found in this survey. Higher platelet levels, lower albumin levels, male sex, incomplete KD, IVIG resistance, and receiving initial IVIG ≤4 days or >10 days, were independently associated with CAL. Results: A total of 4,452 cases were enrolled. Male-to-female ratio was 1.7:1. The incidence of KD was 68.8 to 107.3 per 100,000 children aged <5 years from 2013 to 2017. Age at onset ranged from 15 days to 14.0 years (median: 1.8 years). KD occurred more frequently in spring and summer. INTRODUCTION Kawasaki disease (KD) is an acute, self-limited febrile illness that predominantly aﬀects children under 5 years of age. The etiology remains unclear. It is a systemic vasculitis mainly aﬀecting coronary arteries1 and is becoming the leading cause of childhood acquired heart disease both in developed countries and China.2 We have conducted three epidemiologic surveys of KD in Shanghai since 1998 (once every 5 years). Considering that epidemiologic features of KD keeps changing since our ﬁrst survey, especially the incidence of KD and the proportion of CAL, we continued to carry out the fourth survey to keep track of the latest changes. We aimed to provide the descriptive epidemiology of KD in Shanghai from 2013 through 2017 and identify independent risk factors for CAL. The incidence of KD remains highest in Japan,3 followed by South Korea,4 Taiwan,5 and mainland China,6,7 whereas it was much lower in European8 and American countries.9,10 Actually, the incidence of KD has recently shown an increasing trend in many regions, especially in Asian countries.3–5,7 In Shanghai, a well-developed city in China, the incidence of KD increased from 16.8 to 55.5 per 100,000 children <5 years old from 1998 through 2012.6,11 MATERIALS AND METHODS Statistical analysis Data were collected by pediatricians, including demographic information, clinical manifestations, prognosis, laboratory in- dices, treatment, and echocardiographic ﬁndings. The contents of the questionnaire were mildly modiﬁed, only with addition of erythema and induration at Bacillus Calmette-Guérin (BCG) sites, sites of valvular regurgitation, the presence of coronary stenosis and thrombosis, and some new laboratory indices including neutrophil count (NEUT), lymphocyte count (LYM), aspartate transaminase (AST), serum sodium (Na), and total bilirubin (TB). After all questionnaires were returned, two senior pediatric cardiologists were responsible for further veriﬁcation. All eligible cases were entered into the KD database. This study was approved by the institutional review board of Children’s Hospital of Fudan University and requirements for informed consents were waived for this retrospective study. The incidence of KD was calculated by dividing the number of newly diagnosed KD patients <5 years old who inhabited Shanghai by the resident population of the corresponding age group in Shanghai. Census data were acquired from Shanghai Municipal Center for Disease Control and Prevention. Data were presented as mean (standard deviation) or median (interquartile range) for continuous variables, and count (percentage) for categorical variables. Continuous variables were compared between groups using unpaired Student’s t tests or Mann-Whitney U test. Categorical variables were compared using χ2 test or Fisher exact test. Univariate analysis was performed to explore potential risk factors for CAL, including age of onset, male sex, KD type, the dose and fever days of initial IVIG, and all 13 laboratory indices. Among them, CRP was divided into four groups based on quartiles due to the presence of truncated values. Multivariate logistic regressions were performed to identify risk factors independently associated with CAL. Demographic data and prognosis g g Fifty hospitals (100% response rate) returned the questionnaires by the deadline, with 4,533 cases reported. Among them, 37 non- acute cases and 44 repeated cases were excluded. Ultimately, 4,452 cases were enrolled, of which 2,824 (63.4%) were male and 1,628 (36.6%) were female. Male-to-female ratio was 1.7:1. Sixty patients (1.3%) recurred during this study period, of which 54 (1.2%) relapsed for the ﬁrst time, 5 (0.1%) for the second, and 1 for the third. The average interval of the ﬁrst, second, and third recurrence was 13.7 months (range: 0.5–60 months), 8.4 months (range: 1–21 months) and 32 months, respectively. No death was found in this survey. Data collection Coronary artery lesions (CAL) are still the main factors aﬀecting the long-term prognosis and mortality of KD. Fortunately, the proportion of CAL among KD patients showed As in our previous three surveys, a set of questionnaires and diagnostic guidelines were sent to all hospitals providing pediatric DOI https://doi.org/10.2188/jea.JE20190065 HOMEPAGE http://jeaweb.jp/english/journal/index.html 429 Epidemiology of Kawasaki Disease in Shanghai medical care in Shanghai (50 in total). All patients discharged from these hospitals from January 1, 2013 through December 31, 2017 with an International Classiﬁcation of Diseases (ICD) code for KD or mucocutaneous lymph node syndrome (ICD9 446.1 and ICD10 M30.3) were enrolled. Coronary thrombosis was assessed using echocardiography or angiography and coronary stenosis was assessed only using coronary angiography. Incidence of KD IVIG resistance was deﬁned as recurrent or persistent fever (>38°C) after 36 hours of completion of initial IVIG infusion.1 Laboratory indices before initial IVIG infusion were recorded, including the highest values of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), platelet count (PLT), white blood cell count (WBC), creatine kinase-muscle=brain (CK-MB), alanine aminotransferase (ALT), AST, and TB, and the lowest values of hemoglobin (HB), serum albumin (ALB), and Na if laboratory tests were performed more than once before initial IVIG. NEUT and LYM of the same blood test with highest WBC were also recorded. The incidence of KD was 68.8–107.3 per 100,000 children aged <5 years from 2013 to 2017, with an average of 94.7 (112.7 for males and 75.4 for females; Table 1). It had increased substantially from 2013 to 2015, but remained stable in 2016 and 2017. Generally, the incidence of KD in Shanghai was on the rise since our ﬁrst survey in 1998 (Figure 1). Age of onset Age at onset of KD ranged from 15 days to 14.0 years (median: 1.8 years), peaking at 1 year of age. A total of 464 cases (10.4%) occurred before 6 months old, 1,133 cases (25.4%) before 1 year old, 2,460 cases (55.3%) before 2 years old, and 3,985 cases (89.5%) before 5 years old. Only one patient, a 15-day-old boy, was a neonate. Males and females shared similar age distribution with the total population (Figure 2). CAL, deﬁned as coronary dilation or aneurysm, was evaluated using two-dimensional echocardiography. A patient was consid- ered to have CAL if the luminal diameter of a coronary artery was >3.0 mm in children aged <5 years or >4.0 mm in those aged ≥5 years, or when the internal diameter of a segment was ≥1.5 times that of an adjacent segment.14 A medium aneurysm was deﬁned as an internal luminal diameter from 4 to 8 mm and a giant aneurysm was deﬁned as an internal luminal diameter ≥8 mm. Coronary measurements of the most severe echocardiography were recorded if the measurement was performed more than once. Patients and deﬁnitions KD was diagnosed according to the guidelines released by the Japanese Circulation Society.12 The diagnostic criteria were as follows: (1) fever persisting for 5 days or longer (including cases in whom the fever has subsided before the 5th day in response to therapy); (2) bilateral conjunctival congestion; (3) polymorphous exanthema; (4) changes of lips and oral cavity: reddening of lips, strawberry tongue, diﬀuse injection of oral and pharyngeal mucosa; (5) changes of peripheral extremities: reddening of palms and soles in acute phase and membranous desquamation from ﬁngertips in convalescent phase; and (6) acute non-purulent cervical lymphadenopathy. KD was diagnosed in the presence of at least 5 of the 6 principle symptoms. Incomplete KD, referring to KD patients with fewer than 5 principle symptoms, was diagnosed according to the algorithm in AHA guidelines.1,13 Between-hospital transfers during the same episode were identiﬁed and repeated cases were excluded. Patients who hospitalized only for further evaluation and treatment of cardiac sequelae (non-acute cases) were also excluded. Seasonal distribution Similar to our previous surveys, KD occurred more frequently in spring (March to May; 1,137 cases, 25.5%) and summer (June to August; 1,235 cases, 27.7%), as compared with autumn (September to November; 1,045 cases, 23.5%) and winter 430 j J Epidemiol 2020;30(10):429-435 Xie L, et al. Table 1. Incidence of Kawasaki Disease in Shanghai, China, from 2013 through 2017 (per 100,000 children younger than 5 years) Year Males Females Total Number of KD cases Total population Annual incidence (=100,000) Number of KD cases Total population Annual incidence (=100,000) Number of KD cases Total population Annual incidence (=100,000) 2013 240 256,338 93.6 103 241,937 42.6 343 498,275 68.8 2014 247 244,986 100.8 154 231,271 66.6 401 476,257 84.2 2015 321 262,719 122.2 216 246,978 87.5 537 509,697 105.4 2016 324 274,484 118.0 247 257,505 95.9 571 531,989 107.3 2017 370 293,919 125.9 225 274,904 81.8 595 568,823 104.6 Total 1,502 1,332,446 112.7 945 1,252,595 75.4 2,447 2,585,041 94.7 KD, Kawasaki disease. i Disease in Shanghai, China, from 2013 through 2017 (per 100,000 children younger than 5 years) Incidence of Kawasaki Disease in Shanghai, China, from 2013 through 2017 (per 100,000 children younger tha Figure 1. Incidence rate of Kawasaki disease in Shanghai, China from 1998 through 2017. Figure 2. Age distribution of onset of Kawasaki disease in Shanghai, China from 2013 through 2017. Figure 1. Incidence rate of Kawasaki disease in Shanghai, China from 1998 through 2017. Figure 1. Incidence rate of Kawasaki disease in Shanghai, China from 1998 through 2017. Figure 1. Incidence rate of Kawasaki disease in Shanghai, China from 1998 through 2017. Figure 2. Age distribution of onset of Kawasaki disease in Shanghai, China from 2013 through 2017. Figure 2. Age distribution of onset of Kawasaki disease in Shanghai, China from 2013 through 2017. (December to February; 1,035 cases, 23.2%). There was a clear peak in May, which fell to baseline in September after maintaining for 3 months (Figure 3). There were 1,713 cases (38.5%) occurring between May and August. Clinical manifestations The most common manifestation was fever persisting for 5 days or longer (4,304 cases, 96.7%), followed by bilateral conjunctival congestion (3,933 cases, 88.3%), changes in the lips and oral cavity (3,757 cases, 84.4%), polymorphous exanthema (3,502 J Epidemiol 2020;30(10):429-435 j 431 Epidemiology of Kawasaki Disease in Shanghai Figure 3. Seasonal distribution of Kawasaki disease in Shanghai, China from 2013 through 2017. Figure 3. Seasonal distribution of Kawasaki disease in Shanghai, China from 2013 through 2017. cases, 78.7%), cervical lymphadenopathy (3,256 cases, 73.1%), and changes in peripheral extremities (2,221 cases, 49.9%). Crissum desquamate (667 cases, 15.0%), and erythema and indu- ration at BCG sites (191 cases, 4.3%) occurred less frequently. BCG reaction occurred more frequently in children under 1 year old (113=1,131 cases, 10.0%) than in those over 1 year old (78=3,321 cases, 2.3%; OR 4.62; 95% conﬁdence interval [CI], 3.43–6.21; P < 0.001). giant aneurysms. Furthermore, the left main coronary artery (LMCA; 287 cases, 75.5%) was the most commonly involved, followed by the right coronary artery (RCA; 187 cases, 49.2%), the left anterior descending artery (LAD; 150 cases, 39.5%), and the left circumﬂex artery (LCX; 28 cases, 7.4%). Two hundred and ten cases (55.3%) involved one site, 89 cases (23.4%) involved two sites, 60 cases (15.8%) involved three sites, and 21 cases (5.5%) involved four sites simultaneously. Nine (0.2%) of 4,447 patients had coronary thrombosis, including the involvement of RCA in 6 cases, LMCA in 3 cases, LAD in 2 cases, and LCX in 2 cases. Five of the 9 patients had giant aneurysms and 4 had moderate aneurysms. Coronary stenosis occurred in 14 (0.3%) of 4,447 patients, including the involvement of RCA in 9 cases, LMCA in 7 cases, LAD in 7 cases, and LCX in 1 case. Three patients had severe stenosis, all accompanied with giant aneurysms. Among the patients, 1,564 (35.1%) presented as incomplete KD, of which 1,020 (65.2%) had four principal signs, 352 (22.5%) had three, and 192 (12.3%) had two. Manifestations of other systems mainly included the involvement of respiratory system (1,291 cases, 29.0%), digestive system (681 cases, 15.3%), hematologic system (159 cases, 3.6%), urinary system (99 cases, 2.2%), skin (71 cases, 1.6%), joint (13 cases, 0.3%), and nervous system (33 cases, 0.7%). Other cardiovascular abnormalities included cardiomegaly (402 cases, 9.0%), cardiac dysfunction (19 cases, 0.4%), valvular regurgitation (653 cases, 14.7%), and pericardial eﬀusion (408 cases, 9.2%). Risk factors for coronary artery lesions However, 362 (8.4%) patients failed to respond to the initial treatment. For secondary treatment (354 patients in total because of 8 patients missing further treatment information), a second dose of IVIG (314 cases, 88.7%) was administered most, followed by steroids (41 cases, 11.6%; 1–2 mg=kg, 68.3%) and inﬂiximab (2 cases, 0.6%). Seventy-seven patients (21.8%) were not sensitive to secondary treatment. For third-line or subsequent treatment, steroids (63 cases, 81.8%; ≥10 mg=kg, 55.6%) were administered most, followed by IVIG (28 cases, 36.4%) and inﬂiximab (3 cases, 3.9%). Fever eventually resolved in all patients. y y The laboratory and clinical ﬁndings of KD patients with and without CAL were summarized in Table 2. Compared with the none-CAL group, the CAL group had higher levels of CRP, ESR, PLT, and LYM, and lower levels of HB, ALB, and CK-MB. Male sex, incomplete KD, and IVIG resistance were more common in the CAL group. The dose and fever days of initial IVIG also diﬀered between the two groups. WBC, NEUT, ALT, AST, Na, TB, and age of onset were not signiﬁcantly diﬀerent. The laboratory and clinical ﬁndings of KD patients with and without CAL were summarized in Table 2. Compared with the none-CAL group, the CAL group had higher levels of CRP, ESR, PLT, and LYM, and lower levels of HB, ALB, and CK-MB. Male sex, incomplete KD, and IVIG resistance were more common in the CAL group. The dose and fever days of initial IVIG also diﬀered between the two groups. WBC, NEUT, ALT, AST, Na, TB, and age of onset were not signiﬁcantly diﬀerent. PLT, and LYM, and lower levels of HB, ALB, and CK-MB. Male sex, incomplete KD, and IVIG resistance were more common in the CAL group. The dose and fever days of initial IVIG also diﬀered between the two groups. WBC, NEUT, ALT, AST, Na, TB, and age of onset were not signiﬁcantly diﬀerent. Therefore, 12 signiﬁcant variables from the univariate analysis, including CRP, ESR, PLT, LYM, HB, ALB, CK-MB, male sex, incomplete KD, the dose and fever days of initial IVIG, and IVIG resistance, were included in the multivariate analysis to identify risk factors for CAL. It was shown that a higher platelet level, a lower albumin level, male sex, incomplete KD, IVIG resistance, and receiving initial IVIG within 4 days or over 10 days, were independent risk factors for CAL (Table 3). Clinical manifestations More speciﬁcally, 638 patients (97.7%) had mitral regurgitation (630 mild and 8 moderate cases) and 16 patients (2.5%) had moderate tricuspid regurgitation. Treatment IVIG was administrated to 4,320 cases (97.0%; 6.9 ± 3.4 days, range: 2–55 days), including regimens of 2 g=kg once (2,232 cases, 51.7%), 1 g=kg for two consecutive days (1,671 cases, 38.7%), 1 g=kg once (394 cases, 9.1%) and irregular usage (23 cases, 0.5%). Cardiovascular abnormalities A total of 4,447 patients (99.9%) underwent echocardiography within 1 month of onset, of which 406 (9.1%) developed CAL. Except for 26 cases missing coronary measurements, the remaining 380 cases included 231 (5.2%) of dilation and small aneurysms, 118 (2.7%) of medium aneurysms, and 31 (0.7%) of 432 j J Epidemiol 2020;30(10):429-435 Xie L, et al. Xie L, et al. Table 2. Univariate analysis of relevant factors for coronary artery lesions Table 2. Cardiovascular abnormalities ALB, serum albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; CAL, coronary artery lesions; CK-MB, creatine kinase-muscle=brain; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; HB, hemoglobin; IVIG, intravenous immunoglobulin; KD, Kawasaki disease; LYM: lymphocyte count; Na, serum sodium; NEUT, neutrophil count; PLT, blood platelet count; Q1, quartile 1; Q3, quartile 3; SD, standard deviations; TB, total bilirubin; WBC, white blood cell count. blood cell count. aCRP was converted to a dichotomous variable by the upper quartile due to the presence of truncated values. Table 3. Multivariate logistic regression analysis of risk factors for coronary artery lesions (3,627 cases) Table 3. Multivariate logistic regression analysis of risk factors for coronary artery lesions (3,627 cases) Variables Odds Ratio 95% Conﬁdence Interval P PLT, 1012=L 1.002 1.001–1.003 <0.001 ALB, g=L 0.93 0.90–0.95 <0.001 Male sex 1.64 1.24–2.15 <0.001 Incomplete KD 1.63 1.26–2.09 <0.001 IVIG resistance 2.44 1.72–3.46 <0.001 Fever days at initial IVIG <0.001 ≤4 days 1.62 1.11–2.44 0.014 5–10 days Ref. Ref. >10 days 4.44 3.14–6.28 <0.001 ALB, serum albumin; IVIG, intravenous immunoglobulin; KD, Kawasaki disease; PLT, blood platelet count. Compared with the high prevalence of fever and ﬁve principal symptoms, crissum desquamate (15.0%) and erythema and induration at BCG site (4.3%) occurred less frequently. Not all patients had been checked for BCG site, which might be a reason for the lower rate of BCG site reaction in our study. Changes at the BCG site is increasingly recognized as a speciﬁc diagnostic tool for KD in Japan, especially in infants with incomplete KD.15,16 In our study, we also found that BCG site reaction was more prevalent in infants (59.2% for age <1 year). In addition, of 191 patients with positive BCG site reaction, 151 (79.1%) had polymorphous exanthema whereas 40 (20.9%) did not have. Considering the high speciﬁcity of BCG reaction in KD patients, we suggest that BCG reaction be added as a supplementary criterion for polymorphous exanthema when diagnosing KD. p y p g g The proportion of IVIG administration increased from 71.8% to 97.0% since our ﬁrst survey in 1998.6,11,17 In addition, 90.3% of patients received 2 g=kg IVIG as initial treatment and 93.0% received IVIG within 10 days, indicating the standard treatment has been universally adopted in Shanghai. The proportion of IVIG resistance reported so far in Shanghai was 4.9% in 2008–20126 and 8.4% in 2013–2017. The only two reports were not enough to show the trend in our view. Cardiovascular abnormalities Therefore, continuous surveillance is needed. However, the proportion of IVIG resistance was lower than that in Japan3 (17.8%), South Korea4 (11.8%), and other reported countries (16.3% in the United States,18 9–14% in Canada,19 16.5% in Australia,20 26.8% in German,21 and 23.1% in the Netherlands22), which may partially be attributed to genetic factors. In addition, studies have shown Cardiovascular abnormalities Univariate analysis of relevant factors for coronary artery lesions Variables Total cases (CAL=none CAL) CAL None CAL P CRP ≥105 mg=L, n (%)a 387=4,022 118 (30.5) 1,004 (25.0) 0.017 ESR, mm=h, mean (SD) 358=3,823 73.9 (34.1) 70.3 (31.1) 0.041 PLT, 1012=L, mean (SD) 387=4,018 465.5 (206.7) 393.7 (163.9) <0.001 WBC, 109=L, mean (SD) 388=4,027 15.0 (7.4) 14.7 (6.1) 0.679 NEUT, 109=L, mean (SD) 385=3,928 9.8 (7.0) 9.5 (5.5) 0.310 LYM, 109=L, mean (SD) 377=3,883 3.9 (2.1) 3.7 (2.0) 0.032 HB, g=L, mean (SD) 384=4,012 106.4 (14.2) 108.8 (12.4) <0.001 ALB, g=L, mean (SD) 374=3,902 33.7 (4.5) 35.2 (4.2) <0.001 ALT, U=L, median (Q1, Q3) 376=3,956 25.0 (13.0, 50.5) 22.0 (12.0, 51.0) 0.175 AST, U=L, median (Q1, Q3) 375=3,942 27.0 (18.0, 42.5) 26.0 (18.0, 39.0) 0.277 CK-MB, IU=L, median (Q1, Q3) 351=3,686 16.0 (12.0, 21.0) 18.0 (13.5, 24.0) <0.001 Na, mmol=L, mean (SD) 366=3,845 136.6 (3.9) 136.2 (5.5) 0.171 TB, µmol=L, median (Q1, Q3) 366=3,868 6.1 (3.9, 10.0) 6.1 (4.1, 9.4) 0.688 Age of onset, years, median (Q1, Q3) 406=4,041 1.8 (0.9, 3.5) 1.8 (1.0, 3.2) 0.611 Male, n (%) 406=4,041 291 (71.7) 2,530 (62.6) <0.001 Incomplete KD, n (%) 406=4,041 181 (44.6) 1,381 (34.2) <0.001 IVIG resistance, n (%) 385=3,931 63 (16.4) 299 (7.6) <0.001 Fever days at initial IVIG, n (%) 385=3,931 <0.001 ≤4 days 45 (11.7) 389 (9.9) 5–10 days 253 (65.7) 3,328 (84.7) >10 days 87 (22.6) 214 (5.4) Initial IVIG treatment dose, n (%) 385=3,931 <0.001 2 g=kg × 1 197 (51.2) 2,034 (51.7) 1 g=kg × 2 141 (36.6) 1,527 (38.8) 1 g=kg × 1 35 (9.1) 359 (9.1) Irregular usage 12 (3.1) 11 (0.3) ALB, serum albumin; ALT, alanine aminotransferase; AST, aspartate transaminase; CAL, coronary artery lesions; CK-MB, creatine kinase-muscle=brain; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; HB, hemoglobin; IVIG, intravenous immunoglobulin; KD, Kawasaki disease; LYM: lymphocyte count; Na, serum sodium; NEUT, neutrophil count; PLT, blood platelet count; Q1, quartile 1; Q3, quartile 3; SD, standard deviations; TB, total bilirubin; WBC, white blood cell count. aCRP was converted to a dichotomous variable by the upper quartile due to the presence of truncated values. DISCUSSION The incidence of KD in Shanghai was on the rise since 1998, similar to the other Asian countries and regions.3–5,7 The highest ever record in Shanghai was 107.3 per 100,000 children aged <5 years in 2016, which was still much lower than that in Japan in 20153 (330.2 per 100,000 children aged <5 years) and in South Korea in 20134 (194.9 per 100,000 children aged <5 years), but was for the ﬁrst time higher than that in Taiwan in 20105 (82.8 per 100,000 children aged <5 years). The incidence of KD in Shanghai remained stable in 2016 and 2017. It requires continuous surveillance, whether or not a 3-year plateau period implies no further increase. 433 J Epidemiol 2020;30(10):429-435 j 433 Epidemiology of Kawasaki Disease in Shanghai its decline in the current survey. Second, epidemiological surveys in Shanghai could to some extent but not completely reﬂect epidemiologic features of KD in China. We hope to conduct a nationwide sample survey in the future. that administration of IVIG with chemical modiﬁcations (such as β-propiolactone and enzyme digestion) led to a higher incidence of IVIG nonresponse than with other manufacturing techni- ques.23,24 The IVIG used in Shanghai were all pH4 and prepared with acidiﬁcation rather than chemical modiﬁcations, which may also explain the lower incidence of IVIG nonresponse. Given that the mechanism of IVIG action and IVIG resistance remains unclear, it is really diﬃcult to explain the reasons for these discrepancies. More genetic and basic research is needed to clarify it. that administration of IVIG with chemical modiﬁcations (such as β-propiolactone and enzyme digestion) led to a higher incidence of IVIG nonresponse than with other manufacturing techni- ques.23,24 The IVIG used in Shanghai were all pH4 and prepared with acidiﬁcation rather than chemical modiﬁcations, which may also explain the lower incidence of IVIG nonresponse. Given that the mechanism of IVIG action and IVIG resistance remains unclear, it is really diﬃcult to explain the reasons for these discrepancies. More genetic and basic research is needed to clarify it. In conclusion, the incidence of KD in Shanghai has substantially increased over 20 years. Although the occurrence of CAL has substantially decreased, the proportion of giant aneurysms has remained the same. Higher platelet levels, lower albumin levels, male sex, incomplete KD, IVIG resistance, and receiving initial IVIG within 4 days or beyond 10 days, were independent risk factors for CAL. Continuous epidemiological investigations are still necessary. ACKNOWLEDGEMENTS We would like to express our sincere gratitude to all of the hospitals and investigators involved in this survey, including Children’s Hospital of Fudan University (Guo-ying Huang, Fang Liu, and Li-ping Xie), Children’s Hospital of Shanghai Jiaotong University (Min Huang, Ting-ting Xiao, Li-jian Xie, Wei Liu and Xun-wei Jiang), Shanghai Children’s Medical Center (Mei-rong Huang, Shubao Chen, Ying Guo and Xin-yi Xu), Xinhua Hospital of Shanghai Jiaotong University (Sun Chen, Jia Shen and Hui-ying Wang), Ruijin Hospital (Zhi-ya Dong and Cai-ping Zhang), The ﬁrst People’s Hospital of Shanghai (Jian-guo Hong and Xiao-jian Zhou), The sixth People’s Hospital of Shanghai (Fen Liu, Liang-xia Wu and Fu-juan Wang), The ﬁfth People’s Hospital of Shanghai (Xiao-ming Wang and Fei Wang), Changhai Hospital (Shi-yan Cao), Shanghai Tongji Hospital (Xiao-xun Zhou), Shanghai tenth People’s Hospital (Li Gu and Yan-jie Chen), Renji Hospital (Yu Zhao), Jinshan Hospital of Fudan University (Ye Chen), Shanghai Jiangwan Hospital (Yu Chen), Shanghai Putuo District Center Hospital (Yu Xie), American-Sino Women’s and Children’s Hospital (Guo-liang Teng and Ming-jun Zhang). In addition, the authors appreciate greatly the kind help from Conway Niu for preparing the manuscript and Shanghai Municipal Center for Disease Control and Prevention for providing the population data required to calculate the incidence rate of Kawasaki disease. Previous frequently reported risk factors for CAL, such as a higher platelet level,27,28 a lower albumin level,28,29 male sex,6,28–30 IVIG resistance6,29,31 and receiving initial IVIG beyond 10 days,6,27,28,32 were also demonstrated in our ﬁndings. As for incomplete KD, study results were contradictory even in multivariate analyses. One study showed that incomplete KD was more common in the CAL group due to delayed diagnosis rather than the low number of presenting symptoms.33 However, consistent with other two studies,31,34 our study identiﬁed incomplete KD itself as an independent risk factor for CAL, despite the presence of delayed IVIG treatment. The last risk factor, receiving initial IVIG within 4 days, was identiﬁed beyond expectations. ACKNOWLEDGEMENTS Nomura et al reported that KD patients treated with IVIG before 5 days had a higher incidence of aneurysms than those treated at the ﬁfth day or after (15.2% vs 1.3%; P = 0.004).35 We also found that patients treated within 4 days had a higher incidence of IVIG resistance (21.4% vs 9.1%; P < 0.001), a younger age of onset (1.4 vs 1.8 years; P < 0.001), higher levels of CRP (71 vs 62 mg=L; P = 0.001), ALT (28 vs 22 U=L; P < 0.001), AST (29 vs 26 U=L; P < 0.001) and TB (7.3 vs 6.0 µmol=L; P < 0.001), and lower levels of Na (135.7 vs 136.2 mmol=L; P < 0.001). Considering these, although infusion of IVIG within 4 days was identiﬁed as an independent risk factor for CAL in our analysis, it should be interpreted as a sign of severity of KD, which is consistent with Nomura’s views.35 Therefore, KD patients diagnosed within 4 days should be treated as early as possible, and may need more aggressive treatment (such as combination with steroids or inﬂiximab). Further prospective studies are needed to conﬁrm this. This work was supported by Joint Research Projects of Emerging Cutting-edge Technologies of Shanghai Municipal Hospitals [SHDC12013106]. Funded by Shanghai Natural Science Foundation [15ZR1404300]. Conﬂicts of interest: None declared. Conﬂicts of interest: None declared. DISCUSSION Fortunately, the proportion of CAL has substantially decreased from 19.8% to 9.1% since 1998,11 possibly attributed to the increasing proportion of IVIG administration and early treatment during the acute phase. Similarly, the proportion of coronary dilation and aneurysm showed a decreasing trend in Japan3,25 (from 10.0% to 6.5%, 2007–2016) and in South Korea4,26 (from 18.5% to 12.5%, 2006–2014). However, the proportion of both medium aneurysms and giant aneurysms lack a signiﬁcant decrease in Shanghai. In this survey, severe coronary stenosis (all 3 cases) and thrombosis (5 of 9 cases) mainly occurred on the basis of giant aneurysms, increasing the risk of myocardial ischemia and sudden death. Therefore, further prevention of the development of giant aneurysms remains the top priority in the future. J. 2017;36(5):482–485. 2006;16(1):9–14. 28. Honkanen VE, McCrindle BW, Laxer RM, Feldman BM, Schneider R, Silverman ED. Clinical relevance of the risk factors for coronary artery inﬂammation in Kawasaki disease. Pediatr Cardiol. 2003; 24(2):122–126. 12. JCS Joint Working Group. Guidelines for diagnosis and manage- ment of cardiovascular sequelae in Kawasaki disease (JCS 2013). Digest version. Circ J. 2014;78(10):2521–2562. 29. Hua W, Ma F, Wang Y, et al. A new scoring system to predict Kawasaki disease with coronary artery lesions. Clin Rheumatol. 2019;38(4):1099–1107. 13. Newburger JW, Takahashi M, Gerber MA, et al; Committee on Rheumatic Fever, Endocarditis and Kawasaki Disease; Council on Cardiovascular Disease in the Young; American Heart Association; American Academy of Pediatrics. Diagnosis, treatment, and long- term management of Kawasaki disease: a statement for health professionals from the Committee on Rheumatic Fever, Endocarditis and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association. Circulation. 2004;110(17): 2747–2771. 30. Kuwabara M, Yashiro M, Ae R, Yanagawa H, Nakamura Y. The eﬀects of early intravenous immunoglobulin therapy for Kawasaki disease: the 22nd nationwide survey in Japan. Int J Cardiol. 2018; 269:334–338. 31. Kim JJ, Hong YM, Yun SW, et al; Korean Kawasaki Disease Genetics Consortium. Assessment of risk factors for Korean children with Kawasaki disease. Pediatr Cardiol. 2012;33(4):513–520. 32. Qiu H, He Y, Rong X, et al. Delayed intravenous immunoglobulin treatment increased the risk of coronary artery lesions in children with Kawasaki disease at diﬀerent status. Postgrad Med. 2018; 130(4):442–447. 14. JCS Joint Working Group. Guidelines for diagnosis and manage- ment of cardiovascular sequelae in Kawasaki disease (JCS 2008)— digest version. Circ J. 2010;74(9):1989–2020. g 15. Loh ACE, Kua PHJ, Tan ZL. Erythema and induration of the Bacillus Calmette-Guerin site for diagnosing Kawasaki disease. Singapore Med J. 2019;60(2):89–93. 33. Kim T, Choi W, Woo CW, et al. Predictive risk factors for coronary artery abnormalities in Kawasaki disease. Eur J Pediatr. 2007; 166(5):421–425. 16. Araki T, Kodera A, Kitada K, et al. Analysis of factors associated with development of Bacille Calmette-Guerin inoculation site change in patients with Kawasaki disease. J Int Med Res. 2018;46(4):1640– 1648. 34. Song D, Yeo Y, Ha K, et al. Risk factors for Kawasaki disease- associated coronary abnormalities diﬀer depending on age. Eur J Pediatr. 2009;168(11):1315–1321. 17. Ma XJ, Yu CY, Huang M, et al; Shanghai Kawasaki Research Group. Epidemiologic features of Kawasaki disease in Shanghai from 2003 through 2007. Chin Med J (Engl). 2010;123(19):2629– 2634. 35. J. 2017;36(5):482–485. 21. Jakob A, Whelan J, Kordecki M, et al. Kawasaki disease in Germany: a prospective, population-based study adjusted for under- reporting. Pediatr Infect Dis J. 2016;35(2):129–134. 5. Lin MC, Lai MS, Jan SL, Fu YC. Epidemiologic features of Kawasaki disease in acute stages in Taiwan, 1997–2010: eﬀect of diﬀerent case deﬁnitions in claims data analysis. J Chin Med Assoc. 2015;78(2):121–126. 22. Tacke CE, Breunis WB, Pereira RR, Breur JM, Kuipers IM, Kuijpers TW. Five years of Kawasaki disease in the Netherlands: a national surveillance study. Pediatr Infect Dis J. 2014;33(8):793– 797. 6. Chen JJ, Ma XJ, Liu F, et al; Shanghai Kawasaki Disease Research Group. Epidemiologic features of Kawasaki disease in Shanghai from 2008 through 2012. Pediatr Infect Dis J. 2016;35(1):7–12. 23. Tsai MH, Huang YC, Yen MH, et al. Clinical responses of patients with Kawasaki disease to diﬀerent brands of intravenous immuno- globulin. J Pediatr. 2006;148(1):38–43. g f 7. Du ZD, Zhao D, Du J, et al; Beijing Kawasaki Research Group. Epidemiologic study on Kawasaki disease in Beijing from 2000 through 2004. Pediatr Infect Dis J. 2007;26(5):449–451. 24. Lin MC, Fu YC, Jan SL, Lai MS. Comparative eﬀectiveness of intravenous immunoglobulin for children with Kawasaki disease: a nationwide cohort study. PLoS One. 2013;8(5):e63399. 8. Tulloh RMR, Mayon-White R, Harnden A, et al. Kawasaki disease: a prospective population survey in the UK and Ireland from 2013 to 2015. Arch Dis Child. 2019;104(7):640–646. 25. Nakamura Y, Yashiro M, Uehara R, et al. Epidemiologic features of Kawasaki disease in Japan: results of the 2007–2008 nationwide survey. J Epidemiol. 2010;20(4):302–307. 9. Manlhiot C, O’Shea S, Bernknopf B, et al. Epidemiology of Kawasaki disease in Canada 2004 to 2014: comparison of sur- veillance using administrative data vs periodic medical record review. Can J Cardiol. 2018;34(3):303–309. y p ( ) 26. Park YW, Han JW, Hong YM, et al. Epidemiological features of Kawasaki disease in Korea, 2006–2008. Pediatr Int. 2011;53(1):36– 39. 10. Okubo Y, Nochioka K, Sakakibara H, Testa M, Sundel RP. National survey of pediatric hospitalizations due to Kawasaki disease and coronary artery aneurysms in the USA. Clin Rheumatol. 2017;36 (2):413–419. 27. Chantasiriwan N, Silvilairat S, Makonkawkeyoon K, Pongprot Y, Sittiwangkul R. Predictors of intravenous immunoglobulin resist- ance and coronary artery aneurysm in patients with Kawasaki disease. Paediatr Int Child Health. 2018;38(3):209–212. 11. Huang GY, Ma XJ, Huang M, et al. Epidemiologic pictures of Kawasaki disease in Shanghai from 1998 through 2002. J Epidemiol. REFERENCES 1. McCrindle BW, Rowley AH, Newburger JW, et al; American Heart Association Rheumatic Fever, Endocarditis, and Kawasaki Disease Committee of the Council on Cardiovascular Disease in the Young; Council on Cardiovascular and Stroke Nursing; Council on Cardiovascular Surgery and Anesthesia; Council on Epidemiology and Prevention. Diagnosis, treatment, and long-term management of Kawasaki disease: a scientiﬁc statement for health professionals from the American Heart Association. Circulation. 2017;135(17): e927–e999. There were a couple of limitations in our current study. First, Japanese diagnostic criteria for CAL could lead to underdiagnosis due to the failure of taking patients’ size into account.36 Z score, a normalization of coronary dimensions for body surface area, maybe more suitable to evaluate CAL.37,38 Since Z score has not been widely implemented in Shanghai, we still use coronary diameters to evaluate CAL in this survey and the proportion of CAL could be underestimated, although it is encouraging to see 2. Taubert KA. Epidemiology of Kawasaki disease in the United States and worldwide. Prog Pediatr Cardiol. 1997;6(3):181–185. 3. Makino N, Nakamura Y, Yashiro M, et al. The nationwide epidemiologic survey of Kawasaki disease in Japan, 2015–2016. Pediatr Int. 2019;61(4):397–403. 4. Kim GB, Park S, Eun LY, et al. Epidemiology and clinical features of Kawasaki disease in South Korea, 2012–2014. Pediatr Infect Dis 434 j J Epidemiol 2020;30(10):429-435 Xie L, et al. J. 2017;36(5):482–485. Nomura Y, Masuda K, Yoshinaga M, Sameshima K, Miyata K. Patients diagnosed with Kawasaki disease before the ﬁfth day of illness have a higher risk of coronary artery aneurysm. Pediatr Int. 2002;44(4):353–357. 18. Moﬀett BS, Syblik D, Denﬁeld S, Altman C, Tejtel-Sexson K. Epidemiology of immunoglobulin resistant Kawasaki disease: results from a large, national database. Pediatr Cardiol. 2015;36(2):374– 378. 36. de Zorzi A, Colan SD, Gauvreau K, Baker AL, Sundel RP, Newburger JW. Coronary artery dimensions may be misclassiﬁed as normal in Kawasaki disease. J Pediatr. 1998;133(2):254–258. 37. Dallaire F, Dahdah N. New equations and a critical appraisal of coronary artery Z scores in healthy children. J Am Soc Echocardiogr. 2011;24(1):60–74. 19. Lin YT, Manlhiot C, Ching JC, et al. Repeated systematic surveillance of Kawasaki disease in Ontario from 1995 to 2006. Pediatr Int. 2010;52(5):699–706. 38. Kobayashi T, Fuse S, Sakamoto N, et al; Z Score Project Investigators. A new Z score curve of the coronary arterial internal diameter using the lambda-mu-sigma method in a pediatric popula- tion. J Am Soc Echocardiogr. 2016;29(8):794–801.e29. 20. Saundankar J, Yim D, Itotoh B, et al. The epidemiology and clinical features of Kawasaki disease in Australia. Pediatrics. 2014;133(4): e1009–e1014. J Epidemiol 2020;30(10):429-435 j 43
https://openalex.org/W4245916503	https://link.springer.com/content/pdf/10.1007%2F978-3-319-70815-7_22.pdf	English	null	(De)briefing	Fundamental theories of physics	2,018	cc-by	3,326	Chapter 22 (De)brieﬁng This chapter is for those who neither want to bother with details nor have time for the expositions and the explanatory rants of previous parts; or as a teaser to go deeper into the subject and to want to know more. © The Author(s) 2018 K. Svozil, Physical (A)Causality, Fundamental Theories of Physics 192, https://doi.org/10.1007/978-3-319-70815-7_22 22.1 Provable Unknowables First and foremost, from the rational, scientiﬁc point of view, there is and never will be anything like “absolute randomness.” Knowledge of absolute randomness, if it “exists” in some platonic realm of ideas, is ineffable and thus strictly metaphysical and metamathematical, as it is blocked by various theorems about the impossibility of induction (cf. Chap.7, p. 35ff), forecasting (cf. Chap.6, p. 29ff), and representation (cf. Sect. A.4, p. 173ff). Any proof of such theorems, and thus their validity, is only relative to the assumptions made. Claims regarding “absolute randomness” – and, for the same reasons, “absolute determinism” – in physics should therefore be met with utmost skepticism. Such postures might serve as a heuristic principle, a sign-post, but they do not signify anything beyond the contemporary, most likely transient (some might even say spu- rious), worldview, as well as the personal and subjective preferences of the individual issuing them. Like all constructions of the mind and society, physical theories are suspended in free thought – an echo chamber of sorts. Whoever trusts a physical random number generator has to trust the assurances of the physical authorities that it indeed performs as claimed – in this case, that it produces random numbers. The authorities in turn base their judgement on personal inclinations [68, p. 866] and in metaphysical assertions [589]; as well as on their trust on the theories and models of functioning of such devices. Theories and models are considered trustworthy if they satisfy a “reasonable” and “meaningful” catalogue of criteria; but never more than that. 163 164 22 (De)brieﬁng 22 (De)brieﬁng One such reasonable criterion is the requirement that it should at least in principle be possible to locate the (re)source for randomness or indeterminism. Unfortunately both in quantum mechanics, as well as for classical systems, there is no such consol- idated agreement about the physical resources of randomness. Therefore, whenever a physical random number generator is employed, one has to bear in mind the inse- cure, and means relative, performance of this device. It is not that, pragmatically and for all practical purposes, it would not be usable. But it could fail in particular circumstances one has little idea about, and control of. 22.2 Quantum (In)Determinism There are three classes or types of quantum indeterminism: complementarity (cf. Sect.12.3 and Chap.19), value indeﬁniteness (often, referred to as contextuality after the realist Bell; cf. Sect.12.9.8.7, p. 97ff), as well as single measurement outcomes and events; all of them tied to the quantum measurement problem (cf. Sect.12.10, p. 118ff). Thus quantum random number generators are subject to some form of the quantum measurement problem, which lies at the heart of an ongoing debate – a debate which has been declared (re)solved or superﬂuous by various self-proclaimed authorities for a variety of conﬂicting reasons. Alas, quantum mechanics, despite being immensely useful for the prediction and comprehension of certain phenom- ena, formally operates with an inconsistent set of rules; in particular, pertaining to measurement. As has already been pointed out by von Neumann, the assumption of irreversible measurements contradicts the unitary deterministic evolution of the quantum state. (Inconsistencies, even in the core of mathematics, such as in Cantor’s set theory, should be rather a reason for consideration and prudence but not cause too much panic – after all, as noted earlier, those constructions of our mind are suspended in our free thought.) Some supposedly “active” elements such as beam splitters are represented by perfectly deterministic (unitary, that is, distance preserving permutations, such as the Hadamard gate) evolutions (cf. Chap.11 and Sect.12.5). Therefore they cannot be directly identiﬁed as quantum resource for indeterminism. The measurement process in quantum mechanics appears to be related to entan- glement and individuation: in order to be able to know from each other, the mea- surement apparatus has to acquire knowledge about the object; and in order to do so, the former has to interact with the latter. Thereby entanglement in the form of relational properties of object and apparatus is created. Because of the permutativity (one-to-one-ness) of the entire process (resulting in a sort of zero-sum game) these relationally deﬁnite properties (or, by another term, statistical correlations) come at the price of the indeﬁniteness of the individual, constituent parts – the original object as well as the measurement device are in no deﬁnite individual state any longer. If one forces individuality upon them (by some later measurement on the individual parts), then the outcome cannot be totally (but may be partly) pre-determined by the state of the constituent parts before that measurement. 22.3 Classical (In)Determinism Classical (in)determinism depends on its deﬁnition, and on the assumptions made. One of these assumptions is the existence of the continuum – not only as formal con- venience but as a physical entity. Almost all elements of a continuum are random (cf. Sect. A.2, p. 171ff). Any computable form of evolution “revealing” the algorithmic information content “buried” in a single supposedly random real physical entity (e.g., initial values) corresponds to a form of deterministic chaos (cf. Chap.18, p. 141ff). If the assumption of the physical existence of the continuum is dropped in favour of constructive, computable entities, then what remains from these indeterministic scenarios is the high sensitivity of the system behaviour on variations of initial states. Another form of model-induced classical indeterminism is due to representation and formalization of classical physical systems in terms of differential equations. In such cases the question of uniqueness of its solutions is crucial. Nonunique solutions indicate indeterminism. However, the requirement of Lipschitz continuity guarantees uniqueness in many cases which appear to be indeterministic (due to the possibility of weak solutions) without this property (cf. Sect.17.4, p. 137ff). Classical (in)determinism depends on its deﬁnition, and on the assumptions made. One of these assumptions is the existence of the continuum – not only as formal con- venience but as a physical entity. Almost all elements of a continuum are random (cf. Sect. A.2, p. 171ff). Any computable form of evolution “revealing” the algorithmic information content “buried” in a single supposedly random real physical entity (e.g., initial values) corresponds to a form of deterministic chaos (cf. Chap.18, p. 141ff). If the assumption of the physical existence of the continuum is dropped in favour of constructive, computable entities, then what remains from these indeterministic scenarios is the high sensitivity of the system behaviour on variations of initial states. Another form of model-induced classical indeterminism is due to representation and formalization of classical physical systems in terms of differential equations. In such cases the question of uniqueness of its solutions is crucial. Nonunique solutions indicate indeterminism. However, the requirement of Lipschitz continuity guarantees uniqueness in many cases which appear to be indeterministic (due to the possibility of weak solutions) without this property (cf. Sect.17.4, p. 137ff). 22.2 Quantum (In)Determinism Thereby it may be justiﬁed to say Quantum (In)Determinism 165 22.2 that “the measurement creates the outcome which is indeterminate before.” But this is a rather trivial statement expressing the fact that the outside environment with its supposedly huge number of degrees of freedom, in particular also the measurement device, has contributed to the outcome. The author’s impression is that Bohr and his followers may never have understood the true reason for value indeﬁniteness: the scarcity and constancy of information encoded into the quantum state; and the entanglement across the Heisenberg cut between object and measurement device. This scarcity also shows up in “static” Kochen–Specker type theorems [6, 314, 401] expressing the fact that only a single maximal observable or context is deﬁned at any time. 22.5 Perception and Forward Tactics Toward Unknowns Whatever one’s personal inclinations toward (in)determinism may be – one might characterize our situation either as an ocean of unknowns with a few islands of pre- liminary predictables; or, conversely, as a sea of determinism with the occasional islands or gaps of an otherwise lawful behaviour – every such inclination remains strictly means relative, metaphysical and subjective. Maybe such preferences says more about the person than the situation; because a person’s stance is often deter- mined by the subconscious desires, hopes and fears driving that individual. Choose one, and choose wisely for your needs; or even better, “if you can possibly avoid it [211, p. 129],” choose none, and remain conscious about the impossibility to know. Let me ﬁnally quote the late Planck [410] concluding that [409, p. 539] (see also Earman [186, p. 1372]) “. . . the law of causality is neither right nor wrong, it can be neither generally proved nor generally disproved. It is rather a heuristic principle, a sign-post (and to my mind the most valuable sign-post we possess) to guide us in the motley confusion of events and to show us the direction in which scientiﬁc research must advance in order to attain fruitful results. As the law of causality immediately seizes the awakening soul of the child and causes him indefatigably to ask “Why?” so it accompanies the investigator through his whole life and incessantly sets him Whatever one’s personal inclinations toward (in)determinism may be – one might characterize our situation either as an ocean of unknowns with a few islands of pre- liminary predictables; or, conversely, as a sea of determinism with the occasional islands or gaps of an otherwise lawful behaviour – every such inclination remains strictly means relative, metaphysical and subjective. Maybe such preferences says more about the person than the situation; because a person’s stance is often deter- mined by the subconscious desires, hopes and fears driving that individual. Choose one, and choose wisely for your needs; or even better, “if you can possibly avoid it [211, p. 129],” choose none, and remain conscious about the impossibility to know. Let me ﬁnally quote the late Planck [410] concluding that [409, p. 539] (see also Earman [186, p. 1372]) “. . . the law of causality is neither right nor wrong, it can be neither generally proved nor generally disproved. 22.4 Comparison with Pseudo-randomness Should one prefer physical (re)sources of randomness over mathematical pseudo- random? Of course, “anyone who considers arithmetical methods of producing ran- dom digits is . . . in a state of sin” [553, p. 768]. And yet, some desired features of randomness can be formally certiﬁed even for such computable entities. For instance, take Borel normality; that is, the property that every subsequence of length n occurs in a “large” b-ary sequence with frequency b−n. Almost all real numbers are normal to a given base b; in particular, all random sequences are Borel normal [102]. Yet, individual (even computable) numbers are hard to “pin down” as being normal; and no well-known mathematical constant, such as e or log 2, is known to be normal to any integer base. Also the normality of π, the ratio of 22 (De)brieﬁng 166 (De)brieﬁng the circumference to the diameter of a “perfect” (platonic) circle, remains conjec- tural [26], although particular digits are directly computable [25]. Von Neumann’s paper [553] quoted earlier contains a way to eliminate bias (and thus establish Borel normality up to length 1) of a binary sequence (essentially a partitioning of the sequence into subsequences of length 2, followed by a mapping of 00, 11 →∅, 01 →0, 10 →1); but only if this sequence is generated by independent physical events. Physical independence may be easy to obtain for all practical purposes, but difﬁcult in principle. On the other hand, Champernowne’s number 0.12345678910. . ., obtained by concatenating the decimal representations of the natural numbers in order, as well as the Copeland–Erdos constant 0.2357111317192329. . ., obtained by concatenating the prime numbers in order, are both Borel normal in base 10. So, if Borel normality sufﬁces for the particular task, then it might be better to consider such carefully chosenpseudo-randomnumbers(cf.Ref.[110]forcomparisonswithcertainquantum random sources). There exist situations which are perplexing yet not very helpful for practical purposes: Chaitin’s Ω (cf. Sect. A.6, p. 176ff) is also Borel normal in any base, and additionally it is provable random. Algorithms for computing the very ﬁrst couple of digits of Ω [109, 111] exist; alas the rate of convergence of the sum yielding Ω is so bad (in terms of time and other computational capacities worse than any computable function of the d-ary place) it is incomputable. 22.5 Perception and Forward Tactics Toward Unknowns It is rather a heuristic principle, a sign-post (and to my mind the most valuable sign-post we possess) to guide us in the motley confusion of events and to show us the direction in which scientiﬁc research must advance in order to attain fruitful results. As the law of causality immediately seizes the awakening soul of the child and causes him indefatigably to ask “Why?” so it accompanies the investigator through his whole life and incessantly sets him Perception and Forward Tactics Toward Unknowns 22.5 167 new problems. For science does not mean contemplative rest in possession of sure knowledge, it means untiring work and steadily advancing development.”1 new problems. For science does not mean contemplative rest in possession of sure knowledge, it means untiring work and steadily advancing development.”1 Planck also emphasized the joy of and the motivation from the unknown [411]: “We will never come to a completion, to the ﬁnal. Scientiﬁc work will never cease. It would be bad if it stopped. For if there were no more problems one would put his hands in his lap and his head to rest, and would not work anymore. And rest is stagnation, and rest is death – in a scientiﬁc sense. The fortune of the investigator is not to have the truth, but to gain the truth. And in this progressive successful search for truth, lies the real satisfaction. Of course, the search for itself is not satisfactory. It must be successful. And this successful research is the source of every effort, and also the source of every spiritual enjoyment. When the source dries up, when the truth is found, then it is over, then one can fall asleep mentally and physically. But that is taken care of, that we don’t experience this, and therein persists our happiness.”2 2German original: “. . . zum Abschluss, zum Endgültigen, werden wir nie kommen. Das wis- senschaftliche Arbeiten wird nie aufhören – es wäre schlimm, wenn es aufhören würde. Denn wenn es keine Probleme mehr gäbe, dann würde man die Hände in den Schoß legen und den Kopf zur Ruhe und würde nicht mehr arbeiten. Und Ruhe ist Stillstand, und Ruhe ist Tod – in wissenschaftlicher Beziehung. Das Glück des Forschers besteht nicht darin, eine Wahrheit zu besitzen, sondern die Wahrheit zu erringen. Und in diesem fortschreitenden erfolgreichen Suchen nach der Wahrheit, da liegt die eigentliche Befriedigung. 1In German [410, p. 26]: “. . . das Kausalgesetz ist weder richtig noch falsch, es ist vielmehr ein heuristisches Prinzip, ein Wegweiser, und zwar nach meiner Meinung der wertvollste Wegweiser, den wir besitzen, um uns in dem bunten Wirrwarr der Ereignisse zurechtzuﬁnden und die Richtung anzuzeigen, in der die wissenschaftliche Forschung vorangehen muss, um zu fruchtbaren Ergebnis- sen zu gelangen. Wie das Kausalgesetz schon die erwachende Seele des Kindes sogleich in Beschlag nimmt und ihm die unermüdliche Frage “warum ?” in den Mund legt, so begleitet es den Forscher durch sein ganzes Leben und stellt ihm unaufhörlich neue Probleme. Denn die Wissenschaft bedeutet nicht beschauliches Ausruhen im Besitz gewonnener sicherer Erkenntnis, sondern sie bedeutet rast- lose Arbeit und stets vorwärtsschreitende Entwicklung, nach einem Ziel, das wir wohl dichterisch zu ahnen, aber niemals verstandesmäßig voll zu erfassen vermögen”. 2German original: “. . . zum Abschluss, zum Endgültigen, werden wir nie kommen. Das wis- senschaftliche Arbeiten wird nie aufhören – es wäre schlimm, wenn es aufhören würde. Denn wenn es keine Probleme mehr gäbe, dann würde man die Hände in den Schoß legen und den Kopf zur Ruhe und würde nicht mehr arbeiten. Und Ruhe ist Stillstand, und Ruhe ist Tod – in wissenschaftlicher Beziehung. 1In German [410, p. 26]: “. . . das Kausalgesetz ist weder richtig noch falsch, es ist vielmehr ein heuristisches Prinzip, ein Wegweiser, und zwar nach meiner Meinung der wertvollste Wegweiser, den wir besitzen, um uns in dem bunten Wirrwarr der Ereignisse zurechtzuﬁnden und die Richtung anzuzeigen, in der die wissenschaftliche Forschung vorangehen muss, um zu fruchtbaren Ergebnis- sen zu gelangen. Wie das Kausalgesetz schon die erwachende Seele des Kindes sogleich in Beschlag nimmt und ihm die unermüdliche Frage “warum ?” in den Mund legt, so begleitet es den Forscher durch sein ganzes Leben und stellt ihm unaufhörlich neue Probleme. Denn die Wissenschaft bedeutet nicht beschauliches Ausruhen im Besitz gewonnener sicherer Erkenntnis, sondern sie bedeutet rast- lose Arbeit und stets vorwärtsschreitende Entwicklung, nach einem Ziel, das wir wohl dichterisch zu ahnen, aber niemals verstandesmäßig voll zu erfassen vermögen”. 2German original: “. . . zum Abschluss, zum Endgültigen, werden wir nie kommen. Das wis- senschaftliche Arbeiten wird nie aufhören – es wäre schlimm, wenn es aufhören würde. Denn wenn es keine Probleme mehr gäbe, dann würde man die Hände in den Schoß legen und den Kopf zur Ruhe und würde nicht mehr arbeiten. Und Ruhe ist Stillstand, und Ruhe ist Tod – in wissenschaftlicher Beziehung. Das Glück des Forschers besteht nicht darin, eine Wahrheit zu besitzen, sondern die Wahrheit zu erringen. Und in diesem fortschreitenden erfolgreichen Suchen nach der Wahrheit, da liegt die eigentliche Befriedigung. Das Suchen an sich befriedigt natürlich noch nicht. Es muss erfolgreich sein. Aber dieses erfolgreiche Arbeiten, das ist dasjenige, was den Quell jeder Anstrengung und auch den Quell eines jeden geistigen Genusses darstellt. Wenn der Quell versiegt, wenn die Wahrheit gefunden ist, dann ist es zu Ende, dann kann man sich geistig und körperlich schlafen legen. Aber dafür ist gesorgt, dass wir das nicht erleben, und darin besteht unser Glück”. nicht beschauliches Ausruhen im Besitz gewonnener sicherer Erkenntnis, sondern sie bedeutet rast- lose Arbeit und stets vorwärtsschreitende Entwicklung, nach einem Ziel, das wir wohl dichterisch zu ahnen, aber niemals verstandesmäßig voll zu erfassen vermögen”. 2German original: “. . . zum Abschluss, zum Endgültigen, werden wir nie kommen. Das wis- senschaftliche Arbeiten wird nie aufhören – es wäre schlimm, wenn es aufhören würde. Denn wenn es keine Probleme mehr gäbe, dann würde man die Hände in den Schoß legen und den Kopf zur Ruhe und würde nicht mehr arbeiten. Und Ruhe ist Stillstand, und Ruhe ist Tod – in wissenschaftlicher Beziehung. Das Glück des Forschers besteht nicht darin, eine Wahrheit zu besitzen, sondern die Wahrheit 22.5 Perception and Forward Tactics Toward Unknowns Das Suchen an sich befriedigt natürlich noch nicht. Es muss erfolgreich sein. Aber dieses erfolgreiche Arbeiten, das ist dasjenige, was den Quell jeder Anstrengung und auch den Quell eines jeden geistigen Genusses darstellt. Wenn der Quell versiegt, wenn die Wahrheit gefunden ist, dann ist es zu Ende, dann kann man sich geistig und körperlich schlafen legen. Aber dafür ist gesorgt, dass wir das nicht erleben, und darin besteht unser Glück”. 22 (De)brieﬁng 168 Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
https://openalex.org/W2622733533	https://www.nature.com/articles/s41598-017-03621-z.pdf	English	null	Clostridium difficile flagella induce a pro-inflammatory response in intestinal epithelium of mice in cooperation with toxins	Scientific reports	2,017	cc-by	9,381	Clostridium difficile flagella induce a pro-inflammatory response in intestinal epithelium of mice in cooperation with toxins Received: 3 January 2017 Accepted: 2 May 2017 Published: xx xx xxxx Jameel Batah1, Hussein Kobeissy1, Phuong Trang Bui Pham1, Cécile Denève-Larrazet1, Sarah Kuehne2, Anne Collignon1, Claire Janoir-Jouveshomme1, Jean-Christophe Marvaud1 & Imad Kansau1 Clostridium difficile is the most important enteropathogen involved in gut nosocomial post-antibiotic infections. The emergence of hypervirulent strains has contributed to increased mortality and morbidity of CDI. The C. difficile toxins contribute directly to CDI-associated lesions of the gut, but other bacterial factors are needed for the bacteria to adhere and colonize the intestinal epithelium. The C. difficile flagella, which confer motility and chemotaxis for successful intestinal colonization, could play an additional role in bacterial pathogenesis by contributing to the inflammatory response of the host and mucosal injury. Indeed, by activating the TLR5, flagella can elicit activation of the MAPK and NF-κB cascades of cell signaling, leading to the secretion of pro-inflammatory cytokines. In the current study, we demonstrate, by using an animal model of CDI, a synergic effect of flagella and toxins in eliciting an inflammatory mucosal response. In this model, the absence of flagella dramatically decreases the degree of mucosal inflammation in mice and the sole presence of toxins without flagella was not enough to elicit epithelial lesions. These results highlight the important role of C. difficile flagella in eliciting mucosal lesions as long as the toxins exert their action on the epithelium. The Gram-positive anaerobic bacterium Clostridium difficile is responsible for intestinal nosocomial post-antibiotic infections in developed countries. The clinical features of C. difficile infection (CDI) include diar- rhea, moderately serious disease, and severe pseudomembranous colitis. The major risk factors associated with CDI are antibiotic exposure, hospitalization, and advanced aged1. A dramatic increase of severe disease and mor- tality of CDI have been observed in North America, Europe and Australia2, 3, mainly resulting from the emer- gence of highly virulent and epidemic C. difficile strains3, 4.hfi g g y ffi The C. difficile toxins TcdA and TcdB are largely involved in lesions of the gut observed during CDI5, 6, but other factors such as adhesins7–10, hydrolytic enzymes11, 12, the S-layer proteins13, and cell wall proteins9, are needed for the bacteria to adhere and colonize the gut. The C. difficile flagella confer motility and chemotaxis for successful intestinal colonization following disruption of the bacterial microbiota14, 15. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Clostridium difficile flagella induce a pro-inflammatory response in intestinal epithelium of mice in cooperation with toxins However, flagella could play an additional role in bacterial pathogenesis by contributing to the inflammatory response of the host and mucosal injury. Indeed, flagellin, the principal component of bacterial flagella, is recognized by the Toll-like receptor 5 (TLR5)16, one of the Pattern Recognition Receptor (PRR) involved in innate immune response, which is mostly localized at the basolateral pole of intestinal cells. The TLR5-flagellin interaction triggers activation of the MAPK and NF-κB cascades of cell signaling, leading to the secretion of pro-inflammatory cytokines17, 18.filfil l To date, few studies have addressed this role for C. difficile flagella. Yoshino et al. showed that C. difficile flagel- lin induces activation of NF-κB and the p38 MAPK, thus promoting the synthesis of IL-8 and CCL20 in intesti- nal epithelial cells19. These authors showed that a pretreatment with toxin TcdB enhances the flagellin-induced cytokine production by cells. Recently, we reported that the interaction of C. difficile flagellin and TLR5 1Faculté de Pharmacie, “Unité Bactéries Pathogènes et Santé” (UBaPS), Université Paris-Sud, Université Paris- Saclay, 92296, Châtenay-Malabry Cedex, France. 2School of Dentistry, College of Medical and Dental Sciences, The University of Birmingham, Birmingham, B5 7EG, UK. Jameel Batah and Hussein Kobeissy contributed equally to this work. Correspondence and requests for materials should be addressed to I.K. (email: imad.kansau@u-psud.fr) Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 1 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 1. C. difficile R20291 flagella are involved in caecal inflammation of mice. C57BL/6 mice (n = 10) were infected or not and caeca were collected at the clinical end point day 2 for histology. (A) Representative histology of caecum of healthy uninfected mice (control) with epithelium integrity; (B) R20291 WT-infected mice with ulcerative colitis, necrosis, desquamation, exudates and necrotic cells in the intestinal lumen, edema, inflammatory submucosal cell infiltration, loss of architecture of epithelium and presence of pseudomembranes in caecum; (C) R20291 fliC and (D) A−B− mutant-infected mice with edema, focal desquamation of necrotic cells in the intestinal lumen and submucosal cell infiltrate; (E) R20291 motB mutant-infected mice with similar degree of mucosal lesion than R20291 WT-infected mice (200X magnification). Bar = 200 µm. (F) Inflammation histological scores for individual mice in each group. The horizontal lines represent the mean scores for each group of animals. (G) Individual evaluation criteria of intestinal inflammation. The bars represent the mean scores for each group of animals and standard deviations. P < 0.01 compared to WT-infected mice. Figure 1. C. Clostridium difficile flagella induce a pro-inflammatory response in intestinal epithelium of mice in cooperation with toxins difficile R20291 flagella are involved in caecal inflammation of mice. C57BL/6 mice (n = 10) were infected or not and caeca were collected at the clinical end point day 2 for histology. (A) Representative histology of caecum of healthy uninfected mice (control) with epithelium integrity; (B) R20291 WT-infected mice with ulcerative colitis, necrosis, desquamation, exudates and necrotic cells in the intestinal lumen, edema, inflammatory submucosal cell infiltration, loss of architecture of epithelium and presence of pseudomembranes in caecum; (C) R20291 fliC and (D) A−B− mutant-infected mice with edema, focal desquamation of necrotic cells in the intestinal lumen and submucosal cell infiltrate; (E) R20291 motB mutant-infected mice with similar degree of mucosal lesion than R20291 WT-infected mice (200X magnification). Bar = 200 µm. (F) Inflammation histological scores for individual mice in each group. The horizontal lines represent the mean scores for each group of animals. (G) Individual evaluation criteria of intestinal inflammation. The bars represent the mean scores for each group of animals and standard deviations. P < 0.01 compared to WT-infected mice. predominantly activates the NF-κB pathway, thus leading to up-regulation of pro-inflammatory gene expression and subsequent synthesis of pro-inflammatory mediators20.hfil l The aim of the current study was to evaluate the role of C. difficile flagella in cooperation with toxins in elicit- ing an inflammatory host response during in vivo infection. By using a conventional mouse model of CDI and dif- ferent C. difficile mutants lacking flagella or toxins, we observed that the absence of flagella dramatically decreases the degree of mucosal inflammation in mice and the sole presence of toxins without flagella was not enough to elicit epithelial lesions as observed in mice infected with wild-type bacteria. These results highlight the important role of C. difficile flagella in eliciting mucosal lesions as long as the toxins exert their action on the epithelium. Results i i In order to test the role of flagellum in the epithelial inflammatory response we also performed in vivo experiments using a motB mutant in R20291 strain which is toxigenic and flagellated but not mobile. The motB gene is involved in synthesis of the flagellar motor. Interestingly, all mice infected with the para- lyzed flagella motB mutant (n = 10) developed CDI with a high (50%) mortality rate, a strong colitis with signifi- cant caecal dilatation, luminal liquid accumulation and wall hyperemia as observed in the WT R20291-infected mice. Indeed, as for WT R20291-infected mice, mice infected with the flagella paralyzed motB mutant showed the highest inflammation score (12 out of 15) in caecal sections compared to uninfected mice (Fig. 1E), with a statistically significant difference (P < 0.01) between the score of the motB mutant-infected mice group (score 12) and those of the other two groups (fliC and A−B− mutant-infected mice, scores 8 and 6 respectively) and control (score 0, Fig. 1F). The individual inflammation criteria were similar to those of the WT-infected mice and a sta- tistically significant difference (P < 0.01) between the motB-infected mice and the other two groups was observed (Fig. 1G).ff To note, in all performed experiments, the differences observed between the different groups of animals were not due to differences in the rate of fecal colonization since the fecal shedding of vegetative forms and spores reached by these strains were similar for the 4 groups of mice at days 1 and 2 post-infection (Supplementary Fig. S1A,B). Moreover, as expected, the feces from WT strain, fliC and motB mutant-infected mice, but not those from A−B− mutant-infected mice, showed a strong and similar level of cytotoxic activity (Supplementary Fig. S2A), thus indicating that in vivo toxin production of the WT R20291 strain and its respective fliC and motB mutants are quite comparable.l Taken together, these observations suggest that, in presence of toxins TcdA and TcdB, the flagellum of the C. difficile R20291 strain plays an important role in amplifying the intestinal inflammatory response during infec- tion, and that these toxins are necessary for the pro-inflammatory activity of flagella in this CDI animal model. Flagella and toxins in non-epidemic C. difficile 630Δerm strain are also necessary to induce a caecal inflammatory response in the CDI mouse model. To evaluate a potential strain-dependent effect for FliC-induced caeca inflammation, we performed the same animal experiments using the non-epidemic C. Results i i difficile 630Δerm strain and its respective fliC and A−B− mutants. As for R20291 derivatives, comparable fecal shedding was observed for all 630Δerm derivative strains (Supplementary Fig. S1C,D) and a cytotoxic activ- ity was detected in feces of the WT 630Δerm- and the fliC mutant-infected mice, while no cytotoxicity was detected in feces of the A−B− mutant-infected animals (Supplementary Fig. S2B). In contrast, in vivo cytotox- icity was dramatically lower in C. difficile WT 630Δerm-infected mice (600 U), compared to that observed in the R20291-infected animals (40 000 U), which is in accordance with the different clinical outcomes observed between these animal sets. Indeed, the flagellated WT 630Δerm induced CDI with soft stools, but a moderated degree of caecal mucosa inflammation compared to uninfected mice (Fig. 2A,B), and the inflammation level was lower (score 8.5, Fig. 2E,F) than that observed in the WT strain R20291-infected mice (score 12). Even if an important submucosa edema was observed, the degree of neutrophil infiltrate, epithelial injury and loss of goblet cells was considerably reduced in caecal sections of 630Δerm WT-infected mice compared to sections of R20291-infected mice. Moreover, as for the fliC mutant in R20291, the 630Δerm fliC and A−B− mutants did not induce disease and elicited only a weak degree of caecal mucosa inflammation (Fig. 2C,D) as measured by histological scores (Fig. 2E,F). These results suggest that flagella of two different epidemic and non-epidemic C. difficile strains contribute with toxins to caecal inflammation during infection in mice, despite major differences in the intensity of the inflammation induced by the two strains. The absence of TLR5 strongly reduces the C. difficile infection-induced caecal inflammation of mice. In order to analyze the role of the mucosal TLR5 in the C. difficile flagella-induced inflammatory response, we tested the CDI model in C57BL/6 tlr5−/− KO mice, which do not express TLR5 in the intestinal mucosa. Mice were infected by oral gavage with the C. difficile R20291 WT strain (n = 10) or its fliC (unflagel- lated but toxigenic) mutant (n = 10), as well as with the 630Δerm WT strain (n = 10) and the clinical outcome and caecal lesions were analyzed. No infected tlr5−/− KO mice developed CDI and all animals survived at day 2 post-infection regardless of the strain. Results i i genic strains, induce a caecal inflammatory response in the CDI mouse model. To study the role of C. difficile flagella in the intestinal inflammatory response in vivo, we used a CDI model in conventional mice21. As expected, all mice infected with the hypervirulent WT R20291 strain (n = 10) developed CDI with diarrhea on the first day post-infection and showed ruffled fur and reduced activity at day 2 after challenge, with a 50% (6/12) mortality rate from day 2 post-challenge. A strong colitis with significant caecal dilatation, luminal liquid accu- mulation and wall hyperemia was observed in all WT-infected mice. To measure the degree of inflammation of the caecal mucosa of mice, a histological score based on 4 criteria (submucosa edema, inflammatory cell infiltrate, epithelial injury and loss of goblet cells) was developed (Supplementary Table S1). Indeed, a high inflammation score of 12 out of 15 was observed in caecal sections of WT R20291-infected mice compared to uninfected mice (Fig. 1A,B).llt In contrast, mice infected with the fliC (unflagellated but toxigenic) mutant (n = 10) presented only soft stools at day 1 or 2 post-infection, and survived the challenge. Similarly, the animals infected with flagellated but non-toxigenic A−B− mutant (n = 10) did not develop CDI and all animals survived at day 2 post-infection. A mild colitis without caecal dilatation or wall hyperemia was observed in the fliC or A−B− mutant-infected mice, compared to the group of non-infected mice (n = 10), suggesting that both TcdA/TcdB toxins and fla- gella are necessary to induce CDI in this mice model. Moreover, a slight inflammation with moderate edema, focal erosion, mild luminal cell desquamation and minimal loss of goblet cells was observed in the fliC or A−B− mutant-infected mice (Fig. 1C,D). A statistically significant difference (P < 0.01) was observed between Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 2 www.nature.com/scientificreports/ the score of the WT-infected mice group (score 12) compared to that of the other two groups (fliC and A−B− mutant-infected mice, scores 8 and 6 respectively) and control (score 0, Fig. 1F). The analysis of each inflamma- tion criteria separately also showed (except for edema) a statistically significant difference (P < 0.01) between the WT-infected mice and the other two groups (Fig. 1G). Toxigenic and flagellum-paralyzed C. difficile R20291 mutant induces a caecal inflammatory response in mice. Results i i Neither colitis was observed in infected mice, compared to the group of non-infected mice (n = 10), and only mucosal edema was detected in histological sections from caecum of ani- mals (Fig. 3B–D), probably induced by C. difficile TcdA/TcdB toxins, produced by all tested strains and detected in feces of all infected animals (Supplementary Fig. S2C). In accordance with the low level of cytotoxic activity detected in WT 630Δerm-infected mice, the lowest inflammation score was observed in animals infected by this non-epidemic strain. Altogether, these results strongly suggest that C. difficile flagella-TLR5 interaction leads to a pro-inflammatory response of the intestinal mucosa. The C. difficile flagella, in presence of toxins, induce the NF-κB activation in the caeca of mice. We previously showed that the NF-κB and MAPKs signaling were activated by C. difficile flagellin via TLR5 in epithelial cells, with a predominant activation of the former20. We thus analyzed the NF-κB activation (IκB-α degradation) in the intestine of mice. Significant IκB-α degradation (2 fold) was observed in the R20291 WT-infected mice compared to the negative control (Fig. 4A). Nevertheless, non-significant IκB-α degradation was observed in the fliC and A−B− mutant-infected animals, compared to R20291 WT-infected mice (Fig. 4A). Interestingly, as for WT-infected mice, a Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 3 www.nature.com/scientificreports/ Figure 2. C. difficile 630Δerm flagella are involved in caecal inflammation of mice. C57BL/6 mice (n = 10) were infected by oral gavage or not and caeca were collected at the clinical end point day 2 for histology. (A) Representative histology of caecum of healthy uninfected mice (control) with epithelium integrity; (B) 630Δerm WT-infected mice with moderate desquamation, exudates and necrotic cells in the intestinal lumen, edema, and inflammatory submucosa cell infiltration; (C) fliC and (D) A−B− mutant-infected mice with normal mucosa (200X magnification). Bar = 200 µm. (E) Inflammation histological scores for individual mice in each group. The horizontal lines represent the mean scores for each group of animals. (F) Individual evaluation criteria of intestinal inflammation. The bars represent the mean scores for each group of animals and standard deviations. P < 0.01 compared to WT-infected mice. ved in caecal inflammation of mice. C57BL/6 mice (n = Figure 2. C. difficile 630Δerm flagella are involved in caecal inflammation of mice. C57BL/6 mice (n = 10) were infected by oral gavage or not and caeca were collected at the clinical end point day 2 for histology. Results i i (A) Representative histology of caecum of healthy uninfected mice (control) with epithelium integrity; (B) 630Δerm WT-infected mice with moderate desquamation, exudates and necrotic cells in the intestinal lumen, edema, and inflammatory submucosa cell infiltration; (C) fliC and (D) A−B− mutant-infected mice with normal mucosa (200X magnification). Bar = 200 µm. (E) Inflammation histological scores for individual mice in each group. The horizontal lines represent the mean scores for each group of animals. (F) Individual evaluation criteria of intestinal inflammation. The bars represent the mean scores for each group of animals and standard deviations. P < 0.01 compared to WT-infected mice. Figure 3. C. difficile flagella and caecal inflammation in C57BL/6 tlr5−/− KO mice. C57BL/6 tlr5−/− KO mice (n = 10) were infected by oral gavage or not and caeca were collected at the clinical end point day 2 for histology. (A) Representative histology of caecum of healthy uninfected mice (control) with epithelium integrity; (B) R20291 WT-infected mice with mild to moderate edema; (C) R20291 fliC mutant and (D) 630Δerm WT- infected mice with mild to moderate edema (200X magnification). Bar = 200 µm. (E) Inflammation histological scores for individual mice in each group. The horizontal lines represent the mean scores for each group of animals. (F) Individual evaluation criteria of intestinal inflammation. The bars represent the mean scores for each group of animals and standard deviations. P < 0.01 compared to R20291 WT-infected mice. Figure 3. C. difficile flagella and caecal inflammation in C57BL/6 tlr5−/− KO mice. C57BL/6 tlr5−/− KO mice (n = 10) were infected by oral gavage or not and caeca were collected at the clinical end point day 2 for histology. (A) Representative histology of caecum of healthy uninfected mice (control) with epithelium integrity; (B) R20291 WT-infected mice with mild to moderate edema; (C) R20291 fliC mutant and (D) 630Δerm WT- infected mice with mild to moderate edema (200X magnification). Bar = 200 µm. (E) Inflammation histological scores for individual mice in each group. The horizontal lines represent the mean scores for each group of animals. (F) Individual evaluation criteria of intestinal inflammation. The bars represent the mean scores for each group of animals and standard deviations. P < 0.01 compared to R20291 WT-infected mice. strong NF-κB activation was observed in caeca of the motB mutant-infected mice compared to the negative control (Fig. 4A). Toxigenic and flagellated C. difficile strains elicit up-regulation of pro-inflammatory cytokine i h f i Toxigenic and flagellated C. difficile strains elicit up-regulation of pro-inflammatory cytokine genes in the caeca of mice. Finally, we investigated the changes in the transcription of genes encoding pro-inflammatory cytokines in the intestinal mucosa of C. difficile-infected mice. We selected a set of genes whose over-expression was observed previously (Table 1)20. As expected, at day 2 post-infection the KC gene encoding keratinocyte derived chemokine (equivalent of the hIL-8 in the mouse) was highly over expressed (114-fold) in the set of R20291 WT-infected mice compared to the control group of non-infected mice (Table 1). A strong up-regulation was also observed for the genes encoding the pro-inflammatory cytokines IL-6 and IL-1β (47- and 49-fold respectively), in accordance to the clinical and histopathological features observed in this group of animals. An up-regulation, but to a lesser extent, was shown for the genes encoding IL-22, TNFα and CXCL- 10 (Table 1). Interestingly, in the fliC mutant-infected mice, expression of KC (hIL-8) was increased by only 17-fold and the increase of other cytokines (IL-6, IL-1β, IL-22, CXCL-10, and TNFα,) was from 2 to 9-fold, compared to the control group, suggesting a role for the flagellum of the strain R20291 in the development of this pro-inflammatory profile in the gut of infected animals. In the non-toxigenic A−B− mutant-infected mice, a lower KC gene up-regulation was also observed (5-fold relative to the control group) as for the other genes, with the exception of the IL -1β encoding gene that showed an over-expression slightly greater than that observed in the fliC mutant-infected mice (Table 1). As expected, and according to clinical and histological observations, a strong up-regulation (except for TNFα gene) was also observed for all these pro-inflammatory cytokines encoding genes in animals infected with the paralyzed flagella motB mutant (Table 1), strongly suggesting that the coexistence of flagella and toxins largely contributes to the mucosal injury.lfi l g g y j y We next tested the role of flagella from the non-epidemic C. difficile 630Δerm strain and its derivative mutants in cytokine gene up-regulation using the CDI mouse model. In accordance with clinical and histopathological observations, the flagellated WT 630Δerm induced a weak up-regulation of KC, IL-6, IL-1β, IL-22, CXCL-10 and TNFα genes compared to the WT R20291 (Table 1). This gene up-regulation was further decreased in the 630Δerm fliC mutant-infected mice and completely null in the 630Δerm A−B− mutant-infected animals, except for TNFα. Results i i We also analyzed the ERK1/2 and JNK MAPKs activation in the intestine of mice. Significant increase of ERK1/2 (1.5 fold) and JNK (2.5 fold) phosphorylation was observed in the R20291 WT- and motB mutant-infected mice compared to the negative control (Supplementary Fig. S3A,B). A similar level of ERK1/2 and JNK activation was also observed in the caecum of the fliC mutant, but not in the non-toxigenic A−B− mutant-infected animals (Supplementary Fig. S3A,B), suggesting a probable action of toxins which are known to also activate MAPKs in intestinal cells22. In accordance to the lesser degree of caecal inflammation observed in the 630Δerm WT-infected mice, a weak but significant activation of NF-κB, ERK1/2 and JNK, in caecal tissue of these animals (Fig. 4B and strong NF-κB activation was observed in caeca of the motB mutant-infected mice compared to the negative control (Fig. 4A). We also analyzed the ERK1/2 and JNK MAPKs activation in the intestine of mice. Significant increase of ERK1/2 (1.5 fold) and JNK (2.5 fold) phosphorylation was observed in the R20291 WT- and motB mutant-infected mice compared to the negative control (Supplementary Fig. S3A,B). A similar level of ERK1/2 and JNK activation was also observed in the caecum of the fliC mutant, but not in the non-toxigenic A−B− mutant-infected animals (Supplementary Fig. S3A,B), suggesting a probable action of toxins which are known to also activate MAPKs in intestinal cells22. In accordance to the lesser degree of caecal inflammation observed in the 630Δerm WT-infected mice, a weak but significant activation of NF-κB, ERK1/2 and JNK, in caecal tissue of these animals (Fig. 4B and Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 4 www.nature.com/scientificreports/ Figure 4. C. difficile flagella-induced degradation of IκB-α. Caecal lysates were prepared as indicated in Material and Methods section and proteins were resolved by SDS-PAGE. (A) R20291 derivative-infected C57BL/6 mice; (B) 630Δerm derivative-infected C57BL/6 mice, and (C) R20291 derivative- or 630Δerm WT-infected C57BL/6 tlr5−/− KO mice. Western blot were then performed using anti-IκB-α, actin antibodies. Western blot cropped pictures (full-length blots are in the Supplementary Information file) show the results of a representative experiment. The density of the bands was measured using Fusion software. Ratios IκB-α/ actin were calculated and for the negative control (C-, uninfected mice), this ratio was normalized to 1. The ratio of the other samples was reported to the negative control. Results i i Results represent the mean (n = 10) ± standard deviations for each group of animals. Statistically significant differences (P < 0.05) compared to the negative control. Figure 4. C. difficile flagella-induced degradation of IκB-α. Caecal lysates were prepared as indicated in M t i l d M th d ti d t i l d b SDS PAGE (A) R20291 d i ti i f t d Figure 4. C. difficile flagella-induced degradation of IκB-α. Caecal lysates were prepared as indicated in Material and Methods section and proteins were resolved by SDS-PAGE. (A) R20291 derivative-infected C57BL/6 mice; (B) 630Δerm derivative-infected C57BL/6 mice, and (C) R20291 derivative- or 630Δerm WT-infected C57BL/6 tlr5−/− KO mice. Western blot were then performed using anti-IκB-α, actin antibodies. Western blot cropped pictures (full-length blots are in the Supplementary Information file) show the results of a representative experiment. The density of the bands was measured using Fusion software. Ratios IκB-α/ actin were calculated and for the negative control (C-, uninfected mice), this ratio was normalized to 1. The ratio of the other samples was reported to the negative control. Results represent the mean (n = 10) ± standard deviations for each group of animals. Statistically significant differences (P < 0.05) compared to the negative control. Supplementary Fig. S3C,D). However, no activation of these effectors was observed in caecal tissue from the fliC and A−B− mutant-infected mice. Moreover, according to the role of TLR5 pro-inflammatory cell-signaling, no NF-κB or MAPK activation was observed in caeca of tlr5−/− KO-infected mice (Fig. 4C and Supplementary Fig. S3E,F). These results are consistent with our previous in vitro experiments and suggest that both TcdA/TcdB and flagella are required for NF-κB activation in caeca of C. difficile infected mice. Supplementary Fig. S3C,D). However, no activation of these effectors was observed in caecal tissue from the fliC and A−B− mutant-infected mice. Moreover, according to the role of TLR5 pro-inflammatory cell-signaling, no NF-κB or MAPK activation was observed in caeca of tlr5−/− KO-infected mice (Fig. 4C and Supplementary Fig. S3E,F). These results are consistent with our previous in vitro experiments and suggest that both TcdA/TcdB and flagella are required for NF-κB activation in caeca of C. difficile infected mice. Discussionl Bacterial flagella can be involved in the mucosal inflammatory injury23. Indeed, although the surface components of bacteria in the intestinal microbiota play a role in modulating the adaptive immune response, some of these components, including flagella, from pathogenic bacteria can reach the mucosa and play a role in amplifying the inflammatory response. Flagella from some pathogens such as Salmonella and Pseudomonas aeruginosa, by specifically recognizing the innate immune pattern-recognition receptor TLR5, which is expressed at the basolat- eral pole of the intestinal epithelium, are responsible for inflammatory lesions in the mucosal epithelium during infection18, 24. The bacterial flagellins-TLR5 interaction elicits the MAPKs and NF-κB TLR5-related signaling pathways, thus inducing a pro-inflammatory response from host16.hfillh l The role of C. difficile flagella in the gut inflammatory response remains to be demonstrated. The predicted TLR5-agonist of C. difficile flagellin, which shares the conserved amino acids recognized by TLR5 with other mucosal pathogens, but not with flagellins known to evade TLR525, 26, was previously confirmed by studies from Yoshino et al. and ourselves19, 20. We observed a predominant role for NF-κB, accounting for the importance of this major transcription factor in the C. difficile flagellin-associated pathogenesis27.fi j p ffil g p g C. difficile toxins have been considered as the only pathogenic factors involved in intestinal lesions observed during CDI, and several studies have shown their role in stimulating pro-inflammatory signaling pathways28. The first evidence of the contribution of C. difficile TcdB to the flagellin-induced inflammatory activity was reported by Yoshino et al.19. They hypothesize that toxin B by affecting tight junctions facilitate the access of flagellin to its receptor in polarized cell monolayers. In a recent report, Kasendra et al.29 suggest that C. difficile toxins facilitate bacterial colonization by disrupting cell polarity allowing pathogen-associated molecular pat- terns (PAMPs) to access the exposed basolateral epithelial surface and trigger the production of inflammatory cytokines. Nevertheless, the synergy between toxins and other bacterial compounds involved in inflammation has never been demonstrated in vivo. For the first time, we show the putative role of flagella in C. difficile pathogenesis in a CDI mouse model21. Toxigenic and flagellated C. difficile strains elicit up-regulation of pro-inflammatory cytokine i h f i These results indicate that despite differences in virulence between R20291 and 630Δerm C. difficile strains, both flagella and toxins contribute to the cytokine gene over-expression from intestinal mucosa in the mouse model, strongly suggesting that toxins are necessary for the contributing pro-inflammatory role of C. difficile flagella. ffil g We also analyzed the pro-inflammatory cytokines gene transcription in the intestinal mucosa of the C. difficile-infected tlr5−/− KO mice. The infection with the R20291 WT strain elicited a 23-fold up-regulation of KC expression compared to non-infected KO mice, which was very similar to the level of up-regulation observed in R20291 fliC mutant-infected conventional mice (Table 1). These results strengthen the role of flagella-TLR5 interaction in the intestinal pro-inflammatory response of the CDI animal model. A lesser up-regulation of this gene was also observed in the R20291 fliC mutant- and 630∆erm WT-infected tlr5−/− KO mice (Table 1), thus suggesting that toxins, and probably other bacterial products, elicit up-regulation of pro-inflammatory cytokines genes, independently of flagella in tlr5−/− KO mouse model. Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 5 www.nature.com/scientificreports/ Set of infected mice Genes* KC IL-6 IL-1β IL-22 CXCL-10 TNFα Conventional mice R20291 WT strain 113, 5 46, 8 49, 3 5, 7 9, 7 8, 2 R20291 fliC− mutant 17, 5 9, 3 7, 9 3, 7 2, 2 2, 9 R20291 A−B− mutant 5, 1 4, 9 10, 0 2, 8 1, 8 1, 6 R20291 motB− mutant 92, 5 39, 0 32, 4 41, 5 18, 0 5, 4 630∆erm WT strain 14, 9 8, 3 12, 9 4, 4 3, 5 4, 6 630∆erm fliC− mutant 8, 6 7, 4 7, 9 1, 7 3, 9 3, 9 630∆erm A−B− mutant 1, 5 1, 7 1, 7 1, 5 1, 7 2, 8 tlr5−/− KO mice R20291 WT strain 23, 0 3, 2 11, 1 15, 4 2, 6 6, 0 R20291 fliC− mutant 14, 9 1, 5 1, 0 6, 8 1, 4 1, 0 630∆erm WT strain 8, 4 2, 1 8, 7 8, 5 2, 2 3, 0 Table 1. Fold expression of pro-inflammatory cytokine encoding genes of C. difficile-infected mice. The mean Ct from each group is reported to those of the negative control and with the expression of the GAPDH gene for each set. Toxigenic and flagellated C. difficile strains elicit up-regulation of pro-inflammatory cytokine i h f i Set of infected mice Genes KC IL-6 IL-1β IL-22 CXCL-10 TNFα Conventional mice R20291 WT strain 113, 5 46, 8 49, 3 5, 7 9, 7 8, 2 R20291 fliC− mutant 17, 5 9, 3 7, 9 3, 7 2, 2 2, 9 R20291 A−B− mutant 5, 1 4, 9 10, 0 2, 8 1, 8 1, 6 R20291 motB− mutant 92, 5 39, 0 32, 4 41, 5 18, 0 5, 4 630∆erm WT strain 14, 9 8, 3 12, 9 4, 4 3, 5 4, 6 630∆erm fliC− mutant 8, 6 7, 4 7, 9 1, 7 3, 9 3, 9 630∆erm A−B− mutant 1, 5 1, 7 1, 7 1, 5 1, 7 2, 8 tlr5−/− KO mice R20291 WT strain 23, 0 3, 2 11, 1 15, 4 2, 6 6, 0 R20291 fliC− mutant 14, 9 1, 5 1, 0 6, 8 1, 4 1, 0 630∆erm WT strain 8, 4 2, 1 8, 7 8, 5 2, 2 3, 0 Table 1. Fold expression of pro-inflammatory cytokine encoding genes of C. difficile-infected mice. The mean Ct from each group is reported to those of the negative control and with the expression of the GAPDH gene for each set. Table 1. Fold expression of pro-inflammatory cytokine encoding genes of C. difficile-infected mice. The mean Ct from each group is reported to those of the negative control and with the expression of the GAPDH gene for each set. Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z Materials and Methods i l i d Bacterial strains and growth conditions R20291 ild i (WT) h fliC d h Bacterial strains and growth conditions. The C. difficile strains used in this study were NAP1/027 R20291 wild-type strain (WT), the fliC and the motB flagellar mutants30, and the tcdA and tcdB double mutant (A−B−)33, and the respective fliC-complemented strain (containing pMTL-SB1 fliC-complementation plasmid) as described previously. These mutants were created from C. difficile R20291 by insertional inactivation using the ClosTron gene knock-out system. For some experiments C. difficile 630Δerm strain and its fliC30 and A−B− 6 isogenic mutants (ClosTron) were also used. C. difficile strains were cultured in Brain Heart Infusion (BHI) agar or broth (Oxoid) in an anaerobic chamber (atmosphere of 90% N2, 5% CO2 and 5% H2) at 37 °C. Strains contain- ing the pMTL plasmid were grown in BHI supplemented with 15 µg/ml thiamphenicol. Escherichia coli Top10 and BL21 were cultured in Luria-Bertani agar or broth (Oxoid) supplemented with 50 µg/ml kanamycin. C. difficile spores for mouse challenge. C. difficile spores were prepared at 37 °C in an anaerobic chamber as previously described by Burns et al.34. Briefly, a preculture was grown overnight in BHI supplemented with yeast extract (5 mg/ml, BD) and L-cysteine (0.1%, Merck) (BHIS) and was used to inoculate a starter preculture in BHIS, which was grown to OD600nm 0.2–0.5. Sporulation was then induced in BHIS by inoculation with the starter culture (1 in 100) and incubation for 7 days. Cultures were then washed in sterile water and heated at 70 °C for 25 min to kill vegetative cells and collect spores. Enumeration of spores was performed on BHI agar supple- mented with the bile salt taurocholate 0.1% (Sigma-Aldrich). Animal model. 6–7 week old female C57BL/6 mice were purchased from Charles River (France). 6–7 week old female C57BL/6 tlr5−/− knock-out (KO) mice were kindly provided by Thierry Pedron (Philippe Sansonetti laboratory, Pasteur Institute, Paris) and were housed and bred at the animal facility of the Faculty of Pharmacy, Paris-Sud University. Mice were housed in groups of 5 in sterile cages containing irradiated food and autoclaved water. To overcome their resistance to CDI, all mice received an antibiotic pretreatment, as previously described21. Mice consumed an antibiotic mixture of kanamycin (0.4 mg/ml), gentamicin (0.035 mg/ml), colistin (850 U/ ml), metronidazole (0.215 mg/ml) and vancomycin (0.045 mg/ml) in the drinking water for 3 days. After this treatment, mice were switched to regular autoclaved water. www.nature.com/scientificreports/ www.nature.com/scientificreports/ toxins exert their action on the epithelium. Interestingly, this is not the case of cholera, which is the archetypal non-inflammatory diarrheal disease. In V. cholerae infection no synergy should occur between flagella and chol- era toxin, which induces non-inflammatory watery diarrhea. In the case of CDI, it seems more evident that tox- ins, by eliciting their actions on the cell cytoskeleton and epithelial tight junctions, could promote the interaction between the FliC monomers and the newly exposed TLR5 and the subsequent pro-inflammatory epithelial host response. According to this role of C. difficile flagella, analysis of the CDI elicited by R20291 and 630Δerm C. difficile strains in the tlr5−/− KO mice, which do not express TLR5 in the intestinal mucosa, strengthens the role of the TLR5-related flagellar signaling in the pathogenesis of C. difficile. Indeed, mice infected by the epidemic R20291 or the non-epidemic 630Δerm strains developed only mucosal edema with a very low inflammation score, without clinical signs of CDI. This mucosal edema was probably induced by TcdA/TcdB toxins, which were produced by all the tested strains.fi p y We previously compared the role of R20291 and 630Δerm C. difficile strains in adhesion and colonization in vitro and in vivo and showed major behavioral differences between these strains in a germ-free (gnotoxenic) mouse model30. In the current study, we show another important role for C. difficile flagella in eliciting an inflam- matory host response shared by both strains, but according to the hypervirulent traits of strain R20291, differ- ences persist in the degree of elicited response which may be linked to the level of in vivo regulation of toxin synthesis. Nevertheless, in the two strains, toxins alone were not enough to induce the clinical manifestations and the mucosal lesions observed in this CDI mouse model. Considering the high homology in TLR5-recognizing motifs in FliC from the two strains, we can suggest a similar role for flagella from the two strains in eliciting a pro-inflammatory immune response during CDI.fi pl y p g Our work contributes to the knowledge about the virulence factors and pathogenesis of C. difficile and high- lights other bacterial and host targets in the control of CDI. C. difficile flagellin might constitute an interesting vaccine candidate leading to a protective immune response against intestinal pathogens. Interestingly, purified Salmonella Typhimurium-derived flagellin protects mice from CDI32. www.nature.com/scientificreports/ In this context, by stimulating TLR5, flagel- lin could contribute to the bacterial clearance. On the other hand, the outcome of the CDI could be modulated by targeting the pro-inflammatory signaling pathways so as to reduce the severity of the lesions induced by a strong inflammatory response. Thereby, by deciphering the regulation mechanisms of the TLR5 signaling cascade, it would be possible to target potential effectors of the pro-inflammatory response in order to reduce the effects of a deleterious response. p Altogether, our results strongly suggest the role of the C. difficile flagella-TLR5 interaction leading to a pro-inflammatory response of the intestinal mucosa in synergy with TcdA and TcdB. Therefore, the C. difficile flagellin would exert its action on the pivotal NF-κB signaling pathway through exposed TLR5, thus resulting in the development of a deleterious innate immune response which contributes to the pathogenesis of this intestinal pathogen. Discussionl In this model, we reproduced pathogenesis observed in human CDI as well as different clinical outcomes depending on two distinct clinical strains and demonstrated that the absence of flagella in toxin producing bacteria dramatically decreases the degree of mucosal inflammation in mice.fl p g y gl We previously reported no difference in the in vitro toxin gene expression between the fliC-R20291 mutant and its respective wild-type strain, but increased toxicity in the fliC-630∆erm flagella mutant compared to its respective WT strain30. In agreement with these results, our previous in vivo transcriptomic analysis performed in the gnotobiotic mouse model, revealed no or weak differences in the toxin gene expression between this fliC-R20291 mutant and its parental wild-type strain31. Therefore, toxin gene regulation in absence of FliC is quite different between strains 630 and R20291 and the high mortality previously observed in the monoxenic mice infected by the fliC-R20291 mutant might be explained not by an increase in toxin gene expression but by the reg- ulation of other genes whose expression is dependent on flagella31. In the present work, we used a different mouse model, the conventional mouse model21, much closer to the human CDI, in which both R20291 and 630∆erm strains and their respective fliC mutants induced similar cytotoxicity titer levels in feces.l p fl y y In this study, the presence of toxins without the flagella was not enough to elicit the high degree of epithelial lesions observed in WT R20291-infected mice. The absence of flagella or motility could prevent the fliC R20291 mutant to reach the mucosa and thus to produce inflammatory lesions induced by toxins alone. However, in our study, we found levels of fecal colonization and cytotoxic activity in feces in the unflagellated (fliC) and non-motile (motB) mutant-infected mice, which were similar to those observed in the WT R20291-infected mice, thus reflecting successful intestinal colonization by these flagellar mutants in this CDI model. Moreover, the flagella-paralyzed motB mutant was not only able to colonize the mucosa, but also to produce toxins and elicit mucosal lesions as the WT strain. Therefore, in the context of the natural CDI, even if other bacterial fac- tors, bacterial fitness or metabolic adaptation capability could play a role in the pathogenesis, flagella highly contribute to the development of a detrimental inflammatory host response and the outcome of CDI as long as Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 6 Materials and Methods i l i d Two days later, all mice received a single dose of clin- damycin (250 µg) by intraperitoneal injection (IP), 24 h prior to challenge. Mice were then infected by oral gavage with 105 spores of C. difficile strains. Mice from the control group were pretreated with antibiotic but were orally gavaged with water. Animals were observed daily for signs of disease (diarrhea, hunched posture and ruffled fur). Two days after challenge, mice were euthanized and caecal tissues were collected, washed with PBS and cut in 3 pieces for further analyses (gene expression, histopathology and immunoblotting). Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 7 www.nature.com/scientificreports/ Animal statements. For the mouse studies, animal care and animal experiments were carried out in strict accordance with the Committee for Research and ethical Issues of the International Association for the study of pain (IASP). The animal experimentation protocol was approved by the Animal Welfare Committee of the Paris-Sud University and animal experiments were performed according to the University Paris-Sud guidelines for the husbandry of laboratory animals. Fecal shedding of C. difficile during infection of mice. Unit forming colonies (UFC)/g feces of vege- tative cells were obtained after serial dilution of feces in PBS and plating in BHI medium supplemented with 3% horse blood and Oxoid C. difficile supplement (250 mg/L D-cycloserine, 8 mg/L Cefoxitin). Plates were incubated in an anaerobic chamber (atmosphere of 90% N2, 5% CO2 and 5% H2) at 37 °C. For spore count, an alcoholic shock was performed and then samples were plated in the same medium supplemented with taurocholate 0.1%. Quantitative histologic assessment of inflammation in the caecum. The caecal segments assigned for histological studies were fixed in 10% neutral formalin for 18 h, transferred into 70% ethanol and paraffin embedded. 3 µm sections were stained with hematoxylin and eosin and analyzed in a blinded manner using a his- topathological scoring scheme. Briefly, caecal inflammation was evaluated with the following criteria measured on 10 different microscopic fields: submucosal edema, degree of lamina propria mononuclear cell infiltration, epithelial damage as previously described35. We added the criteria of loss of goblet cells which was graded as fol- lows: 0: >28; 1: 11–28; 2: 1–10; 3: <1 36 (Supplementary Table S1). SDS-PAGE and Western blot analysis. Materials and Methods i l i d Caecal tissue was disrupted and homogenized in a buffer con- taining 186 mM β-mercaptoethanol, 1% bromophenol blue, 10 mM NaF, 25 mM NaPPi, 1 mM Na3VO4, using a microtube adapted pellet mixer. After lysis, samples were incubated at 100 °C for 10 min and centrifuged to remove insoluble materials. Proteins were resolved by SDS–PAGE, and gels were transferred to polyvinylidene difluoride (PVDF) membrane (GE Healthcare). For immunoblotting, membranes were washed with TBS 0.1% Tween 20, blocked in TBS (0.1% Tween 20, 5% milk) and probed overnight with the following specific anti- bodies: anti-ERK1/2, anti-JNK, anti-IκBα, anti-phosphorylated (anti-p)ERK1/2 and anti-pJNK (Cell Signaling Technology), or anti-actin (Millipore). Blots were then incubated with HRP-linked secondary antibodies (Cell Signaling Technology), followed by chemiluminescence detection with the ECL Plus kit (Millipore) according to the manufacturer’s instructions. Chemiluminescence signals were detected with a Fusion FX (Vilber Lourmat) and analyzed densitometrically with Fusion-CAPT software (Vilber Lourmat). Quantitative real-time reverse transcription PCR (qRT-PCR). Tissue from the caecum was sub- merged in RNAlater RNA Stabilization Reagent (Qiagen) to protect the RNA. After stabilization for 15 min at room temperature, tissue was lysed using Lysing Matrix D and a FastPrep apparatus (MP Biomedicals). Total RNA was then isolated using the RNeasy Mini Kit (Qiagen). RNA quantification and quality were assessed by a 2100 Bioanalyzer Agilent. RNA was reverse-transcribed to first strand cDNA using the RT² First Strand Kit (Qiagen). cDNA was prepared from 1 µg RNA using SuperScript™ III Reverse Transcriptase (Invitrogen) with random primers as described by the manufacturer. qPCR was performed in a 10 µl reaction volume containing 4 ng of cDNA, 5 µl of SSo Advanced™ SYBR Green Supermix (Bio-Rad) and 500 nM gene-specific primers. The primers designed with Primer3 software, not to amplify genomic DNA, are listed in Supplementary Table S2. Reactions were run on a CFX96 Real-time system (Bio-Rad) with the following cycling conditions: 30 s polymer- ase activation at 95 °C and 40 cycles at 95 °C for 5 s and 60 °C for 10 s. An additional step from a start at 65 °C to 95 °C (0.5 °C/0.5 s) was performed to establish a melting curve in order to verify the specificity of the real-time PCR reaction for each primer pair. The results were normalized using the geometric averaging of the GAPDH as reference gene. Normalized relative quantities were calculated using the ΔΔCT method37, 38. Materials and Methods i l i d g q g Data were collected from three independent experiments and expressed as down- or up-regulation fold. The statistical differences were determined by Student’s t-test. P values < 0.05 were considered statistically significant. Toxicity assay on Vero cell line. Vero kidney epithelial cell line (kindly donated by I. Beau, INSERM UMR-S 984, France) were grown in DMEM with L-glutamine (Invitrogen) supplemented with 10% heat-inactivated fetal calf serum at 37 °C in an atmosphere containing 5% CO2. Cell lines were harvested from the flask with trypsin (0.5 mg/ml)/EDTA (0.2 mg/ml), washed once with medium and seeded into culture plates (TPP, ATGC Biotechnologies) at the desired cell densities and incubated at 37 °C in an atmosphere containing 5% CO2 before experiments. Quantitative detection of toxin in fecal samples was performed by cytotoxicity assay. Briefly, feces from each mouse were collected and diluted in PBS at 1 mg/ml. After thorough mixing, the samples were centrifuged at 12000 g for 5 min, and the supernatants were collected and then sterilized through a 0.22 µm filter. Confluent Vero cells in 96-well plates were exposed to 1:2 to 1:262144 (vol/vol) dilutions of the supernatants in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco) containing 10% heat-inactivated Fetal Bovine Serum (FBS) (Gibco). The plates were incubated at 37 °C in 5% CO2 and morphological changes were observed by microscopy after 24 h. The endpoint titers were expressed as the reciprocal of the highest dilution giving a 50% cytopathic effect (CPE). The assays were performed in triplicate on independent culture supernatants. Statistical analysis. Results from histologic assessment, fecal shedding, Western blot analysis, and toxicity assay are expressed as means ± sem. Different groups were compared by using Mann and Whitney test performed using StatEL software. P values < 0.05 were considered statistically significant. 8 Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z www.nature.com/scientificreports/ References Importance of toxin A, toxin B, and CDT in virulence of an epidemic Clostridium difficile strain. J Infect Dis 83–86, doi:10.1093/infdis/jit426 (2014). j 34. Burns, D. A., Heeg, D., Cartman, S. T. & Minton, N. P. Reconsidering the sporulation characteristics of hypervirulent Clostridium difficile BI/NAP1/027. PLoS ONE 6, e24894, doi:10.1371/journal.pone.0024894 (2011).hffi j 34. Burns, D. A., Heeg, D., Cartman, S. T. & Minton, N. P. Reconsidering the sporulation characteristics of hypervirulent Clostridium difficile BI/NAP1/027. PLoS ONE 6, e24894, doi:10.1371/journal.pone.0024894 (2011).hffi fi 35. Theriot, C. M. et al. Cefoperazone-treated mice as an experimental platform to assess differential virulence of Clostridium dif strains. Gut Microbes 2, 326–334, doi:10.4161/gmic.19142 (2011).l g 6. Lai, M. A. et al. Innate immune detection of flagellin positively and negatively regulates salmonella infection. PLoS ONE 8, e72047 doi:10.1371/journal.pone.0072047 (2013).it 37. Hellemans, J., Mortier, G., De Paepe, A., Speleman, F. & Vandesompele, J. qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol 8, R19 (2007). g y q 38. Livak, K. J. & Schmittgen, T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta D C(T)) Method. Methods 25, 402–408 (2001). References Cwp84, a surface-associated protein of Clostridium dif protease with degrading activity on extracellular matrix proteins J Bacteriol 189 7174–7180 (2007) fi 12. Janoir, C., Pechine, S., Grosdidier, C. & Collignon, A. Cwp84, a surface-associated protein of Clostridium difficile, is a cysteine protease with degrading activity on extracellular matrix proteins. J Bacteriol 189, 7174–7180 (2007).fi 12. Janoir, C., Pechine, S., Grosdidier, C. & Collignon, A. Cwp84, a surface-associated protein of Clostridium difficile, is a cysteine protease with degrading activity on extracellular matrix proteins. J Bacteriol 189, 7174–7180 (2007).fi 3. Calabi, E., Calabi, F., Phillips, A. D. & Fairweather, N. F. Binding of Clostridium difficile surface layer proteins to gastrointestina tissues. Infect Immun 70, 5770–5778 (2002).i 14. Eaton, K. A., Suerbaum, S., Josenhans, C. & Krakowka, S. Colonization of gnotobiotic piglets by Helicobacter pylori deficient in two flagellin genes. Infect Immun 64, 2445–2448 (1996).lfi l g g f 15. Tasteyre, A., Barc, M. C., Collignon, A., Boureau, H. & Karjalainen, T. Role of FliC and FliD flagellar proteins of Clostridium difficil in adherence and gut colonization. Infect Immun 69, 7937–7940 (2001). l g g f 15. Tasteyre, A., Barc, M. C., Collignon, A., Boureau, H. & Karjalainen, T. Role of FliC and FliD flagellar in adherence and gut colonization. Infect Immun 69, 7937–7940 (2001). g f 16. Prince, A. Flagellar activation of epithelial signaling. Am J Respir Cell Mol Biol 34, 548–551 (2006). g g g 17. Abreu, M. T., Fukata, M. & Arditi, M. TLR signaling in the gut in health and disease. J Immunol 174, 4453–4460 (2005).l 17. Abreu, M. T., Fukata, M. & Arditi, M. TLR signaling in the gut in health and disease. J Immunol 18. Gewirtz, A. T., Navas, T. A., Lyons, S., Godowski, P. J. & Madara, J. L. Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression. J Immunol 167, 1882–1885 (2001).fill p pl y g p 19. Yoshino, Y. et al. Clostridium difficile flagellin stimulates toll-like receptor 5, and toxin B promot production via TLR5. Life Sci 92, 211–217, doi:10.1016/j.lfs.2012.11.017 (2013).fil l 19. Yoshino, Y. et al. Clostridium difficile flagellin stimulates toll-like receptor 5, and toxin B promotes flagellin-induced chemokine production via TLR5. Life Sci 92, 211–217, doi:10.1016/j.lfs.2012.11.017 (2013).fil fil production via TLR5. Life Sci 92, 211–217, doi:10.1016/j.lfs.201 p f j 0. Batah, J. et al. Clostridium difficile flagella predominantly activate TLR5-linked NF-kappaB pathway in epithelial cells. References 1. Cloud, J. & Kelly, C. P. Update on Clostridium difficile associated disease. Curr Opin Gastroenterol 23, 4–9 (2007).fi 1. Cloud, J. & Kelly, C. P. Update on Clostridium difficile associated disease. Curr Opin Gastroenterol 23, 4–9 (2007).fi 2. Cohen, S. H. et al. Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the society for healthcare epidemiology of America (SHEA) and the infectious diseases society of America (IDSA). Infect Control Hosp Epidemiol 31, 431–455 (2010). 3. Barbut, F., Jones, G. & Eckert, C. Epidemiology and control of Clostridium difficile infections in healthcare settings: an update. Cur Opin Infect Dis 24, 370–376 (2011).fi p f 4. Lo Vecchio, A. & Zacur, G. M. Clostridium difficile infection: an update on epidemiology, risk factors, and therapeutic options. Cur Opin Gastroenterol 28, 1–9 (2012).fi ial for virulence of Clostridium difficile. Nature 458, 1176–1179 (20 yfi 6. Kuehne, S. A. et al. The role of toxin A and toxin B in Clostridium difficile infection. Nature 467, 711–713 (2010).ifi hfi 7. Eveillard, M. et al. Identification and characterization of adhesive factors of Clostridium difficile involved in adhesion to h colonic enterocyte-like Caco-2 and mucus-secreting HT29 cells in culture. Mol Microbiol 7, 371–381 (1993).fi hfi veillard, M. et al. Identification and characterization of adhesive factors of Clostridium difficile involved in adhesion to huma y g 8. Hennequin, C. et al. GroEL (Hsp60) of Clostridium difficile is involved in cell adherence. Microbiology 147, 87–96 (2001).fi q , ( p )fi gy , ( ) 9. Waligora, A. J. et al. Characterization of a cell surface protein of Clostridium difficile with adhesive properties. Infect Immun 69 2144–2153 (2001).i 0. Barketi-Klai, A., Hoys, S., Lambert-Bordes, S., Collignon, A. & Kansau, I. Role of fibronectin-binding protein A in Clostridium difficile intestinal colonization. J Med Microbiol 60, 1155–1161, doi:10.1099/jmm.0.029553-0 (2011).fi & Borriello, S. P. Proteolytic activity of Clostridium difficile. J Med M fi 11. Seddon, S. V. & Borriello, S. P. Proteolytic activity of Clostridiu fi noir, C., Pechine, S., Grosdidier, C. & Collignon, A. Cwp84, a surface-associated protein of Clostridium difficile, is a cysteine t ith d di ti it t ll l t i t i J B t i l 189 7174 7180 (2007) fi 12. Janoir, C., Pechine, S., Grosdidier, C. & Collignon, A. References Anaerobe 38 116–124, doi:10.1016/j.anaerobe.2016.01.002 (2016). 20. Batah, J. et al. Clostridium difficile flagella predominan fil 116–124, doi:10.1016/j.anaerobe.2016.01.002 (2016) 21. Chen, X. et al. A mouse model of Clostridium difficile-associated disease. Gastroenterology 135, 1984–1992, doi:10.1053/j. gastro.2008.09.002 (2008).f 22. Lee, J. Y. et al. Effects of transcription factor activator protein-1 on interleukin-8 expression and enteritis in response to Clostridium difficile toxin A. Journal of molecular medicine 85, 1393–1404, doi:10.1007/s00109-007-0237-7 (2007).l fi 23. Ramos, H. C., Rumbo, M. & Sirard, J. C. Bacterial flagellins: mediators of pathogenicity and host immune responses in mu Trends Microbiol 12, 509–517, doi:10.1016/j.tim.2004.09.002 (2004). j 24. Sierro, F. et al. Flagellin stimulation of intestinal epithelial cells triggers CCL20-mediated migration of dendritic cells. Proc Natl Sci USA 98, 13722–13727 (2001). 5. Andersen-Nissen, E. et al. Evasion of Toll-like receptor 5 by flagellated bacteria. Proc Natl Acad Sci USA 102, 9247–9252 doi:10.1073/pnas.0502040102 (2005).l 26. Lee, S. K. et al. Helicobacter pylori flagellins have very low intrinsic activity to stimulate human gastric epithelial cells via T Microbes and infection/Institut Pasteur 5, 1345–1356 (2003). f 27. Hayden, M. S. & Ghosh, S. Shared principles in NF-kappaB signaling. Cell 132, 344–362 (2008).fil 27. Hayden, M. S. & Ghosh, S. Shared principles in NF-kappaB signaling. Cell 132, 344–362 (2008).fil 28. Shen, A. Clostridium difficile toxins: mediators of inflammation. Journal of innate immunity 4, 149–158, doi:10.1159/00033 (2012).fi 29. Kasendra, M., Barrile, R., Leuzzi, R. & Soriani, M. Clostridium difficile toxins facilitate bacterial colonization by modulatin fence and gate function of colonic epithelium. J Infect Dis 209, 1095–1104, doi:10.1093/infdis/jit617 (2014).hlfi 30. Baban, S. T. et al. The role of flagella in Clostridium difficile pathogenesis: comparison between a non-epidemic and an epid strain. PLoS ONE 8, e73026, doi:10.1371/journal.pone.0073026 (2013).hlfi j p 31. Barketi-Klai, A. et al. The flagellin FliC of Clostridium difficile is responsible for pleiotropic gene regulation during in vivo infec PLoS ONE 9, e96876, doi:10.1371/journal.pone.0096876 (2014).fi j p ( ) 32. Jarchum, I., Liu, M., Lipuma, L. & Pamer, E. G. Toll-like receptor 5 stimulation protects mice from acute Clostridium difficile colitis. Infect Immun 79, 1498–1503 (2011). h l f d l f d l d d ffi l f j p ( ) 32. Jarchum, I., Liu, M., Lipuma, L. & Pamer, E. G. Toll-like receptor 5 stimulation protects mice from acute Clostridium difficile colitis. Infect Immun 79, 1498–1503 (2011).fi f 33. Kuehne, S. A. et al. Author Contributions J.B. and H.K. contributed equally to this work. J.B., H.K., C.D.L., P.T.B.P., S.K., J.C.M. and I.K. conducted experiments and analyzed data; A.C., and C.J.J. advised on experimental designs; C.D.L., J.C.M. and I.K. designed experiments, I.K. wrote the manuscript with contributions from all other authors. Acknowledgementsh This study was in part supported by the European Union (HEALTH-F3-2008-223585) and by the University Paris-Sud “Attractivité-2012 and 2013” grants. We sincerely thank the technical assistance of Claudine Delomenie (Region Ile de France, IFR141 IPSIT) for qRT-PCR analyses, as well as Valérie Domergue (Region Ile de France, IFR141 IPSIT) for animal facilities. We also thank Sylvie Lambert-Bordes and Sandra Hoys for technical support. 9 9 Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z www.nature.com/scientificreports/ Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z Additional Information Supplementary information accompanies this paper at doi:10.1038/s41598-017-03621-z Competing Interests: The authors declare that they have no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2017 Scientific Reports \| 7: 3256 \| DOI:10.1038/s41598-017-03621-z 10
https://openalex.org/W4392905794	https://cardiab.biomedcentral.com/counter/pdf/10.1186/s12933-024-02185-3	English	null	Prevalence and prognostic significance of cardiac autonomic neuropathy in community-based people with type 2 diabetes: the Fremantle Diabetes Study Phase II	Cardiovascular diabetology	2,024	cc-by	12,289	© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Cardiovascular Diabetology Cardiovascular Diabetology Davis et al. Cardiovascular Diabetology (2024) 23:102 https://doi.org/10.1186/s12933-024-02185-3 Open Access Prevalence and prognostic significance of cardiac autonomic neuropathy in community‑based people with type 2 diabetes: the Fremantle Diabetes Study Phase II Timothy M. E. Davis1,2,3, Eva Tan1 and Wendy A. Davis1,3 Timothy M. E. Davis1,2,3, Eva Tan1 and Wendy A. Davis1,3 Abstract Background There is a paucity of contemporary data on the prevalence and prognostic significance of cardiac auto- nomic neuropathy (CAN) from community-based cohorts with type 2 diabetes assessed using gold standard meth- ods. The aim of this study was to assess these aspects of CAN in the longitudinal observational Fremantle Diabetes Study Phase II (FDS2). Methods FDS2 participants were screened at baseline using standardised cardiovascular reflex tests (CARTs) of heart rate variation during deep breathing, Valsalva manoeuvre and standing. CAN (no/possible/definite) was assessed from the number of abnormal CARTs. Multinomial regression identified independent associates of CAN status. Cox proportional hazards modelling determined independent baseline predictors of incident heart failure (HF) and ischae- mic heart disease (IHD), and all-cause mortality. Results Of 1254 participants assessed for CAN, 86 (6.9%) were outside CART age reference ranges and valid CART data were unavailable for 338 (27.0%). Of the remaining 830 (mean age 62.3 years, 55.3% males, median diabetes duration 7.3 years), 51.0%, 33.7% and 15.3% had no, possible or definite CAN, respectively. Independent associates of definite CAN (longer diabetes duration, higher body mass index and resting pulse rate, antidepressant and anti- hypertensive therapies, albuminuria, distal sensory polyneuropathy, prior HF) were consistent with those reported previously. In Kaplan–Meier analysis, definite CAN was associated with a lower likelihood of incident IHD and HF versus no/possible CAN (P < 0.001) and there was a graded increase in all-cause mortality risk from no CAN to possible and definite CAN (P < 0.001). When CAN category was added to the most parsimonious models, it was not a signifi- cant independent predictor of IHD (P ≥ 0.851) or HF (P ≥ 0.342). Possible CAN (hazard ratio (95% CI) 1.47 (1.01, 2.14), P = 0.046) and definite CAN (2.42 (1.60, 3.67), P < 0.001) increased the risk of all-cause mortality versus no CAN. Conclusions Routine screening for CAN in type 2 diabetes has limited clinical but some prognostic value. Keywords Type 2 diabetes, Cardiac autonomic neuropathy, Ischemic heart disease, Heart failure, Mortality Correspondence: Timothy M. E. Davis tim.davis@uwa.edu.au Full list of author information is available at the end of the article Background data relating to income, employment, housing, trans- portation and other variables in the study area show an average Index of Relative Socio-economic Advantage and Disadvantage of 1033 with a range by postcode of 977 to 1113, figures similar to the Australian national mean ± SD of 1000 ± 100 [16]. Descriptions of FDS2 recruitment, sample characteristics and details of non- recruited people with diabetes have been published [15]. Individuals resident in the catchment area with a clini- cian-verified diagnosis of diabetes (excluding gestational diabetes) were identified through available hospital and community sources. Of 4639 with known diabetes found between 2008 and 2011, 1668 (36.0%) were recruited. Sixty-four FDS Phase I participants recruited between 1993 and 1996 who had moved out of the catchment area were also enrolled (total cohort 1732, of whom 1551 had clinically diagnosed type 2 diabetes). For the purposes of the present study, there were 1254 participants (89.9% of the FDS2 type 2 diabetes cohort) who were eligible for CAN testing after it first became available in May 2009 as part of baseline assessment. g Cardiovascular autonomic neuropathy (CAN) is charac- terised by orthostatic hypotension, resting tachycardia, impaired exercise tolerance and abnormal blood pressure regulation [1], but it may also remain asymptomatic and thus elude timely diagnosis [2]. It is a common chronic complication of type 2 diabetes with a prevalence esti- mated at between 9 and 78% from studies conducted in primary but mainly secondary care [3, 4]. Among a num- ber of available CAN diagnostic tests, the gold standard comprises several standardised cardiovascular reflex tests (CARTs) including the electrocardiographic R-R inter- val response to deep breathing, the Valsalva manoeuvre and postural changes in blood pressure [5]. CARTs are, however, not widely available, time-consuming, and dif- ficult to perform in people with mobility challenges and in whom forceful breathing is difficult or even contra- indicated [6]. These considerations may underlie the wide range of CAN prevalence estimates in type 2 diabetes, but may also have implications for assessment of the relationship between CAN and both cardiovascular disease (CVD) and death. A recent meta-analysis of unadjusted data suggested that CAN increases the risk of CVD events and all-cause mortality in type 2 diabetes more than three- fold [7], but there was substantial heterogeneity between studies. © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Davis et al. Cardiovascular Diabetology (2024) 23:102 Davis et al. Cardiovascular Diabetology (2024) 23:102 Page 2 of 16 Page 2 of 16 Clinical and laboratory assessments All FDS2 participants were invited to face-to-face assess- ments at entry and then biennially [15]. Each assessment included a standardised comprehensive questionnaire and physical examination, and fasting biochemical tests performed in a single nationally accredited laboratory. Participants were requested to bring all medications and/or prescriptions to each visit. Racial/ethnic back- ground was categorised based on self-selection, country/ countries of birth and parents’/grandparents’ birth, and language(s) spoken at home as Anglo-Celt, Southern European, Other European, Asian, Aboriginal or mixed/ other. Body mass index (BMI) was determined together with a body shape index (ABSI) which represents a more reliable estimate of visceral adiposity [17]. Orthostatic hypotension was defined as a fall in systolic blood pres- sure of ≥ 20 mmHg or in diastolic blood pressure of ≥ 10 mmHg within three minutes of standing [18].h Although testing for autonomic dysfunction has been recommended as part of routine early screening in dia- betes [2, 7], its role as an independent predictor of CVD events and mortality needs to be established in contem- porary cohorts of people with type 2 diabetes. The aim of this study was, therefore, to assess the prevalence and prognostic significance of CAN in well characterised, representative, community-based participants from the Fremantle Diabetes Study Phase II (FDS2). The CARTs were performed on each eligible partici- pant in the morning after an overnight fast, and com- prised measurement and analysis of heart rate variation during deep breathing, the Valsalva manoeuvre, and on standing by electrocardiography using the ANS 2000 system (Hokanson Inc, Bellevue, Washington, US) [19– 21]. The deep breathing test was performed with the participant supine and breathing at a paced rate of five breaths/minute for six minutes, as recommended by the manufacturer. The ratio of the mean of the shortest R-R interval during inspiration to the mean of the long- est R-R wave during expiration (E:I ratio) was calculated, and the MCR determined by vector analysis of the R–R Background Indeed, the risk of all-cause death in the high CVD risk Action to Control Cardiovascular Risk in Dia- betes (ACCORD) trial sample was lower, if still signifi- cantly increased, in fully adjusted statistical models [8]. Furthermore, CVD risk factors are closely associated with the development of CAN [3]. A number of stud- ies included in the meta-analysis were conducted before the publication of the results of intervention trials sup- porting more intensive management of hypertension and dyslipidaemia in type 2 diabetes which have resulted in improved CVD risk factor management [9, 10] and a reduction in CVD events [11–14] over recent decades. Study site, participants and approvalsh The FDS2 is an observational study conducted in a post- code-defined urban community of 157,000 people in the state of Western Australia (WA) [15]. Socio-economic Davis et al. Cardiovascular Diabetology (2024) 23:102 Page 3 of 16 (HF) were defined as hospitalisations or death with/for/of IHD, cerebrovascular disease or HF, respectively, before the first CAN assessment. Incident IHD was defined as hospitalisations or death with/for/of IHD or cardiac/sud- den death, and incident HF as hospitalisations or death with/for/of HF, both endpoints being ascertained from the first CAN assessment to end-December 2021. Causes of death on the death certificate or coroner’s report were reviewed independently by two study physicians and classified under the system used in the UK Prospec- tive Diabetes Study [11]. In the case of discrepant cod- ing, case notes were consulted and a consensus obtained. Death from IHD was defined as death from non-HF car- diac or sudden death, and death from HF was defined as cardiac death in which HF dominated. All endpoints were ascertained from the first CAN assessment to end- December 2021. (HF) were defined as hospitalisations or death with/for/of IHD, cerebrovascular disease or HF, respectively, before the first CAN assessment. Incident IHD was defined as hospitalisations or death with/for/of IHD or cardiac/sud- den death, and incident HF as hospitalisations or death with/for/of HF, both endpoints being ascertained from the first CAN assessment to end-December 2021. Causes of death on the death certificate or coroner’s report were reviewed independently by two study physicians and classified under the system used in the UK Prospec- tive Diabetes Study [11]. In the case of discrepant cod- ing, case notes were consulted and a consensus obtained. Death from IHD was defined as death from non-HF car- diac or sudden death, and death from HF was defined as cardiac death in which HF dominated. All endpoints were ascertained from the first CAN assessment to end- December 2021. intervals [22]. During the Valsalva manoeuvre, partici- pants performed continuous forced exhalation to a pres- sure of 40 mmHg for 15 s. The ratio of the longest R–R interval after the manoeuvre to the shortest R–R interval during the manoeuvre was calculated (the Valsalva ratio) [21]. Study site, participants and approvalsh Evaluation of changes in heart rate was performed during the initial phase of adaptation to orthostasis (first 45 s), and the 30:15 stand ratio calculated from the maxi- mal (around 30th heart beat) to minimal (near 15th heart beat) R-R interval [22]. Abnormalities in the three CART components (one or other of E:I ratio and MCR in the case of the deep breath- ing CART) were identified from age-corrected normal ranges [22] and given a score of 1. Since the age of the healthy individuals used to derive the normal ranges spanned 15–67 years [22], linear (for E:I ratio, MCR and Valsalva ratio; r2 ≥ 0.983) and quadratic (for 30:15 stand ratio; r2 = 0.996) equations were derived from the table of age versus the 2.3 centiles of each CART [22] and extrap- olated to age 80 years to better capture reflect the age range of the FDS2 type 2 diabetes cohort and to paral- lel other studies with age-specific reference ranges which did not include very elderly participants [23, 24]. Diagno- sis of CAN and its stage was determined from modified Toronto Consensus Panel criteria [25] as no CAN (total score = 0), possible CAN (total score = 1) or definite CAN (total score ≥ 2).i Statistical analysis Th y The computer packages IBM SPSS Statistics 29 (IBM Corporation, Armonk, NY, USA) and StataSE 15 (Col- lege Station, TX: StataCorp LP) were used for statistical analysis. Data are reported as percentage, mean ± SD, geometric mean (SD range), or, when variables are not normally distributed, median [interquartile range]. Two- way comparisons were performed using Fisher’s exact test for independent samples, the normally distributed variables compared using Student’s t-test, and the non- normally distributed variables using Mann–Whitney U-test. Comparisons between multiple groups for cat- egoric variables were by Fisher-Freeman-Halton exact or Chi-squared tests, for normally/log-normally distributed continuous variables by one-way ANOVA, and for vari- ables not conforming to normal/log-normal distribution by Kruskal–Wallis test. Where the overall trend for these multiple comparisons was statistically significant, post- hoc Bonferroni-corrected pairwise comparisons were performed. Chronic complications of diabetes were identified using standard definitions [15]. Albuminuria was assessed by early morning spot urinary albumin:creatinine ratio (uACR) measurement and renal impairment from the estimated glomerular filtration rate (eGFR) [26]. Distal symmetrical polyneuropathy (DSPN) was defined using the vibration perception threshold [27]. Retinopathy was defined as one microaneurysm in either eye or worse and/ or previous laser treatment on fundus photography and/ or ophthalmologist assessment. Peripheral arterial dis- ease (PAD) was defined as an ankle brachial index ≤ 0.90 or a diabetes-related lower extremity amputation. Multinomial regression was used to identify independ- ent associates of CAN status with the no CAN group as reference. Clinically relevant and biologically plau- sible variables with bivariable P < 0.20 were considered for model entry. Variables were removed sequentially if P ≥ 0.050 for every CAN group (relative to the reference category), the least significant being removed first, until all variables in the model were significant in at least one group. Baseline participant characteristics Of the 1254 FDS2 participants who underwent CAN testing at baseline, 86 (6.9%) were excluded because they were aged < 20 years or > 80 years and so their CART data could not be assessed against extrapolated age-spe- cific normal ranges [22]. Of the remaining 1168 partici- pants, a further 338 (27.0%) were excluded because they could not perform all the tests according to protocol, they had poor quality electrocardiographic recordings that were unsuitable for analysis, or they had a signifi- cant cardiac arrhythmia that confounded interpretation of the results. Compared to the 830 with complete CART data required for CAN categorisation, the 424 who were excluded were significantly older (age 62.3 ± 10.5 versus 71.0 ± 10.6 years, P < 0.001), less likely to be males (56.0% versus 45.8%, P < 0.001), had longer diabetes duration (7.1 versus 11.2 years, P < 0.001), and were significantly more likely to have chronic complications (see Additional file 1: Table S1).h Incident heart failureh The characteristics of the 119 FDS2 participants (15.0%) who had a HF event during follow-up and those who did not are summarised in Table 5. The Kaplan–Meier curves of HF events during a mean ± SD follow-up period of 10.1 ± 2.8 (range 0.0–12.6) years (equivalent to 7992 person-years) are shown in Fig. 1 (middle panel). There was a significant difference between the three groups (log rank test P < 0.001), with definite CAN different from both the no CAN group (P < 0.001) and possible CAN group (P = 0.002) in unadjusted pairwise comparisons. In Cox proportional hazards modelling, increasing age, Aboriginal ethnic background, longer diabetes duration, a higher uACR, DSPN, PAD and a prior history of IHD were independent predictors of incident HF events but CAN category did not add significantly to the model (see Table 4). The baseline characteristics of included participants are summarised by CAN status in Table 1. Approximately 15% had definite CAN, one third had possible CAN and around one half had no CAN. Compared with the other two groups, those with definite CAN were more likely to have an Aboriginal background, to be diagnosed with diabetes at a younger age and to have longer diabetes duration, to be obese, and to have a higher HbA1c despite a greater likelihood of insulin therapy. They were also more likely to be treated with antidepressants, to have a higher resting pulse rate in the presence of greater beta blocker and calcium channel blocker use, to have hyper- triglyceridemia and microalbuminuria, to have greater degrees of renal impairment, to have DSPN, and to have a prior history of IHD and HF. Those with possible CAN had diabetes duration, and prevalences of IHD and HF, that were intermediate between those in the no CAN and definite CAN groups. Ascertainment of cardiovascular outcomes and deathsh Ascertainment of cardiovascular outcomes and deaths The Hospital Morbidity Data Collection (HMDC) con- tains validated information regarding all public/private hospitalisations in WA since 1970 and the Death Register contains information on all deaths in WA [28]. The FDS2 database has been linked to these databases through the WA Data Linkage System (WADLS), as approved by the WA Department of Health Human Research Ethics Committee. The HMDC was used to supplement data obtained through FDS assessments relating to preva- lent/prior complications/conditions during the five years prior to study entry. A prior history of ischaemic heart disease (IHD), cerebrovascular disease or heart failure Cox proportional hazards modelling (backward con- ditional variable selection with P < 0.050 for entry and ≥ 0.050 for removal) was used to determine independent baseline predictors of incident HF and IHD, and all-cause mortality. All clinically plausible variables with bivariable P ≤ 0.20 were considered for entry into these models in Davis et al. Cardiovascular Diabetology (2024) 23:102 Davis et al. Cardiovascular Diabetology (2024) 23:102 Page 4 of 16 definite CAN were more likely to be treated with beta blockers, calcium channel blockers and antidepressants, and to have a history of HF. In addition, those with pos- sible CAN were more likely to be Aboriginal, while those with definite CAN were had a higher BMI and resting pulse rate, longer diabetes duration, a greater uACR, and a higher likelihood of DSPN. a backward stepwise manner and included demographic and diabetes-related factors, the presence of other com- plications and cardiovascular risk factors. Aboriginal status was also considered for entry since Aboriginal participants were significantly younger than other eth- nic groups. After the most parsimonious model in each instance was defined, CAN status was entered. The pro- portional hazards assumption was assessed and, if vio- lated, adjusted for by adding significant time-varying covariates. A two-tailed significance level of P < 0.05 was used throughout. Incident ischaemic heart diseaseh The characteristics of the 142 FDS2 participants (21.6%) who had an IHD event during follow-up and those who did not are summarised in Table 3. The Kaplan–Meier curves of IHD events during a mean ± SD follow-up period of 9.7 ± 3.2 (range 0.0–12.6) years (equivalent to 6354 person-years) are shown in Fig. 1 (upper panel). There was a statistically significant difference between the three groups (log rank test P = 0.030) with definite CAN significantly different from both the no CAN group (P = 0.009) and possible CAN group (P = 0.039) in unad- justed pairwise comparisons. In Cox proportional haz- ards modelling, longer diabetes duration, a higher heart rate and uACR, and DSPN were independent predictors of incident IHD events, but CAN category did not add significantly to the model (see Table 4). All‑cause mortality Th h The characteristics of the 162 FDS2 participants (19.5%) who died during follow-up and those who did not are summarised in Table 6. The Kaplan–Meier curves of HF events during a mean ± SD follow-up period of 10.5 ± 2.4 (range 0.2–12.6) years (equivalent to 8,684 person-years) are shown in Fig. 1 (lower panel). There was a signifi- cant difference between the three groups (log rank test P < 0.001), with both possible (P = 0.015) and definite (P < 0.001) CAN significantly different from the no CAN i The independent associates of CAN group identified by multinomial modelling are shown in Table 2. Com- pared to the group without CAN, those with possible or Davis et al. All‑cause mortality Th h Cardiovascular Diabetology (2024) 23:102 Page 6 of 16 a 1 standard drink = 10 U ethanol; P < 0.05, P < 0.01, P < 0.001 vs no CAN, †P < 0.05, ††P < 0.01, †††P < 0.001 vs possible CAN in Bonferroni-corrected multiple comparisons Table 1 (continued) No CAN Possible CAN Definite CAN P-value ≥ 90 mL/min/1.73m2 43.0 47.9 30.7 60–89 mL/min/1.73m2 49.9 39.6 44.9 45–59 mL/min/1.73m2 5.2 6.4 13.4 30–44 mL/min/1.73m2 1.4 4.3 6.3 < 30 mL/min/1.73m2 0.5 1.8 4.7 Distal symmetrical polyneuropathy (%) 29.4 32.1 44.4* 0.007 Peripheral arterial disease (%) 18.2 20.1 27.6 0.076 Prior IHD hospitalisation (%) 15.1 23.2* 32.3*** < 0.001 Prior cerebrovascular disease hospitalisation (%) 4.5 4.6 7.9 0.309 Prior HF hospitalisation (%) 1.7 5.7* 8.7 < 0.001 Table 1 (continued) Table 2 Independent associates of CAN category in FDS2 participants with type 2 diabetes Multinomial regression was used with no prevalent CAN as the reference. Data are odds ratios (OR) and 95% confidence intervals (CIs). Values are provided for both CAN categories where variables were significant for at least category a A 2.72-fold increase in urinary albumin:creatinine ratio corresponds to an increased risk of 1 in ln(urinary albumin:creatinine) Possible CAN Definite CAN OR (95% CI) OR (95% CI) Aboriginal descent 1.91 (1.01, 3.62) 1.65 (0.72, 3.78) Diabetes duration (increase of 1 year) 1.01 (0.98, 1.03) 1.05 (1.02, 1.08) Body mass index (increase of 1 kg/m2) 1.02 (0.99, 1.05) 1.04 (1.005, 1.08) Pulse rate (increase of 1 beat/min) 1.01 (0.996, 1.03) 1.04 (1.02, 1.06) On beta-blockers 1.95 (1.24, 3.09) 2.86 (1.60, 4.93) On calcium channel blockers 1.70 (1.15, 2.50) 2.44 (1.47, 4.04) On antidepressant medication 1.67 (1.06, 2.64) 2.81 (1.60, 4.93) Ln [urinary albumin:creatinine ratio (mg/mmol)]a 1.09 (0.95, 1.24) 1.29 (1.10, 1.53) Distal symmetrical polyneuropathy 1.12 (0.79, 1.58) 1.65 (1.04, 2.61) Prior hospitalisation for heart failure 2.96 (1.11, 7.92) 3.46 (1.15, 10.4) Table 2 Independent associates of CAN category in FDS2 participants with type 2 diabetes Table 2 Independent associates of CAN category in FDS2 participants with type 2 diabetes the first involved participants in whom CAN category was determined (n = 830) and the second utilised avail- able data from the 1254 participants who underwent CAN testing. None of the CART variables was a sig- nificant predictor of incident IHD or HF after adjusting for the most parsimonious model. The results of anal- yses for all-cause mortality are shown in Table 7. All‑cause mortality Th h Cardiovascular Diabetology (2024) 23:102 Page 5 of 16 Table 1 Baseline characteristics of FDS2 participants categorised by CAN status No CAN Possible CAN Definite CAN P-value N (%) 423 (51.0) 280 (33.7) 127 (15.3) Orthostatic hypotension (%) 12.3 11.5 11.1 0.945 Age (years) 62.2 ± 10.5 62.1 ± 10.2 63.2 ± 11.4 0.555 Male (%) 57.9 54.6 52.8 0.497 Education beyond primary level (%) 92.3 94.2 90.3 0.366 Not fluent in English (%) 6.6 7.1 6.3 0.946 Ethnic background (%) 0.053 Anglo-Celt 55.1 51.4 48.8 Southern European 10.6 11.1 10.2 Other European 5.9 8.9 10.2 Asian 4.0 5.7 3.1 Aboriginal 5.2 10.0 11.8 Mixed/other 19.1 12.9 15.7 Smoking status (%) 0.262 Never 53.5 50.9 50.4 Ex- 37.7 36.5 34.1 Current 8.8 12.6 15.4 Alcohol use (standard drinksa/day) 0.3 [0–1.5] 0.3 [0–1.5] 0.1 [0–1.0] 0.455 Antidepressant use (%) 10.6 16.4 24.4* < 0.001 Tricyclic antidepressants 2.6 1.8 8.7,† 0.003 Selective serotonin reuptake inhibitors 6.1 11.1 7.1 0.061 Age at diabetes diagnosis (years) 53.8 ± 11.0 53.1 ± 10.7 49.4 ± 13.9,†† < 0.001 Diabetes duration (years) 6.0 [2.6–11.9] 8.0 [3.0–15.1] 10.0 [4.0–17.2],††† < 0.001 Diabetes treatment (%) ,††† < 0.001 Diet/exercise alone 26.5 22.7 12.7 Oral glucose lowering agents 55.7 55.6 46.0 Insulin alone 3.3 2.5 8.7 Insulin + oral agents 14.5 19.1 32.5 HbA1c (%) 6.8 [6.2–7.8] 6.9 [6.2–7.8] 7.3 [6.6–8.4],†† < 0.001 HbA1c (mmol/mol) 50 [44–58] 51 [44–61] 57 [49–74] < 0.001 Fasting serum glucose (mmol/L) 7.2 [6.3–8.6] 7.4 [6.4–9.3] 8.2 [6.7–10.3]*,† < 0.001 ABSI (m11/6 kg−2/3) 0.081 ± 0.005 0.081 ± 0.005 0.082 ± 0.005 0.051 BMI (kg/m2) 31.2 ± 5.9 32.4 ± 6.2 33.5 ± 7.1 < 0.001 Heart rate (beats/min) 69 ± 11 70 ± 11 74 ± 13,†† < 0.001 Supine systolic blood pressure (mmHg) 141 ± 19 143 ± 21 143 ± 22 0.478 Supine diastolic blood pressure (mmHg) 81 ± 11 81 ± 11 78 ± 14 0.071 Antihypertensive medication (%) 66.0 72.1 78.7 0.013 Angiotensin converting enzyme inhibitors 32.9 33.6 38.6 0.479 Angiotensin receptor blockers 30.5 36.4 38.6 0.119 Beta-blockers 10.9 21.4* 29.9* < 0.001 Calcium channel blockers 16.8 26.1 36.2* < 0.001 Diuretics 20.1 29.3* 37.8* < 0.001 Total serum cholesterol (mmol/L) 4.3 ± 0.9 4.4 ± 1.1 4.5 ± 1.6 0.215 Serum HDL-cholesterol (mmol/L) 1.21 ± 0.32 1.16 ± 0.28 1.17 ± 0.32 0.064 Serum triglycerides (mmol/L) 1.5 (0.9–2.4) 1.6 (0.9–2.7) 1.7 (1.1–2.8) 0.004 Lipid-lowering medication (%) 67.8 67.5 69.3 0.947 uACR (mg/mmol) 2.2 (0.7–6.7) 2.9 (0.8–10.4)* 5.2 (0.9–30.5)*,††† < 0.001 eGFR (mL/min/1.73m2) 85 ± 17 84 ± 20 77 ± 23,†† < 0.001 eGFR categories (%) **,† < 0.001 Table 1 Baseline characteristics of FDS2 participants categorised by CAN status Davis et al. All‑cause mortality Th h The proportional hazards assumption was violated in the model involving CAN-categorised participants, with the effects of Aboriginal descent and antidepressant therapy attenuating over time. For the second model, there were 1,101 participants with an MCR and E:I ratio (age range 17–95 years), 1100 with a 30:15 stand ratio (age range 17–95 years), and 904 with a Valsalva ratio (age range 17–89 years). The proportional hazards assumption was violated in this latter model, with the effect of age strengthening over time. In both models, group, and possible CAN significantly different from def- inite CAN (P < 0.001), in unadjusted pairwise compari- sons. In Cox proportional hazards modelling, increasing age, male sex, Aboriginal descent, Other European ethnic background, use of antidepressant therapy and of angio- tensin receptor blockers, an eGFR < 30 mL/min/1.73 m2, DSPN, and a prior history of both cerebrovascular dis- ease and HF were independent predictors of mortality. CAN category added significantly to the model at the expense of DSPN (see Table 4). group, and possible CAN significantly different from def- inite CAN (P < 0.001), in unadjusted pairwise compari- sons. In Cox proportional hazards modelling, increasing age, male sex, Aboriginal descent, Other European ethnic background, use of antidepressant therapy and of angio- tensin receptor blockers, an eGFR < 30 mL/min/1.73 m2, DSPN, and a prior history of both cerebrovascular dis- ease and HF were independent predictors of mortality. CAN category added significantly to the model at the expense of DSPN (see Table 4). Relationship between individual CAN tests and outcome Individual CART test results were included as con- tinuous variables in separate Cox models of the three outcomes in place of CAN category. This allowed use of data from participants of all ages and those whose incomplete CART testing precluded CAN categorisa- tion. Two models for each variable were constructed, Davis et al. All‑cause mortality Th h Cardiovascular Diabetology (2024) 23:102 Page 8 of 16 Table 3 (continued) Table 3 (continued) No hospitalisation for/with IHD or non-HF cardiac/sudden death Incident hospitalisation for IHD or non-HF cardiac/sudden deathb P-value Calcium channel blockers 20.2 21.1 0.814 Diuretics 24.0 28.9 0.274 Total serum cholesterol (mmol/L) 4.3 ± 1.0 4.5 ± 1.2 0.040 Serum HDL-cholesterol (mmol/L) 1.21 ± 0.33 1.19 ± 0.29 0.534 Serum triglycerides (mmol/L) 1.5 (0.9–2.4) 1.6 (1.0–2.7) 0.028 Lipid-lowering medication (%) 64.1 61.3 0.555 uACR (mg/mmol) 2.3 (0.7–7.5) 4.1 (0.8–21.6) < 0.001 eGFR (mL/min/1.73m2) 86 ± 18 82 ± 21 0.048 eGFR categories (%) 0.254 ≥ 90 mL/min/1.73m2 47.2 41.1 60–89 mL/min/1.73m2 43.7 44.7 45–59 mL/min/1.73m2 6.0 7.8 30–44 mL/min/1.73m2 1.7 4.3 < 30 mL/min/1.73m2 1.4 2.1 Distal symmetrical polyneuropathy (%) 28.7 38.3 0.031 Peripheral arterial disease (%) 17.9 23.2 0.184 Prior hospitalisation for cerebrovascular disease (%) 2.9 6.3 0.073 Prior hospitalisation for HF (%) 1.4 2.8 0.264 a 1 standard drink = 10 U ethanol; bthree participants with unknown cause of death excluded from this analysis No hospitalisation for/with IHD or non-HF cardiac/sudden death Incident hospitalisation for IHD or non-HF cardiac/sudden deathb P-value a 1 standard drink = 10 U ethanol; bthree participants with unknown cause of death excluded from this an much lower prevalence of 9% after 6 years of screening- detected diabetes [4]. Our prevalence of definite CAN was intermediate between these values at 15.3%. Since almost all of the studies in the narrative review were con- ducted in secondary care [3], it is likely that our com- munity-based cohort had less at-risk participants than secondary care studies with referred patients but more than in a pure primary care context [4]. The CAN risk factor profile in our FDS2 participants included inde- pendent variables that have been reported previously including BMI, longer diabetes duration, resting tachy- cardia (reflecting increased sympathetic tone [1]), as well as a prior history of HF (another potential manifestation of sympathetic overactivity and neurohormonal dysregu- lation [29]). A previously recognised positive association with antidepressant use in the general population [30] was confirmed in the present case of type 2 diabetes. All‑cause mortality Th h These considerations suggest that, despite the exclusion of FDS2 participants who were recruited before CAN testing was available as part of baseline assessment, those (largely elderly) whose CART data could not be assessed against reference ranges and those in whom valid CART data could not be collected, our final sample of 830 gen- erated representative data and was amongst the larger of studies reporting CAN prevalence [3, 4] and prognosis [7]. MCR showed a significant inverse association with all- cause death. All‑cause mortality Th h Cardiovascular Diabetology (2024) 23:102 Page 7 of 16 Page 7 of 16 Table 3 Baseline characteristics of FDS2 participants with type 2 diabetes by incident IHD status, excluding those with pre- recruitment hospitalisation for/with IHD p p yp y g p recruitment hospitalisation for/with IHD No hospitalisation for/with IHD or non-HF cardiac/sudden death Incident hospitalisation for IHD or non-HF cardiac/sudden deathb P-value N (%) 516 (78.4) 142 (21.6) CAN group (%) 0.099 None 55.6 50.0 Possible 32.9 31.7 Definite 11.4 18.3 MCR (n = 658) 21.4 [12.4–34.8] 22.1 [11.9–31.0] 0.578 E:I ratio (n = 658) 1.13 [1.09–1.20] 1.13 [1.08–1.18] 0.198 Valsalva ratio (n = 642) 1.51 [1.33–1.73] 1.44 [1.28–1.58] 0.009 30:15 Stand ratio (n = 657) 1.16 [1.09–1.25] 1.12 [1.07–1.21] < 0.001 Orthostatic hypotension (%) 14.8 8.5 0.069 Age (years) 60.9 ± 10.5 62.0 ± 10.5 0.257 Male (%) 53.1 54.9 0.705 Education beyond primary level (%) 93.7 92.9 0.702 Not fluent in English (%) 6.4 6.3 > 0.999 Ethnic background (%) 0.046 Anglo-Celt 54.1 50.0 Southern European 11.0 10.6 Other European 7.2 7.7 Asian 4.8 2.8 Aboriginal 5.6 14.1 Mixed/other 17.2 14.8 Smoking status (%) 0.225 Never 54.5 46.4 Ex- 34.9 40.7 Current 10.6 12.9 Alcohol consumption (standard drinksa/day) 0.3 [0–1.5] 0.1 [0–0.8] 0.034 Antidepressant use (%) 14.1 15.5 0.687 Tricyclic antidepressants 2.5 4.2 0.267 Selective serotonin reuptake inhibitors 8.9 6.3 0.394 Age at diabetes diagnosis (years) 52.8 ± 11.2 50.9 ± 12.3 0.076 Diabetes duration (years) 5.7 [2.6–13.0] 10.0 [4.2–16.1] < 0.001 Diabetes treatment (%) 0.019 Diet/exercise 25.5 17.0 Oral agents ± non-insulin injectables 56.1 53.9 Insulin alone 2.7 5.0 Insulin ± oral agents ± non-insulin injectables 15.6 24.1 HbA1c (%) 6.8 [6.2–7.6] 7.1 [6.3–8.3] 0.005 HbA1c (mmol/mol) 51 [44–60] 54 [45–67] 0.005 Fasting serum glucose (mmol/L) 7.3 [6.4–8.9] 7.6 [6.5–10.6] 0.035 ABSI (m11/6 kg−2/3) 0.080 ± 0.005 0.082 ± 0.005 0.004 BMI (kg/m2) 32.2 ± 6.2 31.8 ± 7.0 0.502 Pulse rate (beats/min) 70 ± 11 73 ± 12 0.004 Supine systolic blood pressure (mmHg) 141 ± 20 146 ± 20 0.003 Supine diastolic blood pressure (mmHg) 81 ± 12 81 ± 13 0.851 Antihypertensive medication (%) 64.3 65.5 0.843 Angiotensin converting enzyme inhibitors 30.2 33.1 0.539 Angiotensin receptor blockers 31.2 35.2 0.363 Beta-blockers 8.9 9.9 0.743 recruitment hospitalisation for/with IHD No hospitalisation for/with IHD non-HF cardiac/sudden death N (%) 516 (78.4) CAN group (%) None 55.6 Possible 32.9 Definite 11.4 MCR (n = 658) 21.4 [12.4–34.8] E:I ratio (n = 658) 1.13 [1.09–1.20] Valsalva ratio (n = 642) 1.51 [1.33–1.73] 30:15 Stand ratio (n = 657) 1.16 [1.09–1.25] Orthostatic hypotension (%) 14.8 Age (years) 60.9 ± 10.5 Male (%) 53.1 Education beyond primary level (%) 93.7 Not fluent in English (%) 6.4 Ethnic background (%) Anglo-Celt 54.1 Southern European 11.0 Other European 7.2 Asian 4.8 Aboriginal 5.6 Mixed/other 17.2 Smoking status (%) Never 54.5 Ex- 34.9 Current 10.6 Alcohol consumption (standard drinksa/day) 0.3 [0–1.5] Antidepressant use (%) 14.1 Tricyclic antidepressants 2.5 Selective serotonin reuptake inhibitors 8.9 Age at diabetes diagnosis (years) 52.8 ± 11.2 Diabetes duration (years) 5.7 [2.6–13.0] Diabetes treatment (%) Diet/exercise 25.5 Oral agents ± non-insulin injectables 56.1 Insulin alone 2.7 Insulin ± oral agents ± non-insulin injectables 15.6 HbA1c (%) 6.8 [6.2–7.6] HbA1c (mmol/mol) 51 [44–60] Fasting serum glucose (mmol/L) 7.3 [6.4–8.9] ABSI (m11/6 kg−2/3) 0.080 ± 0.005 BMI (kg/m2) 32.2 ± 6.2 Pulse rate (beats/min) 70 ± 11 Supine systolic blood pressure (mmHg) 141 ± 20 Supine diastolic blood pressure (mmHg) 81 ± 12 Antihypertensive medication (%) 64.3 Angiotensin converting enzyme inhibitors 30.2 Angiotensin receptor blockers 31.2 Beta-blockers 8.9 Davis et al. Discussionh The present study involving representative, community- based people with type 2 diabetes followed for an average of 10 years, showed that definite CAN was significantly associated with incident IHD and HF compared to both no CAN and possible CAN in survival analyses. How- ever, when CAN category was included in multivari- able models of these two incident events, it did not add to other independent predictors. Survival analysis also showed that there was a graded increase in risk of all- cause mortality from no CAN through possible CAN to definite CAN which was observed in multivariable analy- sis after adjustment for confounders. The only individual CART test predictive of all-cause death was MCR. Taken together, these findings question the need for screening for CAN, as has been suggested [2, 7], as part of routine care of type 2 diabetes, especially since the gold stand- ard CART evaluation is demanding for both patients and staff [6]. In addition, the present data suggest that around one-third of patients will either be ineligible because of age or they will, for various reasons, be unable to com- plete a valid CART assessment. In a recent narrative review, the prevalence of CAN in type 2 diabetes was reported as between 31 and 73% [3], but a subsequent Danish primary care study found a Our Kaplan–Meier analyses showed a significant rela- tionship between CAN category and incident IHD. This Davis et al. Cardiovascular Diabetology (2024) 23:102 Page 9 of 16 Page 9 of 16 Fig. 1 Kaplan–Meier plots of incident IHD, HF and all-cause mortality for FDS2 participants with no CAN (green square), possible CAN (red circle) and definite CAN (blue up-pointing triangle). P-values are from log-rank tests Table 4 Most parsimonious Cox models of independent predictors of IHD, HF and all-cause death in the FDS2 cohort with CAN category defined (n = 830) and added g yi Hazard ratios (HR) and 95% confidence intervals (95% CI) are shown. Discussionh The proportional hazards assumption was violated in the model for all-cause mortality but not for IHD or HF a A 2.72-fold increase in urinary albumin:creatinine ratio corresponds to an increased risk of 1 in ln(urinary albumin:creatinine) HR (95% CI) P-value IHD (n = 651) Diabetes duration (increase of 1 year) 1.02 (1.004, 1.04) 0.021 Heart rate (increase of 1 beat/min) 1.02 (1.005, 1.03) 0.009 Ln(uACR)a 1.21 (1.08, 1.35) 0.001 Distal symmetrical polyneuropathy 1.29 (1.05, 1.57) 0.014 CAN category: none 1.00 Possible 0.96 (0.66, 1.41) 0.851 Definite 1.00 (0.60, 1.68) 0.993 HF (n = 788) Age (increase of 1 year) 1.05 (1.02, 1.07) < 0.001 Aboriginal descent 2.79 (1.51, 5.18) 0.001 Diabetes duration (increase of 1 year) 1.03 (1.01, 1.05) 0.011 Ln(uACR)a 1.33 (1.18, 1.49) < 0.001 Distal symmetrical polyneuropathy 1.49 (1.19, 1.85) < 0.001 Peripheral arterial disease 1.66 (1.11, 2.48) 0.013 Ischaemic heart disease 1.95 (1.32, 2.88) 0.001 CAN category: None 1.00 Possible 0.94 (0.61, 1.46) 0.789 Definite 1.27 (0.78, 2.06) 0.342 All-cause mortality (n = 827) Age (increase of 1 year) 1.07 (1.05, 1.09) < 0.001 Male sex 1.55 (1.12, 2.15) 0.008 Aboriginal descent 8.82 (3.20, 24.3) < 0.001 Other European ethnic background 0.36 (0.16, 0.79) 0.011 Antidepressant therapy 3.06 (1.29, 7.26) 0.011 Angiotensin receptor blocker use 1.42 (1.03, 1.95) 0.031 eGFR < 30 mL/min/1.73m2 2.82 (1.40, 5.69) 0.004 Distal symmetrical polyneuropathy 1.14 (0.94, 1.38) 0.174 Cerebrovascular disease 1.67 (1.03, 2.73) 0.039 HF 2.42 (1.42, 4.13) 0.001 CAN category: None 1.00 Possible 1.47 (1.01, 2.14) 0.046 Definite 2.42 (1.60, 3.67) < 0.001 Time-varying covariates Aboriginal descent 0.42 (0.24, 0.75) 0.003 Antidepressant therapy 0.62 (0.39, 0.99) 0.047 Fig. 1 Kaplan–Meier plots of incident IHD, HF and all-cause mortality for FDS2 participants with no CAN (green square), possible CAN (red circle) and definite CAN (blue up-pointing triangle). P-values are from log-rank tests was consistent with the results of a recent meta-analysis [7] in which there was significant heterogeneity, reflect- ing a variety of sample sizes, participant sources includ- ing people with type 1 diabetes and those with type 2 diabetes selected for clinical trials [8, 31, 32], and meth- ods of diagnosing CAN which ranged from full CARTs to change in heart rate on standing [32] and heart rate variability (HRV) and QT index on resting electrocardi- ography [8]. Discussionh Our multivariable analysis showed HRs for possible and definite CAN that were close to unity in the presence of other recognised predictors of incident IHD (longer diabetes duration, higher heart rate and increased uACR [2]). It is possible that relatively intensive CVD risk factor management in FDS2 paralleling trends in other high income countries [9, 10] (for example, approxi- mately two-thirds of our participants were taking renin- angiotensin blocking drugs and statins) attenuated both the risk of CAN and its effect on CVD outcomes found in was consistent with the results of a recent meta-analysis [7] in which there was significant heterogeneity, reflect- ing a variety of sample sizes, participant sources includ- ing people with type 1 diabetes and those with type 2 diabetes selected for clinical trials [8, 31, 32], and meth- ods of diagnosing CAN which ranged from full CARTs to change in heart rate on standing [32] and heart rate variability (HRV) and QT index on resting electrocardi- ography [8]. Our multivariable analysis showed HRs for possible and definite CAN that were close to unity in the presence of other recognised predictors of incident IHD (longer diabetes duration, higher heart rate and increased uACR [2]). It is possible that relatively intensive CVD risk factor management in FDS2 paralleling trends in other high income countries [9, 10] (for example, approxi- mately two-thirds of our participants were taking renin- angiotensin blocking drugs and statins) attenuated both the risk of CAN and its effect on CVD outcomes found in earlier studies, most of which were published before the first FDS2 patient was assessed for CAN [7]. In addition, we excluded participants with a history of IHD at base- line which may not have been the case in at least some of the studies with consequently higher risk samples in the Davis et al. Discussionh Cardiovascular Diabetology (2024) 23:102 Page 11 of 16 Table 5 (continued) Table 5 (continued) a 1 standard drink = 10 U ethanol; btwo participants with unknown cause of death excluded from this analysis Table 5 (continued) No hospitalisation for/with or death from HF Incident hospitalisation for/with or death from HFb P-value Beta-blockers 13.9 26.1 0.002 Calcium channel blockers 20.9 35.3 < 0.001 Diuretics 24.1 30.3 0.169 Total serum cholesterol (mmol/L) 4.3 ± 1.0 4.4 ± 1.7 0.569 Serum HDL-cholesterol (mmol/L) 1.19 ± 0.30 1.23 ± 0.35 0.143 Serum triglycerides (mmol/L) 1.5 (0.9–2.5) 1.6 (0.9–2.8) 0.349 Lipid-lowering medication (%) 66.7 74.8 0.089 uACR (mg/mmol) 2.3 (0.7–7.1) 5.9 (1.1–32.4) < 0.001 eGFR (mL/min/1.73m2) 85 ± 18 79 ± 21 0.002 eGFR categories (%) 0.003 ≥ 90 mL/min/1.73m2 45.6 31.9 60–89 mL/min/1.73m2 45.2 48.7 45–59 mL/min/1.73m2 5.6 12.6 < 45 mL/min/1.73m2 3.6 6.7 Distal symmetrical polyneuropathy (%) 30.3 45.8 0.001 Peripheral arterial disease (%) 16.8 35.3 < 0.001 Prior hospitalisation for IHD (%) 15.4 36.1 < 0.001 Prior hospitalisation for cerebrovascular disease (%) 3.9 10.9 0.004 No hospitalisation for/with or death from HF Incident hospitalisation for/with or death from HFb P-value a 1 standard drink = 10 U ethanol; btwo participants with unknown cause of death excluded from this analysis In both Kaplan–Meier and Cox proportional hazard analyses, there was a significant and graded relation- ship between CAN category and all-cause mortality, with definite CAN associated with a more than doubling of risk at 10 years after adjustment for the presence of other recognised significant independent predictors of death (increasing age, male sex, Aboriginal descent, anti- depressant therapy, eGFR < 30 mL/min/1.73m2, DSPN, and prior cerebrovascular disease and HF) [36–38]. In a recent meta-analysis, the unadjusted risk ratio for death was more than three-fold increased in people with CAN [7]. By far the largest contributor of participants and events in this study was the ACCORD trial [8] in which CAN was associated with a 1.55–2.14-fold risk of mor- tality after full adjustment for confounders, the risk ratio range reflecting three different methods of CAN ascer- tainment. However, the ACCORD participants were selected as having high CVD risk, the definition and stag- ing of CAN was based on electrocardiographic indices with or without the presence of DSPN, and the follow- up duration was relatively short (3.5 years) [8]. Discussionh Cardiovascular Diabetology (2024) 23:102 Page 10 of 16 Table 5 Baseline characteristics of FDS2 type 2 diabetes participants attending for their first ANS examination between May 2009 and November 2012 by incident hospitalisation for/with heart failure or death from heart failure to 31 December 2021, excluding those with prior hospitalisation for/with heart failure Table 5 Baseline characteristics of FDS2 type 2 diabetes participants attending for their first ANS examination between May 2009 and November 2012 by incident hospitalisation for/with heart failure or death from heart failure to 31 December 2021, excluding those with prior hospitalisation for/with heart failure No hospitalisation for/with or death from HF Incident hospitalisation for/with or death from HFb P-value No hospitalisation for/with or death from HF Incident hospitalisation for/with or death from HFb P-value N (%) 675 (85.0) 119 (15.0) CAN group (%) 0.002 None 53.8 43.7 Possible 33.6 31.1 Definite 12.6 25.2 MCR (n = 794) 21.3 [12.1–33.3] 16.1 [7.9–25.3] < 0.001 E:I ratio (n = 794) 1.13 [1.08–1.19] 1.11 [1.06–1.17] 0.008 Valsalva ratio (n = 775) 1.47 [1.31–1.68] 1.38 [1.22–1.52] < 0.001 30:15 Stand ratio (n = 794) 1.15 [1.08–1.23] 1.12 [1.06–1.18] 0.005 Orthostatic hypotension (%) 12.5 8.4 0.223 Age (years) 61.4 ± 10.2 66.5 ± 11.2 < 0.001 Male (%) 56.4 55.5 0.842 Education beyond primary level (%) 93.2 93.1 > 0.999 Not fluent in English (%) 6.4 8.4 0.425 Ethnic background (%) 0.050 Anglo-Celt 53.3 50.4 Southern European 11.1 9.2 Other European 7.4 6.7 Asian 4.9 3.4 Aboriginal 5.9 15.1 Mixed/other 17.3 15.1 Smoking status (%) 0.306 Never 52.8 51.7 Ex- 37.2 33.6 Current 10.0 14.7 Alcohol consumption (standard drinksa/day) 0.3 [0–1.5] 0.1 [0–0.8] 0.023 Antidepressant use (%) 14.7 15.1 0.889 Tricyclic antidepressants 3.3 4.2 0.583 Selective serotonin reuptake inhibitors 8.3 6.7 0.715 Age at diabetes diagnosis (years) 52.9 ± 11.0 52.5 ± 13.6 0.764 Diabetes duration (years) 6.1 [2.9–13.1] 14.0 [5.8–19.2] < 0.001 Diabetes treatment (%) 0.001 Diet/exercise 24.4 16.2 Oral agents ± non-insulin injectables 56.0 47.9 Insulin alone 3.1 6.8 Insulin ± oral agents ± non-insulin injectables 16.5 29.1 HbA1c (%) 6.9 [6.2–7.7] 7.1 [6.4–8.2] 0.032 HbA1c (mmol/mol) 52 [44–61] 54 [46–66] 0.032 Fasting serum glucose (mmol/L) 7.4 [6.5–9.1] 7.5 [6.3–9.3] 0.805 ABSI (m11/6 kg−2/3) 0.081 ± 0.005 0.082 ± 0.005 < 0.001 BMI (kg/m2) 31.9 ± 6.1 31.6 ± 6.6 0.663 Pulse rate (beats/min) 70 ± 11 70 ± 13 0.573 Supine systolic blood pressure (mmHg) 141 ± 19 150 ± 23 < 0.001 Supine diastolic blood pressure (mmHg) 81 ± 11 80 ± 13 0.796 Antihypertensive medication (%) 68.1 77.3 0.052 Angiotensin converting enzyme inhibitors 31.9 38.7 0.169 Angiotensin receptor blockers 32.9 40.3 0.117 Davis et al. Discussionh Although these considerations complicate comparisons with the present study, the ACCORD findings are largely consist- ent with those of the present study.h meta-analysis, a consideration that may have contributed to the heterogeneity observed. There are limited data assessing the relationship between CAN complicating type 2 diabetes and inci- dent HF. In a report from the ACCORD trial involv- ing trial participants with high CVD risk followed for a mean of 4.9 years, those with CAN defined from quar- tiles of HRV had a 2.7-fold greater risk of HF in adjusted analyses [33]. In our community-based participants assessed using CART, there was a significant relation- ship between definite CAN and incident HF in Kaplan– Meier analysis, with a more than doubling of the risk at 10 years. However, as with incident IHD, multivariable analysis showed HRs for possible and definite CAN that were close to unity in the presence of other recognised significant independent predictors of incident HF (older age, Aboriginal descent, longer diabetes duration, higher uACR, DSPN, and prior IHD) [34, 35]. We hypothesise that relatively intensive CVD risk factor management in FDS2, especially the large proportion of our participants who were taking renin-angiotensin blocking drugs, con- tributed to the lack of a significant association in a cohort studied before the widespread availability in Australia of the newer blood glucose-lowering agents (sodium-glu- cose co-transporter-2 inhibitors and glucagon-like pep- tide 1 receptor agonists) with beneficial effects on HF. As with IHD, we also excluded those with a history of HF at baseline. The use of a representative, community-based sam- ple in the present study may have masked sub-groups of people with type 2 diabetes in whom CAN has independ- ent predictive value for IHD and HF. Such a sub-group Davis et al. Discussionh may comprise those who are at high cardiovascular risk or who have established CVD [8, 33]. Nevertheless, in the Detection of Ischemia in Asymptomatic Diabetics (DIAD) study involving participants with type 2 diabetes without known heart disease [39], the incidence of the composite clinical outcome of cardiac death, acute coro- nary syndromes, HF, or coronary revascularization over 5 years was significantly increased in those in the lowest quartile of the Valsalva heart rate ratio (hazard ratio 1.60) amongst a range of tests of autonomic heart rate/blood pressure responses and power spectral analysis of HRV. The results of these studies should be interpreted against the heterogeneity in meta-analysis of prognostic stud- ies of CAN [7] which highlights the influence of sample selection and CAN assessment methods. In addition, the duration of follow-up may be an important consideration since there is evidence that CART indices of autonomic dysfunction attenuate over time [4]. Our mean 10-year follow-up may have captured the effect of this on incident IHD and HF events compared to the ≤ 5 year follow-up in studies such as ACCORD and DIAD [8, 33, 39]. y g The present study had limitations. As acknowledged, we excluded around one in 14 potentially eligible par- ticipants because of their age. Although we could have included them by using fixed thresholds for the vari- ous CART tests, as has been done in previous stud- ies [20, 21, 44], there are important effects of age and sex on CART reference ranges [23, 45]. In any case, we included CART results as continuous variables and found that they were not predictive of endpoints apart from MCR for all-cause death. The presence of orthostatic hypotension in addition to abnormal heart rate test results identifies severe or advanced CAN [1]. However, probably due to the high percentage of our participants taking at least one antihypertensive medi- cation (69.5%), a recognised confounding variable [1, 25], severe CAN was not an independent predictor in any of our multivariable analyses (data not shown). In addition, anaemia is a recognised complication of type 2 diabetes [46] and can contribute to orthostatic hypo- tension [47]. The major strengths of our study include We found evidence that, of the individual CARTs performed, only reduced HRV as assessed from MCR was independently associated with all-cause mor- tality. Discussionh Cardiovascular Diabetology (2024) 23:102 Page 12 of 16 Table 6 Baseline characteristics of FDS2 participants with type 2 diabetes by all-cause mortality at end of follow-up Alive Deceased P-value N 668 (80.5) 162 (19.5) CAN group (%) < 0.001 None 54.6 35.8 Possible 33.4 35.2 Definite 12.0 29.0 MCR (n = 830) 21.5 [12.9–33.9] 11.8 [6.8–23.7] < 0.001 E:I ratio (n = 830) 1.13 [1.08–1.19] 1.11 [1.06–1.15] < 0.001 Valsalva ratio (n = 809) 1.47 [1.31–1.68] 1.37 [1.22–1.51] < 0.001 30:15 Stand ratio (n = 829) 1.15 [1.09–1.23] 1.12 [1.05–1.18] < 0.001 Orthostatic hypotension (%) 12.6 8.7 0.220 Age (years) 60.9 ± 10.2 68.0 ± 9.7 < 0.001 Male (%) 54.3 63.0 0.052 Education beyond primary level (%) 93.3 90.1 0.178 Not fluent in English (%) 6.9 6.2 0.862 Ethnic background (%) 0.093 Anglo-Celt 51.2 59.9 Southern European 11.4 8.0 Other European 8.4 4.3 Asian 5.1 1.9 Aboriginal/TSI 7.6 8.6 Mixed/other 16.3 17.3 Smoking status (%) 0.359 Never 53.1 48.1 Ex- 35.6 41.8 Current 11.3 10.1 Alcohol consumption (standard drinksa/day) 0.3 [0–1.2] 0.1 [0–1.5] 0.106 Antidepressant use (%) 13.5 19.8 0.048 TCAs 2.8 4.9 0.213 SSRIs 7.9 8.0 > 0.999 Age at diabetes diagnosis (years) 52.2 ± 11.0 55.8 ± 12.8 0.001 Diabetes duration (years) 6.5 [3.0–14.0] 11.0 [4.5–17.1] < 0.001 Diabetes treatment (%) 0.223 Diet/exercise 23.6 21.3 Oral agents ± non-insulin injectables 55.2 50.0 Insulin alone 3.8 4.4 Insulin ± oral agents ± non-insulin injectables 17.4 24.4 HbA1c (%) 6.9 [6.3–7.8] 6.8 [6.3–7.7] 0.347 HbA1c (mmol/mol) 52 [45–62] 51 [45–61] 0.347 Fasting plasma glucose (mmol/L) 7.4 [6.4–9.2] 7.5 [6.2–9.0] 0.199 ABSI (m11/6 kg−2/3) 0.081 ± 0.005 0.082 ± 0.005 < 0.001 BMI (kg/m2) 32.1 ± 6.2 31.3 ± 6.5 0.138 Pulse rate (bpm) 70 ± 11 70 ± 12 0.391 Supine SBP (mmHg) 142 ± 20 145 ± 20 0.032 Supine DBP (mmHg) 81 ± 12 79 ± 12 0.016 On BP-lowering medication (%) 67.5 80.2 0.002 ACE-I 33.2 37.0 0.357 ARB 31 1 44 4 0 002 ype 2 diabetes by all-cause mortality at end of follow-up Alive Deceased P-value 668 (80.5) 162 (19.5) < 0.001 54.6 35.8 33.4 35.2 12.0 29.0 21.5 [12.9–33.9] 11.8 [6.8–23.7] < 0.001 1.13 [1.08–1.19] 1.11 [1.06–1.15] < 0.001 1.47 [1.31–1.68] 1.37 [1.22–1.51] < 0.001 1.15 [1.09–1.23] 1.12 [1.05–1.18] < 0.001 12.6 8.7 0.220 60.9 ± 10.2 68.0 ± 9.7 < 0.001 54.3 63.0 0.052 93.3 90.1 0.178 6.9 6.2 0.862 0.093 51.2 59.9 11.4 8.0 8.4 4.3 5.1 1.9 7.6 8.6 16.3 17.3 0.359 53.1 48.1 35.6 41.8 11.3 10.1 0.3 [0–1.2] 0.1 [0–1.5] 0.106 13.5 19.8 0.048 2.8 4.9 0.213 7.9 8.0 > 0.999 52.2 ± 11.0 55.8 ± 12.8 0.001 6.5 [3.0–14.0] 11.0 [4.5–17.1] < 0.001 0.223 23.6 21.3 55.2 50.0 3.8 4.4 17.4 24.4 6.9 [6.3–7.8] 6.8 [6.3–7.7] 0.347 52 [45–62] 51 [45–61] 0.347 7.4 [6.4–9.2] 7.5 [6.2–9.0] 0.199 0.081 ± 0.005 0.082 ± 0.005 < 0.001 32.1 ± 6.2 31.3 ± 6.5 0.138 70 ± 11 70 ± 12 0.391 142 ± 20 145 ± 20 0.032 81 ± 12 79 ± 12 0.016 67.5 80.2 0.002 33.2 37.0 0.357 31.1 44.4 0.002 15.0 27.2 < 0.001 20.7 32.1 0.002 N CAN group (%) None Possible Definite MCR (n = 830) E:I ratio (n = 830) Valsalva ratio (n = 809) 30:15 Stand ratio (n = 829) Orthostatic hypotension (%) Age (years) Male (%) Education beyond primary level (%) Not fluent in English (%) Ethnic background (%) Anglo-Celt Southern European Other European Asian Aboriginal/TSI Mixed/other Smoking status (%) Never Ex- Current Alcohol consumption (standard drinksa/day) Antidepressant use (%) TCAs SSRIs Age at diabetes diagnosis (years) Diabetes duration (years) Diabetes treatment (%) Diet/exercise Oral agents ± non-insulin injectables Insulin alone Insulin ± oral agents ± non-insulin injectables HbA1c (%) HbA1c (mmol/mol) Fasting plasma glucose (mmol/L) ABSI (m11/6 kg−2/3) BMI (kg/m2) Pulse rate (bpm) Supine SBP (mmHg) Supine DBP (mmHg) On BP-lowering medication (%) ACE-I ARB Beta-blockers Calcium channel blockers Davis et al. Discussionh Cardiovascular Diabetology (2024) 23:102 Page 13 of 16 a 1 standard drink = 10 U ethanol Table 6 (continued) Alive Deceased P-value Diuretics 23.8 34.6 0.007 Total serum cholesterol (mmol/L) 4.3 ± 1.0 4.3 ± 1.5 0.840 HDL-cholesterol (mmol/L) 1.19 ± 0.30 1.19 ± 0.34 0.806 Serum triglycerides (mmol/L) 1.5 (0.9–2.6) 1.5 (0.9–2.6) 0.958 On lipid-lowering medication (%) 67.2 71.0 0.399 Urinary albumin:creatinine ratio (mg/mmol) 2.5 (0.7–9.3) 3.9 (1.0–15.2) < 0.001 eGFR (mL/min/1.73m2) eGFR (CKD-EPI) stages (ml/min/1.73m2) < 0.001 ≥ 90 46.8 25.9 60–89 45.2 47.5 45–59 5.3 13.6 30–44 2.1 7.4 < 30 0.6 5.6 Distal symmetrical polyneuropathy (%) 29.8 44.1 < 0.001 Peripheral arterial disease (%) 17.7 30.9 < 0.001 Prior hospitalisation for IHD (%) 16.8 35.8 < 0.001 Prior hospitalisation for cerebrovascular disease (%) 3.4 11.7 < 0.001 Prior hospitalisation for HF (%) 2.5 10.5 < 0.001 may comprise those who are at high cardiovascular risk or who have established CVD [8, 33]. Nevertheless, in the Detection of Ischemia in Asymptomatic Diabetics (DIAD) study involving participants with type 2 diabetes without known heart disease [39], the incidence of the composite clinical outcome of cardiac death, acute coro- nary syndromes, HF, or coronary revascularization over 5 years was significantly increased in those in the lowest quartile of the Valsalva heart rate ratio (hazard ratio 1.60) amongst a range of tests of autonomic heart rate/blood pressure responses and power spectral analysis of HRV. The results of these studies should be interpreted against the heterogeneity in meta-analysis of prognostic stud- ies of CAN [7] which highlights the influence of sample selection and CAN assessment methods. In addition, the duration of follow-up may be an important consideration since there is evidence that CART indices of autonomic dysfunction attenuate over time [4]. Our mean 10-year follow-up may have captured the effect of this on incident IHD and HF events compared to the ≤ 5 year follow-up in studies such as ACCORD and DIAD [8, 33, 39]. independently of cardiac or all-cause mortality and other clinical covariates [41]. It is a nonspecific pre- dictor of mortality which reflects central-autonomic moment-to-moment adaption of somatic responses and emotional appraisal to maintain homeostasis and adapt to environmental stimuli [42]. In the context of diabe- tes, detection of a low HRV through measures such as MCR should prompt consideration of improved life- style factors including exercise [43]. Discussionh MCR is one of the more robust CARTS as it is not influenced by changes in heart rate and presence of extrasystoles [40]. HRV has been shown to be a strong predictor of death in general population studies Page 14 of 16 Page 14 of 16 Davis et al. Cardiovascular Diabetology (2024) 23:102 Table 7 Most parsimonious Cox models of independent predictors of all-cause death in the FDS2 type 2 diabetes cohort with CAN category defined (n = 830) and valid CARTs variables as continuous variables added and retained if statistically significant (up to n = 1101) all-cause mortality. There was evidence that this was mediated through a reduced MCR which has also been found to be a nonspecific adverse prognostic indica- tor in general population studies. Although the clinical value of routine assessment of CAN in type 2 diabetes is questionable as a result of our findings, the pres- ence of CAN should still be established where typical symptoms (including light-headedness, weakness, pal- pitations and syncope on standing) or other features of autonomic neuropathy such as gastroparesis are pre- sent. The results could guide use of fludrocortisone and midodrine, and help tailor use of established therapies for CVD and glycaemic control [2]. In resource-lim- ited settings, or where there are physical impairments to full CART testing, single diagnostic tests could be employed such as heart rate variation on deep breath- ing [6] or analysis of ten-second resting electrocardio- graphic tracings [8]. Author contributions TMED, the Principal Investigator of the FDS2, conceived the study, provided clinical interpretation and produced the final version of the manuscript. ET extracted CART data, participated in data analysis and produced the first draft of the manuscript. WAD, a Co-Investigator of the FDS2, co-ordinated statistical advice and edited the manuscript. The present study has provided no evidence that either possible or definite CAN assessed by the range of rec- ommended CARTs is an independent predictor of inci- dent IHD or HF during relatively long-term follow-up in community-based people with type 2 diabetes. Pos- sible and especially definite CAN were associated with Acknowledgements The authors wish to thank the FDS2 staff, investigators and participants, the staff at the West Australian Data Linkage Branch, the Hospital Morbidity Data Collection, and the Registry for Births, Deaths and Marriages. Discussionh Hazard ratios (HR) and 95% confidence intervals (95% CI) are shown a A 2.72-fold increase in urinary albumin:creatinine ratio corresponds to an increased risk of 1 in ln (urinary albumin:creatinine) HR (95% CI) P-value CAN category defined and valid CART results All-cause mortality (n = 827) Age (increase of 1 year) 1.06 (1.04, 1.08) < 0.001 Male sex 1.55 (1.12, 2.15) 0.008 Aboriginal descent 8.95 (3.26, 24.6) < 0.001 Other European background 0.41 (0.19, 0.90) 0.026 Antidepressant therapy 3.40 (1.44, 8.00) 0.005 Angiotensin receptor blocker use 1.38 (1.001, 1.89) 0.049 eGFR < 30 mL/min/1.73m2 3.03 (1.51, 6.06) 0.002 Distal symmetrical polyneuropathy 1.15 (0.96, 1.39) 0.130 Cerebrovascular disease 1.70 (1.04, 2.77) 0.033 HF 2.31 (1.36, 3.93) 0.002 MCR (increase of 1.0) 0.98 (0.97, 0.99) 0.001 Time-varying covariates Aboriginal descent 0.43 (0.24, 0.76) 0.004 Antidepressant therapy 0.57 (0.38, 0.98) 0.043 Participants of all ages and valid CART results All-cause mortality (n = 1091) Age (increase of 1 year) 1.02 (0.998, 1.05) 0.070 Male sex 1.45 (1.13, 1.87) 0.004 Aboriginal descent 2.61 (1.58, 4.31) < 0.001 Antidepressant therapy 1.47 (1.08, 2.01) 0.016 Ln(uACR)a 1.10 (1.004, 1.21) 0.042 eGFR < 30 mL/min/1.73m2 3.20 (1.88, 5.44) < 0.001 Distal symmetrical polyneuropathy 1.31 (1.09, 1.57) 0.003 Peripheral arterial disease 1.42 (1.09, 1.85) 0.010 HF 1.67 (1.11, 2.50) 0.014 MCR (increase of 1.0) 0.99 (0.98, 0.998) 0.021 Time-varying covariates Age (increase of 1 year) 1.03 (1.02, 1.05) < 0.001 Abbreviations ABSI A body shape index ACCORD Action to Control Cardiovascular Risk in Diabetes BMI Body mass index CAN Cardiac autonomic neuropathy CVD Cardiovascular disease CART Cardiovascular reflex test CHD Coronary heart disease DIAD Detection of Ischemia in Asymptomatic Diabetics DSPN Distal symmetrical polyneuropathy eGFR Estimated glomerular filtration rate E:I Expiration to inspiration ratio FDS2 Fremantle Diabetes Study Phase II HF Heart failure HMDC Hospital Morbidity Data Collection IHD Ischaemic heart disease MCR Mean circular resultant WA Western Australia WADLS WA Data Linkage System Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12933-024-02185-3. The online version contains supplementary material available at https://doi. org/10.1186/s12933-024-02185-3. Additional file 1: Table S1. Baseline characteristics of FDS2 type 2 diabe- tes participants attending their first ANS examination between May 2009 and November 2012 categorised by eligibility (valid CART data and aged 20 to 80 years old). its relatively large, community-based sample with rich phenotypic data, the use of gold standard CAN tests, and long follow-up for outcomes of interest. References Nutr Metab Cardiovasc Dis. 2011;21(1):69–78. 5. Pop-Busui R, Boulton AJ, Feldman EL, Bril V, Freeman R, Malik RA, Sosenko JM, Ziegler D. Diabetic neuropathy: a position statement by the American Diabetes Association. Diabetes Care. 2017;40(1):136–54. 5. Pop-Busui R, Boulton AJ, Feldman EL, Bril V, Freeman R, Malik RA, Sosenko JM, Ziegler D. Diabetic neuropathy: a position statement by the American Diabetes Association. Diabetes Care. 2017;40(1):136–54. 26. Levey A, Bosch J, Lewis J, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130(6):461–70. 6. Stranieri A, Abawajy J, Kelarev A, Huda S, Chowdhury M, Jelinek HF. An approach for Ewing test selection to support the clinical assessment of cardiac autonomic neuropathy. Artif Intell Med. 2013;58(3):185–93. 27. Davis WA, Hamilton E, Davis TME. Temporal trends in distal symmetric polyneuropathy in type 2 diabetes: the Fremantle Diabetes Study. J Clin Endocrinol Metab. 2023;109:e1083–94. 7. Chowdhury M, Nevitt S, Eleftheriadou A, Kanagala P, Esa H, Cuthbert- son DJ, Tahrani A, Alam U. Cardiac autonomic neuropathy and risk of cardiovascular disease and mortality in type 1 and type 2 diabetes: a meta-analysis. BMJ Open Diabetes Res Care. 2021;9(2): e002480. 28. Holman C, Bass A, Rouse I, Hobbs M. Population-based linkage of health records in Western Australia: development of a health services research linked database. Aust N Z J Public Health. 1999;23(5):453–9. y 8. Pop-Busui R, Evans GW, Gerstein HC, Fonseca V, Fleg JL, Hoogwerf BJ, Genuth S, Grimm RH, Corson MA, Prineas R, et al. Effects of cardiac auto- nomic dysfunction on mortality risk in the Action to Control Cardiovascu- lar Risk in Diabetes (ACCORD) trial. Diabetes Care. 2010;33(7):1578–84. 9. Fang M, Wang D, Coresh J, Selvin E. Trends in Diabetes Treatment and Control in U.S. Adults, 1999–2018. N Engl J Med. 2021;384(23):2219–28. 10. Ford ES. Trends in the control of risk factors for cardiovascular disease among adults with diagnosed diabetes: findings from the National 8. Pop-Busui R, Evans GW, Gerstein HC, Fonseca V, Fleg JL, Hoogwerf BJ, Genuth S, Grimm RH, Corson MA, Prineas R, et al. Effects of cardiac auto- nomic dysfunction on mortality risk in the Action to Control Cardiovascu- lar Risk in Diabetes (ACCORD) trial. Diabetes Care. 2010;33(7):1578–84. 29. Vinik A, Ziegler D. Diabetic cardiovascular autonomic neuropathy. Circula- tion. 2007;115(3):387–97. 30. References 1. Spallone V, Ziegler D, Freeman R, Bernardi L, Frontoni S, Pop-Busui R, Stevens M, Kempler P, Hilsted J, Tesfaye S, et al. Cardiovascular autonomic neuropathy in diabetes: clinical impact, assessment, diagnosis, and man- agement. Diabetes Metab Res Rev. 2011;27:639–53. 1. Spallone V, Ziegler D, Freeman R, Bernardi L, Frontoni S, Pop-Busui R, Stevens M, Kempler P, Hilsted J, Tesfaye S, et al. Cardiovascular autonomic neuropathy in diabetes: clinical impact, assessment, diagnosis, and man- agement. Diabetes Metab Res Rev. 2011;27:639–53. 22. Ziegler D, Laux G, Dannehl K, Spuler M, Muhlen H, Mayer P, Gries F. Assess- ment of cardiovascular autonomic function: age-related normal ranges and reproducibility of spectral analysis, vector analysis, and standard tests of heart rate variation and blood pressure responses. Diabetic Med. 1992;9(2):166–75. 2. Williams S, Raheim SA, Khan MI, Rubab U, Kanagala P, Zhao SS, Marshall A, Brown E, Alam U. Cardiac autonomic neuropathy in type 1 and 2 diabetes: epidemiology, pathophysiology, and management. Clin Ther. 2022;44(10):1394–416. 2. Williams S, Raheim SA, Khan MI, Rubab U, Kanagala P, Zhao SS, Marshall A, Brown E, Alam U. Cardiac autonomic neuropathy in type 1 and 2 diabetes: epidemiology, pathophysiology, and management. Clin Ther. 2022;44(10):1394–416. 23. Gelber DA, Pfeifer M, Dawson B, Schumer M. Cardiovascular autonomic nervous system tests: determination of normative values and effect of confounding variables. J Auton Nerv Syst. 1997;62(1–2):40–4. 3. Fisher VL, Tahrani AA. Cardiac autonomic neuropathy in patients with diabetes mellitus: current perspectives. Diabetes Metab Syndr Obes. 2017;10:419–34. 3. Fisher VL, Tahrani AA. Cardiac autonomic neuropathy in patients with diabetes mellitus: current perspectives. Diabetes Metab Syndr Obes. 2017;10:419–34. 24. Risk M, Bril V, Broadbridge C, Cohen A. Heart rate variability measurement in diabetic neuropathy: review of methods. Diabetes Technol Ther. 2001; 3(1):63–76. 4. Andersen ST, Witte DR, Fleischer J, Andersen H, Lauritzen T, Jorgensen ME, Jensen TS, Pop-Busui R, Charles M. Risk factors for the presence and progression of cardiovascular autonomic neuropathy in type 2 diabetes: ADDITION-Denmark. Diabetes Care. 2018;41(12):2586–94. 4. Andersen ST, Witte DR, Fleischer J, Andersen H, Lauritzen T, Jorgensen ME, Jensen TS, Pop-Busui R, Charles M. Risk factors for the presence and progression of cardiovascular autonomic neuropathy in type 2 diabetes: ADDITION-Denmark. Diabetes Care. 2018;41(12):2586–94. 25. Spallone V, Bellavere F, Scionti L, Maule S, Quadri R, Bax G, Melga P, Viviani G, Esposito K, Morganti R, et al. Recommendations for the use of cardio- vascular tests in diagnosing diabetic autonomic neuropathy. Funding Th The present study was funded by the National Health and Medical Research Council of Australia (project grants 513781 and 1042231). TMED is supported Page 15 of 16 Davis et al. Cardiovascular Diabetology (2024) 23:102 Davis et al. Cardiovascular Diabetology (2024) 23:102 Davis et al. Cardiovascular Diabetology (2024) 23:102 by a Medical Research Future Fund Practitioner Fellowship. The present analyses were supported by Australian Centre for Accelerating Diabetes Innovations. The funding bodies had no involvement in the study design, data collection, analysis and interpretation of results or writing this manuscript. by a Medical Research Future Fund Practitioner Fellowship. The present analyses were supported by Australian Centre for Accelerating Diabetes Innovations. The funding bodies had no involvement in the study design, data collection, analysis and interpretation of results or writing this manuscript. Health and Nutrition Examination Survey 1999–2008. J Diabetes. 2011;3(4):337–47. Health and Nutrition Examination Survey 1999–2008*. J Diabetes. 2011;3(4):337–47. 11. Davis WA, Gregg EW, Davis TME. Temporal trends in cardiovascular complications in people with or without type 2 diabetes: the Fremantle Diabetes Study. J Clin Endocrinol Metab. 2020;105(7):e2471–82. Availability of data and materials 12. Gregg EW, Cheng YJ, Srinivasan M, Lin J, Geiss LS, Albright AL, Imperatore G. Trends in cause-specific mortality among adults with and without diagnosed diabetes in the USA: an epidemiological analysis of linked national survey and vital statistics data. Lancet. 2018;391(10138):2430–40. Some outcome data supporting the findings of this study are available from the Western Australian Department of Health, but restrictions apply to the availability of these data, which were used under strict conditions of confiden- tiality for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of Western Australian Department of Health. 13. Gregg EW, Li Y, Wang J, Burrows NR, Ali MK, Rolka D, Williams DE, Geiss L. Changes in diabetes-related complications in the United States, 1990–2010. N Engl J Med. 2014;370(16):1514–23. 14. Harding JL, Pavkov ME, Magliano DJ, Shaw JE, Gregg EW. Global trends in diabetes complications: a review of current evidence. Diabetologia. 2019;62(1):3–16. Author details 1 Medical School, Fremantle Hospital, University of Western Australia, PO Box 480, Fremantle, WA 6959, Australia. 2 Department of Endocrinology and Diabetes, Fiona Stanley and Fremantle Hospitals, Murdoch, WA, Australia. 3 Australian Centre for Accelerating Diabetes Innovations (ACADI), The Univer- sity of Melbourne, Melbourne, VIC, Australia. 19. Pop-Busui R, Backlund JC, Bebu I, Braffett BH, Lorenzi G, White NH, Lachin JM, Soliman EZ, Group DER. Utility of using electrocardiogram measures of heart rate variability as a measure of cardiovascular autonomic neurop- athy in type 1 diabetes patients. J Diabetes Investig. 2022;13(1):125–33. athy in type 1 diabetes patients. J Diabetes Investig. 2022;13(1):125– 20. Pop-Busui R, Low PA, Waberski BH, Martin CL, Albers JW, Feldman EL, Sommer C, Cleary PA, Lachin JM, Herman WH, et al. Effects of prior intensive insulin therapy on cardiac autonomic nervous system function in type 1 diabetes mellitus: the Diabetes Control and Complications Trial/ Epidemiology of Diabetes Interventions and Complications study (DCCT/ EDIC). Circulation. 2009;119(22):2886–93. Received: 21 December 2023 Accepted: 1 March 2024 Received: 21 December 2023 Accepted: 1 March 2024 21. Ewing D, Campbell I, Clarke B. Assessment of cardiovascular effects in diabetic autonomic neuropathy and prognostic implications. Ann Intern Med. 1980;92(2_Part_2):308–11. Ethics approval and consent to participate The FDS2 protocol by the Human Research Ethics Committee of the Southern Metropolitan Area Health Service (reference 07/397). All participants gave written informed consent before study entry. 16. Socio-economic indexes for areas. http://www.abs.gov.au/websitedbs/ censushome.nsf/home/seifa. 17. Krakauer NY, Krakauer JC. Anthropometrics, metabolic syndrome, and mortality hazard. J Obes. 2018;2018:9241904. Declarations 15. Davis T, Bruce D, Davis W. Cohort profile: the Fremantle Diabetes Study. Int J Epidemiol. 2013;42(2):412–21. 15. Davis T, Bruce D, Davis W. Cohort profile: the Fremantle Diabetes Study. Int J Epidemiol. 2013;42(2):412–21. The authors declare that they have no competing interests. The authors declare that they have no competing interests. Competing interests 18. Freeman R, Wieling W, Axelrod F, Benditt D, Benarroch E, Biaggioni I, Cheshire W, Chelimsky T, Cortelli P, Gibbons C, et al. Consensus statement on the definition of orthostatic hypotension, neurally mediated syncope and the postural tachycardia syndrome. Clin Auton Res. 2011;21(2):69–72. References Hu MX, Milaneschi Y, Lamers F, Nolte IM, Snieder H, Dolan CV, Penninx B, de Geus EJC. The association of depression and anxiety with cardiac autonomic activity: the role of confounding effects of antidepressants. Depress Anxiety. 2019;36(12):1163–72. 9. Fang M, Wang D, Coresh J, Selvin E. Trends in Diabetes Treatment and Control in U.S. Adults, 1999–2018. N Engl J Med. 2021;384(23):2219–28. 10. Ford ES. Trends in the control of risk factors for cardiovascular disease among adults with diagnosed diabetes: findings from the National 31. Pop-Busui R, Braffett BH, Zinman B, Martin C, White NH, Herman WH, Genuth S, Gubitosi-Klug R, Group DER. Cardiovascular autonomic Page 16 of 16 Davis et al. Cardiovascular Diabetology (2024) 23:102 Davis et al. Cardiovascular Diabetology (2024) 23:102 neuropathy and cardiovascular outcomes in the diabetes control and complications trial/epidemiology of diabetes interventions and compli- cations (DCCT/EDIC) study. Diabetes Care. 2017;40(1):94–100. neuropathy and cardiovascular outcomes in the diabetes control and complications trial/epidemiology of diabetes interventions and compli- cations (DCCT/EDIC) study. Diabetes Care. 2017;40(1):94–100. 32. Young LH, Wackers FJ, Chyun DA, Davey JA, Barrett EJ, Taillefer R, Heller GV, Iskandrian AE, Wittlin SD, Filipchuk N, et al. Cardiac outcomes after screening for asymptomatic coronary artery disease in patients with type 2 diabetes: the DIAD study: a randomized controlled trial. JAMA. 2009;301(15):1547–55. 33. Kaze AD, Yuyun MF, Erqou S, Fonarow GC, Echouffo-Tcheugui JB. Cardiac autonomic neuropathy and risk of incident heart failure among adults with type 2 diabetes. Eur J Heart Fail. 2022;24(4):634–41. 34. Woods JA, Katzenellenbogen JM, Davidson PM, Thompson SC. Heart fail- ure among Indigenous Australians: a systematic review. BMC Cardiovasc Disord. 2012;12:99. 35. Zhou L, Deng W, Zhou L, Fang P, He D, Zhang W, Liu K, Hu R. Prevalence, incidence and risk factors of chronic heart failure in the type 2 diabetic population: systematic review. Curr Diabetes Rev. 2009;5(3):171–84. 36. Davis TM, McAullay D, Davis WA, Bruce DG. Characteristics and outcome of type 2 diabetes in urban Aboriginal people: the Fremantle Diabetes Study. Intern Med J. 2007;37(1):59–63. y 37. Raghavan S, Vassy JL, Ho YL, Song RJ, Gagnon DR, Cho K, Wilson PWF, Phil- lips LS. Diabetes mellitus-related all-cause and cardiovascular mortality in a national cohort of adults. J Am Heart Assoc. 2019;8(4): e011295. 38. van Dooren FE, Nefs G, Schram MT, Verhey FR, Denollet J, Pouwer F. Depression and risk of mortality in people with diabetes mellitus: a systematic review and meta-analysis. PLoS ONE. References 2013;8(3): e57058. 39. Chyun DA, Wackers FJ, Inzucchi SE, Jose P, Weiss C, Davey JA, Heller GV, Iskandrian AE, Young LH, Investigators D. Autonomic dysfunction independently predicts poor cardiovascular outcomes in asymptomatic individuals with type 2 diabetes in the DIAD study. SAGE Open Med. 2015;3:2050312114568476. 40. Weinberg CR, Pfeifer MA. An improved method for measuring heart- rate variability: assessment of cardiac autonomic function. Biometrics. 1984;40(3):855–61. 41. Jarczok MN, Weimer K, Braun C, Williams DP, Thayer JF, Gundel HO, Balint EM. Heart rate variability in the prediction of mortality: a systematic review and meta-analysis of healthy and patient populations. Neurosci Biobehav Rev. 2022;143: 104907. 42. Thayer JF, Ahs F, Fredrikson M, Sollers JJ 3rd, Wager TD. A meta-analysis of heart rate variability and neuroimaging studies: implications for heart rate variability as a marker of stress and health. Neurosci Biobehav Rev. 2012;36(2):747–56. 43. Hamasaki H. The effect of exercise on cardiovascular autonomic nervous function in patients with diabetes: a systematic review. Healthcare (Basel). 2023;11(19):2668. 43. Hamasaki H. The effect of exercise on cardiovascular autonomic nervous function in patients with diabetes: a systematic review. Healthcare (Basel). 2023;11(19):2668. 44. Soedamah-Muthu S, Chaturvedi N, Witte D, Stevens L, Porta M, Fuller J. Relationship between risk factors and mortality in type 1 diabetic patients in Europe: the EURODIAB Prospective Complications Study (PCS). Diabetes Care. 2008;31(7):1360–6. 44. Soedamah-Muthu S, Chaturvedi N, Witte D, Stevens L, Porta M, Fuller J. Relationship between risk factors and mortality in type 1 diabetic patients in Europe: the EURODIAB Prospective Complications Study (PCS). Diabetes Care. 2008;31(7):1360–6. 45. Low PA, Denq JC, Opfer-Gehrking TL, Dyck PJ, O’Brien PC, Slezak JM. Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve. 1997;20(12):1561–8. 46. Gauci R, Hunter M, Bruce DG, Davis WA, Davis TME. Anemia complicat- ing type 2 diabetes: prevalence, risk factors and prognosis. J Diabetes Complications. 2017;31(7):1169–74. 47. Liu W, Wang L, Huang X, He W, Song Z, Yang J. Impaired orthostatic blood pressure stabilization and reduced hemoglobin in chronic kidney disease. J Clin Hypertens (Greenwich). 2019;21(9):1317–24. Publisher’s Note S N Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.
https://openalex.org/W4313423902	https://www.researchsquare.com/article/rs-1961964/latest.pdf	English	null	Design and analysis of polycentric prosthetic knee with enhanced kinematics and stability	Physical and Engineering Sciences in Medicine/Physical and engineering sciences in medicine	2,023	cc-by	7,920	Design and analysis of polycentric prosthetic knee with enhanced kinematics and stability ajesh Kumar Mohanty (  rajeshmpo48@gmail.com ) Research Article Keywords: Analysis, design, kinematics, knee, polycentric, simulation Posted Date: August 24th, 2022 DOI: https://doi.org/10.21203/rs.3.rs-1961964/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Abstract This paper describes a continuation of earlier work using the finite element method to conduct an engineering failure analysis of existing polycentric prosthetic knee. The primary purpose of this work is to enhance the quality of the existing knee which has been reported with multiple cases of failure during its clinical practice in India. A modified design of the polycentric knee has been proposed based on the findings of failure analysis. Simulation-based comparative analysis of polycentric knees has been performed as per the ISO 10328:2016 standard in terms of stress distribution, total contour deformation, safety factor, and fatigue life. The upper extension lever is subjected to static and cyclic loads of 4130 and 1230 N, whereas lower plate has translational constraint. The modified polycentric knee prosthesis outperforms static and fatigue strength tests. The standard of the existing knee prosthesis has significantly improved as a result of design variations and integration of high-strength and lightweight aluminium 7075- T6 alloy. The modified polycentric knee prosthesis has a predicted maximum deformation of less than 0.7 mm and a minimum safety factor between 1.7 and 2 compared to 2.66 mm and 1.0 for the existing knee prosthesis. Based on the fatigue simulation results, it is predicted that the modified polycentric knee will have a lifespan of at least ten years indicating a safe design. It has improved alignment stability and kinematics, with a significant weight reduction of 33 g, and a high cost-benefit ratio to reach the maximum amputee population in low-income countries like India. Keywords: Analysis, design, kinematics, knee, polycentric, simulation 1. Introduction According to statistics reported by WHO, there are around 40 million amputees in developing countries. The Census 2011 estimates around 2.68 crore persons with disabilities in India, which is 2.22% of the entire population [1]. A comprehensive prosthetic rehabilitation program results in full independence and desirable functional ambulation [2]. Even though amputation is one of the primary causes of disability in India, only about 5% of the amputee population has access to 1 some sort of prosthetic technology. Further, the need and expectations of Indian amputees for prosthetic devices are not properly fulfilled in most cases which ultimately lead to dissatisfaction among this population. The majority of amputees complain about the durability and reliability of products and materials used in the manufacture of prostheses. There are extremely rare indigenously developed products or genuine inventions in the field of prosthetic knees in India and most of the above products are being imported from the global market. Most amputees in developing countries can't bear the huge cost of knee prostheses due to poor economic conditions. The motivation for the present investigation lies in the design and development of a functionally efficient indigenous knee prosthesis that should possess enhanced stability and durability, at the same time, should be cost-effective to reach maximum individuals with transfemoral amputation. Earlier literature has reported more concentration on new prosthetic knee designs and their simulation studies [3-12]. However, few have examined prosthetic knee strength according to ISO standard [8, 13, 14]. Several 4-bar polycentric knee prostheses have been manufactured utilizing kinematic approaches [15, 16], but few studies have investigated their structural designs using finite element analysis (FEA) [5, 6, 14, 16] and its optimization methods [16]. Investigators have shown interest in impact of static loading on stress distribution, total deformation, and safety factor of knee prostheses. Few researches have examined the material fatigue characteristics by applying cyclic loads to knee prostheses to estimate its life. Prosthetic knee failure analysis based on clinical observations utilizing ISO standard is understudied [17]. In our review, the modular prostheses using polycentric 4-bar knees, developed, and utilized in National Institutes in India often require repair or replacement after 1-2 years of usage (Fig. 1). Contour failure and complete breakage of front joint bars are shown in Figs. 1 (a) and (b) respectively. Fracture of knee joint unit lower can be seen in Fig. 1 (c). 1. Introduction Coronal and sagittal mal-alignment of the knee unit is depicted in Figs. 1 (d) and (e) respectively. Fig. 1.6 (f) shows complete detachment of knee unit with failure of back joint bar assembly. Breakage of the adapter of the knee joint unit lower can be observed in Fig. 1 (g). Failure of knee joint unit upper with deformation of front joint bars is shown in Fig. 1 (h). Such clinical findings may suggest fatigue failure of polycentric knee structural components due to cyclic stress, which occurs at substantially lower stress than typical static mechanical load. Furthermore, neither the suggest fatigue failure of polycentric knee structural components due to cyclic stress, which occurs at substantially lower stress than typical static mechanical load. Furthermore, neither the 2 ISO standard test nor a performance study has been performed on this polycentric knee for its compliance. Fig. 1 Clinical failure cases of knee prosthesis (a) Gross deformation and failure of front joint bars; (b) Complete breakage of front joint bar; (c) Breakage of knee joint unit lower; (d) Gross failure in knee joint assembly with mal- alignment (e) Hyperextension mal-alignment of knee unit; (f) Complete detachment knee unit; (g) Breakage of knee joint unit lower adapter (h) Failure in knee joint unit upper with enlargement of front joint bars. Fig. 1 Clinical failure cases of knee prosthesis (a) Gross deformation and failure of front joint bars; (b) Complete breakage of front joint bar; (c) Breakage of knee joint unit lower; (d) Gross failure in knee joint assembly with mal- alignment (e) Hyperextension mal-alignment of knee unit; (f) Complete detachment knee unit; (g) Breakage of knee joint unit lower adapter (h) Failure in knee joint unit upper with enlargement of front joint bars. This work discusses a continuation of earlier work performed by same authors Mohanty et al. [18] on engineering failure analysis of existing knee prosthesis. It has been observed that the existing knee is mildly strong enough to pass static strength test and the fatigue life fails below ISO 10328:2016 norm. The accuracy of the model has been validated. The mean absolute error between FE simulation and experiment is 10 percent indicating high reliability and validation results. The development of maximum deformation of 2.66 mm on front and back joint bars, mainly at connections and a safety factor of 1 is deemed critical from a design standpoint. 2. Methods This section reports the simulation of static and cyclic structural tests on polycentric knee prosthesis by FEA [19]. A modification in the existing design of polycentric knee prostheses through the incorporation of suitable materials with higher mechanical strength and altering the design of polycentric knee components has been suggested in this modified approach. The results of the modified polycentric knee prosthesis are compared with the existing knee in terms of structural stability, weight, and cost-benefit analyses. 1. Introduction Based on the poor fatigue life results of some common failure components, the existing polycentric knee is predicted to last 1-1.5 years. This expected result is in agreement with the reported clinical cases of failure in prosthetic knee joints which need repair in a few cases or a complete replacement in major cases. This failure analysis has characterized the mechanical and fatigue properties of different knee components for improving the quality of existing knee prostheses. This work discusses a continuation of earlier work performed by same authors Mohanty et al. [18] on engineering failure analysis of existing knee prosthesis. It has been observed that the existing knee is mildly strong enough to pass static strength test and the fatigue life fails below ISO 10328:2016 norm. The accuracy of the model has been validated. The mean absolute error between FE simulation and experiment is 10 percent indicating high reliability and validation results. The development of maximum deformation of 2.66 mm on front and back joint bars, mainly at connections and a safety factor of 1 is deemed critical from a design standpoint. Based on the poor fatigue life results of some common failure components, the existing polycentric knee is predicted to last 1-1.5 years. This expected result is in agreement with the reported clinical cases of failure in prosthetic knee joints which need repair in a few cases or a complete replacement in major cases. This failure analysis has characterized the mechanical and fatigue properties of different knee components for improving the quality of existing knee prostheses. In continuation to our previous work, the novelty of this investigation has been to enhance the currently used polycentric knee through material and design changes to enhance the quality of 3 life for those with transfemoral amputation. This work will establish a good standard of the prosthetic knee joint in terms of function, cosmesis, and cost-effectiveness to reach maximum amputee populations in developing countries like India. 2.1 Overview of modified polycentric knee prosthesis The principal components in the existing and modified prosthetic knee assembly are same with some alterations in the geometry. Thus the modified polycentric knee prosthesis consists of a knee joint unit upper, knee joint unit lower, hyperextension stopper, front joint bars, back joint bar assembly, roller pin (bush) upper and lower, alignment coupling unit, and a knee cap as presented in Fig. 2. The total weight of the polycentric knee joint is found to be 499 g compared to 532 g for the existing one. Fig. 2 Components of existing and modified polycentric knee prosthesis Fig. 2 Components of existing and modified polycentric knee prosthesis 4 4 2.2 Materials used and properties Materials used for knee component are aluminium alloy (AA7075-T6), stainless steel (SS304), and high damping rubber whose mechanical properties are presented in Table 1. AA7075-T6 has been chosen as the base material for the modified polycentric knee prosthesis as it is lighter in weight and possesses significantly improved mechanical properties. In addition to its high strength, it possesses good toughness, fatigue, and corrosion resistance and is most regularly utilized for severely stressed applications [20]. T6 heat treatment is used to improve the mechanical characteristics by integrating thermal treatment and artificial ageing [21, 22]. Furthermore, AA7075 is used in lower-limb prosthetics with light weight components [23]. The chemical composition of AA7075 is presented in Table 2. e 1 Mechanical characteristics of materials employed in modified knee prosthesis Table 1 Mechanical characteristics of materials employed in modified knee prosthesis Component Material Density (kg/m3) Young’s modulus (GPa) Poisson’s ratio Yield stress (MPa) Ultimate stress (MPa) Front joint bars Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Back joint bar assembly Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Main body and adapter Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Allen screws Stainless steel (SS304) 8000 200 0.27 215 505 Hyperextensi on stopper High damping Rubber 1100 0.05 0.49 15 5 Roller pin (Bush) Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Table 1 Mechanical characteristics of materials employed in modified knee prosthesis Component Material Density (kg/m3) Young’s modulus (GPa) Poisson’s ratio Yield stress (MPa) Ultimate stress (MPa) Front joint bars Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Back joint bar assembly Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Main body and adapter Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Allen screws Stainless steel (SS304) 8000 200 0.27 215 505 Hyperextensi on stopper High damping Rubber 1100 0.05 0.49 15 5 Roller pin (Bush) Aluminium alloy (AA7075-T6) 2810 71.7 0.33 503 572 Table 2 Chemical composition of aluminium alloy (AA7075-T6) Element Zinc Magnesium Copper Iron Silicon Chromium Titanium Others Weight (%) 5.1-6.1 2.1-2.9 1.2-2.0 0.2-0.5 0.2-0.4 0.18-2.0 0.1-0.2 0.05-0.2 2 3 Generation of prosthetic knee model Table 2 Chemical composition of aluminium alloy (AA7075-T6) Table 2 Chemical composition of aluminium alloy (AA7075-T6) Element Zinc Magnesium Copper Iron Silicon Chromium Titanium Others Weight (%) 5.1-6.1 2.1-2.9 1.2-2.0 0.2-0.5 0.2-0.4 0.18-2.0 0.1-0.2 0.05-0.2 2.3 Generation of prosthetic knee model The dimensional information is used to create a 2D technical drawing. The design criteria involve giving strength to weaker components in critical areas and reducing base materials at high strength robust areas to achieve the light weight of the product. The 3D CAD model is created in CATIA V5. The components of 4-bar knee is designed and assembled independently. The major design changes incorporated in the modified polycentric knee are based on simulation-based failure analysis reports and reported cases of damage from clinical 5 observations as summarized in Table 3. Different transparent perspective views of the CAD model of the modified polycentric prosthetic knee have been shown in Fig. 3. model of the modified polycentric prosthetic knee have been shown in Fig. 3. Table 3 Major changes incorporated in the modified polycentric knee prosthesis Part Name of the Component Clinical Problem in the existing knee (Fig. 1) Simulation findings in existing knee Design and material changes in modified knee A Knee joint unit upper Overall poor strength and frictional properties Maximum von-Mises stress developed at the connection of joint bars approximates to the yield strength of AA390.0-T6, and the maximum deformation produced is more than 2 mm. 1. Reduction of materials in front, back, and side for weight reduction. 2. Change of material from A390.0-T6 to AA7075-T6. B Knee joint unit lower Failure or breakage in lower pylon receiving adapter and poor strength and material properties Maximum von-Mises stress developed at the connection of joint bars approximates to the yield strength of A390.0-T6 and it has a poor fatigue life 1. Adapter placement is reversed 2. Fillets have been added to protrusions for increasing strength 3. Thickness of the adapter has been increased 4. Chamfers are added 5. Material reduction at the side with holes for weight reduction 6. Change of material from A390.0-T6 to AA7075-T6. D Front joint bar Wear and tear or complete failure Developed von-Mises stress is more than the fatigue strength of A390.0- T5, maximum deformation developed is 2.36 mm, and it has a poor fatigue life 1. Shape is changed to the gentle double curve (S-shaped) 2. Thickness is reduced 3. Multiple holes are incorporated 4. Change of material from A390.0-T5 to AA7075-T6 E Back joint bar assembly Tear or scratch marks, Difficulty in dismantling of the component as both sides are screwing in the same direction. 2.3 Generation of prosthetic knee model Excessive stress concentration at the junction, maximum deformation produced is 2.66 mm and it has a poor fatigue life 1. Stress concentration is reduced by decreasing the angle and improving the kinematics of ICOR. 2. Created an elliptical hole for weight reduction 3. Change of material from AISI 316 to AA7075-T6 for weight reduction 4. Opposite side threading has been included F Roller pins or bush and Allen screws Plastic bush inside the main body lacks durability and fatigue failure of front joint bar due to huge difference in material properties - 1. Use high-grade aluminium AA7075-T6 for bush to improve fatigue damage 2. Conversion of V-head full thread to the square-head half thread of Allen screw. Table 3 Major changes incorporated in the modified polycentric knee prosthesis 1. Adapter placement is reversed 6 Fig. 3 Transparent perspective views of the modified polycentric knee (a) Antero-lateral and (b) Postero-lateral 2 4 Ki ti d d i l i Fig. 3 Transparent perspective views of the modified polycentric knee (a) Antero-lateral and (b) Postero-lateral Fig. 3 Transparent perspective views of the modified polycentric knee (a) Antero-lateral and (b) Postero-lateral 2 4 Ki i d d i l i 2.4 Kinematics and design analysis The kinematic evaluation of a 4-bar polycentric knee is based on Grashof’s law double rocker with length bar condition ( d c b a    ) as shown in Fig. 4(a), where a, b are the longest and shortest link respectively and c, d are the other links. The optimal dimensions of the polycentric knee mechanism is based on the data of the trajectory of the instantaneous center of rotation (ICOR) calculated from the kinematic analysis of the four-bar polycentric knee. This verifies the stability to set the ICOR behind the load line (x = 12 mm) as shown in Fig. 4(b). Fig. 4 (a) Double rocker mechanism in modified polycentric knee prosthesis (b) ICOR of the polycentric knee during extension (x = 12 mm) Fig. 4 (a) Double rocker mechanism in modified polycentric knee prosthesis (b) ICOR of the polycentric knee during extension (x = 12 mm) 7 7 In this scenario, the initial elevation of the ICOR is situated at less height from the center and the downward course of the instant center remains relatively elevated and within the zone of stability for the first few degrees of knee flexion. The instant center is approximately 100 mm above and 12 mm behind the vertical reference line (load line). The kinematic analysis of the modified polycentric knee has been performed in CATIA V5 as shown in Fig. 5. The digital mockup result has revealed smooth motion in the joint assembly without any interference. It is presumed to work well in real-time with true physical functions. Fig. 5 Digital mockup simulation of joint motion (kinematics) in CATIA V5 Fig. 5 Digital mockup simulation of joint motion (kinematics) in CATIA V5 2.5 The finite element analysis The 3D CAD model of the modified polycentric knee prosthesis is imported into ANSYS Workbench R19.2 for simulation. The mesh independence test, grid refinements and mesh quality are ensured as defined in our earlier study [18]. The h-method has been followed for numerical convergence. The FE model has been discretized into 875639 nodes and 602951 elements satisfying the convergence criteria. The elements are 3D tetrahedral in nature and mesh refinement has been performed at appropriate parts as shown in Fig. 6. The procedure 8 performed for conducting ISO 10328:2016 structural standard test for static and fatigue performance in the case of an existing prosthetic knee as discussed in our earlier study [24] is similar to the modified polycentric knee prosthesis. Fig. 6 Discretization of prosthetic knee model (a) Antero-lateral and (b) Postero-lateral Fig. 6 Discretization of prosthetic knee model (a) Antero-lateral and (b) Postero-lateral 3. Results The present investigation relies on static and cyclic structural simulations of modified 4- bar knee prosthesis. The results so obtained are compared with that of the existing polycentric knee prosthesis. Soderberg's theory for fatigue analysis has been used as the design criterion which is related to the yield stress. 3.1 Simulation results of static test 9 (a) Maximum and minimum deformation; Deformation under static load condition I (b) Antero-lateral (c) Postero-lateral Fig. 10 (a) Maximum and minimum deformation; Deformation under static loading condition II (b) Antero-lateral (c) Postero-lateral Fig. 10 (a) Maximum and minimum deformation; Deformation under static loading condition II (b) Antero-lateral (c) Postero-lateral 3.1 Simulation results of static test The static strength simulation findings of the modified knee prosthesis are shown in Figs. 7 and 8. It shows prosthetic knee stress under loading conditions I and II. Stresses in front and back joint bars are 85-107 and 42-64 MPa respectively for load condition I (Figs. 7a and 7b) and 81- 101 and 40-60 MPa for load condition II (Figs. 8a and 8b). The maximum von-Mises stress value of 300 MPa is observed at the front link bush for load conditions I (Fig. 7c) and 284 MPa for load conditions II (Fig. 8c). Comparing these results to corresponding material properties, the maximum stresses are sufficiently below the yield strength of AA7075-T6 aluminium alloy 9 (503 MPa) utilized for the front and back joint bars and bush. These results suggest that the modified knee prosthesis will successfully pass the ISO 10328:2016 static strength test. Fig. 7 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress under static load condition I istribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress under static load cond Fig. 7 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress under static load condition I Fig. 8 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress under static load condition II ion of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress under static load condition II Fig. 8 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress under static The static strength test results for the modified knee are shown in terms of total deformation in Figs. 9 and 10. Maximum deformation is observed at the load application point which is 1.44 and 1.36 mm for the load I and II respectively as shown in Figs. 9a and 10a. The alignment coupling unit of the knee prosthesis experiences the most deformation of 0.7 mm, suggesting the least amount of displacement. The probe points are marked at different parts of modified polycentric knee prosthesis ranging from 0.27 to 0.68 mm as depicted in Figs. 9b, 9c, 10b and 10c. A deformation of less than 2.5 mm during static testing suggests that this 10 modified prosthetic knee has sufficient strain-bearing ability to meet the structural norms, as recommended by the ISO 10328:2016 standard. Fig. 9 (a) Maximum and minimum deformation; Deformation under static load condition I (b) Antero-lateral (c) Postero-lateral Fig. 3.2 Simulation results of cyclic test The distribution of von-Mises stress, maximum deflection, number of cycles before failure, and safety factor of the modified polycentric prosthetic knee are investigated in this fatigue analysis. The results of von-Mises stress distribution during cyclic strength simulation are shown in Figs. 11 and 12. For both loading conditions, the von-Mises stress developed at the front and back 11 joint bars is 18 to 23 MPa as shown in Figs. 11a, 11b, 12a, and 12b. The maximum stresses of 91 and 71 MPa are developed in the knee joint unit lower for load conditions I and II respectively as depicted in Figs. 11c and 12c. Therefore, it can be predicted that the modified polycentric prosthetic knee prosthesis employed in this research can withstand cyclic stress considering the high fatigue strength of AA7075-T6. considering the high fatigue strength of AA7075-T6. Fig. 11 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition I Fig. 12 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition II Fig. 11 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition I Fig. 11 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition I Fig. 12 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition II Fig. 12 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition II Fig. 12 Distribution of stress (a) Antero-lateral (b) Postero-lateral; (c) Maximum stress at critical areas under cyclic load condition II The cyclic strength test results for the modified knee prosthesis are shown in terms of total deformation in Fig. 13. A maximum deformation has been observed at the point of load application which is 0.38 mm as shown in Fig. 13a. The alignment coupling unit of the knee prosthesis experiences the maximum deformation of 0.18 mm as depicted in Figs. 13b and 13c, suggesting the least amount of displacement. A deformation less than 2.5 mm during fatigue 12 testing suggests that this modified prosthetic knee has sufficient strain-bearing ability to meet the structural norms, as recommended by the ISO 10328:2016 standard. Fig. 13 (a) Maximum and minimum deformation; Deformation obtained under cyclic loading (b) Antero-lateral (c) Postero-lateral Fig. 3.2 Simulation results of cyclic test 13 (a) Maximum and minimum deformation; Deformation obtained under cyclic loading (b) Antero-lateral (c) Postero-lateral The maximum stress of 91 MPa is obtained from the simulation at the knee joint unit lower for load conditions I, which is below the fatigue strength of 150 MPa for AA7075-T6. So a minimum design safety factor of 1.7 is computed from the above stress values. This is also in good agreement with the simulation results. The safety factor for modified polycentric knee prosthesis including the critical areas is depicted in Fig 14. An earlier report has suggested that any engineering design with a fatigue safety factor above 1.25 is considered safe [25]. It is observed that the von-Mises stress developed in regions other than critical areas of the knee prosthesis shows stress magnitude below 23.21 MPa. This signifies that the modified knee has sufficient mechanical strength to endure cyclic loads. Fig. 14 Prediction of safety factors (a) Antero-lateral (b) Postero-lateral views; (c) Critical areas during cyclic loading Fig. 14 Prediction of safety factors (a) Antero-lateral (b) Postero-lateral views; (c) Critical areas during cyclic loading 13 The results of cyclic strength simulations for commonly failing components are summarized in Table 4. It is observed that the maximum developed stress is sufficiently below the fatigue strength of AA7075 T6 and the deformation produced is extremely minimal. There is a significant increase in fatigue life (no. of cycles) of most of the knee components. These values indicate that the modified polycentric knee prosthesis qualifies the requirements of ISO 10328:2016 standard. Table 4 Simulation results of common failure knee components Knee component Max. stress (MPa) Max. deflection (mm) Average no. of cycles ISO standard for the cyclic test (pass/fail) Front joint bars 15.01 0.11 240×106 Pass Back joint bar assembly 12.17 0.11 240×106 Pass Knee joint unit upper 7.14 0.14 240×106 Pass Knee joint unit lower 90.88 0.10 240×106 Pass 3 3 Comparison of simulation results between existing and modified knee able 4 Simulation results of common failure knee components 3.3 Comparison of simulation results between existing and modified knee The modified polycentric mechanical knee prosthesis surpasses static and fatigue tests. The design variation and adoption of stronger materials have substantially improved the strength of knee prostheses. The summarized simulation results showing the comparison between existing and modified knee prostheses are presented in Table 5. Table 5 Comparison of simulation results between existing and modified knee prostheses Parameters Structural test Load condition Existing polycentric knee Modified polycentric knee Max. von-Mises stress (MPa) Static Load I 602 300 Load II 569 284 Cyclic Load I 100 91 Load II 100 73 Max deformation (mm) Static - 2.66 0.69 Min safety factor Cyclic - 1 1.7 Min fatigue life (No. of cycles) Cyclic - 6 10 97 .1  6 10 240 ISO standard for the cyclic test (pass/fail) Fail Pass Comparison of simulation results between existing and modified knee prostheses 14 To estimate the fatigue life (number of cycles), the stress distribution data is utilized in the MSC fatigue module to plot the S-N curve as depicted in Fig. 15. A visual representation of the fatigue life of knee components under cyclic load circumstances in isometric views is presented in Fig. 16. The prosthetic fatigue life is indicated by the color legend bar. As illustrated in Fig. 16, the Allen screws made up of stainless steel are found to have an average fatigue life of 6 10 10 . On the other hand, it is observed that the average fatigue life of the overall prosthetic knee is 6 10 240 . Fig. 15 S-N curves for (a) AA7075-T6 (b) SS 304 and (c) Damping rubber Fig. 15 S-N curves for (a) AA7075-T6 (b) SS 304 and (c) Damping rubber Fig. 16 Prediction of fatigue life (a) Antero-lateral (b) Postero-lateral Fig. 16 Prediction of fatigue life (a) Antero-lateral (b) Postero-lateral 15 4. Discussion This research discusses a simulation approach to improve the standard of polycentric knee subjected to static and cyclic loading following recommendation of ISO 10328:2016. The proposed modified design of the polycentric knee has been compared with the existing one in terms of stress distribution, total contour deformation, safety factor, and fatigue life. According to the presented data, it is apparent that this modified polycentric knee has sufficient stability to withstand the static load, has outstanding life prediction for its components and has the potential to outlast the margin of cycles required by the ISO 10328:2016 standard. The final CAD version of the modified polycentric knee is depicted in Fig. 17. Fig. 17 Different views of the modified polycentric knee Fig. 17 Different views of the modified polycentric knee 4.1 Comparison and validation with the previous study The work of Rosas et al. [26] has been considered for validation. This study seeks to computationally investigate the mechanical fatigue of polycentric knee prostheses using the ISO 10328 standard involving fluctuating loads for both load conditions. SolidWorks, which requires a static analysis with the maximum load from each condition and S-N curves of each material, was employed for the investigation. The knee prosthesis was constructed using five links made of aluminium alloy (AA7075 T6) and four axes made of stainless steel (SS304). The findings of this study show that the prosthetic knee has good fatigue life for over 3 million 16 cycles. For steel components, the life was infinite, there was no damage, and the factor of safety was either 3.45 or 2.97, depending on the condition. Under condition I, the minimum life for aluminium parts was 73,262,632 cycles, the maximum damage was 4.095%, and the minimum factor of safety was 1.533. Under condition II, the minimum life for aluminium parts was 223,973,424 cycles, the damage was 1.322 percent, and the minimum factor of safety was 1.779. The polycentric knee was expected to last for a minimum of 5 years and 8 months’ time period. These validation findings support the FE model reliability with a 7.75% error. The same comparison is presented in Table 6. Table 6 Comparison of present work with Rosas et al. [26] Parameters Present work Rosas et al. [26] Error (%) Max. von-Mises stress (MPa) loading I 232.15 224.55 3.2 Max. von-Mises stress (MPa) loading II 184.45 205.79 11.5 Min. safety factor 1.713 1.533 10.5 Min. fatigue life (No. of cycles) 6 10 240 6 10 226  5.8 Average Error (%) 7.75 Table 6 Comparison of present work with Rosas et al. [26] 4.2 Weight comparison analysis The weight comparative analysis between the existing and modified polycentric knee prosthesis is presented in Table 7. Some components such as knee joint unit upper and lower, and back joint bar assembly are reduced in weight whereas alignment coupling unit, and hyperextension stopper are increased in weight. Moreover, components like front joint bars, knee cap, washers, screws, bolts and accessory components are same in weight. It is found that the overall saving in weight by adopting modified design is 33 g. Table 7 Comparative analysis between existing and modified polycentric knee prosthes weight in gram (g) Sl. No. Name of component Existing polycentric knee Modified polycentric knee 1 Knee joint unit upper 130 127 2 Knee joint unit lower 122 106 3 Front joint bars (2 Nos.) 26 26 4 Back joint bar assembly 60 22 5 Alignment coupling unit 94 109 6 Hyperextension stopper 6 15 mparative analysis between existing and modified polycentric knee prosthesis for ( ) Table 7 Comparative analysis between existing and modified polycentric knee weight in gram (g) 17 7 Knee cap 6 6 8 Washers (Flat and spring washer) 14 14 9 Screws Allen screws for joint bars (08) Allen screws for adapter (01) Screws for hyperextension stopper (02) Screws for alignment coupling (04) Screws for knee cap (03) 46 46 10 Bolt M10 (Long) 28 28 Total 532 g 499 g Accessory components 11 Top plate 70 70 12 Intermediate plate 36 36 13 Bolt M10 (Short) 16 16 Grand total 654 g 621 g 4.3 Cost comparison analysis This study develops a framework for constructing possibly low-cost, fully passive polycentric knee prosthesis suitable for transfemoral amputees' daily life activities in poor nations. Acute economic restraints, combined with socioeconomic factors, indicate an urgent need for a low- cost solution capable of delivering high levels of functional performance. Prosthetic knee joints cost several thousand dollars to produce and distribute in the United States and Europe. A popular active transfemoral prosthesis that works exceptionally well can cost more than $50,000 [27]. Even passive knee joints in wealthy nations are too costly to suit the needs of amputees in developing countries [28]. It is important to mention that, most of the polycentric mechanical knee prostheses in India are imported from the global market and the cost ranges from INR 6000 to 50,000. Table 8 presents a comparison of the cost of polycentric mechanical prosthetic knees from Indian manufacturers or suppliers. To determine the price of a potential product that can produce sufficient functional requirements, a full cost reduction analysis has been performed. AA7075 has been selected for this work as it is comparatively lower in cost, and has high strength and corrosion resistance. The suggested cost of the knee has been decreased by taking a more austere approach to design, such as using AA7075 alloy instead of contemporary composites or non-ferrous alloys to fulfill structural design goals without raising the cost of the system. The cost of fabricating the mechanical structure, including the cost of materials, is expected to be around 4,900 INR and the details are presented in Table 9. Though it is not so easy to understand the various aspects that influence the pricing of a product, still the following criteria have been taken into account while calculating the cost of the prosthetic knee model. 4.3 Cost comparison analysis Replacing the components with the use of less expensive polymers [30] or other variants of aluminium alloy 1. Replacing the components with the use of less expensive polymers [30] or other variants of aluminium alloy 2. Planning for commercialization that considers local manufacturing resources, design-for- manufacturing analysis, and design-for-assembly analysis 3. Utilizing the optimized criteria for “Designing for Machinability” 4. Simplifying the design further and using easier-to-machine material that can fulfil the design requirements 4.3 Cost comparison analysis 18 Table 8 Comparison of cost of polycentric mechanical prosthetic knees from Indian manufacturers or suppliers Table 8 Comparison of cost of polycentric mechanical prosthetic knees from Indian manufacturers or suppliers manufacturers or suppliers Type of knee Manufacturer/supplier in India Cost (INR) Cost (USD) Prosthetic multi-axis knee joint Proactive Technical Orthopaedics Private Limited 6000 77 Prosthetic knee joint polycentric or four-bar Active Medical and Rehabilitation Services Private Limited 7500 97 Four-bar mechanical pyramid base Paschim Rehabs 9000 116 Mechanical four-bar knee In Motion Prosthetic & Orthotic 15000 193 Four-bar mechanical knee Evolution Health Care Private Limited 20150 260 Four bar knee joint Bhatt Surgicals 50000 644 Modular four-bar knee joint P & O Rehab India 50054 645 Table 9 Estimated cost of proposed modified polycentric knee prosthesis Sl. No Criteria Cost of modified knee prosthesis (INR) Existing polycentric knee prosthesis (ALIMCO) Advanced polycentric knee prosthesis (ALIMCO) 1 Materials cost (Price for solid cylinder or block of AA7075-T6) 2500 2678 INR (USD 35) 13788 INR (USD 178) 2 CNC Machining cost (Depends on the type of machine) 1500 3 Labour cost (for design and digitalization) 500 4 Additional cost (Tooling and surface finish) 400 Total 4900 INR (USD 68) Table 9 Estimated cost of proposed modified polycentric knee prosthesis The preliminary prototype mentioned in this work is mostly made of AA7075 alloy which is relatively expensive but the overall cost of the proposed modified knee prosthesis is substantially reduced compared to the cost of imported or Indian-made prosthetic knee models as presented in Table 8. The cost of existing prosthetic knees used in National institutes and mass-manufactured by ALIMCO is 2678 INR compared to 4900 INR for the modified model. The increased cost of modified knee prostheses can be neglected because of the excellent mechanical properties and fatigue life of the product. Moreover, the predicted cost is within the budget of amputees from economically disadvantaged backgrounds, and the prosthesis can be provided free of charge to such amputees under the ADIP Scheme of the Government of India [29], which has a cost restriction of 15,000 INR for the entire prosthesis. Though the initial cost of the modified polycentric knee is increased by 83 percent, the same can be substantially reduced by the following means: 19 1. Replacing the components with the use of less expensive polymers [30] or other variants of aluminium alloy 1. 4.4 Technical advantages 1. Simple mechanical components with innovative shapes and ease of manufacturing 2. Light-weighted polycentric knee prosthesis kept under 500 g only 3. Qualifies ISO 10328:2016 standard including ultimate static load and prediction of fatigue life 4. Low profile design to suit subjects with long transfemoral or knee disarticulation 5. Optimized location of the instantaneous center of rotation to encourage voluntary control of knee stability and better swing clearance 6. Made of high strength and lightweight AA7075 T6 alloy 7. Provision of fine-tuning of linear and angular alignment to optimize function specific to subjects' requirements 8. Provides knee flexion of 150 degrees for encouraging squatting activities 9. Suitable in Indian conditions including socioeconomic and environmental requirements 10. Cost-effective to reach the maximum amputee population in low-income countries like India People with transfemoral amputations are estimated to walk on average 3553 steps per day [31] 9. Suitable in Indian conditions including socioeconomic and environmental requirements 10. Cost-effective to reach the maximum amputee population in low-income countries like India People with transfemoral amputations are estimated to walk on average 3553 steps per day [31]. The modified polycentric knee is predicted to last at least 10 years, based on the outstanding fatigue life findings of knee components. However, this study has few shortcomings. The fatigue testing with a knee prosthesis is limited in this study because a specific load can only be People with transfemoral amputations are estimated to walk on average 3553 steps per day [31]. The modified polycentric knee is predicted to last at least 10 years, based on the outstanding fatigue life findings of knee components. However, this study has few shortcomings. The fatigue testing with a knee prosthesis is limited in this study because a specific load can only be 20 administered two times, during heel strike and the toe-off phase of gait, and not applied during the entire walking cycle. Additionally, ISO 10328:2016 conformity on the structural strength of knee components may be too high for the Indian population. Future works may focus on the experimental fatigue and clinical gait analysis using the modified polycentric knee prosthesis; adoption of a comprehensive commercialization strategy for mass production to reduce the unit cost of the knee joint; and study of a dynamic model based on bipedal locomotion theory to obtain a more realistic load condition. 5. Conclusion The modified polycentric mechanical prosthetic knee model outperforms static and fatigue strength tests. The quality of existing knee prostheses has greatly improved due to design alterations and incorporation of new materials with superior mechanical strength. The modified knee prosthesis made from high-strength and lightweight AA7075 T6 alloy has a predicted maximum deformation of less than 1 mm and a minimum safety factor between 1.7 and 2, indicating the design is safe. Based on the remarkable fatigue life findings of knee components, it is anticipated that the newly developed polycentric knee will have a lifespan of at least ten years. This work has resulted in the development of a light-weighted polycentric knee prosthesis weighing 499 g compared to the existing prosthetic knee which is 532 g. The cost of fabricating the mechanical structure, including the cost of materials, is expected to be around 4,900 INR which is within the budget of amputees from economically disadvantaged backgrounds. Moreover, the prosthesis can be provided free of charge to such amputees under the ADIP scheme of Government of India. [1] Seymour R (2002) Introduction of Prosthetics and Orthotics. In, Tim Jules (ed). Prosthetics and Orthotics Lower limb & Spinal. Philadelphia. Lippincott Williams & Wilkins, pp 10-11 [2] Mohanty RK, Kumar JP, Rout, Das SP (2020) Successful prosthetic rehabilitation and gait analysis of individual with bilateral transtibial amputation: A case study with comparison to an able-bodied gait. Journal of Orthopaedics, Trauma, and Rehabilitation 27(1):93–100. doi:10.1177/2210491719893071 References [1] Seymour R (2002) Introduction of Prosthetics and Orthotics. In, Tim Jules (ed). Prosthetics and Orthotics Lower limb & Spinal. Philadelphia. Lippincott Williams & Wilkins, pp 10-11 [2] Mohanty RK, Kumar JP, Rout, Das SP (2020) Successful prosthetic rehabilitation and gait analysis of individual with bilateral transtibial amputation: A case study with comparison to an able-bodied gait. Journal of Orthopaedics, Trauma, and Rehabilitation 27(1):93–100. doi:10.1177/2210491719893071 [2] Mohanty RK, Kumar JP, Rout, Das SP (2020) Successful prosthetic rehabilitation and gait analysis of individual with bilateral transtibial amputation: A case study with comparison to an able-bodied gait. Journal of Orthopaedics, Trauma, and Rehabilitation 27(1):93–100. doi:10.1177/2210491719893071 [3] Mohanty RK, Biswal S, Sahoo PK, Das SP, Mohanty RC, Sabut SK (2020) Clinical gait analysis of subjects with a trans-femoral amputation using polycentric four-bar linkage prosthetic knee joint. Journal of Mechanics in Medicine and Biology 20(5):1-13 21 [4] International Society for Prosthetics and Orthotics (ISPO). Report of ISPO consensus conference on appropriate orthopedic technology for low-income countries; 1995 Jun 5; Phnom Penh, Cambodia. Day HJ, Hughes J, Jacobs NA, editors. ISPO; 1996 [5] Khamar M, Edrisi M (2018) Designing a backstepping sliding mode controller for an assistant human knee exoskeleton based on nonlinear disturbance observer. Mechatronics 54:121–132. doi:10.1016/j.mechatronics.2018.07.010 [6] Phanphet S, Dechjarern S, Jomjanyong S (2017) Above-knee prosthesis design based on fatigue life using finite element method and design of experiment. Medical Engineering & Physics 43:86-91. doi: 10.1016/j.medengphy.2017.01.001 [7] Diaz FR, Altamirano A, Orozco AV (2016) External knee prosthesis with four-bar linkage mechanism. 13th International Conference on Electrical Engineering, Computing Science and Automatic Control (CCE), Mexico City. 2016:1-6 [8] Lapapong S, Sucharitpwatskul S, Pitaksapsin N, Srisurangkul C, Lerspalungsanti S, Naewngerndee R, Sedchaicharn K, Chonnaparamutt W, Pipitpukdee J (2014) Finite element modeling and validation of a four-bar linkage prosthetic knee under static and cyclic strength tests. Journal of Assistive, Rehabilitative & Therapeutic Technologies 2:1. doi: 10.3402/jartt.v2.23211 [9] International Organization for Standardization. Prosthetics structural testing of lower-limb prostheses requirements and test methods. Geneva, Switzerland: International Organization for Standardization. 2006 [10] Kadhim FM, Chiad JS, Takhakh AM (2018) Design and manufacturing knee joint for smart transfemoral prosthetic. IOP Conference Series: Materials Science and Engineering 454:012078. doi: 10.1088/1757- 899X/454/1/012078 [11] Chauhan SS (2019) Kinematic and kinetic analysis of knee joint during squatting. IOP Conference Series: Materials Science and Engineering 691:012020. References doi:10.1088/1757-899X/691/1/012020 [12] Gopinath G, Krishnan H, Bandaru K, Kanamarlapudi SKM, Chittawadigi RG, Kakoty NM (2019) A biomimetic design of an artificial knee for lower limb prosthesis. In Proceedings of the Advances in Robotics 2019 (AIR 2019). Association for Computing Machinery, New York, NY, USA. 51:1–6. doi:10.1145/3352593.3352661 [13] Turner S, McGregor AH (2020) Perceived effect of socket fit on major lower limb prosthetic rehabilitation: A clinician and amputee perspective. Archives of Rehabilitation Research and Clinical Translation 2(3):100059. doi:10.1016/j.arrct.2020.100059 [14] Gard SA (2018) The Influence of prosthetic knee joints on gait. In: Müller B., Wolf S. (eds.) Handbook of Human Motion. Springer, Cham. https://doi.org/10.1007/978-3-319-14418-4_75 [15] Chauhan SS, Bhaduri SC (2020) Structural analysis of a four-bar linkage mechanism of prosthetic knee joint using Finite Element Method. Evergreen 7(2):209-215 22 [16] Afzal MR, Junaid AB, A. Yoon AEJ, Ullah Koreshi Z (2014) A cost effective design and analysis of an active prosthetic knee for trans-femoral amputees. International Conference on Robotics and Emerging Allied Technologies in Engineering (iCREATE) 270-276. doi:10.1109/iCREATE.2014.6828378 [17] Mohanty RK, Mohanty RC, Sabut SK (2020) A systematic review on design technology and application of polycentric prosthetic knee in amputee rehabilitation. Physical and Engineering Sciences in Medicine 43(3):781–798. doi:10.1007/s13246-020-00882-3 [17] Mohanty RK, Mohanty RC, Sabut SK (2020) A systematic review on design technology and application of polycentric prosthetic knee in amputee rehabilitation. Physical and Engineering Sciences in Medicine 43(3):781–798. doi:10.1007/s13246-020-00882-3 [18] Mohanty RK, Mohanty RC, Sabut, Pandey M (2022) Conformity assessment with structural strength requirements of mechanical polycentric prosthetic knee used for amputee rehabilitation. Computer Methods in Biomechanics and Biomedical Engineering 1–13. doi:10.1080/10255842.2022.2088233 [18] Mohanty RK, Mohanty RC, Sabut, Pandey M (2022) Conformity assessment with structural strength requirements of mechanical polycentric prosthetic knee used for amputee rehabilitation. Computer Methods in Biomechanics and Biomedical Engineering 1–13. doi:10.1080/10255842.2022.2088233 [19] Mohanty RK, Sabut SK, Mohanty RC (2017) Polycentric Prosthetic Knee Joint: A Review. International Journal of Engineering, Science and Mathematics 6(8):503-510. [19] Mohanty RK, Sabut SK, Mohanty RC (2017) Polycentric Prosthetic Knee Joint: A Review. International Journal of Engineering, Science and Mathematics 6(8):503-510. [20] Mahathaninwong N, Plookphol T, Wannasin J, Wisutmethangoon S (2011) T6 heat treatment of rheocasting 7075 Al alloy. Materials Science and Engineering A -structural Materials Properties Microstructure and Processing 532: 91-99. 10.1016/j.msea.2011.10.068. References [21] Clark R, Coughran B, Traina I, Hernandez A, Scheck T, Etuk C, Peters J, Lee EW, Ogren J, Es-Said OS (2005) On the correlation of mechanical and physical properties of 7075-T6 Al alloy. Engineering Failure Analysis 12(4):520-526. https://doi.org/10.1016/j.engfailanal.2004.09.005 [22] Song M, & Chen K (2008) Effects of the enhanced heat treatment on the mechanical properties and stress corrosion behavior of an Al–Zn–Mg alloy. Journal of Materials Science 43:5265-5273. 10.1007/s10853- 008-2773-0 [23] Wannasin J, Janudom S, Rattanochaikul T, Canyook R, Burapa R, Chucheep T, Thanabumrungkul S (2010) Research and development of gas induced semi-solid process for industrial applications. Trans. Nonferrous Met. Soc. China 20: s1010–s1015. [24] Mohanty RK, Mohanty RC, Sabut SK (2022) Finite element analysis and experimental validation of polycentric prosthetic knee. Materials Today: Proceedings 63: 207-214. doi:10.1016/j.matpr.2022.02.509 [25] El-Sayed AM, Hamzaid NA, Abu Osman NA (2014) Technology efficacy in active prosthetic knees for trans-femoral amputees: a quantitative evaluation. The Scientific World Journal 297431. doi: 10.1155/2014/297431 [26] Rosas, MA, Sánchez E, Amador B, Müller-Karger C (2019) Computational mechanical fatigue analysis in external prosthetic knee. Artículo de Investigación. Revista Ciencia e Ingeniería 40(1): 87-96 [27] https://bionicsforeveryone.com/microprocessor-knee-price-list/ [28] Arelekatti VNM, Winter AGV (2017) Design and preliminary field validation of a fully passive prosthetic knee mechanism for users with transfemoral amputation in India. ASME J Mech Robot 10(3):031007 [29] https://disabilityaffairs.gov.in/upload/uploadfiles/files/Revised%20ADIP%20Scheme%202022%20Englis h.pdf 23 [30] Hamner SR, Narayan VG, Donaldson KM (2013) Designing for Scale: Development of the ReMotion Knee for Global Emerging Markets. Annals of Biomedical Engineering 41(9):1851–9 [30] Hamner SR, Narayan VG, Donaldson KM (2013) Designing for Scale: Development of the ReMotion Knee for Global Emerging Markets. Annals of Biomedical Engineering 41(9):1851–9 [31] Turner S, McGregor AH (2020) Perceived effect of socket fit on major lower limb prosthetic rehabilitation: A clinician and amputee perspective. Archives of Rehabilitation Research and Clinical Translation 2(3):100059. doi:10.1016/j.arrct.2020.100059 [31] Turner S, McGregor AH (2020) Perceived effect of socket fit on major lower limb prosthetic rehabilitation: A clinician and amputee perspective. Archives of Rehabilitation Research and Clinical Translation 2(3):100059. doi:10.1016/j.arrct.2020.100059 [31] Turner S, McGregor AH (2020) Perceived effect of socket fit on major lower limb prosthetic rehabilitation: A clinician and amputee perspective. Archives of Rehabilitation Research and Clinical Translation 2(3):100059. doi:10.1016/j.arrct.2020.100059 Authors' contributions: All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Rajesh Kumar Mohanty, Ramesh Chandra Mohanty and Sukanta Kumar Sabut. The first draft of the manuscript was written by Rajesh Kumar Mohanty and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Competing Interests: The authors declare that there are no conflicts of interest. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Ethical approval: This article does not contain any studies with human participants or animals performed by any of the authors. Consent to participate Not applicable Not applicable Consent to publish Not applicable ORCiD ID: https://orcid.org/0000-0003-1433-9884 24
https://openalex.org/W2155671194	https://europepmc.org/articles/pmc3697414?pdf=render	English	null	Innate Immunity and Neuroinflammation	Mediators of inflammation	2,013	cc-by	19,027	1. Introduction recognise not only exogenous pathogen-associated molecular pattern1 (PAMP) but also endogenous modified molecules called damage-associated molecular pattern2 (DAMP). The innate immune system launches inflammatory and regula- tory responses via PRRs, phagocytes (macrophages), comple- ment system, cytokines, and chemokines in order to counter- act infection, injury, and maintenance of tissue homeostasis. Here, we discuss the role of innate immune players involved in neuroinflammation. Neuroinflammation is the mechanism of CNS inflamma- tion that occurs in response to trauma, infections, and/ or neurodegenerative diseases. In neuroinflammation, cel- lular and molecular immune components such as spe- cialised macrophages (microglia), cytokines, complement, and pattern-recognition receptors are the contributing play- ers. These proinflammatory mediators are either produced locally within the CNS or recruited from the peripheral system following disruption of the blood-brain barrier. This in turn leads to the activation of the glial cells, such as microglia and astrocytes. The effect of neuroinflammation is considered neuroprotective when the inflammatory activity is for a shorter period of time whereas chronic neuroinflam- mation is associated with harmful consequences for the CNS. Innate immunity is the first line of defence against the invading pathogens. Some of the components of first line of defence include epithelium (skin, gut, and lungs) that acts as a physical barrier and also produces several kinds of antimi- crobial enzymes and peptides, namely, lysozyme, defensins, mucin, lectin [1]. Other components of innate immunity include the pattern-recognition receptors (PRRs) such as toll- like receptors (TLRs), nucleotide-binding, and oligomeri- sation domain, leucine-rich repeats containing (NOD)-like receptors (NLRs); and Scavenger receptors (SRs). Present on phagocytic and antigen-presenting cells, these receptors Neuroinflammation is the mechanism of CNS inflamma- tion that occurs in response to trauma, infections, and/ or neurodegenerative diseases. In neuroinflammation, cel- lular and molecular immune components such as spe- cialised macrophages (microglia), cytokines, complement, and pattern-recognition receptors are the contributing play- ers. These proinflammatory mediators are either produced locally within the CNS or recruited from the peripheral system following disruption of the blood-brain barrier. This in turn leads to the activation of the glial cells, such as microglia and astrocytes. The effect of neuroinflammation is considered neuroprotective when the inflammatory activity is for a shorter period of time whereas chronic neuroinflam- mation is associated with harmful consequences for the CNS.i Abhishek Shastri,1 Domenico Marco Bonifati,2 and Uday Kishore1 1 Centre for Infection, Immunity and Disease Mechanisms, Heinz Wolff Building, Brunel University, London UB8 3PH, UK 2 Unit of Neurology, Department of Neurological Disorders, Santa Chiara Hospital, Largo Medaglie d’oro 1, 38100 Trento, Italy Correspondence should be addressed to Uday Kishore; uday.kishore@brunel.ac.uk Received 28 March 2013; Accepted 15 May 2013 Academic Editor: Hidde Bult Copyright © 2013 Abhishek Shastri et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Inflammation of central nervous system (CNS) is usually associated with trauma and infection. Neuroinflammation occurs in close relation to trauma, infection, and neurodegenerative diseases. Low-level neuroinflammation is considered to have beneficial effects whereas chronic neuroinflammation can be harmful. Innate immune system consisting of pattern-recognition receptors, macrophages, and complement system plays a key role in CNS homeostasis following injury and infection. Here, we discuss how innate immune components can also contribute to neuroinflammation and neurodegeneration. Hindawi Publishing Corporation Mediators of Inflammation Volume 2013, Article ID 342931, 19 pages http://dx.doi.org/10.1155/2013/342931 Hindawi Publishing Corporation Mediators of Inflammation Volume 2013, Article ID 342931, 19 pages http://dx.doi.org/10.1155/2013/342931 Hindawi Publishing Corporation Mediators of Inflammation Volume 2013, Article ID 342931, 19 pages http://dx.doi.org/10.1155/2013/342931 2. Microglia In order to regulate the immune responses, anti-inflammatory cytokines IL-10 and transforming growth factor beta are produced by microglia [18–20]. Microglia also produce inhibitor of nuclear factor 𝜅𝛽(NF-𝜅𝛽), mitogen-activated protein kinase (MAPK) phosphatases, and suppressor of cytokine signalling proteins [21], which help in immune activation regulation. Glucocorticoids have also been considered to play a reg- ulatory role for innate immunity in CNS by regulation of microglial TNF-𝛼[22, 23] although there are debatable views to the same [24].h Figure 1: Schematic diagram showing structure of TLR and NLR family. TLR: toll-like receptor; NLRP: NOD-like receptor containing pyrin domain; NLRC: NOD-like receptor containing NLR-containing caspase activation and recruitment domain; NLRB: NOD-like receptor containing baculovirus inhibitor of apoptosis protein repeat domain; LRR: leucine-rich repeat; TIR: toll/il-1 recep- tor; PYD: pyrin domain; CARD: caspase activation and recruitment domain; BIR: baculovirus inhibitor of apoptosis protein repeat. The figure shows the structure of a TLR containing a TIR domain present inside nucleus which is involved in signalling pathway and an LRR domain present in the cytoplasm which is involved in pathogen recognition. NLR are intracellular receptors containing a C-terminal LRR domain, a central NACHT domain, and a variable N-terminal domain which can be a PYD, a CARD, or a BIR domain. Figure 1: Schematic diagram showing structure of TLR and NLR family. TLR: toll-like receptor; NLRP: NOD-like receptor containing pyrin domain; NLRC: NOD-like receptor containing NLR-containing caspase activation and recruitment domain; NLRB: NOD-like receptor containing baculovirus inhibitor of apoptosis protein repeat domain; LRR: leucine-rich repeat; TIR: toll/il-1 recep- tor; PYD: pyrin domain; CARD: caspase activation and recruitment domain; BIR: baculovirus inhibitor of apoptosis protein repeat. The figure shows the structure of a TLR containing a TIR domain present inside nucleus which is involved in signalling pathway and an LRR domain present in the cytoplasm which is involved in pathogen recognition. NLR are intracellular receptors containing a C-terminal LRR domain, a central NACHT domain, and a variable N-terminal domain which can be a PYD, a CARD, or a BIR domain. and even long-term memory formation [33]. Any insult to the CNS is associated with changes in the structure, morphology, and hypertrophy of astrocytes, followed by cytokine and C1q secretion, leading to scar formation, collectively termed as reactive astrogliosis [34]. There are a variety of receptors expressed on microglia related to the different functions of these cells. 2. Microglia Some of the receptors associated with innate immunity are listed in Table 1. TLR 1–9 receptors are known to be expressed by microglial cells (discussed in detail later). NLR form com- plexes called inflammasomes (for a detailed review see [25]) that have been shown to activate and recruit microglia in response to amyloid-𝛽(A𝛽) [26] and prion peptide [27]. Some of the other innate immune receptors expressed on microglia surface are CD14, CD18, CD36, CD68, mannose, and lectin (Dendritic Cell-Specific Intercellular adhesion molecule-3-Grabbing nonintegrin or DC-SIGN) receptors. Complement receptors found on microglia are C3a, C5a, and C1q receptors [28]. Like microglia, astrocytes have been shown to express innate immune PRR like TLR, NLR, scavenger, complement, and mannose receptors [35]. They have also been shown to release cytokines like TNF, IL-6, IL-1, Interferon-𝛾, and chemokines when stimulated with lipopolysaccharide (LPS) [36, 37]. Reactive astrogliosis is associated with a number of CNS diseases such as AD [38, 39], PD, autism, and prion diseases [40, 41]. 4. Toll-Like Receptors (TLR) 4.1. Structure and Signalling Pathway. TLRs are expressed on microglia, neurons, and astrocytes similar to dendritic cells, B cells, neutrophils, epithelia, and fibroblast [42]. TLR is a type 1 membrane protein containing an extracellular leucine-rich repeat (LRR) domain and a Toll/IL-1 receptor (TIR) domain in the cytoplasmic region (Figure 1). LRR domain is involved in specific pathogen recognition [43] and TIR domain is involved in the signalling pathway. TLRs are considered to exist as dimers and bind to various ligands [44, 45]. For example, TLR2 heterodimerises with TLR1 [46] and also with TLR6 [44] and recognises bacterial lipopro- teins. Upon sensing ligands, recruitment of adaptor proteins takes place which is necessary for signal transduction [47]. 2. Microglia Upon environmental stimula- tion/challenges, the microglia become “activated” and the morphology changes to an amoeboid appearance where they retract the ramifications [9]. Activation of microglia by TLRs and NLRs is considered to be “classical” form of microglial activation where innate immune responses include pro- duction of proinflammatory cytokines like tumour necrosis factor (TNF)-𝛼, interleukin (IL)-1 and IL-6, and chemokines. Classical activation also leads to adaptive immune response by expressing major histocompatibility class II molecules and interaction with T cells [10]. TNF-𝛼stimulation increases phagocytic activity of microglia [11], and deficiency of TNF receptors has been found to reduce microglial activation [12]. TNF-𝛼is associated with activation of microglial cells involved in pathogenesis of neurodegenerative diseases like Alzheimer’s disease (AD) [13] and Parkinson’s disease (PD) [14]. IL-1 induces expression of TNF-𝛼and IL-6 [15] and is implicated in neuroinflammatory processes in traumatic brain injury (TBI), AD, and PD [16]. Activated microglia have also been implicated in neurotransmission [17]. In order to regulate the immune responses, anti-inflammatory cytokines IL-10 and transforming growth factor beta are produced by microglia [18–20]. Microglia also produce inhibitor of nuclear factor 𝜅𝛽(NF-𝜅𝛽), mitogen-activated protein kinase (MAPK) phosphatases, and suppressor of cytokine signalling proteins [21], which help in immune activation regulation. Glucocorticoids have also been considered to play a reg- ulatory role for innate immunity in CNS by regulation of microglial TNF-𝛼[22, 23] although there are debatable views to the same [24].h complement dependent [8]. Upon environmental stimula- tion/challenges, the microglia become “activated” and the morphology changes to an amoeboid appearance where they retract the ramifications [9]. Activation of microglia by TLRs and NLRs is considered to be “classical” form of microglial activation where innate immune responses include pro- duction of proinflammatory cytokines like tumour necrosis factor (TNF)-𝛼, interleukin (IL)-1 and IL-6, and chemokines. Classical activation also leads to adaptive immune response by expressing major histocompatibility class II molecules and interaction with T cells [10]. TNF-𝛼stimulation increases phagocytic activity of microglia [11], and deficiency of TNF receptors has been found to reduce microglial activation [12]. TNF-𝛼is associated with activation of microglial cells involved in pathogenesis of neurodegenerative diseases like Alzheimer’s disease (AD) [13] and Parkinson’s disease (PD) [14]. IL-1 induces expression of TNF-𝛼and IL-6 [15] and is implicated in neuroinflammatory processes in traumatic brain injury (TBI), AD, and PD [16]. Activated microglia have also been implicated in neurotransmission [17]. 2. Microglia Microglial cells are the specialised resident macrophages of the CNS. The origin of these innate immune cells is debatable but it is now widely believed that they are of myeloid lineage [2]. In mice studies, it has been found that microglia originate from primitive (yolk sac) myeloid progenitors that migrate to CNS independent of definitive progenitors and circulation (i.e., bone marrow) [3]. These cells are found in brain, spinal cord, retina, and optic nerve. Their morphology differs from “conventional” macrophages by the presence of branch- like processes (ramified appearance). This is the shape they have when in “resting” state. In this state, these cells con- stantly monitor and survey their area [4]. The microglial cells in resting form have been shown to be involved in other functions such as neurogenesis [5], neuroprotection [6] and synaptic pruning [7], which has been found to be Innate immunity is the first line of defence against the invading pathogens. Some of the components of first line of defence include epithelium (skin, gut, and lungs) that acts as a physical barrier and also produces several kinds of antimi- crobial enzymes and peptides, namely, lysozyme, defensins, mucin, lectin [1]. Other components of innate immunity include the pattern-recognition receptors (PRRs) such as toll- like receptors (TLRs), nucleotide-binding, and oligomeri- sation domain, leucine-rich repeats containing (NOD)-like receptors (NLRs); and Scavenger receptors (SRs). Present on phagocytic and antigen-presenting cells, these receptors 2 2 Mediators of Inflammation Cytoplasm Nucleus TLR NLRP NLRC NLRB LRR TIR domain NACHT domain PYD CARD BIR Cell Membrane Figure 1: Schematic diagram showing structure of TLR and NLR family. TLR: toll-like receptor; NLRP: NOD-like receptor containing pyrin domain; NLRC: NOD-like receptor containing NLR-containing caspase activation and recruitment domain; NLRB: NOD-like receptor containing baculovirus inhibitor of apoptosis protein repeat domain; LRR: leucine-rich repeat; TIR: toll/il-1 recep- tor; PYD: pyrin domain; CARD: caspase activation and recruitment domain; BIR: baculovirus inhibitor of apoptosis protein repeat. The figure shows the structure of a TLR containing a TIR domain present inside nucleus which is involved in signalling pathway and an LRR domain present in the cytoplasm which is involved in pathogen recognition. NLR are intracellular receptors containing a C-terminal LRR domain, a central NACHT domain, and a variable N-terminal domain which can be a PYD, a CARD, or a BIR domain. Cytoplasm Nucleus TLR NLRP NLRC NLRB LRR TIR domain NACHT domain PYD CARD BIR Cell Membrane complement dependent [8]. 3. Astrocytes Astrocytes are specialised glial cells and the most abundant cells of the CNS. Morphologically, astrocytes are of two types: protoplasmic (found in grey matter) and fibrous (found in white matter). The basic astrocyte morphology resembles that of a star (with multiple processes). Protoplasmic astrocytes have undistinguishable dense processes while fibrous astro- cytes have clearly distinguishable processes [29]. Astrocytes have conventionally been considered to be supporting cells to the neurons. However, recently they have been shown to play an active part in the modulation of neural activity [30], poten- tiation of synaptic transmission [31], sleep homeostasis [32], 3 Mediators of Inflammation 3 Table 1: Innate immune receptors on microglia. Table 1: Innate immune receptors on microglia. Receptor Functions/comments References TLR Pattern-recognition receptors that respond to self (DAMPs) and nonself (PAMPs) activators. Microglia are known to express TLR1-9. TLRs are implicated in neuroinflammation in response to bacterial and viral infections, Alzheimer’s disease, prion diseases, and amyotrophic lateral sclerosis. [59, 69] NLR Cytoplasmic pattern-recognition receptors. Microglia are known to express NOD2 in response to CNS infection and NALP3 inflammasome in Alzheimer’s disease. [109, 110] Scavenger Another group of pattern-recognition receptors. The receptors expressed on microglia are Class A, CD36, and RAGE. [111, 112] RLR RIG-I is a pattern-recognition receptor that is expressed by microglia in response to viral infections. [110, 113] Complement Complement receptors expressed include CR1, CR3 and CR4. These receptors bind complement proteins and activate complement pathway which is considered to be both beneficial and detrimental depending on the level of activation. [114] Cytokines Some of the cytokine receptors expressed in microglia are IL-1R, TNFR (responsible for proinflammatory actions of cytokines IL-1 and TNF-𝛼resp.), IL-10R, TGFR (responsible for the anti-inflammatory cytokines IL-10 and TGF-𝛽), and CCR1-5 responsible for actions of chemokines. These are expressed and produced in neuroinflammation. [115, 116] TLR: toll-like receptor; DAMP: damage-associated molecular pattern; PAMP: pattern-associated molecular pattern; NLR: NOD-like receptors; NOD: nucleotide-binding and oligomerisation domain; RLR: RIG-like receptors; RIG: retinoic acid-inducible gene; CR: complement receptor; IL: interleukin; TNF: tumour necrosis factor; TGF: transforming growth factor. endocytosis of TLR4 [65]. Injection of LPS directly into brain has been shown to produce an increased expression of genes of proinflammatory cytokines, chemokines, and complement proteins and receptors such as CD14 [66, 67]. Production of TNF by microglial cells upon LPS stimulation has been found to cause death of dopaminergic neurons [68]. 3. Astrocytes TLR2 ligands stimulation of microglial and astrocytic cells leads to an increase in production of IL-6, chemokines, and IFN-𝛽[69]. In mice studies, TLR9 ligand CpG has been found to be neuroprotective in cerebral ischaemia [70] while similar findings have been reported in TLR4 knockout mice [71]. TLR2 activation has been shown to be involved in neurogenesis [72] while TLR8 induces apoptosis of neurons [73]. TLR3 impairs plasticity and working memory [74] while TLR7 and TLR9 have been found to be associated with the development of mouse brain [75]. Interestingly, increased peripheral responses of TLR2, TLR4, TLR8, and TLR9 have been detected in psychosis [76] while TLR9 is associated with posttraumatic anxiety [77]. The adaptor proteins are (i) myeloid differentiating factor 88 (MyD88); (ii) MyD88 adaptor-like protein (Mal); (iii) TIR domain-containing adaptor inducing interferon-𝛽(TRIF); (iv) TRIF-related adaptor molecule; and (v) sterile-𝛼and armadillo-motif-containing protein. These adaptor proteins are recruited by TIR domain leading to activation of NF- 𝜅𝛽. NF-𝜅𝛽then induces production of proinflammatory cytokines such as TNF-𝛼, IL-1𝛽, and IL-6, and chemokines. All TLRs are activated by MyD88 except TLR3; instead MyD88 may be restricting TLR3 signalling [48]. Some of the other adaptors investigated in detail include major histocom- patibility complex class II molecules [49], small heterodimer partner [50], and Dedicator of Cytokinesis 8 (DOCK8) [51]. It has recently been shown that oligomerisation of TLR4 with myeloid differentiation protein-2 by morphine causes neuroinflammation [52]. Necrotic neurons have been shown to activate microglia via MyD88 pathway leading to increased neuroinflammation [53]. In mouse models, both MyD88 and TRIF pathways have been implicated in regulation of IL-6 and IL-10 after cerebral ischaemia [54] as well as regulation of IL6, TNF-𝛼, and IL-1𝛽following intracerebral haemorrhage [55]. MyD88 pathway also plays an important role in CNS infection and consequent astrocyte activation [56]. MyD88 pathway may also be involved in PD [57] and optic nerve injury [58]. 4.4. TLR Response to Pathogens. Pneumococcal infection leads to innate immune response in brain and this depends on TLR2 and TLR4 [78]. Deficiency of TLR2 causes an increased TNF gene expression in the brain [79]. TLRs have been found to be involved in pneumococcal infection in HIV- associated neurocognitive disorders [80]. TLR signalling is also associated with virulence of intracellular pathogens [81]. TLR2 and TLR9 initiate immune response against herpes simplex virus (HSV) [82] and also control HSV infection in the brain [83]. 3. Astrocytes TLR3 is protective for the CNS in HSV1 infection [84]. In mice models, TLR3 in astrocytes may be protective in HSV2 infection [85] and has been reported to mediate entry of West Nile virus (WNV) into the CNS, causing encephalitis [86]. TLRs have also been implicated 4.2. Ligands. Some of the exogenous and endogenous ligands of TLR are listed in Table 2 [59–62]. 4.3. Response in CNS to Ligands of TLR. In vivo stud- ies have shown that the administration of LPS (periph- eral/intraperitoneal) leads to expression of genes coding for proinflammatory cytokines in the microglial cells [63, 64]. CD14 has been found to be required for LPS-induced 4 Mediators of Inflammation 4 Table 2: Exogenous and endogenous ligands of toll-like receptors. Table 2: Exogenous and endogenous ligands of toll-like receptors. Ligand TLR Implications/comments References Lipopolysaccharide TLR4 Recognition of Gram (−) bacteria [117] Triacylated lipopeptides TLR1 and TLR2 Recognition of Gram (−) bacteria and mycobacteria [118] Diacylated lipopeptides TLR2 and TLR6 Recognition of Gram (+) bacteria and mycoplasma [119, 120] Lipoteichoic acid TLR2 Recognition of Gram (+) bacteria [121] Zymosan TLR2 Recognition of fungi [122] Double-stranded RNA TLR3 Recognition of virus [123] Single-stranded RNA TLR7 and TLR8 Recognition of virus [124, 125] Flagellin TLR5 Recognition of Gram (−) bacteria [126] Unmethylated CpG DNA TLR9 Recognition of bacteria and virus [127, 128] 𝛽-amyloid TLR2; TLR4; TLR4 and TLR6 Neuroinflammation in Alzheimer’s disease [95, 96, 129, 130] Mitochondrial DNA TLR9 Pathogenesis of myocarditis and heart failure [128] Lung surfactant protein-A and -D TLR4 TLR2 Innate immune component of lung. Act as opsonin and macrophage activator. Physiological implications of excessive activation by TLR is not known [131–133] Tenascin-C TLR4 Maintenance and pathogenesis of inflammation in rheumatoid arthritis [134, 135] Fibrinogen TLR4 Present normally in serum and activation has been implicated in rheumatoid arthritis and atherosclerosis [136, 137] Oxidised low-density lipoprotein TLR4 Pathogenesis of atherosclerosis [95] MicroRNA let-7 TLR7 Pathogenesis of neurodegeneration [138] in CNS parasitic infections like toxoplasmosis,3 sleeping sickness,4 cerebral malaria,5 and neurocysticercosis6 [87]. TLR2 is associated with protection from cerebral malaria [88] and therapeutic targeting of TLRs has been shown to prevent experimental cerebral malaria [89, 90]. mice leads to delayed onset of the disease [107]. However, MyD88 knockout mice (lacking TLR signalling) were found to develop prion disease similar to wild-type mice both in terms of time and severity [108]. 5. NOD-Like Receptors 4.5. Neurodegenerative Diseases. In mouse model of AD, MyD88 has been found to prevent memory [91] and cognitive deficits [92] while another study found MyD88 deficiency to improve AD-related pathology [93]. TLR2 clears A𝛽and delays cognitive decline, again in mouse model of disease [94]. TLR4 causes A𝛽-induced microglial activation [95] and A𝛽-induced neuronal apoptosis [96]. A loss-of-function mutation of TLR4 has been found to reduce microglial activation and increase A𝛽deposits with increased cog- nitive deficits [97]. Intracranial injection of LPS (a TLR4 ligand) reduces A𝛽levels in brain [98]. TLR9 may have a protective role in AD by improving cognitive functions [99], reducing A𝛽-toxicity [100], and clearing A𝛽[101]. In amyotrophic lateral sclerosis7 (ALS), MyD88 has been shown to activate microglia due to mutant SOD1 [102] and in vitro studies show enhanced microglial activation and neurotoxicity when stimulated with TLR2 and TLR4 ligands [103, 104]. MyD88 pathway may also be involved in PD [57] where 𝛼-synuclein directly activates microglia and alters expression of TLRs [105]. TLR signalling has been found to interfere with prion disease pathogenesis. Studies involving mice possessing mutant gene which prevents TLR4 signalling was found to have a shorter time for scrapie pathogenesis [106] while administration of TLR9 agonist in prion-infected 5.1. Structure. Like TLRs, NOD-like receptors (NLRs) also detect PAMPs and DAMPs. NLRs are intracellular recep- tors thereby monitoring intracellular environment. They consist of a central nucleotide-binding and oligomerisation (NACHT) domain and a C-terminal LRRs. Their N-terminal component may be variable based on which NLRs are further subdivided. It can be caspase activation and recruitment domain (CARD); a pyrin domain (PYD), or baculovirus inhibitor of apoptosis protein repeat (BIR) termed, respec- tively, as NLRC, NLRP, and NLRB [139]. Upon binding to agonists, NLR can lead to the activation of NF-𝜅𝛽or MAPK signalling pathways and production of cytokines and chemokines. NLR binding to agonist also causes the activation of procaspase-1 leading to inflammasome forma- tion; pyroptosis; autophagy; and IFN-1 signalling [140–145] (Figure 1) [141–145]. 5.2. Inflammasomes. Inflammasomes are multiprotein com- plexes that activate caspase-1, which in turn leads to pro- cessing and secretion of proinflammatory cytokines such as IL-1𝛽and IL-18. The members of NLR family that are capable of forming inflammasomes are PYD-containing NLRP1, NLRP3, NLRP6, and CARD-containing NLRC4 Mediators of Inflammation 5 of SR-A in AD mouse models was not found to impact amyloid plaque deposition or clearance [181]. 7. Complement 7.1. Three Activation Pathways of the Complement System. The complement system comprises of more than 30 proteins in the serum as well as membrane-bound receptors and regula- tors. The complement system consists of 3 different initiating or activation pathways culminating into a final common lytic pathway, leading to the formation of membrane attack complex (MAC) (Figure 2). MAC are pores that penetrate cell membrane (lipid bilayers) of pathogens or abnormal cells, thereby causing their lysis. The three initiating pathways are called (i) classical pathway which is mostly antibody medi- ated (C1q being the first subcomponent) and is activated by C1 complex (C1q-C1r-C1s); (ii) alternative pathway (AP) which is activated spontaneously involving low-level hydrolysis of C3 to C3 (H20); and (iii) lectin pathway where activation occurs through binding of a carbohydrate pattern present on microorganisms called mannan, with mannan-binding lectin (MBL) and Ficolins (ficolin-1, -2 and -3). They circulate in the serum in combination with zymogen serine proteases called MBL-associated serine proteases (MASPs) [191–196]. All the 3 pathways ultimately converge to lead to formation of C3 convertase. C3 convertases then cleaves C3 into C3a and C3b. This C3b binds to C3 convertase and leads to the formation of C5 convertase. This C5 convertase cleaves C5 into C5a and C5b. C3a and C5a are called anaphylatoxins and are chemoattractants. The C5b formed associates with C6, C7, C8, and C9 to form MAC [197]. The functions of the complement system include opsonisation of pathogens, direct lysis of foreign cells, chemotaxis and activation of leukocytes, and clearance of apoptotic cells. The complement system also interacts with TLRs [198] and plays a role in the regulation of humoral immunity [199]. The complement system is kept in check by regulators in order to prevent overactivation leading to damage to tissues and autoimmune diseases. The regulators can be grouped into fluid-phase: factor H (fH) and properdin for alternative pathway, C1 inhibitor and C4b-binding protein (C4BP) for classical and MBL pathway; host cell membrane- bound: CR1, CR2, CD55, CD46, CD59; cell surface-attached 5. NOD-Like Receptors In vitro studies have shown that astrocytes express SR-A and thus play a role in neuroinflammation [166]. Class B SR Type I (SR- BI) has been shown to be produced in vivo in AD brains [182] with increased expression being observed in cerebellum and cortex [183]. In mice studies, SR-BI has also been shown to impair perivascular macrophages leading to AD pathology such as increased amyloid deposition, cerebral amyloid angiopathy (deposition of A𝛽in cerebral arteries), and memory deficits [184]. CD36 appears to be involved in neurovascular dysfunction due to A𝛽[185] and promotes cerebral amyloid angiopathy leading to cognitive deficits [186]. RAGE is a receptor for A𝛽and expressed on neurons, microglia, astrocytes, and endothelial cells [187]. RAGE signalling in microglia due to p38 MAPK signalling pathway leads to neuroinflammation and cognitive disturbances in AD [188] as well as synaptic [189] and neuronal [190] dys- function. [146]. Inflammasome complex formation occurs when a ligand binds to NLR and thereby induces a conformational change, leading to ATP binding at NACHT domain which causes receptor oligomerisation and recruitment of other complex members [141]. Inflammasomes have been impli- cated in various diseases such as gout, pseudogout, contact dermatitis, allergic dermatitis, vitiligo, hydatidiform mole8 [147], Muckle-Wells syndrome9 [148], atherosclerosis, type 2 diabetes mellitus, obesity [149], metabolic syndrome10 [150], acute myocardial infarction [151], coeliac disease, inflamma- tory bowel disease [152], asthma, pulmonary fibrosis [153], and viral [154] and bacterial infections [155]. 5.3. Role in Neuroinflammation. NLRP3 inflammasome is involved in the innate immune response to A𝛽[156] leading to AD pathology. In multiple sclerosis (MS), NLRP3 knockout mice model of disease shows reduced demyelination [157], while another study shows NLRP3 involvement in migration of T-helper cells into CNS [158]. IFN-𝛽therapy is effective in treating inflammasome-dependent disease in mouse models of MS [159]. NLRP1 has been found to be involved in TBI and neutralising its effect or formation was found to have beneficial effects [160]. Inflammasome complex inhibition has also been found to reduce inflammation and improve pathology in mouse models of stroke [161]. NLRP3 inflam- masome contributes to brain injury in pneumococcal menin- gitis [162] and is associated with inflammation in Japanese encephalitis [163]. Both NLRP1 and NLRP3 are increased in postmortem alcoholic human brains and inhibition of these inflammasomes was found to be beneficial in reversing ethanol-mediated neuroinflammation [164]. 6. Scavenger Receptors The classical pathway is usually activated by antigen-antibody complexes, the lectin pathway is activated by microbes with MBL-MASP complex, and the alternative pathway is activated spontaneously by hydrolysis of C3 to C3(H2O). All 3 pathways lead to formation of C3 convertase, followed by C5 convertase, ultimately leading to formation of membrane attack complex. In this process, anaphylatoxins C3a and C5a are also released. The complement system is kept in check by a number of regulators. Figure 2: The complement system. Complement regulators are indicated in red. MBL: mannan-binding lectin; MASP: MBL-associated serine protease; C4BP: C4b-binding protein; CR1: complement receptor 1. The complement system consists of 3 initiating pathways: classical pathway, lectin pathway, and alternative pathway. The classical pathway is usually activated by antigen-antibody complexes, the lectin pathway is activated by microbes with MBL-MASP complex, and the alternative pathway is activated spontaneously by hydrolysis of C3 to C3(H2O). All 3 pathways lead to formation of C3 convertase, followed by C5 convertase, ultimately leading to formation of membrane attack complex. In this process, anaphylatoxins C3a and C5a are also released. The complement system is kept in check by a number of regulators. complement regulators: fH, factor H-like protein 1 (FHL-1), C4BP and clusterin [200, 201]. For certain ligands, factor H can also regulate C1q-mediated classical pathway [202–205]. in neural progenitor cells and also negatively regulates hip- pocampal neurogenesis [216]. An emerging area for comple- ment involvement in CNS is in relation to synapse (reviewed in [217]). C1q, initiating component of classical pathway and widely expressed by postnatal neurons and immature astrocytes [218], mediates the elimination of synapse [219, 220]. C1q knockout mice show increased synaptic connec- tivity and spontaneous epilepsy [221]. Synapse remodelling by microglia involves CR3 [8]. In vitro studies show that C1q also promotes neuronal viability and survival [222]. In vitro and in vivo studies implicate a role for C3aR and C5aR in the development of cerebellum [223]. Many other in vitro and in vivo studies show neuroprotective functions for C3a and C5a that include protection against NMDA-induced apoptosis [224] and protection against glutamate-induced apoptosis [225] via MAPK-dependent inhibition of caspase 3 [226] as well as regulation of glutamate receptor subunit 2 [227]. 7.2. Role in CNS Physiology. Complement is produced mainly in the liver and, over the years, it was thought that the brain was an immune-privileged organ due to the presence of blood-brain barrier. 6. Scavenger Receptors 6.1. Types. Scavenger receptors (SRs) are members of PRRs and are transmembrane glycoprotein PRRs [165]. SRs are expressed on macrophages, dendritic cells, microglia, and endothelial cells [111, 112]. Recently, SR expression on astro- cytes has been reported [166]. The family of SRs include class A (macrophage receptors, MARCO), class B (CD36, SR-BI), CD68 and endothelial or LOX-1, CD163, and receptor for advanced glycation end products (RAGE) [167, 168]. Some of the ligands that SRs bind to are pathogen-specific: LPS, lipote- ichoic acid, Streptococcus pneumoniae, Staphylococcus aureus, Mycoplasma pneumoniae, Neisseriia meningitides, Escherichia coli [169], apoptotic cells [170], and erythrocytes infected with Plasmodium [171–173]. SRs have been implicated in atherosclerosis [174], lung inflammation [175], cystic fibrosis [176], SLE [170], and AD [112]. 6.2. Role in Neuroinflammation. Microglia express SR and thus bind to A𝛽fibrils [177] which is associated with AD plaques [178]. Class A SR (SR-A) has also been shown to play an important role in cerebral injury due to ischemia. Mice deficient in SR-A showed reduced expression levels of TNF-𝛼and IL-1𝛽as well as decreased infarct size [179]. In experimental model of MS, SR-A knockout mice showed significantly reduced demyelination as well as reduced proin- flammatory cytokines production [180]. However, deficiency 6 Mediators of Inflammation Classical pathway Lectin pathway Alternative pathway complex Mannose binding on microbes Spontaneous hydrolysis of C3 MBL-MASP C2 C2 C4 C4 Membrane attack complex Factor H Lectin pathway is activated with the help of MASP following MBL binding to PAMPs. Alternative pathway is activated by Antigen-antibody C1q-C1r-C1s C3(H2O) C1 inhibitor C4BP C4BP C4BP C4b2a C3bBb (C3 convertase) (C3 convertase) C4b2aC3b (C5 convertase) (C5 convertase) C3bBbC3b C4BP, CR1, CD55, CD46 C4BP, CR1, CD55, CD46 C3b C3 C3 C3a C3a C5b C5 C5 C5a C5a C6, C7, C8, C9 Classical pathway is activated when antigen-antibody complex bind to C1q leading to activation of C1r and C1s (together constitute C1). spontaneous low-level hydrolysis of C3. All 3 pathways lead to conversion of C3 into C3b by forming C3 convertase. The C3 convertase for classical and lectin pathway is C4b2a while that of alternative pathway is C3bBb. This step also leads to the production of C3a anaphylatoxin. Then, C3b facilitates conversion of C5 into C5b by being involved in the formation of C5 convertase. The C5 convertase for classical and lectin pathway is C4b2aC3b while that for alternative pathway is C3bBbC3b. This step also leads to the formation of C5a anaphylatoxin. 6. Scavenger Receptors Following this, C5b forms a complex with C6, C7, C8, and C9 called membrane attack complex. CD59 Figure 2: The complement system. Complement regulators are indicated in red. MBL: mannan-binding lectin; MASP: MBL-associated serine protease; C4BP: C4b-binding protein; CR1: complement receptor 1. The complement system consists of 3 initiating pathways: classical pathway, ectin pathway, and alternative pathway. The classical pathway is usually activated by antigen-antibody complexes, the lectin pathway is ctivated by microbes with MBL-MASP complex, and the alternative pathway is activated spontaneously by hydrolysis of C3 to C3(H2O). All pathways lead to formation of C3 convertase, followed by C5 convertase, ultimately leading to formation of membrane attack complex. In his process, anaphylatoxins C3a and C5a are also released. The complement system is kept in check by a number of regulators. Classical pathway Lectin pathway Alternative pathway complex Mannose binding on microbes Spontaneous hydrolysis of C3 MBL-MASP C2 C2 C4 C4 Factor H Antigen-antibody C1q-C1r-C1s C3(H2O) C1 inhibitor C4BP C4BP C4BP C4b2a C3bBb (C3 convertase) (C3 convertase) C3b C3 C3 C3a C3a Classical pathway Classical pathway is activated when antigen-antibody complex bind to C1q leading to activation of C1r and C1s (together constitute C1). Alternative pathway complex Antigen-antibody Mannose binding on microbes All 3 pathways lead to conversion of C3 into C3b by forming C3 convertase. The C3 convertase for classical and lectin pathway is C4b2a while that of alternative pathway is C3bBb. This step also leads to the production of C3a anaphylatoxin. C3bBb (C3 convertase) C4b2a (C3 convertase) C3b Then, C3b facilitates conversion of C5 into C5b by being involved in the formation of C5 convertase. The C5 convertase for classical and lectin pathway is C4b2aC3b while that for alternative pathway is C3bBbC3b. This step also leads to the formation of C5a anaphylatoxin. Membrane attack complex C4b2aC3b (C5 convertase) (C5 convertase) C3bBbC3b C4BP, CR1, CD55, CD46 C4BP, CR1, CD55, CD46 C5b C5 C5 C5a C5a C6, C7, C8, C9 CD59 Following this, C5b forms a complex with C6, C7, C8, and C9 called membrane attack complex. Following this, C5b forms a complex with C6, C7, C8, and C9 called membrane attack complex. Figure 2: The complement system. Complement regulators are indicated in red. MBL: mannan-binding lectin; MASP: MBL-associated serine protease; C4BP: C4b-binding protein; CR1: complement receptor 1. The complement system consists of 3 initiating pathways: classical pathway, lectin pathway, and alternative pathway. 6. Scavenger Receptors Upon exposure to A𝛽, C1q is expressed in neurons (hippocampus) [273], and it has been found that inhibiting the binding of C1q to A𝛽leads to protection of hippocampal cells [274]. In mouse models of AD, absence of C1q shows less neuropathology [275]. Complement regula- tors factor H, FHL-1, and C4BP have also been localised in amyloid plaques and fH and C4BP have been shown to bind A𝛽in vitro [276–278]. These regulators could be involved in regulation of excessive complement activation. Another interesting feature is the presence of microglia expressing complement receptors found in close proximity to plaques. Microglia are found in and around plaques of AD brains [279] and are found to express C1q [280] and complement receptors C1qR, CR3, CR4, and C5aR, which help in the phagocytosis of A𝛽[281, 282]. Complement activation is therefore also considered to be neuroprotective [266]. C3 deficiency in mouse model shows accelerated amyloid plaque deposition [283]. Furthermore, inhibition of complement was found to be associated with an increased formation of plaque and neurodegeneration [284]. Amyloid precursor protein transgenic mouse models of AD that lack the ability to acti- vate classical pathway (APPQ−/−) (i.e., C1q−/−phenotype) show less neuropathology as compared to APPQ+/+ mice. However, APPQ−/−mice also show increased C3 levels, providing evidence for alternative pathway activation in AD [285]. In mice models, deficiency of sCrry increases tau pathology [286]. Genetic association of AD and complement involves complement genes CR1 and CLU [287]. Micro- RNAs11 (miRNAs) −9, −125b, −146a, and −155 are found to be upregulated in AD and these miRNAs target gene encoding fH [288]. y g Viruses have also evolved mechanisms to evade comple- ment system [235]. Gamma-herpesvirus encode for proteins that regulate and inhibit host C3-mediated resistance [236]. Complement controls antibody response in WNV infection [237] with lectin pathway activation being found to be protec- tive in WNV infection [238]. C3 has been found to participate in seizure induction during viral encephalitis [239]. Increased MBL is seen in postmortem HIV encephalitis brains [240]. p p [ ] Fungal infection like cerebral aspergillosis leads to increased complement production seen in astrocytes, neu- rons, and oligodendrocytes, especially C1q production by infiltrating macrophages [241]. Some of the defence mech- anisms developed by Aspergillus fumigatus to avoid com- plement include secreting fungal proteases [242] as well as production and recruitment of complement inhibitors [243]. 6. Scavenger Receptors Now, it is well known that com- ponents of innate immunity like complement are present and even produced within the CNS. Neuronal cells [206– 209], astrocytes [210, 211], and microglia [212–214] have been shown to produce complement and also express complement receptors. Role of complement in CNS is considered to be dual-neurotoxic and/or neuroprotective, depending on the level of its activation. Complement has been shown to play a role in adult neurogenesis. Complement receptors C3aR and C5aR are expressed on neural stem cells and reduced neurogenesis is observed in the absence of C3aR signalling [215]. Another complement receptor CR2 has been found to be expressed 7.3. Role in CNS Pathology. CNS can be infected by bacteria, virus, fungus, or protozoa. Deficiency of C3 is associated with Mediators of Inflammation 7 low levels of MBL activity and mice lacking MBL [260]. Immunohistochemistry on brains of stroke patients shows C1q deposition while complement regulator CD59 was found to be absent [261]. Studies involving C5- [262] and C3- deficient mice [263] as well as C1 inhibition [264] have been successful in having beneficial effects in stroke therapy by targeting complement [256, 265]. increased susceptibility to meningococcal and pneumococcal infections [228]. Meningococcus binds to Factor H (fH), a negative regulator of alternative pathway, and evades host innate immune system [229, 230]. Neisseria meningitidis recruits host fH using protein mimicry [231]. Individuals with deficiency of properdin (positive regulator of alternative pathway) are susceptible to meningitis and individuals with combined properdin and MBL deficiency are at increased risk of infection with Neisseria meningitidis [232]. Streptococcus pneumonia infection of CNS is kept in check by complement system (mainly C1q and C3) [233]. C1q and C3 genetically deficient mice each showed considerably high bacterial titres in CNS as compared to wild-type mice. Escherichia coli, a cause for neonatal meningitis, crosses the blood-brain barrier by surviving in the serum where it binds to C4BP [234]. g g A major role for complement is also seen in neurode- generative diseases like AD. The neuropathology in AD includes loss of neurons, extracellular amyloid plaques, and intracellular neurofibrillary tangles consisting of abnormally phosphorylated tau protein [266]. A𝛽activates complement [267], most notably via the classical pathway. Activated com- plement components C1q, C3d, and C4d have been detected in amyloid plaques [268, 269] by immunohistochemistry. C1q binds to A𝛽[270, 271] and modulates phagocytosis of A𝛽 by microglia [272]. 6. Scavenger Receptors In cerebral malaria, C1q and C5 levels have been found to be increased in mice studies [244] while another murine study also points to the requirement of MAC in the pathogenesis of cerebral malaria [245]. Infectious particles called prions cause CNS disorders like Creutzfeldt-Jakob disease and Bovine Spongiform Encephalopathy. These prion particles which activate classical complement pathway [246] are thought to bind to C1q and subsequently transported to the CNS [247]. C1q, C3b have been detected in postmortem brains of individuals with prion diseases [248], and MAC deposition was found to co-relate with prion disease severity [249]. Complement activation occurs in TBI and act as medi- ators of secondary brain injury [250, 251]. Following injury, levels of MAC corelate with blood-brain barrier (BBB) dis- ruption [252]. In mice studies, absence of CD59 (a regulator of MAC formation) leads to increased neuropathology [253]. Postmortem studies on TBI brains show upregulation of C1q, C3b, and MAC [251]. Studies involving mice overex- pressing complement inhibitor CR-related protein y (Crry) show reduced neurological impairment following TBI [254]. Hence, targeting complement activity in TBI may have therapeutic implications [255]. An emerging role for complement in MS has become evident recently [289]. C3d is localized along with microglia in MS tissues [290]. Priming of microglia in MS has been found to be C3-dependent and, in the same study, it was found that in animal model of MS, Crry-deficient mice show exacerbated and accelerated disease progression [291]. Serum factor H has been found to be a useful biomarker for MS [292]. Pathological studies of MS lesions have found presence of complement components C3d, C4b, C1q, and MAC on myelin sheath, surrounding vessel walls, microglia, and astrocytes [293–296]. Cerebral ischemia can lead to the activation of the complement cascade leading to inflammation [256]. Sys- temic complement activity is also found to be enhanced in ischaemic stroke [257]. Complement system is implicated in ischemia reperfusion injury [258]. Ischaemic neurons have been found to produce C5a which causes apoptosis of neurons [259]. Better outcome is seen in individuals with 8 Mediators of Inflammation There is evidence for neuroinflammation in PD as well [297] with the presence of reactive microglia and activated components of complement. Elevated mRNA levels of acti- vated complement and markers of reactive microglia are also seen in PD [298]. Pathological studies show the presence of MAC components intracellularly on the characteristic Lewy Bodies [299, 300]. 6. Scavenger Receptors The cerebrospinal fluid levels of C3 and factor H have been observed to correlate with severity of PD [301]. An interesting study found a role for C1q in PD. Neuromelanin (NM) is a pigment that accumulates in dopaminergic neurons in normal aging process. In PD, these dopaminergic neurons are susceptible to degeneration [302] which is thought to be caused by activation of microglia by NM [303]. Furthermore, this NM pigment is found to be opsonised by C1q and phagocytosed by C1q-positive microglia [304]. its effects. Modulation of cellular processes, phenotypes, and functions looks increasingly likely to be a way forward in combating CNS disorders. Abbreviations A𝛽: Amyloid-𝛽 AD: Alzheimer’s disease BIR: Baculovirus inhibitor of apoptosis protein repeat C4BP: C4b-binding protein CARD: Caspase activation and recruitment domain CNS: Central nervous system Crry: Complement receptor 1-related protein-y DAMP: Damage-associated molecular pattern DOCK8: Dedicator of cytokinesis 8 HSV: Herpes simplex virus HD: Huntington’s disease IL: Interleukin LPS: Lipopolysaccharide MAPK: Mitogen-activated protein kinase MBL: Mannan-binding lectin MASP: MBL-associated serine protease MyD88: Myeloid differentiating factor 88 NF-𝜅𝛽: Nuclear factor-𝜅𝛽 NOD: Nucleotide-binding and oligomerisation domain NLR: NOD-like receptors NM: Neuromelanin PAMP: Pathogen-associated molecular pattern PD: Parkinson’s disease PRR: Pattern-recognition receptor PYD: Pyrin domain RAGE: Receptor for advanced glycation endproducts SR: Scavenger receptor SR-BI: Class B SR type I TBI: Traumatic brain injury TLRs : Toll-like receptors TNF: Tumour necrosis factor WNV: West Nile virus. Huntington’s disease (HD) is another neurodegenerative disorder and a genetic cause of dementia. It is inherited as an autosomal-dominant trait characterised by abnormal (at least 36) CAG repeats on the coding sequence of huntingtin gene [305]. Neuropathological studies in HD brains show presence of complement components C1q, C4, and C3 along with upregulation of complement regulators C1 inhibitor, clusterin, CD59, and CD55. In this study, microglial expres- sion of higher levels of C3 and C9 was also observed [306].h g There has been increasing evidence for involvement of complement in schizophrenia. Schizophrenia is a psychiatric illness characterised by thought insertion, thought with- drawal, hallucinations, delusions, and negative symptoms such as apathy, speech problems, and slow cognition. There is an increase in serum levels of classical pathway complement proteins such as C1q, C1, C3, and C4; increased total comple- ment activity (CH50), CR1 levels; and decreased C4BP levels [307–309]. The alternative pathway is also involved with increased factor B levels and increased activity in serum [310]. MBL pathway shows increased activity as well (increased MBL and MASP-2 levels) [311, 312]. Genetic studies have shown C1QB gene polymorphism, CSMD1 and CSMD2 (code for complement regulatory proteins), C3, MBL2, and MASP2 gene association [313–316]. 8. Conclusion 1. PAMPs are conserved sequences or structural fragments on pathogens (nonself) that are recognised by PRRs. Examples of PAMP include bacterial, viral, fungal, and parasitic-derived lipids (lipopolysaccharide, lipotei- choic acid), proteins (flagellin), carbohydrates (mannan, zymosan), and nucleic acids (dsRNA, CpG). A role for innate immunity in inflammation of CNS is being increasingly evidenced. Cells of the CNS such as neurons, astrocytes, and microglia along with pattern recognition receptors, cytokines, chemokines, complement, peripheral immune cells, and signal pathways form the basis for neuroinflammation. Local synthesis of a number of innate immune humoral factors within CNS offers an opportunity for therapeutic intervention. Furthermore, excessive acti- vation of immune system is thought to be destructive to tissues whereas, simultaneously, it opens up possibilities to harness this activation in a controlled manner to obtain desired therapeutic or preventive strategies in CNS diseases. A detailed understanding of the processes and mechanisms involved in the etiopathogenesis of CNS diseases as well as normal functioning of CNS immunity is essential and can pave the way for reducing excessive neuroinflammation and 2. DAMPs are endogenous molecules released from dam- aged cells (altered self). Examples of DAMP include heat shock proteins, ATP, and uric acid. 3. Toxoplasmosis is caused by Toxoplasma gondii. Cats are the definitive hosts and humans being intermediate hosts of T. gondii. Infection to humans spreads with contamination of food and water by cat faeces as well as eating undercooked meat infected with the parasitic cyst. Clinically, swelling of lymph nodes may occur but, inter- estingly, toxoplasmosis is associated with psychiatric Mediators of Inflammation 9 disorders like schizophrenia, bipolar disorder, anxiety, and personality disorders. disorders like schizophrenia, bipolar disorder, anxiety, and personality disorders. [2] K. Saijo and C. K. Glass, “Microglial cell origin and phenotypes in health and disease,” Nature Reviews Immunology, vol. 11, no. 11, pp. 775–787, 2011. 4. Sleeping sickness is also known as Human African trypanosomiasis. It is caused by Trypanosoma brucei and is transmitted by tsetse fly. Prevalence is mainly in West, Central, and East Africa. It is characterised by intermittent fever and CNS manifestations in late stages including tremors, encephalopathy, and sleep disturbances which is mainly daytime somnolence. [3] F. Ginhoux, M. Greter, M. Leboeuf et al., “Fate mapping analysis reveals that adult microglia derive from primitive macrophages,” Science, vol. 330, no. 6005, pp. 841–845, 2010. [4] A. Nimmerjahn, F. Kirchhoff, and F. 8. Conclusion Helmchen, “Neuroscience: resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo,” Science, vol. 308, no. 5726, pp. 1314–1318, 2005. [5] A. Sierra, J. M. Encinas, J. J. P. Deudero et al., “Microglia shape adult hippocampal neurogenesis through apoptosis-coupled phagocytosis,” Cell Stem Cell, vol. 7, no. 4, pp. 483–495, 2010. 5. Cerebral malaria is encephalopathy caused by sequelae of Plasmodium falciparum infection. Neurological fea- tures include coma, seizures, and upper-motor neuron lesion features (muscle spasticity and rigidity). 5. Cerebral malaria is encephalopathy caused by sequelae of Plasmodium falciparum infection. Neurological fea- tures include coma, seizures, and upper-motor neuron lesion features (muscle spasticity and rigidity). [6] J. Vinet, H. R. Weering, A. Heinrich, R. E. Kalin, A. Wegner, N. Brouwer et al., “Neuroprotective function for ramified microglia in hippocampal excitotoxicity,” Journal of Neuroin- flammation, vol. 9, p. 27, 2012. 6. Neurocysticercosis is an infection of the CNS caused by the tapeworm Taenia solium. Pig is the intermediate host while humans are the definitive hosts of T. solium. Most common clinical presentation is seizures (an important and leading cause for acquired epilepsy) and focal neurological signs depending on the site and localisation of the cysts. 6. Neurocysticercosis is an infection of the CNS caused by the tapeworm Taenia solium. Pig is the intermediate host while humans are the definitive hosts of T. solium. Most common clinical presentation is seizures (an important and leading cause for acquired epilepsy) and focal neurological signs depending on the site and localisation of the cysts. [7] R. C. Paolicelli, G. Bolasco, F. Pagani, L. Maggi, M. Scianni, P. Panzanelli et al., “Synaptic pruning by microglia is necessary for normal brain development,” Science, vol. 333, no. 6048, pp. 1456–1458, 2011. [8] D. P. Schafer, E. K. Lehrman, A. G. Kautzman, R. Koyama, A. R. Mardinly, R. Yamasaki et al., “Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner,” Neuron, vol. 74, no. 4, pp. 691–705, 2012. 7. ALS is also known as motor neurone disease and Lou Gehrig’s disease. Majority of the cases are idiopathic with however a small percentage (5–10%) being familial. Mutations in genes SOD1 (codes for Superoxide dismu- tase 1, an antioxidant); TARDBP (codes for Transactive response DNA-binding protein 43, a nuclear protein); and FUS (codes for Fused in Sarcoma, another cellular protein) are involved in familial ALS. 8. Conclusion It is a fatal, progres- sive neurodegenerative disease characterised by muscle spasticity, wasting and fasciculations as well as dysphagia and dysarthria. Interestingly, ALS is associated with frontotemporal dementia and this lead to discovery of mutation in C9ORF72 gene (abnormal nucleotide repeats) in familial and sporadic forms of ALS [317–319]. [9] H. Kettenmann, U. K. Hanisch, M. Noda, and A. Verkhratsky, “Physiology of microglia,” Physiological Reviews, vol. 91, no. 2, pp. 461–553, 2011. [10] J. K. Olson and S. D. Miller, “Microglia initiate central nervous system innate and adaptive immune responses through multiple TLRs,” The Journal of Immunology, vol. 173, no. 6, pp. 3916–3924, 2004. [11] J. Von Zahn, T. M¨oller, H. Kettenmann, and C. Nolte, “Microglial phagocytosis is modulated by pro-and anti- inflammatory cytokines,” NeuroReport, vol. 8, no. 18, pp. 3851– 3856, 1997. [12] K. Sriram, J. M. Matheson, S. A. Benkovic, D. B. Miller, M. I. Luster, and J. P. O’Callaghan, “Deficiency of TNF receptors suppresses microglial activation and alters the susceptibility of brain regions to MPTP-induced neurotoxicity: role of TNF-𝛼,” The FASEB Journal, vol. 20, no. 6, pp. 670–682, 2006. 8. Hydatidiform mole is a gestational trophoblastic disease. Trophoblasts are precursors to placental cells. The prod- ucts of conception will completely or partially comprise of grape-like vesicles or sacs (villous trophoblast). Most pregnancies are not viable with presenting symptom being vaginal bleeding. Early diagnosis can be estab- lished by ultrasonography (“snowstorm” appearance). [13] C. K. Combs, J. Colleen Karlo, S. C. Kao, and G. E. Landreth, “𝛽-amyloid stimulation of microglia anti monocytes results in TNF𝛼-dependent expression of inducible nitric oxide synthase and neuronal apoptosis,” The Journal of Neuroscience, vol. 21, no. 4, pp. 1179–1188, 2001. 9. Muckle-Wells syndrome is an autosomal dominant dis- ease characterised by the presence of intermittent fevers, rashes, sensorineural hearing loss, and amyloidosis. Mutation occurs in gene CIAS1. [14] C. Barcia, C. M. Ros, V. Annese et al., “IFN-𝛾signaling, with the synergistic contribution of TNF-𝛼, mediates cell specific microglial and astroglial activation in experimental models of Parkinson’s disease,” Cell Death and Disease, vol. 2, no. 4, article e142, 2011. 10. Metabolic syndrome refers to a combination of hyper- glycemia, obesity, dyslipidaemia, and hypertension. [15] A. Basu, J. K. Krady, and S. W. Levison, “Interleukin-1: a mas- ter regulator of neuroinflammation,” Journal of Neuroscience Research, vol. 78, no. 2, pp. 151–156, 2004. 11. 8. Conclusion MicroRNAs are 22 nucleotide RNAs that are noncoding and repress expression of mRNAs. [16] S. S. Shaftel, W. S. T. Griffin, and K. M. Kerry, “The role of interleukin-1 in neuroinflammation and Alzheimer disease: an evolving perspective,” Journal of Neuroinflammation, vol. 5, article 7, 2008. [1] K. Murphy, Innate Immunity: The First Lines of Defense. Jane- way’s Immunobiology, Garland Science, Taylor & Francis Group, Abingdon, UK, 8th edition, 2012. References [1] K. Murphy, Innate Immunity: The First Lines of Defense. Jane- way’s Immunobiology, Garland Science, Taylor & Francis Group, Abingdon, UK, 8th edition, 2012. [17] O. Pascual, S. Ben Achour, P. Rostaing, A. Triller, and A. Bessis, “Microglia activation triggers astrocyte-mediated modulation of excitatory neurotransmission,” Proceedings of the National Mediators of Inflammation 10 Academy of Sciences of the United States of America, vol. 109, no. 4, pp. E197–E205, 2012. astrocytes stimulated with lipopolysaccharide or a neurotropic virus,” Proceedings of the National Academy of Sciences of the United States of America, vol. 86, no. 16, pp. 6348–6352, 1989. [18] [18] D. B. Constam, J. Philipp, U. V. Malipiero, P. Ten Dijke, M. Schachner, and A. Fontana, “Differential expression of trans- forming growth factor-𝛽1, -𝛽2, and -𝛽3 by glioblastoma cells, astrocytes, and microglia,” The Journal of Immunology, vol. 148, no. 5, pp. 1404–1410, 1992. [37] S. van Neerven, A. Nemes, P. Imholz et al., “Inflammatory cytokine release of astrocytes in vitro is reduced by all-trans retinoic acid,” Journal of Neuroimmunology, vol. 229, no. 1-2, pp. 169–179, 2010. [38] M. Johnstone, A. J. H. Gearing, and K. M. Miller, “A central role for astrocytes in the inflammatory response to 𝛽- amyloid; chemokines, cytokines and reactive oxygen species are pro- duced,” Journal of Neuroimmunology, vol. 93, no. 1-2, pp. 182– 193, 1999. [19] P. A. Lodge and S. Sriram, “Regulation of microglial activation by TGF-𝛽, IL-10, and CSF-1,” Journal of Leukocyte Biology, vol. 60, no. 4, pp. 502–508, 1996. [20] K. Williams, N. Dooley, E. Ulvestad, B. Becher, and J. P. Antel, “IL-10 production by adult human derived microglial cells,” Neurochemistry International, vol. 29, no. 1, pp. 55–64, 1996. [39] C. J. Garwood, A. M. Pooler, J. Atherton, D. P. Hanger, and W. Noble, “Astrocytes are important mediators of A𝛽-induced neurotoxicity and tau phosphorylation in primary culture,” Cell Death and Disease, vol. 2, no. 6, article e167, 2011. [21] S. Rivest, “Regulation of innate immune responses in the brain,” Nature Reviews Immunology, vol. 9, no. 6, pp. 429–439, 2009. [22] I. Glezer and S. Rivest, “Glucocorticoids: protectors of the brain during innate immune responses,” Neuroscientist, vol. 10, no. 6, pp. 538–552, 2004. [40] M. V. Sofroniew and H. V. Vinters, “Astrocytes: biology and pathology,” Acta Neuropathologica, vol. 119, no. 1, pp. 7–35, 2010. [41] G. E. Barreto, J. Gonzalez, F. Capani, and L. References Morales, “Role of astrocytes in neurodegenerative diseases,” in Neurodegenerative Diseases—Processes, Prevention, Protection and Monitoring, R. C. C. Chang, Ed., InTech, 2011, http://www.intechopen.com/ books/neurodegenerative-diseases-processes-prevention- protection-and-monitoring/role-of-astrocytes-in-neurode- generative-diseases. [23] S. Nadeau and S. Rivest, “Glucocorticoids play a fundamental role in protecting the brain during innate immune response,” The Journal of Neuroscience, vol. 23, no. 13, pp. 5536–5544, 2003. [24] S. F. Sorrells, J. R. Caso, C. D. Munhoz, and R. M. Sapolsky, “The stressed CNS: when glucocorticoids aggravate inflammation,” Neuron, vol. 64, no. 1, pp. 33–39, 2009. [25] K. Schroder and J. Tschopp, “The Inflammasomes,” Cell, vol. 140, no. 6, pp. 821–832, 2010. [42] T. Kawai and S. Akira, “Signaling to NF-𝜅B by toll-like recep- tors,” Trends in Molecular Medicine, vol. 13, no. 11, pp. 460–469, 2007. [26] A. Halle, V. Hornung, G. C. Petzold et al., “The NALP3 inflammasome is involved in the innate immune response to amyloid-𝛽,” Nature Immunology, vol. 9, no. 8, pp. 857–865, 2008. [43] S. Akira, S. Uematsu, and O. Takeuchi, “Pathogen recognition and innate immunity,” Cell, vol. 124, no. 4, pp. 783–801, 2006. [27] F. Shi, L. Yang, M. Kouadir, Y. Yang, J. Wang, X. Zhou et al., “The NALP3 inflammasome is involved in neurotoxic prion peptide- induced microglial activation,” Journal of Neuroinflammation, vol. 9, p. 73, 2012. [44] L. A. J. O’Neill and A. G. Bowie, “The family of five: TIR- domain-containing adaptors in Toll-like receptor signalling,” Nature Reviews Immunology, vol. 7, no. 5, pp. 353–364, 2007. [45] A. Ozinsky, D. M. Underhill, J. D. Fontenot et al., “The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors,” Proceedings of the National Academy of Sciences of the United States of America, vol. 97, no. 25, pp. 13766–13771, 2000. [28] S. D. Webster, M. Park, M. I. Fonseca, and A. J. Tenner, “Structural and functional evidence for microglial expression of C1qR(p), the C1q receptor that enhances phagocytosis,” Journal of Leukocyte Biology, vol. 67, no. 1, pp. 109–116, 2000. [29] V. Matyash and H. Kettenmann, “Heterogeneity in astrocyte morphology and physiology,” Brain Research Reviews, vol. 63, no. 1-2, pp. 2–10, 2010. [46] M. Oosting, H. Ter Hofstede, P. Sturm, G. J. Adema, B. J. Kull- berg, J. W. van der Meer et al., “TLR1/TLR2 heterodimers play an important role in the recognition of borrelia spirochetes,” PLoS ONE, vol. 6, no. 10, Article ID e25998, 2011. [30] M. M. References Halassa and P. G. Haydon, “Integrated brain circuits: astrocytic networks modulate neuronal activity and behavior,” Annual Review of Physiology, vol. 72, pp. 335–355, 2009. [47] F. Leulier and B. Lemaitre, “Toll-like receptors—taking an evolutionary approach,” Nature Reviews Genetics, vol. 9, no. 3, pp. 165–178, 2008. [31] C. Henneberger, T. Papouin, S. H. R. Oliet, and D. A. Rusakov, “Long-term potentiation depends on release of d-serine from astrocytes,” Nature, vol. 463, no. 7278, pp. 232–236, 2010. [48] J. Siednienko, T. Gajanayake, K. A. Fitzgerald, P. Moynagh, and S. M. Miggin, “Absence of MyD88 results in enhanced TLR3- dependent phosphorylation of IRF3 and increased IFN-𝛽and RANTES production,” The Journal of Immunology, vol. 186, no. 4, pp. 2514–2522, 2011. [32] M. M. Halassa, C. Florian, T. Fellin et al., “Astrocytic modulation of sleep homeostasis and cognitive consequences of sleep loss,” Neuron, vol. 61, no. 2, pp. 213–219, 2009. [33] A. Suzuki, S. A. Stern, O. Bozdagi et al., “Astrocyte-neuron lactate transport is required for long-term memory formation,” Cell, vol. 144, no. 5, pp. 810–823, 2011. [49] X. Liu, Z. Zhan, D. Li et al., “Intracellular MHC class II molecules promote TLR-triggered innate immune responses by maintaining activation of the kinase Btk,” Nature Immunology, vol. 12, no. 5, pp. 416–424, 2011. [34] M. V. Sofroniew, “Molecular dissection of reactive astrogliosis and glial scar formation,” Trends in Neurosciences, vol. 32, no. 12, pp. 638–647, 2009. [50] J. M. Yuk, D. M. Shin, H. M. Lee, J. J. Kim, S. W. Kim, H. S. Jin et al., “The orphan nuclear receptor SHP acts as a negative regulator in inflammatory signaling triggered by toll- like receptors,” Nature Immunology, vol. 12, no. 8, pp. 742–751, 2011. [35] C. Farina, F. Aloisi, and E. Meinl, “Astrocytes are active players in cerebral innate immunity,” Trends in Immunology, vol. 28, no. 3, pp. 138–145, 2007. [36] A. P. Lieberman, P. M. Pitha, H. S. Shin, and M. L. Shin, “Production of tumor necrosis factor and other cytokines by [51] H. H. Jabara, D. R. McDonald, E. Janssen, M. J. Massaad, N. Ramesh, A. Borzutzky et al., “DOCK8 functions as an adaptor Mediators of Inflammation 11 that links TLR-MyD88 signaling to B cell activation,” Nature Immunology, vol. 2012 13, no. 6, pp. 612–620. [67] I. Glezer, A. R. Simard, and S. Rivest, “Neuroprotective role of the innate immune system by microglia,” Neuroscience, vol. 147, no. 4, pp. 867–883, 2007. [52] X. References [57] J. Drouin-Ouellet, C. Gibrat, M. Bousquet, F. Calon, J. Kriz, and F. Cicchetti, “The role of the MYD88-dependent pathway in MPTP-induced brain dopaminergic degeneration,” Journal of Neuroinflammation, vol. 8, p. 137, 2011. [74] E. Okun, K. Griffioen, B. Barak et al., “Toll-like receptor 3 inhibits memory retention and constrains adult hippocampal neurogenesis,” Proceedings of the National Academy of Sciences of the United States of America, vol. 107, no. 35, pp. 15625–15630, 2010. [58] Z. Zheng, R. Yuan, M. Song, Y. Huo, W. Liu, X. Cai et al., “The toll-like receptor 4-mediated signaling pathway is activated following optic nerve injury in mice,” Brain Research, vol. 1489, pp. 90–97, 2012. [59] D. Trudler, D. Frenkel, and D. Farfara, “Toll-like receptors expression and signaling in glia cells in neuro-amyloidogenic diseases: towards future therapeutic application,” Mediators of Inflammation, vol. 2010, Article ID 497987, 12 pages, 2010. [75] D. Kaul, P. Habbel, K. Derkow, C. Kruger, E. Franzoni, F. G. Wulczyn et al., “Expression of toll-like receptors in the developing brain,” PLoS ONE, vol. 7, no. 5, Article ID e37767, 2012. [60] J. E. Cole, E. Georgiou, and C. Monaco, “The expression and functions of toll-like receptors in atherosclerosis,” Mediators of Inflammation, vol. 2010, Article ID 393946, 18 pages, 2010. [76] D. P. McKernan, U. Dennison, G. Gaszner, J. F. Cryan, and T. G. Dinan, “Enhanced peripheral toll-like receptor responses in psychosis: further evidence of a pro-inflammatory phenotype,” Transl Psychiatry, vol. 1, no. 8, p. e36, 2011. [61] M. F. Tsan and B. Gao, “Endogenous ligands of Toll-like receptors,” Journal of Leukocyte Biology, vol. 76, no. 3, pp. 514– 519, 2004. [77] G. Zimmerman, G. Shaltiel, S. Barbash, J. Cohen, C. J. Gasho, S. Shenhar-Tsarfaty et al., “Post-traumatic anxiety associates with failure of the innate immune receptor TLR9 to evade the pro- inflammatory NF𝜅B pathway,” Transl Psychiatry, vol. 2, p. e78, 2012. [62] A. A. Beg, “Endogenous ligands of Toll-like receptors: Impli- cations for regulating inflammatory and immune responses,” Trends in Immunology, vol. 23, no. 11, pp. 509–512, 2002. [63] S. Lacroix, D. Feinstein, and S. Rivest, “The bacterial endotoxin lipopolysaccharide has the ability to target the brain in upreg- ulating its membrane CD14 receptor within specific cellular populations,” Brain Pathology, vol. 8, no. 4, pp. 625–640, 1998. [78] M. Klein, B. Obermaier, B. References Wang, L. C. Loram, K. Ramos, A. J. de Jesus, J. Thomas, K. Cheng et al., “Morphine activates neuroinflammation in a manner parallel to endotoxin,” Proceedings of the National Academy of Sciences of the United States of America, vol. 109, no. 16, pp. 6325–6330, 2012. [68] A. S. Harms, J. K. Lee, T. A. Nguyen, J. Chang, K. M. Ruhn, I. Trevino et al., “Regulation of microglia effector functions by tumor necrosis factor signaling,” Glia, vol. 60, no. 2, pp. 189–202, 2012. [69] C. S. Jack, N. Arbour, J. Manusow et al., “TLR signaling tailors innate immune responses in human microglia and astrocytes,” The Journal of Immunology, vol. 175, no. 7, pp. 4320–4330, 2005. [53] T. F. Pais, C. Figueiredo, R. Peixoto, M. H. Braz, and S. Chatterjee, “Necrotic neurons enhance microglial neurotoxicity through induction of glutaminase by a MyD88-dependent pathway,” Journal of Neuroinflammation, vol. 5, article 43, 2008. [70] A. E. Packard, P. Y. Leung, K. B. Vartanian, S. L. Stevens, F. R. Bahjat, and M. P. Stenzel-Poore, “TLR9 bone marrow chimeric mice define a role for cerebral TNF in neuroprotection induced by CpG preconditioning,” Journal of Cerebral Blood Flow & Metabolism, vol. 32, no. 12, pp. 2193–2200, 2012. [54] F. Bolanle, M. Yongshan, S. Maria, L. Modinat, and J. Hal- lenbeck, “Downstream toll-like receptor signaling mediates adaptor-specific cytokine expression following focal cerebral ischemia,” Journal of Neuroinflammation, vol. 9, p. 174, 2012. [71] K. Hyakkoku, J. Hamanaka, K. Tsuruma et al., “Toll-like re- ceptor 4 (TLR4), but not TLR3 or TLR9, knock-out mice have neuroprotective effects against focal cerebral ischemia,” Neuroscience, vol. 171, no. 1, pp. 258–267, 2010. [55] S. Lin, Q. Yin, Q. Zhong, F. L. Lv, Y. Zhou, J. Q. Li et al., “Heme activates TLR4-mediated inflammatory injury via MyD88/TRIF signaling pathway in intracerebral hemorrhage,” Journal of Neuroinflammation, vol. 9, p. 46, 2012. [72] A. Rolls, R. Shechter, A. London et al., “Toll-like receptors mod- ulate adult hippocampal neurogenesis,” Nature Cell Biology, vol. 9, no. 9, pp. 1081–1088, 2007. [56] S. Liu and T. Kielian, “MyD88 is pivotal for immune recognition of Citrobacter koseri and astrocyte activation during CNS infection,” Journal of Neuroinflammation, vol. 8, article 35, 2011. [73] Y. Ma, J. Li, I. Chiu et al., “Toll-like receptor 8 functions as a negative regulator of neurite outgrowth and inducer of neuronal apoptosis,” The Journal of Cell Biology, vol. 175, no. 2, pp. 209– 215, 2006. References [99] H. Scholtzova, R. J. Kascsak, K. A. Bates et al., “Induction of toll-like receptor 9 signaling as a method for ameliorating Alzheimer’s disease-related pathology,” The Journal of Neuro- science, vol. 29, no. 6, pp. 1846–1854, 2009. [85] L. S. Reinert, L. Harder, C. K. Holm, M. B. Iversen, K. A. Horan, F. Dagnaes-Hansen et al., “TLR3 deficiency renders astrocytes permissive to herpes simplex virus infection and facilitates establishment of CNS infection in mice,” The Journal of Clinical Investigation, vol. 122, no. 4, pp. 1368–1376, 2012. [100] Y. Doi, T. Mizuno, Y. Maki et al., “Microglia activated with the toll-like receptor 9 ligand CpG attenuate oligomeric amyloid 𝛽neurotoxicity in in vitro and in vivo models of Alzheimer’s disease,” American Journal of Pathology, vol. 175, no. 5, pp. 2121– 2132, 2009. [86] T. Wang, T. Town, L. Alexopoulou, J. F. Anderson, E. Fikrig, and R. A. Flavell, “Toll-like receptor 3 mediates west nile virus entry into the brain causing lethal encephalitis,” Nature Medicine, vol. 10, no. 12, pp. 1366–1373, 2004. [101] P. Iribarren, K. Chen, J. Hu et al., “CpG-containing oligodeox- ynucleotide promotes microglial cell uptake of amyloid 𝛽1- 42 peptide by up-regulating the expression of the G-protein- coupled receptor mFPR2,” The FASEB Journal, vol. 19, no. 14, pp. 2032–2034, 2005. [87] B. B. Mishra, U. M. Gundra, and J. M. Teale, “Toll-like receptors in CNS parasitic infections,” Current Topics in Microbiology and Immunology, vol. 336, no. 1, pp. 83–104, 2009. [102] J. Kang and S. Rivest, “MyD88-deficient bone marrow cells accelerate onset and reduce survival in a mouse model of amyotrophic lateral sclerosis,” The Journal of Cell Biology, vol. 179, no. 6, pp. 1219–1230, 2007. [88] J. A. Greene, N. Sam-Agudu, C. C. John, R. O. Opoka, P. A. Zim- merman, and J. W. Kazura, “Toll-like receptor polymorphisms and cerebral malaria: TLR2 Δ22 polymorphism is associated with protection from cerebral malaria in a case control study,” Malaria Journal, vol. 11, p. 47, 2012. [103] Y. Liu, W. Hao, A. Dawson, S. Liu, and K. Fassbender, “Expres- sion of amyotrophic lateral sclerosis-linked SOD1 mutant increases the neurotoxic potential of microglia via TLR2,” The Journal of Biological Chemistry, vol. 284, no. 6, pp. 3691–3699, 2009. [89] B. S. Franklin, S. T. Ishizaka, M. Lamphier et al., “Therapeutical targeting of nucleic acid-sensing toll-like receptors prevents experimental cerebral malaria,” Proceedings of the National Academy of Sciences of the United States of America, vol. References Angele et al., “Innate immunity to pneumococcal infection of the central nervous system depends on toll-like receptor (TLR) 2 and TLR4,” The Journal of Infectious Diseases, vol. 198, no. 7, pp. 1028–1036, 2008. [64] N. Quan, M. Whiteside, L. Kim, and M. Herkenham, “Induction of inhibitory factor 𝜅B𝛼mRNA in the central nervous system after peripheral lipopolysaccharide administration: an in situ hybridization histochemistry study in the rat,” Proceedings of the National Academy of Sciences of the United States of America, vol. 94, no. 20, pp. 10985–10990, 1997. [79] M. Letiembre, H. Echchannaoui, F. Ferracin, S. Rivest, and R. Landmann, “Toll-like receptor-2 deficiency is associated with enhanced brain TNF gene expression during pneumococcal meningitis,” Journal of Neuroimmunology, vol. 168, no. 1-2, pp. 21–33, 2005. [80] R. Dutta, A. Krishnan, J. Meng, S. Das, J. Ma, S. Banerjee et al., “Morphine modulation of toll-like receptors in microglial cells potentiates neuropathogenesis in a HIV-1 model of coinfection with pneumococcal pneumoniae,” The Journal of Neuroscience, vol. 32, no. 29, pp. 9917–9930, 2012. [65] I. Zanoni, R. Ostuni, L. R. Marek, S. Barresi, R. Barbalat, G. M. Barton et al., “CD14 controls the LPS-induced endocytosis of toll-like receptor 4,” Cell, vol. 147, no. 4, pp. 868–880, 2011. [66] I. Glezer, A. Chernomoretz, S. David, M. M. Plante, and S. Rivest, “Genes involved in the balance between neuronal survival and death during inflammation,” PLoS ONE, vol. 2, no. 3, article e310, 2007. [81] N. Arpaia, J. Godec, L. Lau et al., “TLR signaling is required for salmonella typhimurium virulence,” Cell, vol. 144, no. 5, pp. 675–688, 2011. Mediators of Inflammation 12 [82] L. A. Morrison, “The Toll of herpes simplex virus infection,” Trends in Microbiology, vol. 12, no. 8, pp. 353–356, 2004. cognitive deficits in a mouse model of Alzheimer’s disease,” Journal of Neuroinflammation, p. 92, 2011. [83] L. N. Sørensen, L. S. Reinert, L. Malmgaard, C. Bartholdy, A. R. Thomsen, and S. R. Paludan, “TLR2 and TLR9 synergistically control herpes simplex virus infection in the brain,” The Journal of Immunology, vol. 181, no. 12, pp. 8604–8612, 2008. [98] D. L. Herber, L. M. Roth, D. Wilson et al., “Time-dependent reduction in A𝛽levels after intracranial LPS administration in APP transgenic mice,” Experimental Neurology, vol. 190, no. 1, pp. 245–253, 2004. [84] S. Y. Zhang, E. Jouanguy, S. Ugolini et al., “TLR3 deficiency in patients with herpes simplex encephalitis,” Science, vol. 317, no. 5844, pp. 1522–1527, 2007. References 108, no. 9, pp. 3689–3694, 2011. [104] W. Zhao, D. R. Beers, J. S. Henkel et al., “Extracellular mutant SOD1 induces microglial-mediated motoneuron injury,” Glia, vol. 58, no. 2, pp. 231–243, 2010. [90] X. Zhu, Y. Pan, Y. Li, Y. Jiang, H. Shang, D. C. Gowda et al., “Targeting toll-like receptors by chloroquine protects mice from experimental cerebral malaria,” International Immunopharma- cology, vol. 13, no. 4, pp. 392–397, 2012. [105] D. Beraud and K. A. Maguire-Zeiss, “Misfolded alpha-synuclein and toll-like receptors: therapeutic targets for parkinson’s dis- ease,” Parkinsonism & Related Disorders, 1, pp. S17–S20, 2012. [91] J. P. Michaud, K. L. Richard, and S. Rivest, “MyD88-adaptor protein acts as a preventive mechanism for memory deficits in a mouse model of Alzheimer’s disease,” Molecular Neurodegen- eration, vol. 6, no. 1, article 5, 2011. [106] D. S. Spinner, S. C. In, Y. P. Seung et al., “Accelerated prion disease pathogenesis in toll-like receptor 4 signaling-mutant mice,” Journal of Virology, vol. 82, no. 21, pp. 10701–10708, 2008. [92] J. P. Michaud, K. L. Richard, and S. Rivest, “Hematopoietic MyD88-adaptor protein acts as a natural defense mechanism for cognitive deficits in Alzheimer’s disease,” Stem Cell Reviews and Reports, vol. 8, no. 3, pp. 898–904, 2012. [107] S. Sethi, G. Lipford, H. Wagner, and H. Kretzschmar, “Postex- posure prophylaxis against prion disease with a stimulator of innate immunity,” The Lancet, vol. 360, no. 9328, pp. 229–230, 2002. [93] W. Hao, Y. Liu, S. Liu et al., “Myeloid differentiation factor 88-deficient bone marrow cells improve Alzheimer’s disease- related symptoms and pathology,” Brain, vol. 134, no. 1, pp. 278– 292, 2011. [108] M. Prinz, M. Heikenwalder, P. Schwarz, K. Takeda, S. Akira, and A. Aguzzi, “Prion pathogenesis in the absence of toll-like receptor signalling,” The EMBO Reports, vol. 4, no. 2, pp. 195– 199, 2003. [94] K. L. Richard, M. Filali, P. Pr´efontaine, and S. Rivest, “Toll-like receptor 2 acts as a natural innate immune receptor to clear amyloid 𝛽1−42 and delay the cognitive decline in a mouse model of Alzheimer’s disease,” The Journal of Neuroscience, vol. 28, no. 22, pp. 5784–5793, 2008. [109] V. S. Chauhan, D. G. Sterka, S. R. Furr, and I. Marriott, “NOD2 plays an important role in the inflammatory responses of microglia and astrocytes to bacterial CNS pathogens,” Glia, vol. 57, no. 4, pp. 414–423, 2009. [110] J. J. References Tambuyzer, P. Ponsaerts, and E. J. Nouwen, “Microglia: gatekeepers of central nervous system immunology,” Journal of Leukocyte Biology, vol. 85, no. 3, pp. 352–370, 2009. [117] A. Poltorak, X. He, I. Smirnova et al., “Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene,” Science, vol. 282, no. 5396, pp. 2085–2088, 1998. [133] M. Ohya, C. Nishitani, H. Sano et al., “Human pulmonary surfactant protein D binds the extracellular domains of Toll-like receptors 2 and 4 through the carbohydrate recognition domain by a mechanism different from its binding to phosphatidylinos- itol and lipopolysaccharide,” Biochemistry, vol. 45, no. 28, pp. 8657–8664, 2006. [118] O. Takeuchi, S. Sato, T. Horiuchi et al., “Cutting edge: role of Toll-like receptor 1 in mediating immune response to microbial lipoproteins,” The Journal of Immunology, vol. 169, no. 1, pp. 10– 14, 2002. [134] K. Midwood, S. Sacre, A. M. Piccinini et al., “Tenascin-C is an endogenous activator of Toll-like receptor 4 that is essential for maintaining inflammation in arthritic joint disease,” Nature Medicine, vol. 15, no. 7, pp. 774–780, 2009. [119] O. Takeuchi, K. Hoshino, T. Kawai et al., “Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram- positive bacterial cell wall components,” Immunity, vol. 11, no. 4, pp. 443–451, 1999. [135] A. M. Piccinini and K. S. Midwood, “Endogenous control of immunity against infection: tenascin-C regulates TLR4- mediated inflammation via microRNA-155,” Cell Reports, vol. 2, no. 4, pp. 914–926, 2012. [120] O. Takeuchi, T. Kawai, P. F. M¨uhlradt et al., “Discrimination of bacterial lipoproteins by Toll-like recepttor 6,” International Immunology, vol. 13, no. 7, pp. 933–940, 2001. [121] R. Schwandner, R. Dziarski, H. Wesche, M. Rothe, and C. J. Kirschning, “Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by Toll-like receptor 2,” The Journal of Biological Chemistry, vol. 274, no. 25, pp. 17406–17409, 1999. [136] J. Sokolove, X. Zhao, P. E. Chandra, and W. H. Robinson, “Immune complexes containing citrullinated fibrinogen cos- timulate macrophages via toll-like receptor 4 and Fc𝛾receptor,” Arthritis and Rheumatism, vol. 63, no. 1, pp. 53–62, 2011. [122] M. Sato, H. Sano, D. Iwaki et al., “Direct binding of toll- like receptor 2 to Zymosan, and Zymosan-induced NF-𝜅B activation and TNF-𝛼secretion are down-regulated by lung collectin surfactant protein A,” The Journal of Immunology, vol. 171, no. 1, pp. 417–425, 2003. [137] C. P. Hodgkinson, K. Patel, and S. References Ye, “Functional Toll-like receptor 4 mutations modulate the response to fibrinogen,” Thrombosis and Haemostasis, vol. 100, no. 2, pp. 301–307, 2008. [138] S. M. Lehmann, C. Kruger, B. Park, K. Derkow, K. Rosenberger, J. Baumgart et al., “An unconventional role for miRNA: let- 7 activates toll-like receptor 7 and causes neurodegeneration,” Nature Neuroscience, vol. 15, no. 6, pp. 827–835, 2012. [123] L. Alexopoulou, A. C. Holt, R. Medzhitov, and R. A. Flavell, “Recognition of double-stranded RNA and activation of NF-𝜅B by Toll-like receptor 3,” Nature, vol. 413, no. 6857, pp. 732–738, 2001. [139] P. Rosenstiel and S. Schreiber, “NOD-like receptors—pivotal guardians of the immunological integrity of barrier organs,” in Target Pattern Recognition in Innate Immunity, U. Kishore, Ed., pp. 35–47, Landes Bioscience, Austin, Tex, USA, 2009. [124] F. Heil, H. Hemmi, H. Hochrein et al., “Species-specific recog- nition of single-stranded RNA via toll-like receptor 7 and 8,” Science, vol. 303, no. 5663, pp. 1526–1529, 2004. [140] T. Langefeld, W. Mohamed, R. Ghai, and T. Chakraborty, “Toll- like receptors and NOD-like receptors: domain architecture and cellular signalling,” in Target Pattern Recognition in Innate Immunity, U. Kishore, Ed., pp. 48–57, Landes Bioscience, Austin, Tex, USA, 2009. [125] S. S. Diebold, T. Kaisho, H. Hemmi, S. Akira, and C. Reis E Sousa, “Innate antiviral responses by means of TLR7-mediated recognition of single-stranded RNA,” Science, vol. 303, no. 5663, pp. 1529–1531, 2004. [126] F. Hayashi, K. D. Smith, A. Ozinsky et al., “The innate immune response to bacterial flagellin is mediated by toll-like receptor 5,” Nature, vol. 410, no. 6832, pp. 1099–1103, 2001. [141] K. Kersse, M. J. Bertrand, M. Lamkanfi, and P. Vandenabeele, “NOD-like receptors and the innate immune system: coping with danger, damage and death,” Cytokine & Growth Factor Reviews, vol. 22, no. 5-6, pp. 257–276, 2011. [127] H. Hemmi, O. Takeuchi, T. Kawai et al., “A Toll-like receptor recognizes bacterial DNA,” Nature, vol. 408, no. 6813, pp. 740– 745, 2000. [142] T. Bergsbaken, S. L. Fink, and B. T. Cookson, “Pyroptosis: host cell death and inflammation,” Nature Reviews Microbiology, vol. 7, no. 2, pp. 99–109, 2009. [128] T. Oka, S. Hikoso, O. Yamaguchi, M. Taneike, T. Takeda, T. Tamai et al., “Mitochondrial DNA that escapes from autophagy causes inflammation and heart failure,” Nature, vol. 485, no. 7397, pp. 251–255, 2012. [143] R. Cooney, J. Baker, O. References Bajramovic, “Regulation of innate immune responses in the central nervous system,” CNS & Neurological Disorders Drug Targets, vol. 10, no. 1, pp. 4–24, 2011. [95] E. G. Reed-Geaghan, J. C. Savage, A. G. Hise, and G. E. Landreth, “CD14 and toll-like receptors 2 and 4 are required for fibrillar A𝛽-stimulated microglial activation,” The Journal of Neuroscience, vol. 29, no. 38, pp. 11982–11992, 2009. [111] T. Areschoug and S. Gordon, “Scavenger receptors: role in innate immunity and microbial pathogenesis,” Cellular Micro- biology, vol. 11, no. 8, pp. 1160–1169, 2009. [96] S. C. Tang, J. D. Lathia, P. K. Selvaraj et al., “Toll-like receptor-4 mediates neuronal apoptosis induced by amyloid 𝛽-peptide and the membrane lipid peroxidation product 4-hydroxynonenal,” Experimental Neurology, vol. 213, no. 1, pp. 114–121, 2008. [112] K. Wilkinson and J. El Khoury, “Microglial scavenger receptors and their roles in the pathogenesis of Alzheimer’s disease,” International Journal of Alzheimer’s Disease, vol. 2012, Article ID 489456, 10 pages, 2012. [97] M. Song, J. Jin, J. E. Lim et al., “TLR4 mutation reduces microglial activation, increases A𝛽deposits and exacerbates Mediators of Inflammation 13 [113] S. R. Furr, V. Chauhan, D. Sterka, V. Grdzelishvili, and I. Marriott, “Characterization of retinoic acid-inducible gene-I expression in primary murine glia following exposure to vesic- ular stomatitis virus,” Journal of NeuroVirology, vol. 14, no. 6, pp. 503–513, 2008. [130] C. R. Stewart, L. M. Stuart, K. Wilkinson et al., “CD36 ligands promote sterile inflammation through assembly of a Toll-like receptor 4 and 6 heterodimer,” Nature Immunology, vol. 11, no. 2, pp. 155–161, 2010. [131] L. N. Henning, A. K. Azad, K. V. L. Parsa, J. E. Crowther, S. Tridandapani, and L. S. Schlesinger, “Pulmonary surfactant protein a regulates TLR expression and activity in human macrophages,” The Journal of Immunology, vol. 180, no. 12, pp. 7847–7858, 2008. [114] H. Crehan, J. Hardy, and J. Pocock, “Microglia, alzheimer’s disease, and complement,” International Journal of Alzheimer’s Disease, vol. 2012, Article ID 983640, 2012. [115] Y. B. Lee, A. Nagai, and S. U. Kim, “Cytokines, chemokines, and cytokine receptors in human microglia,” The Journal of Neuroscience Research, vol. 69, no. 1, pp. 94–103, 2002. [132] S. Murakami, D. Iwaki, H. Mitsuzawa et al., “Surfactant pro- tein a inhibits peptidoglycan-induced tumor necrosis factor- 𝛼secretion in U937 cells and alveolar macrophages by direct interaction with toll-like receptor 2,” The Journal of Biological Chemistry, vol. 277, no. 9, pp. 6830–6837, 2002. [116] B. R. Mediators of Inflammation 14 a complex that regulates type I interferon and autophagy,” Immunity, vol. 36, no. 6, pp. 933–946, 2012. [160] J. P. De Rivero Vaccari, G. Lotocki, O. F. Alonso, H. M. Bramlett, W. D. Dietrich, and R. W. Keane, “Therapeutic neutralization of the NLRP1 inflammasome reduces the innate immune response and improves histopathology after traumatic brain injury,” Journal of Cerebral Blood Flow and Metabolism, vol. 29, no. 7, pp. 1251–1261, 2009. [145] S. L. Masters, M. Gerlic, D. Metcalf, S. Preston, M. Pellegrini, J. A. O’Donnell et al., “NLRP1 inflammasome activation induces pyroptosis of hematopoietic progenitor cells,” Immunity, vol. 37, no. 6, pp. 1009–1023, 2012. [161] D. P. Abulafia, J. P. De Rivero Vaccari, J. D. Lozano, G. Lotocki, R. W. Keane, and W. D. Dietrich, “Inhibition of the inflammasome complex reduces the inflammatory response after thromboembolic stroke in mice,” Journal of Cerebral Blood Flow and Metabolism, vol. 29, no. 3, pp. 534–544, 2009. [146] M. Lamkanfi and V. M. Dixit, “Inflammasomes and their roles in health and disease,” Annual Review of Cell and Developmental Biology, vol. 28, pp. 137–161, 2012. [147] F. Martinon, A. Mayor, and J. Tschopp, “The inflammasomes: guardians of the body,” Annual Review of Immunology, vol. 27, pp. 229–265, 2009. [162] T. Hoegen, N. Tremel, M. Klein, B. Angele, H. Wagner, C. Kirschning et al., “The NLRP3 inflammasome contributes to brain injury in pneumococcal meningitis and is activated through ATP-dependent lysosomal cathepsin B release,” The Journal of Immunology, vol. 187, no. 10, pp. 5440–5451, 2011. [148] L. Agostini, F. Martinon, K. Burns, M. F. McDermott, P. N. Hawkins, and J. Tschopp, “NALP3 forms an IL-1𝛽-processing inflammasome with increased activity in Muckle-Wells autoin- flammatory disorder,” Immunity, vol. 20, no. 3, pp. 319–325, 2004. [163] D. K. Kaushik, M. Gupta, K. L. Kumawat, and A. Basu, “NLRP3 inflammasome: key mediator of neuroinflammation in murine japanese encephalitis,” PLoS ONE, vol. 7, no. 2, Article ID e32270, 2012. [149] H. Wen, J. P. Ting, and L. A. O’Neill, “A role for the NLRP3 inflammasome in metabolic diseases—did warburg miss inflammation?” Nature Immunology, vol. 13, no. 4, pp. 352– 357, 2012. [164] J. Zou and F. T. Crews, “Inflammasome-IL-1𝛽signaling medi- ates ethanol inhibition of hippocampal neurogenesis,” Frontiers in Neuroscience, vol. 6, p. 77, 2012. [150] T. Strowig, J. Henao-Mejia, E. Elinav, and R. Flavell, “Inflam- masomes in health and disease,” Nature, vol. 481, no. 7381, pp. 278–286, 2012. [165] J. L. Mediators of Inflammation Goldstein, Y. K. Ho, S. K. Basu, and M. S. Brown, “Binding site on macrophages that mediates uptake and degra- dation of acetylated low density lipoprotein, producing massive cholesterol deposition,” Proceedings of the National Academy of Sciences of the United States of America, vol. 76, no. 1, pp. 333– 337, 1979. [151] E. Mezzaroma, S. Toldo, D. Farkas, I. M. Seropian, B. W. Van Tassell, F. N. Salloum et al., “The inflammasome promotes adverse cardiac remodeling following acute myocardial infarc- tion in the mouse,” Proceedings of the National Academy of Sciences of the United States of America, vol. 108, no. 49, pp. 19725–19730, 2011. [166] B. Godoy, P. Murgas, J. Tichauer, and R. Von Bernhardi, “Scavenger receptor class a ligands induce secretion of IL1𝛽and exert a modulatory effect on the inflammatory activation of astrocytes in culture,” Journal of Neuroimmunology, vol. 251, no. 1-2, pp. 6–13, 2012. [152] P. J. Shaw, M. F. McDermott, and T. D. Kanneganti, “Inflamma- somes and autoimmunity,” Trends in Molecular Medicine, vol. 17, no. 2, pp. 57–64, 2011. [167] S. Mukhopadhyay, A. Pluddemann, and S. Gordon, “Macro- phage pattern recognition receptors in immunity, homeostasis and self tolerance,” in Target Pattern Recognition in Innate Immunity, U. Kishore, Ed., pp. 1–14, Landes Bioscience, Austin, Tex, USA, 2009. [153] G. dos Santos, M. A. Kutuzov, and K. M. Ridge, “The inflamma- some in lung diseases,” American Journal of Physiology, vol. 303, no. 8, pp. L627–L633, 2012. [154] T. D. Kanneganti, “Central roles of NLRs and inflammasomes in viral infection,” Nature Reviews Immunology, vol. 10, no. 10, pp. 688–698, 2010. [168] Y. Yamada, T. Doi, T. Hamakubo, and T. Kodama, “Scavenger receptor family proteins: roles for atherosclerosis, host defence and disorders of the central nervous system,” Cellular and Molecular Life Sciences, vol. 54, no. 7, pp. 628–640, 1998.h [155] M. Lamkanfi and V. M. Dixit, “Modulation of inflammasome pathways by bacterial and viral pathogens,” The Journal of Immunology, vol. 187, no. 2, pp. 596–602, 2011. [169] D. R. Greaves and S. Gordon, “The macrophage scavenger receptor at 30 years of age: current knowledge and future challenges,” Journal of Lipid Research, vol. 50, pp. S282–S286, 2009. [156] A. Halle, V. Hornung, G. C. Petzold et al., “The NALP3 inflammasome is involved in the innate immune response to amyloid-𝛽,” Nature Immunology, vol. 9, no. 8, pp. 857–865, 2008. [157] S. Jha, S. Y. Srivastava, W. J. References Brain et al., “NOD2 stimulation induces autophagy in dendritic cells influencing bacterial handling and antigen presentation,” Nature Medicine, vol. 16, no. 1, pp. 90–97, 2010. [129] S. Liu, Y. Liu, W. Hao, L. Wolf, A. J. Kiliaan, B. Penke et al., “TLR2 is a primary receptor for alzheimer’s amyloid beta peptide to trigger neuroinflammatory activation,” The Journal of Immunology, vol. 188, no. 3, pp. 1098–1107, 2012. [144] Y. Lei, H. Wen, Y. Yu, D. J. Taxman, L. Zhang, D. G. Widman et al., “The mitochondrial proteins NLRX1 and TUFM form Mediators of Inflammation Mediators of Inflammation Brickey et al., “The inflammasome sensor, NLRP3, regulates CNS inflammation and demyelination via caspase-1 and interleukin-18,” The Journal of Neuroscience, vol. 30, no. 47, pp. 15811–15820, 2010. [170] F. Wermeling, Y. Chen, T. Pikkarainen et al., “Class A scav- enger receptors regulate tolerance against apoptotic cells, and autoantibodies against these receptors are predictive of systemic lupus,” Journal of Experimental Medicine, vol. 204, no. 10, pp. 2259–2265, 2007. [158] M. Inoue, K. L. Williams, M. D. Gunn, and M. L. Shino- hara, “NLRP3 inflammasome induces chemotactic immune cell migration to the CNS in experimental autoimmune encephalomyelitis,” Proceedings of the National Academy of Sciences of the United States of America, vol. 109, no. 26, pp. 10480–10485, 2012. [171] P. Oquendo, E. Hundt, J. Lawler, and B. Seed, “CD36 directly mediates cytoadherence of plasmodium falciparum parasitized erythrocytes,” Cell, vol. 58, no. 1, pp. 95–101, 1989. [172] B. C. Urban, D. J. P. Ferguson, A. Pain et al., “Plasmodium falciparuminfected erythrocytes modulate the maturation of dendritic cells,” Nature, vol. 400, no. 6739, pp. 73–77, 1999. [159] M. Inoue, K. L. Williams, T. Oliver, P. Vandenabeele, J. V. Rajan, E. A. Miao et al., “Interferon-beta therapy against EAE is effective only when development of the disease depends on the NLRP3 inflammasome,” Science Signaling, vol. 5, no. 225, p. ra38, 2012. [173] B. C. Urban, N. Willcox, and D. J. Roberts, “A role for CD36 in the regulation of dendritic cell function,” Proceedings of the National Academy of Sciences of the United States of America, vol. 98, no. 15, pp. 8750–8755, 2001. Mediators of Inflammation 15 [174] S. L. Stephen, K. Freestone, S. Dunn, M. W. Twigg, S. Homer- Vanniasinkam, J. H. Walker et al., “Scavenger receptors and their potential as therapeutic targets in the treatment of cardiovas- cular disease,” International Journal of Hypertension, vol. 2010, Article ID 646929, 21 pages, 2010. induced cellular perturbation?” Journal of Alzheimer’s Disease, vol. 16, no. 4, pp. 833–843, 2009. [188] Fang, L. F. Lue, S. Yan et al., “RAGE-dependent signaling in microglia contributes to neuroinflammation, A𝛽accumu- lation, and impaired learning/memory in a mouse model of Alzheimer’s disease,” The FASEB Journal, vol. 24, no. 4, pp. 1043– 1055, 2010. [175] S. A. Thakur, R. F. Hamilton, and A. Holian, “Role of scavenger receptor a family in lung inflammation from exposure to environmental particles,” Journal of Immunotoxicology, vol. 5, no. 2, pp. 151–157, 2008. [189] N. Origlia, M. Righi, S. Mediators of Inflammation Capsoni et al., “Receptor for advanced glycation end product-dependent activation of p38 mitogen- activated protein kinase contributes to amyloid-𝛽-mediated cortical synaptic dysfunction,” The Journal of Neuroscience, vol. 28, no. 13, pp. 3521–3530, 2008. [176] A. K. A. Wright, S. Rao, S. Range et al., “Pivotal advance: expansion of small sputum macrophages in CF: failure to express MARCO and mannose receptors,” Journal of Leukocyte Biology, vol. 86, no. 3, pp. 479–489, 2009. [190] K. Takuma, F. Fang, W. Zhang et al., “RAGE-mediated signal- ing contributes to intraneuronal transport of amyloid-𝛽and neuronal dysfunction,” Proceedings of the National Academy of Sciences of the United States of America, vol. 106, no. 47, pp. 20021–20026, 2009. [177] J. El Khoury, S. E. Hickman, C. A. Thomas, L. Cao, S. C. Silver- stein, and J. D. Loike, “Scavenger receptor-mediated adhesion of microglia to 𝛽-amyloid fibrils,” Nature, vol. 382, no. 6593, pp. 716–719, 1996. [191] R. R. Porter and K. B. M. Reid, “Activation of the complement system by antibody-antigen complexes: the classical pathway,” Advances in Protein Chemistry, vol. 33, pp. 1–71, 1979. [178] R. H. Christie, M. Freeman, and B. T. Hyman, “Expression of the macrophage scavenger receptor, a multifunctional lipopro- tein receptor, in microglia associated with senile plaques in Alzheimer’s disease,” The American Journal of Pathology, vol. 148, no. 2, pp. 399–403, 1996. [192] M. R. Dahl, S. Thiel, M. Matsushita et al., “MASP-3 and its association with distinct complexes of the mannan-binding lectin complement activation pathway,” Immunity, vol. 15, no. 1, pp. 127–135, 2001. [179] Y. Xu, L. Qian, G. Zong, K. Ma, X. Zhu, H. Zhang et al., “Class A scavenger receptor promotes cerebral ischemic injury by pivoting microglia/macrophage polarization,” Neuroscience, vol. 218, pp. 35–48, 2012. [193] S. Thiel, T. Vorup-Jensen, C. M. Stover et al., “A second serine protease associated with mannan-binding lectin that activates complement,” Nature, vol. 386, no. 6624, pp. 506–510, 1997. [180] H. Levy-Barazany and D. Frenkel, “Expression of scavenger receptor A on antigen presenting cells is important for CD4+ T-cells proliferation in EAE mouse model,” Journal of Neuroin- flammation, vol. 9, 120 pages, 2012. [194] M. V. Carroll and R. B. Sim, “Complement in health and disease,” Advanced Drug Delivery Reviews, vol. 63, no. 12, pp. 965–975, 2011. [195] Y. H. Kang, L. A. Tan, M. V. Carroll, M. E. Gentle, and R. B. Mediators of Inflammation Sim, “Target pattern recognition by complement proteins of the classical and alternative pathways,” Advances in Experimental Medicine and Biology, vol. 653, pp. 117–128, 2009. [181] F. Huang, M. Buttini, T. Wyss-Coray et al., “Elimination of the class a scavenger receptor does not affect amyloid plaque formation or neurodegeneration in transgenic mice expressing human amyloid protein precursors,” The American Journal of Pathology, vol. 155, no. 5, pp. 1741–1747, 1999. [196] S. Thiel and M. Gadjeva, “Humoral pattern recognition molecules: mannan-binding lectin and ficolins,” Advances in Experimental Medicine and Biology, vol. 653, pp. 58–73, 2009.h [182] J. Husemann and S. C. Silverstein, “Expression of scavenger receptor class B, type I, by astrocytes and vascular smooth muscle cells in normal adult mouse and human brain and in Alzheimer’s disease brain,” The American Journal of Pathology, vol. 158, no. 3, pp. 825–832, 2001. [197] K. B. Reid and R. R. Porter, “The proteolytic activation systems of complement.,” Annual Review of Biochemistry, vol. 50, pp. 433–464, 1981. [198] G. Hajishengallis and J. D. Lambris, “Crosstalk pathways between toll-like receptors and the complement system,” Trends in Immunology, vol. 31, no. 4, pp. 154–163, 2010. [183] R. A. K. Srivastava and J. C. Jain, “Scavenger receptor class B type I expression and elemental analysis in cerebellum and parietal cortex regions of the Alzheimer’s disease brain,” Journal of the Neurological Sciences, vol. 196, no. 1-2, pp. 45–52, 2002. [199] M. C. Carroll and D. E. Isenman, “Regulation of humoral immunity by complement,” Immunity, vol. 37, no. 2, pp. 199–207, 2012. [184] K. Thanopoulou, A. Fragkouli, F. Stylianopoulou, and S. Geor- gopoulos, “Scavenger receptor class B type i (SR-BI) regulates perivascular macrophages and modifies amyloid pathology in an Alzheimer mouse model,” Proceedings of the National Academy of Sciences of the United States of America, vol. 107, no. 48, pp. 20816–20821, 2010. [200] S. V. Petersen, S. Thiel, L. Jensen, T. Vorup-Jensen, C. Koch, and J. C. Jensenius, “Control of the classical and the MBL pathway of complement activation,” Molecular Immunology, vol. 37, no. 14, pp. 803–811, 2001. [185] L. Park, G. Wang, P. Zhou, J. Zhou, R. Pitstick, M. L. Previti et al., “Scavenger receptor CD36 is essential for the cerebrovascular oxidative stress and neurovascular dysfunction induced by amyloid-𝛽,” Proceedings of the National Academy of Sciences of the United States of America, vol. 108, no. 12, pp. 5063–5068, 2011. [201] P. F. Zipfel and C. Mediators of Inflammation 75–82, 1995. [223] M. B´enard, E. Raoult, D. Vaudry et al., “Role of complement anaphylatoxin receptors (C3aR, C5aR) in the development of the rat cerebellum,” Molecular Immunology, vol. 45, no. 14, pp. 3767–3774, 2008. [208] R. Veerhuis, I. Janssen, J. J. M. Hoozemans, C. J. A. De Groot, C. E. Hack, and P. Eikelenboom, “Complement C1-inhibitor expression in Alzheimer’s disease,” Acta Neuropathologica, vol. 96, no. 3, pp. 287–296, 1998. [224] J. Van Beek, O. Nicole, C. Ali et al., “Complement anaphylatoxin c3a is selectively protective against nmda-induced neuronal cell death,” NeuroReport, vol. 12, no. 2, pp. 289–293, 2001. [209] A. Thomas, P. Gasque, D. Vaudry, B. Gonzalez, and M. Fontaine, “Expression of a complete and functional complement system by human neuronal cells in vitro,” International Immunology, vol. 12, no. 7, pp. 1015–1023, 2000. [225] H. Osaka, P. Mukherjee, P. S. Aisen, and G. M. Pasinetti, “Complement-derived anaphylatoxin C5a protects against glutamate-mediated neurotoxicity,” Journal of Cellular Biochem- istry, vol. 73, no. 3, pp. 303–311, 1999. [210] S. R. Barnum, J. L. Jones, and E. N. Benveniste, “Interleukin- 1 and tumor necrosis factor-mediated regulation of C3 gene expression in human astroglioma cells.,” Glia, vol. 7, no. 3, pp. 225–236, 1993. [226] P. Mukherjee and G. M. Pasinetti, “Complement anaphylatoxin C5a neuroprotects through mitogen-activated protein kinase- dependent inhibition of caspase 3,” Journal of Neurochemistry, vol. 77, no. 1, pp. 43–49, 2001.h [211] P. Gasque, P. Chan, C. Mauger et al., “Identification and char- acterization of complement C3 receptors on human astrocytes,” The Journal of Immunology, vol. 156, no. 6, pp. 2247–2255, 1996. [227] P. Mukherjee, S. Thomas, and G. M. Pasinetti, “Complement anaphylatoxin C5a neuroprotects through regulation of glu- tamate receptor subunit 2 in vitro and in vivo,” Journal of Neuroinflammation, vol. 5, article 5, 2008. [212] D. G. Walker, S. U. Kim, and P. L. McGeer, “Complement and cytokine gene expression in cultured microglia derived from postmortem human brains,” The Journal of Neuroscience Research, vol. 40, no. 4, pp. 478–493, 1995. l [228] S. Ram, L. A. Lewis, and P. A. Rice, “Infections of people with complement deficiencies and patients who have undergone splenectomy,” Clinical Microbiology Reviews, vol. 23, no. 4, pp. 740–780, 2010. [213] R. Veerhuis, I. Janssen, C. J. A. De Groot, F. L. Van Muiswinkel, C. E. Hack, and P. Mediators of Inflammation Skerka, “Complement regulators and inhibitory proteins,” Nature Reviews Immunology, vol. 9, no. 10, pp. 729–740, 2009. [202] U. Kishore and R. B. Sim, “Factor H as a regulator of the classical pathway activation,” Immunobiology, vol. 217, no. 2, pp. 162–168, 2012. [186] L. Park, J. Zhou, P. Zhou, R. Pistick, S. El Jamal, L. Younkin et al., “Innate immunity receptor CD36 promotes cerebral amyloid angiopathy,” Proceedings of the National Academy of Sciences of the United States of America, vol. 110, no. 8, pp. 3089–3094, 2013. [203] L. A. Tan, A. C. Yang, U. Kishore, and R. B. Sim, “Interactions of complement proteins C1q and factor H with lipid A and Escherichia coli: further evidence that factor H regulates the classical complement pathway,” Protein & Cell , vol. 2, no. 4, pp. 320–332, 2011. [187] D. Y. Shi, A. Bierhaus, P. P. Nawroth, and D. M. Stern, “RAGE and Alzheimer’s disease: a progression factor for amyloid-𝛽- Mediators of Inflammation 16 [204] L. A. Tan, B. Yu, F. C. J. Sim, U. Kishore, and R. B. Sim, “Com- plement activation by phospholipids: the interplay of factor H and C1q,” Protein & Cell, vol. 1, no. 11, pp. 1033–1049, 2010. [220] A. Nayak, J. Ferluga, A. G. Tsolaki, and U. Kishore, “The non- classical functions of the classical complement pathway recog- nition subcomponent C1q,” Immunology Letters, vol. 131, no. 2, pp. 139–150, 2010. [205] Y. H. Kang, B. C. Urban, R. B. Sim, and U. Kishore, “Human complement factor H modulates C1q-mediated phagocytosis of apoptotic cells,” Immunobiology, vol. 217, no. 4, pp. 455–464, 2012. [221] Y. Chu, X. Jin, I. Parada et al., “Enhanced synaptic connectivity and epilepsy in C1q knockout mice,” Proceedings of the National Academy of Sciences of the United States of America, vol. 107, no. 17, pp. 7975–7980, 2010. [206] D. G. Walker and P. L. McGeer, “Complement gene expression in neuroblastoma and astrocytoma cell lines of human origin,” Neuroscience Letters, vol. 157, no. 1, pp. 99–102, 1993. [222] M. E. Benoit and A. J. Tenner, “Complement protein C1q- mediated neuroprotection is correlated with regulation of neuronal gene and microRNA expression,” The Journal of Neuroscience, vol. 31, no. 9, pp. 3459–3469, 2011. [207] D. G. Walker, O. Yasuhara, P. A. Patston, E. G. McGeer, and P. L. McGeer, “Complement C1 inhibitor is produced by brain tissue and is cleaved in Alzheimer disease,” Brain Research, vol. 675, no. 1-2, pp. Mediators of Inflammation Eikelenboom, “Cytokines associated with amyloid plaques in Alzheimer’s disease brain stimulate human glial and neuronal cell cultures to secrete early complement proteins, but not C1-inhibitor,” Experimental Neurology, vol. 160, no. 1, pp. 289–299, 1999. [229] E. Kugelberg, B. Gollan, and C. M. Tang, “Mechanisms in Neis- seria meningitidis for resistance against complement-mediated killing,” Vaccine, vol. 26, no. 8, pp. I34–I39, 2008. [230] M. C. Schneider, R. M. Exley, H. Chan et al., “Functional significance of factor H binding to Neisseria meningitidis,” The Journal of Immunology, vol. 176, no. 12, pp. 7566–7575, 2006. [214] P. Gasque, S. K. Singhrao, J. W. Neal et al., “The receptor for complement anaphylatoxin C3a is expressed by myeloid cells and nonmyeloid cells in inflamed human central nervous system: analysis in multiple sclerosis and bacterial meningitis,” The Journal of Immunology, vol. 160, no. 7, pp. 3543–3554, 1998. [231] M. C. Schneider, B. E. Prosser, J. J. E. Caesar et al., “Neisseria meningitidis recruits factor H using protein mimicry of host carbohydrates,” Nature, vol. 458, no. 7240, pp. 890–893, 2009. [232] L. Bathum, H. Hansen, B. Teisner et al., “Association between combined properdin and mannose-binding lectin deficiency and infection with Neisseria meningitidis,” Molecular Immunol- ogy, vol. 43, no. 5, pp. 473–479, 2006. [215] Y. Rahpeymai, M. A. Hietala, U. Wilhelmsson et al., “Comple- ment: a novel factor in basal and ischemia-induced neurogene- sis,” The EMBO Journal, vol. 25, no. 6, pp. 1364–1374, 2006. [233] T. A. Rupprecht, B. Angele, M. Klein et al., “Complement C1q and C3 are critical for the innate immune response to Streptococcus pneumoniae in the central nervous system,” The Journal of Immunology, vol. 178, no. 3, pp. 1861–1869, 2007. [216] M. Moriyama, T. Fukuhara, M. Britschgi et al., “Complement receptor 2 is expressed in neural progenitor cells and regulates adult hippocampal neurogenesis,” The Journal of Neuroscience, vol. 31, no. 11, pp. 3981–3989, 2011. [234] D. G. Wooster, R. Maruvada, A. M. Blom, and N. V. Prasadarao, “Logarithmic phase Escherichia coli K1 efficiently avoids serum killing by promoting C4bp-mediated C3b and C4b degrada- tion,” Immunology, vol. 117, no. 4, pp. 482–493, 2006. [217] A. H. Stephan, B. A. Barres, and B. Stevens, “The complement system: an unexpected role in synaptic pruning during develop- ment and disease,” Annual Review of Neuroscience, vol. 35, pp. 369–389, 2012. [235] K. A. Stoermer and T. E. Morrison, “Complement and viral pathogenesis,” Virology, vol. 411, no. 2, pp. Mediators of Inflammation Otten et al., “The role of complement in stroke therapy,” Advances in Experimental Medicine and Biology, vol. 632, pp. 23–33, 2008. [241] G. Rambach, H. Maier, G. Vago et al., “Complement induction and complement evasion in patients with cerebral aspergillosis,” Microbes and Infection, vol. 10, no. 14-15, pp. 1567–1576, 2008. [257] M. Di Napoli, “Early inflammatory response in ischemic stroke,” Thrombosis Research, vol. 103, no. 3, pp. 261–264, 2001. [242] G. Rambach, D. Dum, I. Mohsenipour et al., “Secretion of a fungal protease represents a complement evasion mechanism in cerebral aspergillosis,” Molecular Immunology, vol. 47, no. 7- 8, pp. 1438–1449, 2010. [258] T. V. Arumugam, I. A. Shiels, T. M. Woodruff, D. N. Granger, and S. M. Taylor, “The role of the complement system in ischemia-reperfusion injury.,” Shock, vol. 21, no. 5, pp. 401–409, 2004. [243] C. Speth and G. Rambach, “Complement attack against aspergillus and corresponding evasion mechanisms,” Interdisci- plinary Perspectives on Infectious Diseases, vol. 2012, Article ID 463794, 9 pages, 2012. [259] D. Pavlovski, J. Thundyil, P. N. Monk, R. A. Wetsel, S. M. Taylor, and T. M. Woodruff, “Generation of complement component C5a by ischemic neurons promotes neuronal apoptosis,” The FASEB Journal, vol. 26, no. 9, pp. 3680–3690, 2012. [244] P. Lackner, C. Hametner, R. Beer et al., “Complement factors C1q, C3 and C5 in brain and serum of mice with cerebral malaria,” Malaria Journal, vol. 7, article 207, 2008. [260] A. Cervera, A. M. Planas, C. Justicia et al., “Genetically- defined deficiency of mannose-binding lectin is associated with protection after experimental stroke in mice and outcome in human stroke,” PLoS ONE, vol. 5, no. 2, Article ID e8433, 2010. [245] T. N. Ramos, M. M. Darley, X. Hu et al., “Cutting edge: the membrane attack complex of complement is required for the development of murine experimental cerebral malaria,” The Journal of Immunology, vol. 186, no. 12, pp. 6657–6660, 2011. [261] E. D. Pedersen, E. M. Løberg, E. Vege, M. R. Daha, J. Mæhlen, and T. E. Mollnes, “In situ deposition of complement in human acute brain ischaemia,” Scandinavian Journal of Immunology, vol. 69, no. 6, pp. 555–562, 2009. [246] D. A. Mitchell, L. Kirby, S. M. Paulin, C. L. Villiers, and R. B. Sim, “Prion protein activates and fixes complement directly via the classical pathway: implications for the mechanism of scrapie agent propagation in lymphoid tissue,” Molecular Immunology, vol. 44, no. 11, pp. 2997–3004, 2007. Mediators of Inflammation 362–373, 2011. [218] L. M. Boulanger, “Immune proteins in brain development and synaptic plasticity,” Neuron, vol. 64, no. 1, pp. 93–109, 2009. [236] S. B. Kapadia, B. Levine, S. H. Speck, and H. W. Virgin, “Critical role of complement and viral evasion of complement in acute, persistent, and latent 𝛾-herpesvirus infection,” Immunity, vol. 17, no. 2, pp. 143–155, 2002. [219] B. Stevens, N. J. Allen, L. E. Vazquez et al., “The classical complement cascade mediates CNS synapse elimination,” Cell, vol. 131, no. 6, pp. 1164–1178, 2007. Mediators of Inflammation 17 [237] E. Mehlhop, K. Whitby, T. Oliphant, A. Marri, M. Engle, and M. S. Diamond, “Complement activation is required for induc- tion of a protective antibody response against west nile virus infection,” Journal of Virology, vol. 79, no. 12, pp. 7466–7477, 2005. [252] P. F. Stahel, M. C. Morganti-Kossmann, D. Perez et al., “Intrathecal levels of complement-derived soluble membrane attack complex (sC5b-9) correlate with blood-brain barrier dysfunction in patients with traumatic brain injury,” Journal of Neurotrauma, vol. 18, no. 8, pp. 773–781, 2001. [253] P. F. Stahel, M. A. Flierl, B. P. Morgan et al., “Absence of the complement regulatory molecule CD59a leads to exacerbated neuropathology after traumatic brain injury in mice,” Journal of Neuroinflammation, vol. 6, article 2, 2009. [238] A. Fuchs, A. K. Pinto, W. J. Schwaeble, and M. S. Diamond, “The lectin pathway of complement activation contributes to protection from west nile virus infection,” Virology, vol. 412, no. 1, pp. 101–109, 2011. [254] M. Rancan, M. C. Morganti-Kossmann, S. R. Barnum et al., “Central nervous system-targeted complement inhibition mediates neuroprotection after closed head injury in transgenic mice,” Journal of Cerebral Blood Flow and Metabolism, vol. 23, no. 9, pp. 1070–1074, 2003. [239] J. E. Libbey, N. J. Kirkman, K. S. Wilcox, H. S. White, and R. S. Fujinami, “Role for complement in the development of seizures following acute viral infection,” Journal of Virology, vol. 84, no. 13, pp. 6452–6460, 2010. [240] K. K. Singh, S. Nathamu, A. Adame, T. U. Alire, W. Dumaop, B. Gouaux et al., “Expression of mannose binding lectin in HIV- 1-infected brain: implications for HIV-related neuronal damage and neuroAIDS,” Neurobehavioral HIV Medicine, vol. 3, pp. 41– 52, 2011. [255] V. Yanamadala and R. M. Friedlander, “Complement in neu- roprotection and neurodegeneration,” Trends in Molecular Medicine, vol. 16, no. 2, pp. 69–76, 2010. [256] R. J. Komotar, G. H. Kim, M. L. Mediators of Inflammation 241–253, 2004. [288] W. J. Lukiw, B. Surjyadipta, P. Dua, and P. N. Alexandrov, “Common micro RNAs (miRNAs) target complement factor H, (CFH) regulation in alzheimer’s disease (AD) and in age- related macular degeneration (AMD),” International Journal of Biochemistry and Molecular Biology, vol. 3, no. 1, pp. 105–116, 2012. [274] M. Sarvari, I. Vago, C. S. Weber, J. Nagy, P. Gal, M. Mak et al., “Inhibition of C1q-beta-amyloid binding protects hippocampal cells against complement mediated toxicity,” Journal of Neu- roimmunology, vol. 137, no. 1-2, pp. 12–18, 2003. [275] M. I. Fonseca, J. Zhou, M. Botto, and A. J. Tenner, “Absence of C1q leads to less neuropathology in transgenic mouse models of Alzheimer’s disease,” The Journal of Neuroscience, vol. 24, no. 29, pp. 6457–6465, 2004. [289] G. Ingram, S. Hakobyan, N. P. Robertson, and B. P. Morgan, “Complement in multiple sclerosis: its role in disease and poten- tial as a biomarker,” Clinical and Experimental Immunology, vol. 155, no. 2, pp. 128–139, 2009. [276] R. Strohmeyer, Y. Shen, and J. Rogers, “Detection of com- plement alternative pathway mRNA and proteins in the Alzheimer’s disease brain,” Molecular Brain Research, vol. 81, no. 1-2, pp. 7–18, 2000. [290] M. H. Barnett, J. D. E. Parratt, E. S. Cho, and J. W. Prineas, “Immunoglobulins and complement in postmortem multiple sclerosis tissue,” Annals of Neurology, vol. 65, no. 1, pp. 32–46, 2009. [277] R. Strohmeyer, M. Ramirez, G. J. Cole, K. Mueller, and J. Rogers, “Association of factor H of the alternative pathway of complement with agrin and complement receptor 3 in the Alzheimer’s disease brain,” Journal of Neuroimmunology, vol. 131, no. 1-2, pp. 135–146, 2002. [291] V. Ramaglia, T. R. Hughes, R. M. Donev, M. M. Ruseva, X. Wu, I. Huitinga et al., “C3-dependent mechanism of microglial prim- ing relevant to multiple sclerosis,” Proceedings of the National Academy of Sciences of the United States of America, vol. 109, no. 3, pp. 965–970, 2012. [278] L. A. Trouw, H. M. Nielsen, L. Minthon et al., “C4b-binding protein in Alzheimer’s disease: binding to A𝛽1-42 and to dead cells,” Molecular Immunology, vol. 45, no. 13, pp. 3649–3660, 2008. [292] G. Ingram, S. Hakobyan, C. L. Hirst et al., “Complement regulator factor H as a serum biomarker of multiple sclerosis disease state,” Brain, vol. 133, no. 6, pp. 1602–1611, 2010. [279] S. Haga, K. Akai, and T. Mediators of Inflammation [262] T. V. Arumugam, S. C. Tang, J. D. Lathia et al., “Intravenous immunoglobulin (IVIG) protects the brain against experimen- tal stroke by preventing complement-mediated neuronal cell death,” Proceedings of the National Academy of Sciences of the United States of America, vol. 104, no. 35, pp. 14104–14109, 2007. [247] R. B. Sim, U. Kishore, C. L. Villiers, P. N. Marche, and D. A. Mitchell, “C1q binding and complement activation by prions and amyloids,” Immunobiology, vol. 212, no. 4-5, pp. 355–362, 2007. [263] J. Mocco, W. J. Mack, A. F. Ducruet et al., “Complement component C3 mediates inflammatory injury following focal cerebral ischemia,” Circulation Research, vol. 99, no. 2, pp. 209– 217, 2006. [248] T. Ishii, S. Haga, S. Yagishita, and J. Tateishi, “The presence of complements in amyloid plaques of creutzfeldt-Jakob disease and gerstmann-straussler-scheinker disease,” Applied Pathol- ogy, vol. 2, no. 6, pp. 370–379, 1984. [264] A. Heimann, T. Takeshima, G. Horstick, and O. Kempski, “C1- esterase inhibitor reduces infarct volume after cortical vein occlusion,” Brain Research, vol. 838, no. 1-2, pp. 210–213, 1999. [249] G. G. Kovacs, P. Gasque, T. Str¨obel et al., “Complement activation in human prion disease,” Neurobiology of Disease, vol. 15, no. 1, pp. 21–28, 2004. [265] T. V. Arumugam, T. M. Woodruff, J. D. Lathia, P. K. Selvaraj, M. P. Mattson, and S. M. Taylor, “Neuroprotection in stroke by complement inhibition and immunoglobulin therapy,” Neuro- science, vol. 158, no. 3, pp. 1074–1089, 2009. [250] B. M. Bellander, I. H. Olafsson, P. H. Ghatan et al., “Secondary insults following traumatic brain injury enhance complement activation in the human brain and release of the tissue damage marker S100B,” Acta Neurochirurgica, vol. 153, no. 1, pp. 90–100, 2011. [266] D. M. Bonifati and U. Kishore, “Role of complement in neurode- generation and neuroinflammation,” Molecular Immunology, vol. 44, no. 5, pp. 999–1010, 2007. [251] B. M. Bellander, S. K. Singhrao, M. Ohlsson, P. Mattsson, and M. Svensson, “Complement activation in the human brain after traumatic head injury,” Journal of Neurotrauma, vol. 18, no. 12, pp. 1295–1311, 2001. [267] J. Rogers, N. R. Cooper, S. Webster et al., “Complement activation by 𝛽-amyloid in Alzheimer disease,” Proceedings of the National Academy of Sciences of the United States of America, vol. 89, no. 21, pp. 10016–10020, 1992. Mediators of Inflammation 18 [268] P. Eikelenboom, C. E. Hack, J. M. Rozemuller, and F. C. Mediators of Inflammation Stam, “Complement activation in amyloid plaques in Alzheimer’s dementia,” Virchows Archiv Abteilung B Cell Pathology, vol. 56, no. 4, pp. 259–262, 1989. [283] M. Maier, Y. Peng, L. Jiang, T. J. Seabrook, M. C. Carroll, and C. A. Lemere, “Complement C3 deficiency leads to acceler- ated amyloid 𝛽plaque deposition and neurodegeneration and modulation of the microglia/macrophage phenotype in amyloid precursor protein transgenic mice,” The Journal of Neuroscience, vol. 28, no. 25, pp. 6333–6341, 2008. [269] P. L. McGeer, H. Akiyama, S. Itagaki, and E. G. McGeer, “Activation of the classical complement pathway in brain tissue of Alzheimer patients,” Neuroscience Letters, vol. 107, no. 1–3, pp. 341–346, 1989. [284] T. Wyss-Coray, F. Yan, A. H. T. Lin et al., “Prominent neu- rodegeneration and increased plaque formation in comple- ment-inhibited Alzheimer’s mice,” Proceedings of the National Academy of Sciences of the United States of America, vol. 99, no. 16, pp. 10837–10842, 2002. [270] S. Webster, B. Bonnell, and J. Rogers, “Charge-based binding of complement component C1q to the Alzheimer amyloid 𝛽- peptide,” The American Journal of Pathology, vol. 150, no. 5, pp. 1531–1536, 1997. [285] J. Zhou, M. I. Fonseca, K. Pisalyaput, and A. J. Tenner, “Complement C3 and C4 expression in C1q sufficient and deficient mouse models of Alzheimer’s disease,” Journal of Neurochemistry, vol. 106, no. 5, pp. 2080–2092, 2008. [271] U. Kishore, S. K. Gupta, M. V. Perdikoulis, M. S. Kojouharova, B. C. Urban, and K. B. M. Reid, “Modular organization of the carboxyl-terminal, globular head region of human C1q A, B, and C chains,” The Journal of Immunology, vol. 171, no. 2, pp. 812–820, 2003. [286] M. Britschgi, Y. Takeda-Uchimura, E. Rockenstein, H. Johns, E. Masliah, and T. Wyss-Coray, “Deficiency of terminal comple- ment pathway inhibitor promotes neuronal tau pathology and degeneration in mice,” Journal of Neuroinflammation, vol. 9, 220 pages, 2012. [272] S. D. Webster, A. J. Yang, L. Margol, W. Garzon-Rodriguez, C. G. Glabe, and A. J. Tenner, “Complement component C1q modulates the phagocytosis of A𝛽by microglia,” Experimental Neurology, vol. 161, no. 1, pp. 127–138, 2000. [287] J. C. Lambert, S. Heath, G. Even et al., “Genome-wide associ- ation study identifies variants at CLU and CR1 associated with Alzheimer’s disease,” Nature Genetics, vol. 41, no. 10, pp. 1094– 1099, 2009. [273] R. Fan and A. J. Tenner, “Complement C1q expression induced by A𝛽in rat hippocampal organotypic slice cultures,” Experi- mental Neurology, vol. 185, no. 2, pp. Mediators of Inflammation McGeer, “Lewy bodies in Parkinson’s disease are recognized by antibodies to complement proteins,” Acta Neuropathologica, vol. 84, no. 1, pp. 100–104, 1992. [315] C. Rudduck, L. Beckman, G. Franzen, and L. Lindstrom, “C3 and C6 complement types in schizophrenia,” Human Heredity, vol. 35, no. 4, pp. 255–258, 1985. [300] D. A. Loeffler, D. M. Camp, and S. B. Conant, “Comple- ment activation in the Parkinson’s disease substantia nigra: an immunocytochemical study,” Journal of Neuroinflammation, vol. 3, article 29, 2006. [316] C. Rudduck, L. Beckman, G. Franzen, and L. Lindstrom, “C3 and C6 complement types in schizophrenia,” Human Heredity, vol. 35, no. 4, pp. 255–258, 1985. [301] Y. Wang, A. M. Hancock, J. Bradner et al., “Complement 3 and factor H in human cerebrospinal fluid in Parkinson’s disease, Alzheimer’s disease, and multiple-system atrophy,” The American Journal of Pathology, vol. 178, no. 4, pp. 1509–1516, 2011. [317] A. E. Renton, E. Majounie, A. Waite, J. Simon-Sanchez, S. Rollinson, J. R. Gibbs et al., “A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS- FTD,” Neuron, vol. 72, no. 2, pp. 257–268, 2011. [318] M. DeJesus-Hernandez, I. R. Mackenzie, B. F. Boeve, A. L. Boxer, M. Baker, N. J. Rutherford et al., “Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS,” Neuron, vol. 72, no. 2, pp. 245–256, 2011. [302] E. Hirsch, A. M. Graybiel, and Y. A. Agid, “Melanized dopamin- ergic neurons are differentially susceptible to degeneration in Parkinson’s disease,” Nature, vol. 334, no. 6180, pp. 345–348, 1988. [303] W. Zhang, K. Phillips, A. R. Wielgus et al., “Neuromelanin activates microglia and induces degeneration of dopaminergic neurons: implications for progression of parkinson’s disease,” Neurotoxicity Research, vol. 19, no. 1, pp. 63–72, 2011. [319] A. Shastri, D. M. Bonifati, and U. Kishore, “Other Dementias,” in Neurodegenerative Diseases, U. Kishore, Ed., InTech, 2013, http://www.intechopen.com/books/neurodegenerative-dis- eases/other-dementias. [304] C. Depboylu, M. K. H. Sch¨afer, O. Arias-Carri´on, W. H. Oertel, E. Weihe, and G. U. H¨oglinger, “Possible involvement of complement factor C1q in the clearance of extracellular neuromelanin from the substantia nigra in Parkinson disease,” Journal of Neuropathology and Experimental Neurology, vol. 70, no. 2, pp. 125–132, 2011. [305] M. E. MacDonald, C. M. Ambrose, M. P. Duyao et al., “A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes,” Cell, vol. 72, no. 6, pp. 971–983, 1993. [306] S. K. Mediators of Inflammation Ishii, “Demonstration of microglial cells in and around senile (neuritic) plaques in the Alzheimer brain. An immunohistochemical study using a novel monoclonal antibody,” Acta Neuropathologica, vol. 77, no. 6, pp. 569–575, 1989. [293] D. A. S. Compston, B. P. Morgen, A. K. Campbell et al., “Immunocytochemical localization of the terminal comple- ment complex in multiple sclerosis,” Neuropathology and Applied Neurobiology, vol. 15, no. 4, pp. 307–316, 1989. [294] J. W. Prineas, E. E. Kwon, E. S. Cho et al., “Immunopathology of secondary-progressive multiple sclerosis,” Annals of Neurology, vol. 50, no. 5, pp. 646–657, 2001. [280] A. R. Korotzer, J. Watt, D. Cribbs et al., “Cultured rat microglia express C1q and receptor for C1q: implications for amyloid effects on microglia,” Experimental Neurology, vol. 134, no. 2, pp. 214–221, 1995. [295] B. P. Brink, R. Veerhuis, E. C. W. Breij, P. Van Der Valk, C. D. Dijkstra, and L. B¨o, “The pathology of multiple sclerosis is location-dependent: no significant complement activation is detected in purely cortical lesions,” Journal of Neuropathology and Experimental Neurology, vol. 64, no. 2, pp. 147–155, 2005. [281] P. L. McGeer and E. G. McGeer, “The possible role of com- plement activation in Alzheimer disease,” Trends in Molecular Medicine, vol. 8, no. 11, pp. 519–523, 2002. [296] E. C. W. Breij, B. P. Brink, R. Veerhuis et al., “Homogeneity of active demyelinating lesions in established multiple sclerosis,” Annals of Neurology, vol. 63, no. 1, pp. 16–25, 2008. [282] J. D. Sokolowski and J. W. Mandell, “Phagocytic clearance in neurodegeneration,” The American Journal of Pathology, vol. 178, no. 4, pp. 1416–1428, 2011. Mediators of Inflammation 19 [297] E. C. Hirsch and S. Hunot, “Neuroinflammation in Parkinson’s disease: a target for neuroprotection?” The Lancet Neurology, vol. 8, no. 4, pp. 382–397, 2009. [313] R. Zakharyan, A. Khoyetsyan, A. Arakelyan, A. Boyajyan, A. Gevorgyan, A. Stahelova et al., “Association of C1QB gene polymorphism with schizophrenia in armenian population,” BMC Medical Genetics, vol. 12, p. 126, 2011. [298] P. L. McGeer and E. G. McGeer, “Inflammation and neurode- generation in Parkinson’s disease,” Parkinsonism and Related Disorders, vol. 10, no. 1, pp. S3–S7, 2004. [314] B. H˚avik, S. Le Hellard, M. Rietschel et al., “The comple- ment control-related genes CSMD1 and CSMD2 associate to schizophrenia,” Biological Psychiatry, vol. 70, no. 1, pp. 35–42, 2011. [299] T. Yamada, P. L. McGeer, and E. G. Mediators of Inflammation Singhrao, J. W. Neal, B. P. Morgan, and P. Gasque, “Increased complement biosynthesis by microglia and comple- ment activation on neurons in Huntington’s disease,” Experi- mental Neurology, vol. 159, no. 2, pp. 362–376, 1999. [307] S. Hakobyan, A. Boyajyan, and R. B. Sim, “Classical pathway complement activity in schizophrenia,” Neuroscience Letters, vol. 374, no. 1, pp. 35–37, 2005. [308] A. Arakelyan, R. Zakharyan, A. Khoyetsyan, D. Poghosyan, R. Aroutiounian, F. Mrazek et al., “Functional characterization of the complement receptor type 1 and its circulating ligands in patients with schizophrenia,” BMC Clinical Pathology, vol. 11, p. 10. [309] K. R. Mayilyan, A. Dodds, A. S. Boyajyan, A. F. Soghoyan, and R. B. Sim, “Complement C4B protein in schizophrenia,” World Journal of Biological Psychiatry, vol. 9, no. 3, pp. 225–230, 2008. [310] A. Boyajyan, A. Khoyetsyan, and A. Chavushyan, “Alterna- tive complement pathway in schizophrenia,” Neurochemical Research, vol. 35, no. 6, pp. 894–898, 2010. [311] K. R. Mayilyan, J. N. Arnold, J. S. Presanis, A. F. Soghoyan, and R. B. Sim, “Increased complement classical and mannan- binding lectin pathway activities in schizophrenia,” Neuro- science Letters, vol. 404, no. 3, pp. 336–341, 2006.h [312] K. R. Mayilyan, D. R. Weinberger, and R. B. Sim, “The comple- ment system in schizophrenia,” Drug News and Perspectives, vol. 21, no. 4, pp. 200–210, 2008.
https://openalex.org/W4390455277	https://publications.muet.edu.pk/index.php/muetrj/article/download/3006/757	English	null	Aerial identification of flashed over faulty insulator using binary image classification	Mehran University research journal of engineering and technology	2,024	cc-by	5,201	A B S T R A C T Flashed over insulator faults are the most significant faults in high voltage line insulators. They are complicated to identify using traditional methods due to their labor-intensive nature. This study proposes a deep learning-based algorithm for detecting flashed over insulator faults in the real time. The algorithm is based on the Resnet 50 architecture, which has been shown to be effective for image classification tasks in the previous studies regarding image analysis. The algorithm is fast, robust and efficient, making it suitable for real-time applications. The algorithm is trained on a dataset of images of flashed over and non-flashed over insulators. This dataset was collected from various transmission lines and National Center of Robotics and Automation, which are located in Pakistan. For validating the effectiveness of the Resnet 50 algorithm, it was compared with the results obtained from the two other widely popular deep learning algorithms, Densenet 121 and VGG 16 (trained and validated on the same dataset). The results showed that the Resnet 50 was able to detect flashed over insulator faults with an accuracy of over 99%. Whereas the Densenet 121 and VGG 16 have achieved an accuracy of less than 51%. prone to failure due to the faults they endure during their service period. The significant percentage of these faults cannot be managed through any operational diagnosis as the faults are related with the uncertain atmospheric factors and the characteristics of the transmission line routes [2]. Considering the causes of power outages, the most outages are caused by transmission wires that sustain overvoltage transients either via thunderstorm over voltages or non-thunderstorm over voltages. In © Mehran University of Engineering and Technology 2024 Aerial identification of flashed over faulty insulator using binary image classification Aerial identification of flashed over faulty insulator using binary ima classification Shafi Muhammad Jiskani a, c, , Tanweer Hussain b, Anwar Ali Sahito c, Faheemullah Shaikh c, A Akbar Shah d a Directorate of Postgraduate Studies, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan b Department of Mechanical Engineering, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan c Department of Electrical Engineering, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan d NCRA-HHR&CM Lab, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan Corresponding author: Shafi Muhammad Jiskani, Email: shafi.jiskani@faculty.muet.edu.pk a Directorate of Postgraduate Studies, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan b Department of Mechanical Engineering, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan c Department of Electrical Engineering, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan d NCRA-HHR&CM Lab, Mehran University of Engineering and Technology Jamshoro-76060 Pakistan * Corresponding author: Shafi Muhammad Jiskani, Email: shafi.jiskani@faculty.muet.edu.pk Received: 01 September 2023, Accepted: 20 December 2023, Published: 01 January 2024 K E Y W O R D S Flashed Over Insulator Faults High Voltage Line Insulators Resnet 50 Image Classification Densenet 121 VGG 16 Mehran University Research Journal of Engineering and Technology Mehran University Research Journal of Engineering and Technology https://doi.org/10.22581/muet1982.2401.3006 2024, 43(1) 225-233 1.1 Power outages due to faults The country’s economy is mainly dependent on the industrial and commercial utilization of electrical energy. Such dependence on electrical energy impels the government agencies to spread new high voltage transmission networks and augment the existing ones [1]. These transmission line networks are very much 225 For the defect detection on the regions that are hard to inspect manually (as evident in Fig. 2), Unmanned Aerial Vehicle (UAV) is preferred due to inexpensiveness, high mobility, much efficiency and flexibility to control [14]. which, the insulator failures are classified, as one of the main form of support structure failures. The reasons of power outages with their significant failure percentage are illustrated in the Fig. 1 [3-9]. Fig. 1. Percentage power outages’ factors in high voltage power lines. f f h h l l l Fig. 2. Conventional Manpower Based Insulator Inspection Methods. Fig. 1. Percentage power outages’ factors in high voltage power lines. © Mehran University of Engineering and Technology 2024 1.3 Inspecting suspended aerial high voltage insulators 1.2 Significance of high voltage line insulators In another research regarding the face recognition technique [35] researchers address the challenge of limited image samples for face verification and recognition, a common issue in practical applications Similarly, [23] and [24] developed fault detection algorithms for high-voltage line insulators using unmanned aerial systems (UASs) illustrated in table 1. Jalil et al. used a multi-modal data technique, while in [24] researchers used You Only Look Once (YOLO). The technique used in [23], the UAS captures images in the visible and infrared domains and transmits them to the ground station. The captured images are then analysed to identify any flaws, in the insulator. In study [31] the researchers worked on vision based fall detection. The research utilizes YoloV5 and YoloV7 along with Deep SORT algorithm. While pointing the accuracy of algorithm Yolo is designed and well suited for object detection goals whereas for classification Resnet 50 is more accurate as it utilizes deep network. Researchers in [32] highlighted the work on x-ray domain imaging domain with multiclass classification. The proposed approach in the study generates highly realistic diversified synthetic X-ray scans with less cost and time that assists human-annotated datasets for training robust neural networks. For colour images researchers in [33] applied a novel approach for multi- level thresholding. Energy Curve and Minimum Fuzzy Entropy (ECFE) function was designed to address the challenges of handling fuzziness and spatial uncertainties in color images, particularly in satellite images. In [34] researchers used another multilevel thresholding technique but the technique particularly addresses for remote sensing image analysis. The primary focus of the study was to address the challenges posed by remote sensing imagery, such as dense features, low illumination, uncertainties, and noise, which make traditional segmentation techniques less effective in identifying multiple regions of interest. In another research regarding the face recognition technique [35] researchers address the challenge of limited image samples for face verification and recognition, a common issue in practical applications However, for vision inspection of the insulator’s fault, a real time or near real time is required that might require high processing speed. Which can be achieved using off-site or offline processing. This off-side or offline processing has large delays associated with it. Due to this sole reason, the essence of real time detection of the insulator’s faults is disregarded. And the delays can ultimately cause a chain of insulators’ failure. 1.2 Significance of high voltage line insulators For increasing high voltage transmission networks, the engineers are required to escalate the use of insulator strings by combining number of disc insulators in a single string. The insulator strings are designed to support the load of transmission conductors and to cease the path of current towards ground for ensuring the stable and safe operation of power grids [10, 11]. Disc insulators in string provides dielectric strength against high voltages and, are susceptible to different faults as these insulators lay bare in the harsh ambience [12]. These faults include broken insulator, cracked or surface damaged insulator, flash over surface insulator, self- shattered insulator etc. The insulator string becomes overstressed due to these faulty insulators and by any means these faulty insulators decrease the overall dielectric strength of insulator string and can cause the string failure that might cause flash over (the event when the capability of disc insulator is compromised due to factors such as pollution on the insulator surface) bridging the conducting paths i.e. conductor with metallic part of tower [13]. Fig. 2. Conventional Manpower Based Insulator Inspection Methods. These systems embed UAVs, with various airborne sensors for inspection of the insulators. Sensors like thermal cameras [15], lidar sensors [16], and vision sensors [17] are employed to address specific inspection requirements. The main purpose of using each of these airborne sensors for the inspection of insulators’ health is mentioned as follow: Thermal cameras capture infrared radiation to detect variations in temperature. This allows UAV based scheme to identify anomalies such as overheating components or hotspots in power transmission systems, which can indicate potential incepting failures or faults [15]. Lidar (Light Detection and Ranging) sensors in UAV based schemes use laser beams to measure distances and generate highly accurate 3D point cloud models of the environment. These sensors can provide detailed © Mehran University of Engineering and Technology 2024 1.3 Inspecting suspended aerial high voltage insulators 226 information about the structural integrity of power transmission infrastructure, such as power lines, towers, and insulators. They are particularly useful in identifying physical damage, wear and tear, or any signs of deformation [16]. and classifying defects in insulator systems in electric power lines by reviewing machine vision-based insulator inspection systems for overhead power distribution systems. In [22], researchers introduced an autonomous landing control technique for UAVs when charging at electric towers based on vision positioning methods. 1.2 Significance of high voltage line insulators Thus, to mitigate these issues, a real time inspection system is developed in this study that is able to diagnose the insulator’s flashover faults using binary classification. For selection of binary image classification, the data acquired is trained using Densenet 121, Resnet 50 and VGG 15. The results of this study, clearly demonstrate the accuracies obtained from these algorithms by virtue of their respective Receiver Operating Characteristics. 2. Existing Machine Learning Processes For Identification of HVLI Faults © Mehran University of Engineering and Technology 2024 1.2 Significance of high voltage line insulators Unmanned aerial vehicles (UAVs) have been widely used for this purpose because they provide a safe way to bring sensors close to power transmission lines and their associated components without halting the equipment during the inspection. This also reduces operational costs (Operational Expenditures - OpEx) and risks. Vision sensors, including high-resolution cameras and video recording systems, are also utilized in UAV- based inspection systems. These sensors capture visual data and imagery of the power transmission infrastructure from various angles and perspectives. This visual information can be analyzed to detect issues such as corrosion, loose connections, vegetation encroachment, or other visible signs of damage or deterioration [17]. Similarly, [23] and [24] developed fault detection algorithms for high-voltage line insulators using unmanned aerial systems (UASs) illustrated in table 1. Jalil et al. used a multi-modal data technique, while in [24] researchers used You Only Look Once (YOLO). The technique used in [23], the UAS captures images in the visible and infrared domains and transmits them to the ground station. The captured images are then analysed to identify any flaws, in the insulator. In study [31] the researchers worked on vision based fall detection. The research utilizes YoloV5 and YoloV7 along with Deep SORT algorithm. While pointing the accuracy of algorithm Yolo is designed and well suited for object detection goals whereas for classification Resnet 50 is more accurate as it utilizes deep network. Researchers in [32] highlighted the work on x-ray domain imaging domain with multiclass classification. The proposed approach in the study generates highly realistic diversified synthetic X-ray scans with less cost and time that assists human-annotated datasets for training robust neural networks. For colour images researchers in [33] applied a novel approach for multi- level thresholding. Energy Curve and Minimum Fuzzy Entropy (ECFE) function was designed to address the challenges of handling fuzziness and spatial uncertainties in color images, particularly in satellite images. In [34] researchers used another multilevel thresholding technique but the technique particularly addresses for remote sensing image analysis. The primary focus of the study was to address the challenges posed by remote sensing imagery, such as dense features, low illumination, uncertainties, and noise, which make traditional segmentation techniques less effective in identifying multiple regions of interest. 2. Existing Machine Learning Processes For Identification of HVLI Faults The use of machine vision and deep learning to detect power insulator faults has been an active area of research in recent years. Researchers have proposed a variety of methods to detect power insulator faults using machine vision and deep learning. These methods include the use of linear, area, silicon, and InGaAs detectors [18]. Some of the notable studies are discussed as follows: In [19], researchers developed an algorithm for inspecting the insulators using deep convolutional neural networks (CNNs) by employing Region of Interest (ROI) in the infrared images of the insulator strings based on Binary Robust Invariant Scalable Keypoints (BRISK) and Vector of Locally Aggregated Descriptors (VLAD). Whereas in [20] and [21], researchers studied the use of vision-based techniques for automatically detecting 227 • The contribution made in this research practical solution for detecting the faulty flashed over insulators. like passport and driver's licenses. Their system utilizes the Tetrolet-LDP descriptor for feature extraction and employs Cat Swarm Optimization (CSO) to train a 2- Dimensional Hidden Markov Model (2DHMM) for classification. In a study [24] an Unmanned Aerial System (UAS) was utilized to detect faults in insulators from images that had complex backgrounds. This detection technique is based on YOLO, a learning algorithm specifically designed for object detection in images. Both of these methods show promise for enhancing the safety and reliability of power transmission systems. However more research is required to assess their effectiveness in real world scenarios. Here's a table summarizing the distinctions, between these two approaches. • The proposed solution has the potential to revolutionize the transmission network monitoring and maintenance practices. 3. Deficiencies Of Existing Machine Learning Algorithms 3. Deficiencies Of Existing Machine Learning Algorithms Detecting faults in high voltage line insulators is a complex task that can benefit significantly from machine learning models. These models contribute to the improvement of detection accuracy and the reduction of failure risk. However, using machine learning for this purpose presents several significant obstacles that must be carefully navigated [26]. Table 1 Key differences between given techniques Feature Jalil et al. (2019) [23] Dhulipudi et al. (2020) [24] Data used Multi-modal (visible and infrared) Visible Algorithm Not specified YOLO Background Simple Complex Advantages Can detect faults in visible and infrared images Can detect faults in images with complex backgrounds Disadvantages Not as accurate as YOLO Not as versatile as multi-modal data technique The potential imbalance of training data is a prevalent concern. Typically, the data contains more instances of "normal" insulators than "faulty" ones. This disparity may bias machine learning models, particularly Support Vector Machines (SVM) and Decision Trees, towards the majority class. Consequently, these models may perform poorly when identifying the minority class, i.e., detecting faults in the insulators, which is the primary objective [27]. Key differences between given techniques Key differences between given techniques For a better understanding of the binary classifier, the algorithm's operation is depicted in the block diagram below. Key differences between given techniques Key differences between given techniques In addition, the input data used to train these models may be noisy or inconsistent. This suggests that the images or sensor readings of insulators may be susceptible to corruption or slight variations due to environmental conditions, operational fluctuations, or inconsistencies in data collection. This variability can make it difficult for models, particularly complex ones like Convolutional Neural Networks (CNN), to learn fault-associated patterns accurately and consistently. Considering the above literature, these discussed algorithms are hard to process and cannot be operated in real time due to their high computing speeds. However, in a recent study by [25], the authors proposed a binary image classification method for detecting power insulator faults. The method is based on the YOLO algorithm and can detect faults at a low latency rate. Authors also evaluate the method on a dataset of real- world images and show that it has achieved the desired high accuracy. The training and operating machine learning models, especially advanced ones such as deep learning models, can be computationally intensive. It may not be feasible for real-time fault detection scenarios or systems with limited computational resources. Moreover, feature selection is an essential step in model training that requires meticulous care. The features used to train models such as k-Nearest Neighbors (k-NN) and support vector machines (SVM) have a significant impact on their accuracy. Incorporating irrelevant features or omitting essential ones may impair the model's ability to distinguish between normal and faulty insulator states, thereby decreasing the overall accuracy of predictions [26] [27]. Following are the contributions of the proposed algorithm as the applications of machine vision and deep learning to detect power insulator faults are vast and have huge potential. • The literature suggests binary image classification is optimal for this task, citing literature for its support. • The proposed Rasnet-50 algorithm high lightens the exceptional accuracy and speed. Despite these challenges, machine learning continues to be a potent tool for improving fault detection in high © Mehran University of Engineering and Technology 2024 228 voltage line insulators. By diligently addressing these challenges and optimizing the ML models, it is possible to design fault detection systems with significantly enhanced accuracy. These innovations contribute to the safe, dependable operation of power systems, reducing the risk of disruptive failures and enhancing overall efficiency. 4.1 Binary Image Classifier Fig. 3. Working of Binary Image Classifier in this study. 4.2 Advantages of Binary Image Classifier This study began with the acquisition of a private dataset collected from various high voltage transmission lines for the diagnosis of insulator faults in high voltage transmission lines. The collected data includes over 100 GB of images of high voltage line ceramic insulators (healthy and faulty, respectively). Initially, it was believed that models would be trained using collected datasets. As the dataset was biased (unbalanced) and the images were of high resolution (which requires time to process), the images were pre- processed to a lower resolution in order for the model to fit in this study. The primary benefit of pre-processing the collected dataset is the algorithm's robustness and durability under various conditions (physical or environmental). Fig. 3. Working of Binary Image Classifier in this study. This research was followed by an equal distribution of data in image datasets of healthy and defective high voltage line insulators. The training and testing datasets were split 80:20 so that the desired deep learning models (namely VGG16, DenseNet 121, and Resnet 50) could be trained and evaluated, respectively. 4.2 Advantages of Binary Image Classifier Binary image classifiers have a number of potential advantages over existing machine learning techniques for detecting faults in high voltage line insulators. These advantages include: • Oversampling and Undersampling are the techniques that can be used in binary image classifier to address the issues of imbalanced data in the dataset. This will assist in creating a balanced dataset from which the model can learn easily hence optimized ability of defective insulator detection can be achieved. The architecture diagrams of Resnet 50, VGG16 and DenseNet 121 are shown in Fig.s given below as. © Mehran University of Engineering and Technology 2024 • Noise and inconsistency in data can be easily coped up with the binary classifier using advanced techniques such as data augmentation. This will generate extra training dataset and thus increased robustness of the classifier can be achieved. Thus overall binary classifier can be programmed against the noisy and inconsistent dataset. • Limited computational resources can be compensated through binary classifier as it requires less computational power as compared to deep learning algorithms. This makes it more applicable to real time fault detection. 4.1 Binary Image Classifier © Mehran University of Engineering and Technology 2024 229 In general using binary image classifiers could offer an effective and efficient method, for identifying faults in high voltage line insulators. The focus of the research is on detecting and categorizing instances where insulators become flashed over. This occurs when dust accumulates on the insulators surface mixes with moisture and forms layers of mud over the surface of insulator. As these layers dry out due to increased leakage current they create bands on the surface of the insulator. These dry bands then produce arcs that result in flashovers, over the insulators surface leading to degradation of its condition. Fig. 7. ROC of Resnet 50 Fig. 8. ROC of Densenet 121 Fig. 9. ROC of VGG16 As evident from the Fig.s, the response of ROC curve obtained from Resnet 50 is desirable and optimal in comparison to the ROCs of Densenet 121 and VGG 16. The Resnet 50 fits in the architectural features for optimal detection of flashed over faults of the high The binary image classifier used in this research work is able to detect flashed over surface insulators with an accuracy of 95%. This shows that binary image classifiers can be a powerful tool for detecting faults in high voltage line insulators. Fig. 7. ROC of Resnet 50 g 5. Results and Discussion The accuracy obtained from Resnet 50 was over 98% with a loss of 0.00084678 for the detection of flashed over insulators. Whereas, the losses obtained from Densenet 121 and VGG 16 were 0.0122 and 0.79. These results were obtained with 10 epochs and 0.1 learning rate but the Resnet 50 achieved an optimal accuracy in just 5 epochs. The results achieved from Resnet 50, Densenet 121 and VGG 16 from 5 epochs at a learning rate of 0.1 are illustrated in the Fig.s 4, 5, and 6. Fig. 8. ROC of Densenet 121 g Fig 9 ROC of VGG16 © Mehran University of Engineering and Technology 2024 Fig. 4. Training Accuracy of Resnet 50 Fig. 5. Training Accuracy of Densenet 121 Fig. 6. Training Accuracy of VGG16 Fig. 4. Training Accuracy of Resnet 50 Fig. 4. Training Accuracy of Resnet 50 Fig. 5. Training Accuracy of Densenet 121 Fig. 6. Training Accuracy of VGG16 Whereas, the Receiver Operating Characteristics (ROC) obtained from these algorithms is illustrated in the Fig.s 7, 8 and 9. Fig. 9. ROC of VGG16 As evident from the Fig.s, the response of ROC curve obtained from Resnet 50 is desirable and optimal in comparison to the ROCs of Densenet 121 and VGG 16. The Resnet 50 fits in the architectural features for optimal detection of flashed over faults of the high voltage line insulators. The accuracy obtained from Resnet 50 is more than 99%, which demonstrates the 230 Aforementioned results demonstrate the effectiveness of Resnet 50 in the identification of the flashed over insulator faults. Thus, the accuracy obtained from the algorithm clearly differentiate its optimal working when it is compared with Densenet 50 and VGG 16. effectiveness of the algorithm. The proposed Resnet 50- based deep learning algorithm has proven highly effective, achieving an accuracy rate of over 99% in fault detection, outperforming Densenet 121 and VGG 16. Its speed and low latency make it suitable for integration into Unmanned Aerial Vehicles (UAVs) for real-time inspections, reducing manual labor and improving safety. Table 2 Validation of Resnet 50 Validation of Resnet 50 Healthy Insulators Flashed over Insulators Remarks Detected Detected Detected Detected Detected Flashed over Insulators Remarks Deep learning techniques were evaluated for insulator fault detection due to their high-level architectural design and hidden layers that are effective in determining the image features because for the image analysis, deep learning algorithms implement 3 or more hidden layers. Detec Detected This study employs three algorithms, all of which were trained on the identical data. In compared to Densenet 121 and VGG 16, Resnet 50’s results showed to be the most accurate which are evident in the Fig.s. Detected The Resnet 50's ability in identifying high voltage line insulator defect diagnosis can be described as fast, accurate, and robust. Because of its fast processing and low latency time, this technique can be used in UAVs for real-time diagnosis of high voltage line insulator defect detection. Moreover a mobile app is in the process that can be integrated with an IP camera/UAV camera for the classification of high voltage line insulators faults. Detected Detected Detected 6. Conclusion This research is focused on determining the best algorithm to use for determining flashed over faults. Because of their labor-intensive nature, flashed over insulator faults are the most significant defects that are complicated to identify using methodologies/algorithms discussed in the literature. Because of the latencies associated with them, the recommended algorithms in the literature frequently implemented offline processing. To reduce latencies, the devised algorithm in this study can detect insulator faults in real time. The algorithm was validated by a dataset that consisted of images of the high voltage line insulators that were taken in various angles and backgrounds as shown in the Table 2. 8. References [1] S. Panigrahy, S. Karmakar, and R. Sahoo, "Condition assessment of high voltage insulator using convolutional neural network," IEEE International Conference on Electronics, Computing and Communication Technologies, 2021. [11] E. Antwi-Bekoe, et al., "A deep learning approach for insulator instance segmentation and defect detection," Neural Computing and Applications, 2022, vol. 34, no. 9, pp. 7253- 7269. [2] I. Breido, V. Kaverin, and V. Ivanov, "Telemetric monitoring insulation condition of high voltage overhead power lines," Annals of DAAAM and Proceedings, 2018, vol. 29. [12] G. Hu, et al., "Automatic classification of insulator by combining k-nearest neighbor algorithm with multi-type feature for the Internet of Things," EURASIP Journal on Wireless Communications and Networking, 2018, vol. 2018, no. 1, pp. 1-10. [3] [3] L. Yakovlev, R. Kaverina, and L. Dubinich, "Complex of works and proposals for increasing reliability of overhead lines at the stage of design and operation," Power transmission lines: design, construction, operating experience and scientific and technical progress, 2008, Novosibirsk. [13] Y. Lv, et al., "Simulation study on pollution accumulation characteristics of XP13-160 porcelain suspension disc insulators," IEEE Transactions on Dielectrics and Electrical Insulation, 2016, vol. 23, no. 4, pp. 2196-2206. [4] [4] S. Kashevarov, "Damages of power transmission lines and review of new technical and organizational solutions for their limitation," Current state and prospects for the development of technical sciences: Collection of articles of the International Scientific and Practical Conference (May 23, 2015. Ufa), 2015. [14] H. Zhang, et al., "An empirical study of multi- scale object detection in high resolution UAV images," Neurocomputing, 2021, vol. 421, pp. 173-182. [15] X. Gong, et al., "A deep learning approach for oriented electrical equipment detection in thermal images," IEEE Access, 2018, vol. 6, pp. 41590-41597. [5] [5] F. L. Kogan and R. S. Kaverina, "Complex of works and proposals for increasing reliability of overhead lines at the stage of their design," Branch Firm ORGRES, Materials of the conference of the CIS Electric Power Council, 2008. [16] J. Yang and Z. Kang, "Voxel-based extraction of transmission lines from airborne lidar point cloud data," IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, vol. 11, no. 10, pp. 3892-3904. [6] [6] S. Merkhalev and E. Solomonik, Selection and operation of insulation in areas with the contaminated atmosphere. L.: Energoatomizdat, 1983. [17] D. Waleed, et al., "Drone-based ceramic insulators condition monitoring," IEEE Transactions on Instrumentation and Measurement, 2021, vol. 70, pp. 7. Acknowledgement Detected The research team would like to thank the officials of National Transmission and Despatch Company LTD. for their support and access to visit different Grid stations located in Sindh Province of Pakistan. The research is made possible through the financial support of Sindh Higher Education Commission (SHEC) and the ICT Endowment Fund for sustainable development of Mehran University of Engineering and Technology Jamshoro. Detected Detected 231 © Mehran University of Engineering and Technology 2024 [10] R. S. Gorur, E. A. Cherney, and J. T. Burnham, Outdoor insulators. 1999: Ravi S. Gorur Phoenix. 8. References © Mehran University of Engineering and Technology 2024 8. References 1-12. [7] L. Thionc, "Dielectric strength of external insulation systems under live working," Proc. 35th CIGRE Session, Paris, 1994, pp. 33-306. [18] J. Dougherty and R. Varghese, "Low SWaP multispectral sensors using dichroic filter arrays," in Next-Generation Spectroscopic Technologies VIII, 2015, SPIE. [8] G. Manone, "Investigation on the dielectric strength of damaged insulation strings of HV overhead lines during repair operations by live working," Proc. 35th CIGRE Session, Paris, 1994, pp. 33-305. [19] Z. Zhao, G. Xu, and Y. Qi, "Representation of binary feature pooling for detection of insulator strings in infrared images," IEEE Transactions on Dielectrics and Electrical Insulation, 2016, vol. 23, no. 5, pp. 2858-2866. [9] [9] B. Hutzler, "Strength of external insulation during live line maintenance and repair work with special reference to transient overvoltage," Proc. 32th SIGRE Session, Paris, 1988, pp. 33- 307. [20] [20] P. S. Prasad and B. P. Rao, "review of machine vision based insulator inspection systems for overhead power distribution system," 2017. © Mehran University of Engineering and Technology 2024 232 [32] K. Chaturvedi, et al., "Automated threat objects detection with synthetic data for real-time X-ray baggage inspection," in 2021 International Joint Conference on Neural Networks, 2021, IEEE. [21] P. S. Prasad and B. P. Rao, "LBP-HF features and machine learning applied for automated monitoring of insulators for overhead power distribution lines," in 2016 international conference on wireless communications, signal processing and networking, 2016, IEEE. [33] S. Pare, et al., "Multilevel color image segmentation using modified fuzzy entropy and cuckoo search algorithm," in 2021 IEEE International Conference on Fuzzy Systems, 2021, IEEE. [22] [22] R. Qiu, et al., "Design and implementation of an autonomous landing control system of unmanned aerial vehicle for power line inspection," in 2017 Chinese Automation Congress, 2017, IEEE. [34] S. Pare, H. Mittal, M. Sajid, J. C. Bansal, A. Saxena, T. Jan, W. Pedrycz, and M. Prasad, "Remote sensing imagery segmentation: a hybrid approach," Remote Sensing, 2021, vol. 13, no. 22, p. 4604. [Online]. Available: https://doi.org/10.3390/rs13224604 [23] B. Jalil, et al., "Fault detection in power equipment via an unmanned aerial system using multi modal data," Sensors, 2019, vol. 19, no. 13, pp. 3014. [24] D. P. Dhulipudi and K. Rajan, "Multiclass geospatial object detection using machine learning-aviation case study," in 2020 AIAA/IEEE 39th Digital Avionics Systems Conference, 2020, IEEE. [35] [35] M. S. © Mehran University of Engineering and Technology 2024 8. References Meena, et al., "A robust face recognition system for one sample problem," in Image and Video Technology: 9th Pacific-Rim Symposium, PSIVT 2019, Sydney, NSW, Australia, November 18–22, 2019, Proceedings 9, 2019, Springer. [25] M. Tomaszewski, R. Gasz, and J. Osuchowski, "Detection of power line insulators in digital images based on the transformed colour intensity profiles," Sensors, 2023, vol. 23, no. 6, p. 3343. [26] J. Chen, et al., "An method for power lines insulator defect detection with attention feedback and double spatial pyramid," Electric Power Systems Research, 2023, vol. 218, p. 109175. [27] B. Gjorgiev, et al., "Simulation-driven deep learning for locating faulty insulators in a power line," Reliability Engineering and System Safety, 2023, vol. 231, p. 108989. [28] C. Liu, et al., "Insulator faults detection in aerial images from high-voltage transmission lines based on deep learning model," Applied Sciences, 2021, vol. 11, no. 10, p. 4647. [29] J. Holtzhausen and W. Vosloo, "The pollution flashover of ac energized post type insulators," IEEE Transactions on Dielectrics and Electrical Insulation, 2001, vol. 8, no. 2, pp. 191-194. [30] Z. Zhang, et al., "Study on AC flashover performance for different types of porcelain and glass insulators with non-uniform pollution," IEEE Transactions on Power Delivery, 2013, vol. 28, no. 3, pp. 1691-1698. [31] X. Zi, et al., "Detecting human falls in poor lighting: object detection and tracking approach for indoor safety," Electronics, 2023, vol. 12, no. 5, p. 1259. 233 © Mehran University of Engineering and Technology 2024
W4283278650.txt	https://bmcoralhealth.biomedcentral.com/counter/pdf/10.1186/s12903-022-02276-1	en		Periodontal conditions in a 65-year-old population and prevalence of periodontitis according to three different bone level thresholds	BMC oral health	2,022	cc-by	6,554	(2022) 22:246 Sødal et al. BMC Oral Health https://doi.org/10.1186/s12903-022-02276-1 Open Access RESEARCH Periodontal conditions in a 65‑year‑old population and prevalence of periodontitis according to three different bone level thresholds Anne Thea Tveit Sødal1, Lene Hystad Hove1, My Tien Diep1, Rasa Skudutyte‑Rysstad1 and Odd Carsten Koldsland2 Abstract Aims: The aims of this cross-sectional study were to describe the prevalence and severity of periodontal disease in a 65-year-old population in Oslo, Norway, and to investigate to what extent the radiographic bone level threshold for periodontitis case definition influences the prevalence. Materials and methods: A random sample of 454 subjects underwent a clinical and radiographic examination and answered a questionnaire regarding general health, medications, and smoking habits. Clinical periodontal parameters (periodontal pocket depths, bleeding on probing, mobility, and furcation involvement) and radiographic bone loss were used to identify periodontitis cases and to assess periodontal stage and grade. Results: Of the 454 participants, 52.6% were defined as “periodontitis cases”. Of the total study population “unstable cases of recurrent periodontitis” were present in 38.1%, 16.5% of the participants were assigned to stage II, 32.8% to stage III, and 3.3% to stage IV. When lowering the radiographic bone loss cutoff from > 3 mm to > 2 mm or > 1 mm the prevalence of periodontitis increased to 91.9% and 99.6%. Conclusions: Periodontitis was common among 65 year-olds living in Oslo, and in the majority of those with perio‑ dontitis, the disease was recurrent and unstable. This study also shows that the choice of bone loss cutoff for defining a periodontitis case affects the prevalence estimates to a large extent. In addition, this study addresses weaknesses in the use of the consensus report of the 2017 World Workshop on the Classification of Periodontal and Peri-implant Diseases and Conditions for epidemiologic studies in its current form. Keywords: Alveolar bone loss, Classification, Periodontal diseases, Periodontitis, Prevalence Introduction Periodontitis is an inflammatory disease affecting the supportive tissues of the teeth, and may lead to tooth loss if left untreated [1–3]. In Norway, patients’ expenses for Correspondence: a.t.t.sodal@odont.uio.no 1 Department of Cariology and Gerodontology, Faculty of Dentistry, University of Oslo, Blindern, P.O. Box 1109, 0317 Oslo, Norway Full list of author information is available at the end of the article periodontal treatment have partly been subsidized by the public health insurance system for two decades [4]. Despite this public financial support and readily accessible dental health services, recent studies from different parts of Norway have shown a periodontitis prevalence ranging from 49 to 72% [5–7] with higher prevalence in older age groups. However, age-specific prevalence in a large sample of young elderly have not been reported. In a national population study from the U.S. the prevalence © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Sødal et al. BMC Oral Health (2022) 22:246 of periodontitis was estimated to be 71.5% among adults 65–74 years of age [8]. Due to a higher proportion of the elderly retaining their natural teeth [9–11], an increased need for periodontal treatment in the population may occur. Therefore, updated prevalence data in the young elderly will be of great significance for planning future dental health services for the growing population of elderly. In addition, periodontitis has been shown to affect oral health-related quality of life [12]. Periodontitis may also increase the risk of complications of systemic diseases [1] and has shown associations to systemic diseases that are highly prevalent with increasing age [13, 14]. A high prevalence of periodontal disease among older adults, along with the increasing proportion of elderly in the world population [15] may therefore have an impact on the need for health care services in the years to come. There is a large variation in definitions and classifications of periodontitis used in previous studies. This might partly explain the differences in reported prevalence data, and makes comparison of results from different studies challenging as previously described [16, 17]. The case definition developed by the Centers for Disease Control (CDC) and Prevention and the American Academy of Periodontology (AAP) [18] has been widely used in epidemiological studies [16]. However, new definitions and diagnostic criteria were presented in the 2017 World Workshop on the Classification of Periodontal and Periimplant Diseases and Conditions [19] organized by AAP and the European Federation of Periodontology (EFP). Ke et al. found a large variation in the periodontitis prevalence estimates when comparing the use of the consensus report from 2017 and other periodontal classifications [20]. However, only CAL cutoffs for stages were used, and further investigation on the epidemiological utility of the consensus report containing more of the parameters required to predict the severity and prognosis of periodontitis was requested. The aims of the present study were therefore to describe the prevalence and severity of periodontal disease in a 65-year-old population in Oslo, Norway based on the 2017 World Workshop on the Classification of Periodontal and Peri-implant Diseases and Conditions, and to investigate to what extent the choice of radiographic bone level cutoffs for periodontitis case definition influences the disease prevalence estimates. Material and methods Participants In this cross-sectional study, oral health in a 65-year-old population in Norway was investigated. A random sample was drawn from the Norwegian Population Register (retrieved from the Norwegian Tax Administration) Page 2 of 9 and invitation letters were sent out. Inclusion criteria were “born in 1954” and “resident in Oslo”. All individuals who received the letter and were reachable by phone were contacted and given the opportunity to participate in the study. The recruitment procedure has been described in detail in a previous publication [21]. The study was approved by the Norwegian Regional Committee for Medical and Health Research Ethics (REK 2018/1383) and performed in compliance with the tenets of the Declaration of Helsinki. A written informed consent was signed by each participant prior to the clinical examination. Questionnaire All participants answered a semi-structured questionnaire prior to the clinical examination using the Nettskjema software (University of Oslo, Norway). The questionnaire contained items regarding general diseases, medication use, and smoking habits. Selfreported diabetes type 1 and 2 were assessed by yes/no questions. Smoking habits were assessed by the three response alternatives: “never smoker”, “former smoker”, and “current smoker”. “Current smoker” was defined as an individual who smoked at least one cigarette daily. Current smokers also reported the number of cigarettes daily consumed. The participants’ country of birth was dichotomized into ‘western’ (Nordic countries, Western Europe, North America, and Australia) and ‘non-western’ (the rest of the world). The level of education was dichotomized into ‘higher education’ (university/college education) and ‘basic education’ (high school, elementary school, or lower). Clinical periodontal examination Two trained, calibrated dentists (ATTS and MTD) performed all clinical examinations at the Research Clinic at the Faculty of Dentistry, University of Oslo from February to December 2019. Periodontal probing depths (PPD), bleeding on probing (BoP) and suppuration were measured on six sites per tooth using LM 52B XSI Perio Probe (LM-Dental™, Planmeca Group, Helsinki, Finland). PPD was recorded to the nearest mm, rounded down. Tooth mobility was recorded as grade 1, 2, or 3 as described by Nyman et al. [22]. Furcation involvement was measured on molars using Nabers Q2N probe (HuFriedy, Chicago, USA) and recorded as grade 0, I, II, or III [23]. Third molars were not included in the clinical registrations. The examiners were trained by a specialist in periodontology (OCK) prior to inter-rater calibration and clinical examination. Inter-rater reliability of the clinical examiners was estimated from pocket depth registrations from seven participants, a total of 336 values per examiner. Sødal et al. BMC Oral Health (2022) 22:246 The intra-class correlation coefficient (ICC) (95% CI) was 0.82 (0.78–0.86). Radiographic examination Orthopantograms (OPG) and horizontal bitewings (BW) were used to assess radiographic bone loss (RBL). The OPGs were obtained using a panoramic imaging unit (ProMax X-ray Dimax 3 and Planmeca ProOne, Planmeca Oy, Helsinki). Two BWs per participant were obtained using an intraoral imaging unit (MINIRAY, SOREDEX, PaloDEx Group Oy, Tuusula, Finland) with a rectangular collimator (length 30.5 cm). Percentage bone loss was calculated for each tooth by dividing the distance from cemento-enamel junction to alveolar crest by the distance from the cemento-enamel junction to apex measured on OPGs. In addition, detectable interproximal bone loss was recorded as the distance in millimeters from the cemento-enamel junction to the alveolar bone crest on the two most severely affected non-adjacent sites on BW. Radiographic registrations were performed by one calibrated dentist (ATTS) in a room with adapted ambient light and by applying measuring instruments in the ImageJ software (ImageJ 1.52a, National Institutes of Health, USA). Calibrations were performed on OPGs from 50 participants. A specialist in periodontology (OCK) and a trained dentist (ATTS) separately evaluated 50 radiographs. The results were evaluated and for the inter-examiner calibration the ICC (95% CI) was 0.79 (0.66–0.86) for % bone loss and the weighted Cohen’s kappa (95% CI) was 0.72 (0.66–0.78) for stage. Intraexaminer calibration was performed (ATTS) on 25 radiographs. For the intra-examiner radiographic registration the ICC (95% CI) was 0.88 (0.86–0.90) for % bone loss and the weighted Cohen’s kappa (95% CI) was 0.90 (0.82– 0.98) for stage. Definitions The results from the clinical and radiographic examination in the present study were classified based on the 2017 World Workshop on the Classification of Periodontal and Peri-implant Diseases and Conditions [19, 24, 25]. A periodontitis case was defined as an individual with detectible RBL (distance from the cemento-enamel junction (CEJ) to the alveolar bone crest (AC) measured on BW radiographs) exceeding 3 mm on ≥ 2 non-adjacent teeth. For further classification of disease activity, the following definitions based on Sanz et al. [26] were used: • Periodontitis case, stable case of periodontal health: PPD ≤ 4 mm, BoP < 10%, no PPD ≥ 4 mm with BoP • Periodontitis case, some gingival inflammation: PPD ≤ 4 mm, BoP ≥ 10%, no PPD = 4 mm with BoP Page 3 of 9 • Periodontitis case, unstable case of recurrent periodontitis: PPD ≥ 5 mm, or PPD ≥ 4 mm with BoP Percentage of radiographic bone loss measured on each tooth was used for evaluating the severity of periodontitis according to Papapanou et al. [19]: < 15%; stage I, 15–33%; stage II, extending to mid-third of root and beyond; stage III. In addition, PPD, furcation involvement, and tooth loss due to periodontitis were investigated as complexity factors and used for further staging of the periodontitis cases as described by Papapanou et al. [19]. Vertical bone loss was registered if the radiographic defect was ≥ 3 mm deep and ≤ 3 mm wide [6]. Teeth were registered as lost due to periodontitis based on clinical judgment if this was considered as the most likely cause for tooth loss, that is, cases where no other reasons for tooth loss such as caries, endodontic lesions, root canal treatments seemed reasonable, and the general bone level in remaining dentition suggested periodontal tooth loss. In cases of doubt, tooth loss was not assigned to periodontitis. For each stage group the extent was described as localized (< 30% of teeth involved) or generalized (≥ 30% of teeth involved). For grading of periodontitis cases, percent bone loss at the most affected tooth divided by age was calculated and categorized as grade A (< 0.25), B (0.25–1.0), or C (> 1.0). In addition, smoking and diabetes were treated as grade modifiers. The grade modifiers could only shift the grade, based on radiographic bone loss, to a higher level. Smokers who smoked < 10 cigarettes daily were categorized as grade B and individuals who smoked ≥ 10 cigarettes daily were categorized as grade C. In lack of hemoglobin A1c measurements, diabetes could only shift the grade from A to B and not from B to C. Staging and grading were only performed on periodontitis cases. For non-periodontitis cases, the following definitions based on Chapple et al. [24] were used: • Cases with periodontal health: No detectable bone loss and BoP < 10% • Gingivitis cases: No detectable bone loss and BoP ≥ 10% In cases with no detectable bone loss, PPD > 3 mm and BoP < 10%, pocket depths were assigned to anatomical causes, and the participants were defined as cases with periodontal health. In cases with PPD > 3 mm, no detectable bone loss on radiographs, and BoP ≥ 10%, pocket depths were considered as pseudo pockets and the participants were assigned to the gingivitis group. Gingivitis was defined as localized when BoP was present in < 10% of sites and generalized when BoP was present in ≥ 10% of sites. Sødal et al. BMC Oral Health (2022) 22:246 In addition to the 3 mm interproximal RBL cutoff used for identifying periodontitis cases, the difference in prevalence, stages, and grades when using RBL > 1 mm and RBL > 2 mm cutoffs were investigated. Statistical analyses Clinical and radiographic registrations were collected in The Oral Data Collector sheet specifically designed for data entry in this study, developed in Microsoft Excel 2016 (Microsoft Corporation, Redmond, Washington, US), and imported into STATA (Stata version 16.1; College Station, TX, USA) for statistical analysis. Data were stored in Service for Sensitive Data (TSD facilities, UiO). Participants with ≤ 1 remaining tooth were excluded from the analyses. The results from the descriptive analyses are presented as percentage distributions or mean and standard deviation (SD). Page 4 of 9 Table 1 Background characteristics of participants Participants’ characteristics n (%) (N = 454) Gender Men 233 (51.3) Women 221 (48.7) Country of birth Western 413 (91.0) Non-western 41 (9.0) Education level Higher education 304 (67.0) Basic education 150 (33.0) Smoking Never-smoker 197 (43.4) Former-smoker 210 (46.3) Smoker < 10/day 18 (4.0) Smoker ≥ 10/day 29 (6.4) No 420 (92.5) Type 1 4 (0.9) Type 2 30 (6.6) Diabetes Results Participants The recruitment process is presented in Fig. 1. Distribution with respect to participants’ characteristics is presented in Table 1. Leers sent 1230 Returned 20 No phone no. 209 Phoned 1001 No answer 205 Invited 796 Declined 336 Agreed 460 Not analyzed 6 <2 teeth n=3 Total 454 Fig. 1 Study participants No quesonnaire n=3 Clinical periodontal parameters In the present population, the average number of teeth present was 25.6 (SD = 3.4), ranging from 4 to 28 remaining teeth. Distributions of participants with respect to BoP and PPD are presented in Table 2. Periodontitis prevalence, periodontal health, and gingivitis Figure 2 shows the prevalence of periodontitis (52.6%) and non-periodontitis cases (47.4%) according to the 3 mm interproximal RBL cutoff. In addition, the distribution of participants with respect to periodontal health Table 2 Distribution of participants with respect to proportion of sites with bleeding on probing (BoP) and the most severe periodontal probing depth (PPD) (N = 454) n (%) % of sites with BoP < 10% 341 (75.1) 10–30% 98 (21.6) > 30% 15 (3.3) Maximum PPD ≤ 3 mm 123 (27.1) 4–5 mm 210 (46.3) 6–8 mm 101 (22.2) 9–10 mm 14 (3.1) ≥ 11 mm 6 (3.1) Sødal et al. BMC Oral Health (2022) 22:246 Page 5 of 9 Fig. 2 Distribution of participants with respect to periodontal health, gingivitis and periodontitis presented as n (%). †Participants with PPD > 3 mm regarded as anatomical pockets (n = 100). ‡Participants with PPD > 3 mm regarded as pseudo pockets (n = 29) (39.4%), gingivitis (8%), and periodontal disease activity are shown in Fig. 2. periodontal disease with respect to stages is shown in Table 3. Staging Grading In the study population, staging based only on RBL, resulted in 125 (27.5%) periodontitis cases assigned to stage II and 114 (25.1%) to stage III. In twenty-seven cases (5.9%) tooth loss was assigned to periodontitis, of which 12 participants (2.6%) had lost 1–4 teeth and 15 participants (3.3%) had lost ≥ 5 teeth. Of the participants classified as periodontitis cases, 96 individuals had PPD of ≤ 5 mm and 143 individuals had PPD ≥ 6 mm. Vertical bone loss ≥ 3 mm was detected in 18 periodontitis cases, while furcation involvement grade 2 or 3 was identified in 81 periodontitis cases. Distribution of periodontitis cases with respect to stage, when complexity factors are taken into account, and also disease activity are shown in Table 3. Of the total study population, 220 participants (48.5%) were assigned to grade B and 19 participants (4.2%) to grade C when grade was calculated from radiographic bone loss only. No periodontitis cases were assigned to grade A. When grade modifiers were included, the number of participants assigned to grade B decreased to 203 (44.7%) while the number assigned to grade C increased to 36 (7.9%) (Table 3). The distribution of study participants with respect to periodontitis presence, stage, grade, and extent is shown in Table 4. Extent In the total study population, 21.1% showed generalized and 31.5% showed localized disease based on radiographic bone loss and complexity factors. The extent of Interproximal RBL cutoff and its impact on prevalence When changing the RBL cutoff level for defining a periodontitis case from > 3 mm to > 2 mm, the prevalence of periodontitis cases increased to 91.9% of the study population. Reducing the cutoff level for periodontitis to > 1 mm RBL, increased the prevalence further to 99.6%. Differences in distribution with respect to stages are shown in Table 5. Sødal et al. BMC Oral Health (2022) 22:246 Page 6 of 9 Table 3 Distribution of periodontitis cases with respect to stage and disease activity, extent, and grade (N = 454) Non-periodontitis cases Stage I Stage II Stage III Stage IV n (%) n (%) n (%) n (%) n (%) 215 (47.4) 0 (0) 75 (16.5) 149 (32.8) 15 (3.3) Periodontitis case with stable periodontal health – 0 (0) 37 (8.1) 26 (5.7) 0 (0) Periodontitis with some gingival inflammation – 0 (0) 3 (0.7) 0 (0) 0 (0) Unstable case of recurrent periodontitis – 0 (0) 35 (7.7) 123 (27.1) 15 (3.3) Localized – 0 (0) 5 (1.1) 138 (30.4) 0 (0) Generalized – 0 (0) 70 (15.4) 11 (2.4) 15 (3.3) A – 0 (0) 0 (0) 0 (0) 0 (0) B – 0 (0) 73 (16.1) 124 (27.3) 6 (1.3) C – 0 (0) 2 (0.4) 25 (5.5) 9 (2.0) Total Disease activity Extent of periodontal stage Grade Table 4 Distribution of study participants by periodontitis presence, stage, grade, and extent (N = 454) Total Nonperiodontitis cases Stage I Stage II Stage III Stage IV n (%) n (%) n (%) n (%) n (%) 215 (47.4) 0 (0) 75 (16.5) 149 (32.8) 15 (3.3) Grade A Localized – 0 (0) 0 (0) 0 (0) 0 (0) Generalized – 0 (0) 0 (0) 0 (0) 0 (0) Grade B Localized – 0 (0) 5 (1.1) 121 (26.7) 0 (0) Generalized – 0 (0) 68 (15.0) 3 (0.7) 6 (1.3) Grade C Localized – 0 (0) 0 (0) 17 (3.8) 0 (0) Generalized – 0 (0) 2 (0.4) 8 (1.8) 9 (2.0) Localized; < 30% of teeth involved. Generalized; ≥ 30% of teeth involved Discussion This study provides comprehensive data regarding the prevalence of periodontitis in a population of 65 yearolds living in Oslo. Of the total study population, the majority had one or more PPDs of 4 mm or more, and BoP was present in 10% or more sites in one-fourth of the participants. In addition, approximately half of the participants were defined as periodontitis cases according to the 3 mm RBL cutoff, and almost three fourth of those with periodontitis had unstable, recurrent periodontitis. The present study also showed that the pre-selected radiographic bone level for defining a periodontitis case had large impact on the prevalence estimate. PPD is an important measurement for evaluating periodontal conditions [19, 24]. In the present study, 72.9% of the participants had at least one site with PPD ≥ 4 mm, and PPD ≥ 6 mm was present in 26.6% of the participants. This is in line with results in the study from Troms county in northern Norway [7]. Holde et al. found that 81% of the participants had PPD ≥ 4 mm and 33% of the participants had PPD ≥ 6 mm in at least one site in the age group 65–75 years [7]. In a Swedish population of 60 year-olds, similar prevalences were found where PPD ≥ 4 mm and ≥ 6 mm was present in 87% and 37% respectively [27]. In cross-sectional studies, where the participants are only examined once, it is not possible to distinguish between pseudo-pockets and pockets caused by periodontitis. However, sites with BoP and PPD > 3 mm indicates a need for improvement of oral hygiene and initiation of preventive measures to prevent progression of periodontal disease. Prevalence data on periodontitis among elderly varies between studies. In accordance with the present study where more than half of the participants were classified as periodontitis cases, a prevalence of 49.7% has Table 5 Periodontal stages distributed by different interproximal radiographic bone loss cutoffs (N = 454) RBL (mm) Non-periodontitis cases n (%) Periodontitis cases n (%) Stage I n (%) Stage II n (%) Stage III n (%) Stage IV n (%) >1 2 (0.4) 452 (99.6) 2 (0.4) 243 (53.5) 191 (42.1) 16 (3.5) >2 37 (8.1) 417 (91.9) 0 (0) 212 (46.7) 189 (41.6) 16 (3.5) >3 215 (47.4) 239 (52.6) 0 (0) 75 (16.5) 149 (32.8) 15 (3.3) Sødal et al. BMC Oral Health (2022) 22:246 been reported in a study population of 18–75 year-old in northern part of Norway using a case definition based on the 2017 World Workshop on the Classification of Periodontal and Peri-implant Diseases and Conditions by measuring bone loss on radiographs [5]. However, in the oldest age group ranging from 65 to 75 years old the prevalence was 80.6%. Other studies from Norway have reported prevalence of periodontitis ranging from 33 to 98% [6, 7, 28] with increasing prevalence in older age groups. These studies were based on radiographic evaluation combined with PPD or calculated CAL. Compared to other countries, the prevalence in the present study is in line with results from a population study in the U.S. where 60% of the participants aged 65 years or older had periodontitis according to the CDC/AAP case definitions [8, 29], and with results from Spain where 50% in the age group 55 years or older was registered with bone loss > 3 mm, however this was measured using CAL and not RBL [30]. In addition to the fact that half of the participants in the present study were classified as periodontitis cases, more than one-third were registered with unstable periodontitis. This indicates a considerable periodontal treatment need among the young elderly in Oslo. Dental health services are readily accessible in Oslo and periodontal treatment expenses are partly subsidized by the Norwegian welfare system, therefore the prevalence of 65-year-olds with unstable periodontitis is surprisingly high. It might be speculated that poor patient compliance with respect to dental hygiene instructions, infrequent dental visits, or underdiagnosed periodontitis might be possible reasons for the high prevalence. These factors should be investigated in future studies. In the present study, 16.5% of the participants were assigned to stage II, 32.8% to stage III, and 3.3% to stage IV periodontitis. Stødle et al. found a comparable prevalence of stage III (29.1%) and stage IV (3.6%) among 60–69 year-olds in a population in Trøndelag county [6], however, the prevalence of stage II was 58.8% which is a high proportion compared to the present study. In the study from Trøndelag, the prevalence of periodontitis was assessed based on the consensus report from the 2017 World Workshop on the Classification of Periodontal and Peri-implant Diseases and Conditions, and a RBL cutoff of 1.5 mm was used for all age groups. This might explain the higher prevalence of stage II periodontitis compared to the present study. When lowering the RBL cutoff for identifying a periodontitis case from > 3 mm to > 2 mm or > 1 mm in the present study the prevalence of periodontitis increased remarkably to 91.9% and 99.6%. This indicates that a universal threshold for the diagnosis of periodontitis in epidemiological studies is required. It has previously been reported a Page 7 of 9 large variation in the prevalence of periodontitis categorized by CAL threshold of 3 mm (≈ 95%), 4 mm (≈ 75%), 5 mm (≈ 50%), 6 mm (≈ 30%), and 7 mm (≈ 20%) in a population of 65 year-olds in the U.S. [29]. The challenge regarding the use of different case definitions and periodontal examination protocols has previously been highlighted [16, 31, 32]. The consensus report from 2017 [19] has tried to alleviate this and presents cutoffs with respect to CAL for staging periodontitis cases. Interproximal CAL of ≥ 2 mm or ≥ 3 mm are presented as commonly used cutoffs for identifying a periodontitis patient [19], however, no exact value of CAL is given in the 2017 World Workshop case definition. Physiological bone level has been defined as bone level 1 mm to 3 mm apical to the cemento-enamel junction [24]. Ke et al. compared the consensus report using CAL cutoffs for stage II and III with other previously used periodontal classifications [20]. In addition to affecting the prevalences reported, the new classification showed strong associations with known risk factors such as smoking. However, Ke et al. used CAL only and requested further investigation of the epidemiological utility [20]. As described, the staging and grading procedure requires several parameters which are rare in epidemiological studies, like longitudinal design and radiographic determination of bone levels in addition to several clinical parameters [20]. The present study used radiographic bone levels to determine percent bone loss per age and thereby calculating the indirect evidence of progression as described in the consensus report [19]. However, to determine the direct evidence of progression, longitudinal data would be required, and results from the present study may serve as a baseline for followup studies in the same study population. The availability of CAL instead of RBL in the present study could give a more precise estimate of the prevalence. In the study by Eke et al. CAL was used and bone loss in frontal areas, which obviously are not included on BWs, could be identified [29]. This might explain the higher prevalence of periodontitis with the 3 mm cutoff in the study from the U.S. [29] compared to the present study. As previously described, the use of RBL for defining a periodontitis case might lead to underestimation of mild to moderate periodontitis [25]. However, measurements necessary for calculating CAL in addition to all the other parameters required for staging and grading according to the consensus report are time consuming and was not used in the present study due to time limitation. The current study indicates that the consensus report is poorly applicable in large epidemiological studies including several aspects of oral health, and may be better suited for clinical practice and studies investigating periodontitis only. Sødal et al. BMC Oral Health (2022) 22:246 Another limitation of the present study was that no dental records regarding participants’ treatment history were available. Teeth were registered as lost due to periodontitis if no other reasons for tooth loss such as caries, endodontic lesions, root canal treatments seemed reasonable, and when the bone level in the remaining dentition suggested loss of periodontal attachment. In addition, in cases where the reason for the bone loss was questionable, the case was discussed until a consensus was made. In uncertain cases, the cause was not assigned to periodontitis, and this may have led to an underestimation of both the prevalence and severity of periodontitis. The response rate in the OM65 study was 58% [21], leading to a sizable proportion of non-respondents and possibility of selection bias. The selection of individuals from the target population was random, however, several factors may have influenced whether individuals agreed to participate or were reachable by phone. Due to ethical considerations, information regarding reasons for not attending the study was not available. Nevertheless, some individuals unsolicited stated periodontitis as a reason for not participating in the study due to tiring and frequent visits to the dentist, which may indicate that the prevalence might be even higher in the target population. However, a lower prevalence of periodontitis among Caucasians compared to individuals of other ethnicities has been reported [33]. Furthermore, low socioeconomic status has also been reported as a risk indicator for attachment loss and increased probing depth [33]. When comparing background characteristics of individuals in the present study with the general population (based on register data from Statistics Norway), western-born and highly educated individuals were overrepresented compared to the target population, and the prevalence might therefore have been underestimated in the present study. In addition, the proportion of current smokers, which is a well-documented risk factor for periodontitis [34, 35], was lower in the present study population compared to the target population which may also have led to an underestimation of periodontitis prevalence. The proportion of individuals with diabetes was similar in the present study population compared to the target population. However, due to the lack of information regarding diseases control in diabetic individuals (HbA1c) in the present study, diabetes was only treated as a grade modifier for shifting the grade from A to B and not from B to C. This may have led to an underestimation of grade C periodontitis. Due to the high proportion of highly educated individuals, low proportion of smokers, public financial support for periodontal treatment, and readily accessible dental health services, one could expect a low risk for periodontitis. Despite this, we found a high prevalence Page 8 of 9 in the present study population. This indicates that the prevalence might be even higher in the target population of 65-year-olds in Oslo and even higher in areas where dental health services are not that readily accessible. Conclusions Periodontitis was common among 65 year-olds living in Oslo, and the majority of those with the disease had unstable, recurrent periodontitis. This indicates a substantial need for periodontal treatment in this population in the years to come. In addition, this study shows that the choice of radiographic bone loss cutoff for defining a periodontitis case affects the prevalence estimates to a large extent. This study also indicates that the staging and grading procedure in the 2018 EFP/AAP classification has major limitations for epidemiologic studies in its current form. Acknowledgements We want to thank The Department of Maxillofacial Radiology at the Faculty of Dentistry, University of Oslo, for organizing and conducting of orthopanto‑ mogram x-rays. In addition, we want to thank all subjects for participating in the study. Author contributions Study concept and design: ATTS, LHH, OCK, RSR, MTD. Subject recruitment: ATTS, MTD, LHH. Clinical data collection: ATTS, MTD, OCK, LHH. Analysis and interpretation of data: ATTS, MTD, OCK, LHH, RSR. Writing the manuscript: ATTS, OCK, LHH, RSR, MTD. Critically evaluating the manuscript: ATTS, OCK, LHH, RSR, MTD. Project leader: LHH. All authors read and approved the final manuscript. Funding The study was funded by the Faculty of Dentistry, University of Oslo, Norway. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate The study was approved by the Norwegian Regional Committee for Medical and Health Research Ethics (REK 2018/1383) and performed in compliance with the tenets of the Declaration of Helsinki. All participants signed a written informed consent prior to examinations. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Author details 1 Department of Cariology and Gerodontology, Faculty of Dentistry, University of Oslo, Blindern, P.O. Box 1109, 0317 Oslo, Norway. 2 Department of Periodon‑ tology, Faculty of Dentistry, University of Oslo, Oslo, Norway. Received: 12 February 2022 Accepted: 9 June 2022 Sødal et al. BMC Oral Health (2022) 22:246 References 1. Pihlstrom BL, Michalowicz BS, Johnson NW. Periodontal diseases. Lancet. 2005;366:1809–20. 2. Helal O, Göstemeyer G, Krois J, Fawzy El Sayed K, Graetz C, Schwendicke F. Predictors for tooth loss in periodontitis patients: systematic review and meta-analysis. J Clin Periodontol. 2019;46:699–712. 3. Ravidà A, Qazi M, Troiano G, Saleh MHA, Greenwell H, Kornman K, et al. Using periodontal staging and grading system as a prognostic factor for future tooth loss: a long-term retrospective study. J Periodontol. 2020;91:454–61. 4. The National Insurance Act chapter 5. The Norwegian Directorate of Health, Oslo. 1997. https://www.helsedirektoratet.no/rundskriv/folke trygdloven-kap-5. Accessed 2 Feb 2022. 5. Bongo AS, Brustad M, Oscarson N, Jönsson B. Periodontal health in an indigenous Sámi population in Northern Norway: a cross-sectional study. BMC Oral Health. 2020;20:104. 6. Stødle IH, Verket A, Høvik H, Sen A, Koldsland OC. Prevalence of peri‑ odontitis based on the 2017 classification in a Norwegian population: The HUNT study. J Clin Periodontol. 2021;48:1189–99. 7. Holde GE, Oscarson N, Trovik TA, Tillberg A, Jonsson B. Periodontitis prevalence and severity in adults: a cross-sectional study in Norwegian Circumpolar Communities. J Periodontol. 2017;88:1012–22. 8. Eke PI, Wei L, Borgnakke WS, Thornton-Evans G, Zhang X, Lu H, et al. Periodontitis prevalence in adults >/= 65 years of age, in the USA. Peri‑ odontol. 2000;2016(72):76–95. 9. Hugoson A, Koch G, Göthberg C, Helkimo AN, Lundin SA, Norderyd O, et al. Oral health of individuals aged 3–80 years in Jönköping, Sweden during 30 years (1973–2003). II. Review of clinical and radiographic find‑ ings. Swed Dent J. 2005;29:139–55. 10. Müller F, Shimazaki Y, Kahabuka F, Schimmel M. Oral health for an ageing population: the importance of a natural dentition in older adults. Int Dent J. 2017;67(Suppl 2):7–13. 11. Gooch BF, Eke PI, Malvitz DM. Public health and aging: retention of natu‑ ral teeth among older adults--United States, 2002. MMWR Morb Mortal Wkly Rep: Centers for Disease Control and Prevention (CDC), Div of Oral Health NCfCDPaHP, CDC; 2003 Dec 19. Report No.: 0149-2195 Contract No. 50. 12. Masood M, Younis LT, Masood Y, Bakri NN, Christian B. Relationship of peri‑ odontal disease and domains of oral health-related quality of life. J Clin Periodontol. 2019;46:170–80. 13. Persson GR. Periodontal complications with age. Periodontol. 2000;2018(78):185–94. 14. Weijdijk LPM, Ziukaite L, Van der Weijden GAF, Bakker EWP, Slot DE. The risk of tooth loss in patients with diabetes: a systematic review and metaanalysis. Int J Dent Hyg. 2022;20:145–66. 15. United Nations DoEaSA, Population Division. World Population Ageing 2019 (ST/ESA/SER.A/444). United Nations, Department of Economic and Social Affairs PD; 2020. 16. Holtfreter B, Albandar JM, Dietrich T, Dye BA, Eaton KA, Eke PI, et al. Stand‑ ards for reporting chronic periodontitis prevalence and severity in epide‑ miologic studies: proposed standards from the Joint EU/USA Periodontal Epidemiology Working Group. J Clin Periodontol. 2015;42:407–12. 17. Leite FRM, Peres KG, Do LG, Demarco FF, Peres MAA. Prediction of peri‑ odontitis occurrence: influence of classification and sociodemographic and general health information. J Periodontol. 2017;88:731–43. 18. Eke PI, Page RC, Wei L, Thornton-Evans G, Genco RJ. Update of the case definitions for population-based surveillance of periodontitis. J Periodon‑ tol. 2012;83:1449–54. 19. Papapanou PN, Sanz M, Buduneli N, Dietrich T, Feres M, Fine DH, et al. Periodontitis: consensus report of workgroup 2 of the 2017 World Work‑ shop on the classification of periodontal and peri-implant diseases and conditions. J Periodontol. 2018;89(Suppl 1):S173–82. 20. Ke L, Nogueira G, Thomson WM. Influence of case definitions on epide‑ miological estimates of periodontitis prevalence and its associations with smoking and OHRQoL. Community Dent Oral Epidemiol. 2022. https:// doi.org/10.1111/cdoe.12726. 21. Diep MT, Jensen JL, Skudutyte-Rysstad R, Young A, Sødal ATT, Petrovski B, et al. Xerostomia and hyposalivation among a 65-yr-old population living in Oslo, Norway. Eur J Oral Sci. 2021;129: e12757. Page 9 of 9 22. Nyman S, Lindhe J, Lundgren D. The role of occlusion for the stability of fixed bridges in patients with reduced periodontal tissue support. J Clin Periodontol. 1975;2:53–66. 23. Hamp SE, Nyman S, Lindhe J. Periodontal treatment of multirooted teeth. Results after 5 years. J Clin Periodontol. 1975;2:126–35. 24. Chapple ILC, Mealey BL, Van Dyke TE, Bartold PM, Dommisch H, Eickholz P, et al. Periodontal health and gingival diseases and conditions on an intact and a reduced periodontium: consensus report of workgroup 1 of the 2017 World Workshop on the classification of periodontal and peri-implant diseases and conditions. J Clin Periodontol. 2018;45(Suppl 20):S68-s77. 25. Tonetti MS, Greenwell H, Kornman KS. Staging and grading of periodon‑ titis: framework and proposal of a new classification and case definition. J Periodontol. 2018;89(Suppl 1):S159–72. 26. Sanz M, Herrera D, Kebschull M, Chapple I, Jepsen S, Beglundh T, et al. Treatment of stage I-III periodontitis—the EFP S3 level clinical practice guideline. J Clin Periodontol. 2020;47(Suppl 22):4–60. 27. Wahlin Å, Papias A, Jansson H, Norderyd O. Secular trends over 40 years of periodontal health and disease in individuals aged 20–80 years in Jönköping, Sweden: repeated cross-sectional studies. J Clin Periodontol. 2018;45:1016–24. 28. Norderyd O, Henriksen BM, Jansson H. Periodontal disease in Norwegian old-age pensioners. Gerodontology. 2012;29:4–8. 29. Eke PI, Thornton-Evans GO, Wei L, Borgnakke WS, Dye BA, Genco RJ. Peri‑ odontitis in US adults: National Health and Nutrition Examination Survey 2009–2014. J Am Dent Assoc. 2018;149:576-88.e6. 30. Carasol M, Llodra JC, Fernández-Meseguer A, Bravo M, García-Margallo MT, Calvo-Bonacho E, et al. Periodontal conditions among employed adults in Spain. J Clin Periodontol. 2016;43:548–56. 31. Costa FO, Guimarães AN, Cota LO, Pataro AL, Segundo TK, Cortelli SC, et al. Impact of different periodontitis case definitions on periodontal research. J Oral Sci. 2009;51:199–206. 32. Manau C, Echeverria A, Agueda A, Guerrero A, Echeverria JJ. Periodontal disease definition may determine the association between periodontitis and pregnancy outcomes. J Clin Periodontol. 2008;35:385–97. 33. Albandar JM. Epidemiology and risk factors of periodontal diseases. Dent Clin N Am. 2005;49(517–32):v–vi. 34. Gelskey SC. Cigarette smoking and periodontitis: methodology to assess the strength of evidence in support of a causal association. Community Dent Oral Epidemiol. 1999;27:16–24. 35. Nociti FH Jr, Casati MZ, Duarte PM. Current perspective of the impact of smoking on the progression and treatment of periodontitis. Periodontol. 2000;2015(67):187–210. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions
https://openalex.org/W3214292763	https://www.frontiersin.org/articles/10.3389/fpsyg.2021.781794/pdf	English	null	Social Support and Subjective Well-Being in Chinese Parents of Children With Autism Spectrum Disorder: The Mediating Role of Perceived Discrimination	Frontiers in psychology	2,021	cc-by	4,664	Social Support and Subjective Well-Being in Chinese Parents of Children With Autism Spectrum Disorder: The Mediating Role of Perceived Discrimination 1 School of Educational Sciences, Anshun University, Anshun, China, 2 Faculty of Education, Yunnan Normal University, Kunming, China The present research was done to examine whether social support was related to subjective well-being on Chinese parents of children with autism spectrum disorder (ASD) and how perceived discrimination affected this relationship. Two hundred four parents with ASD children were investigated by Inventory of Social Support Behavior, Perceived Discrimination Scale for Parents of Children With ASD, Subjective Well-being Scale. The results showed that perceived discrimination was negatively associated with social support and subjective well-being, and social support was positively related to subjective well-being. Furthermore, perceived discrimination played a partial mediating role between social support and subjective well-being. All the ﬁndings suggest that social support can directly inﬂuence subjective well-being of parents of ASD children and indirectly inﬂuence subjective well-being through perceived discrimination. Correspondence: Correspondence: Yongfei Ban byf20200508@163.com ORIGINAL RESEARCH published: 08 November 2021 doi: 10.3389/fpsyg.2021.781794 INTRODUCTION Specialty section: This article was submitted to Health Psychology, a section of the journal Frontiers in Psychology Autism spectrum disorder (ASD) is a kind of developmental disability which has been characterized by persistent impairments in social interaction and the existence of repetitive behaviors, interests, and activities (American Psychiatric Association, 2013). Globally, the number of ASD children increases rapidly, and the core symptoms of ASD often persist in life-long development (Centers for Disease Control and Prevention, 2020). Previous studies have suggested that parents raising a child with ASD may experience enormous challenges associated with intensive caregiving tasks and the accompanied comorbidities (e.g., anxiety and depression). In addition, parents may be constantly exposed in public hospitals and centers and were strongly subjected to stigma. Under the high levels of emotional and ﬁnancial presses, parents with ASD children had low levels of subjective well- being (SWB; Lu et al., 2015; Costa et al., 2017; Garrido et al., 2021). To improving SWB of ASD families, it was essential to explore the potential predictors of SWB for parents with ASD children. Social support (SS) is presented by the perceived comfort, caring, assistance, and esteem one individual receives from others (Wallston et al., 1983). It may be one of the most beneﬁcial and robust predictors of well-being (Shorey and Lakey, 2011; Hassanein et al., 2021). Previous studies have found that SS is signiﬁcantly related to families’ well-being (Lin et al., 2011; Frantz et al., 2018; Received: 23 September 2021 Accepted: 15 October 2021 Published: 08 November 2021 Edited by: Song Wang, Sichuan University, China Reviewed by: Yanhui Xiang, Hunan Normal University, China Anguo Fu, Hainan University, China Correspondence: Yongfei Ban byf20200508@163.com Edited by: Song Wang, Sichuan University, China Reviewed by: Yanhui Xiang, Hunan Normal University, China Anguo Fu, Hainan University China Yanhui Xiang, Hunan Normal University, China Anguo Fu, Keywords: social support, perceived discrimination, subjective well-being, parents with ASD children, mediation analysis Hainan University, China Citation: Ban Y, Sun J and Liu J (2021) Social Support and Subjective Well-Being in Chinese Parents of Children With Autism Spectrum Disorder: The Mediating Role of Perceived Discrimination. Front. Psychol. 12:781794. doi: 10.3389/fpsyg.2021.781794 November 2021 \| Volume 12 \| Article 781794 1 Frontiers in Psychology \| www.frontiersin.org SWB, Social Support, Perceived Discrimination Ban et al. Khusaifan and Keshky, 2020). Furthermore, the parents who had high SS needs reported low levels of well-being (Hassanein et al., 2021). The age of parents ranged from 23 to 48 years (M = 34.23, SD = 5.78). Most of the parents were female (69.7%) and did not have a university degree (83.3%). The children were between 4–18 years old (Mage = 7.25, SDage = 3.53). Demographic characteristics can be found in Table 1. The data of six parents were excluded due to the missing data on demographic information and experimental measures. All participants gave written informed consent before the experiment and received a small gift (e.g., a coloring book) to thank them. After the experimental session was ﬁnished, the participants were told about the purpose of the study. In addition, members of developmental disabilities families may face discrimination in some social interaction contexts (Schmitt et al., 2014). Perceived discrimination (PD) is understood as the subjective interpretation of unfair and inequitable treatment based on societal group membership (Schmitt et al., 2014). Several studies have shown that PD negatively aﬀects caregiver’s well-being for developmental disabilities families (Zhou et al., 2018; Mitter et al., 2019). For parents with ASD children, PD may lead them to accept stigma as a part of their self-concept (Herek, 2007) and feel blame, anxiety, hopelessness (Recio et al., 2021). The phenomenological variant of ecological systems Theory (PVEST; Spencer et al., 1997) may serve as a theoretical perspective to understand the relationships among SS, PD, and SWB. PVEST emphasized the critical role of coping and adaptive processes in resource access and suggested that coping and adaptive processes aﬀected the individuals’ and families’ appropriate behavior and mental health (Spencer et al., 2015). According to PVEST (Spencer et al., 1997), minority families and individuals might experience stress engagement variables (e.g., perceived social support) when linked with intergroup contact contexts. Furthermore, positive cognitive appraisals of SS would lead to adaptive coping responses (e.g., less discrimination perception, self-acceptance) which inﬂuenced behavioral and health-relevant outcomes (e.g., mental health and intimacy). Inventory of Social Support Behavior y pp The 10-item Chinese version of the ISSB was used to assess SS of Chinese parents of children with ASD in this study (Barrera et al., 1981; Luo, 2001). Sample items are as follows: “Talked with you about some interests of yours”, “Listened to you talk about your private feelings”. Items are scored on a 3-point scale from 1 (not satisﬁed) to 3 (satisﬁed). The lower score of all items indicated that parents reported lower levels of SS. Internal consistency coeﬃcients of the scale reached α = 0.91. In the current study, the Chinese version of ISSB resulted in an internal consistency coeﬃcient of α = 0.88. The previous studies conﬁrmed the construct validity of ISSB (Luo, 2001). However, to our knowledge, previous studies have mainly examined the relationships among SS, PD, and SWB of patients with mental disorders (Jasinskaja-Lahti et al., 2006; Schauman et al., 2019), people with physical disabilities (Ji et al., 2019), and immigrants (Fernández et al., 2014; Hashemi et al., 2019, 2020). Few studies have focused on members of ASD families, especially parents. Thus, it is crucial to understand whether SS is associated with SWB and how PD inﬂuence this relationship in the Chinese sample of parents with ASD children. In summary, this study aims to examine (a) the relationship between SS and SWB and (b) the mediating role of PD in Chinese parents with ASD children. We used measures to assess the levels of SS, PD, and SWB for parents with ASD children. Then, we conducted mediating analysis to examine the mediating role of PD in the relationship between SS and SWB. TABLE 1 \| Demographic characteristics of participants. Variable Mean SD Range Parent’s age 34.23 5.78 23–48 Child’s age 7.25 3.53 4–18 Variable % (n) Parent’s gender Male 30.3(60) Female 69.7(138) Parent’s educational level Primary school and below 11.1(22) Middle school 27.3(54) High school degree 23.2(46) College degree 21.7(43) University degree and above 16.7(33) November 2021 \| Volume 12 \| Article 781794 TABLE 1 \| Demographic characteristics of participants. Citation: Based on the PVEST, it seems reasonable to assume that PD may mediate the relationship between SS and SWB in parents with ASD children. The Perceived Discrimination Scale for Parents of Children With ASD (PDS-FP) ( ) The PDS-FP was used to measure PD of Chinese parents with ASD children (Zhao, 2018). The scale includes 10 items and uses 4-point scale from 1 (strongly disagree) to 4 (strongly agree). The scale consists of two subscales: discrimination perception and discrimination attribution. Sample items are as follows: “It is hard to make friends”, “Being blamed for not teaching your children well”. The lower scores reﬂect lower levels of PD. The Cronbach’s α of the scale was 0.91. Internal consistency coeﬃcients of the scale were α = 0.88 in the present study. The previous studies provided good evidence of construct validity for PDS-FP (Zhao, 2018). November 2021 \| Volume 12 \| Article 781794 Frontiers in Psychology \| www.frontiersin.org The Brief Subjective Well-Being Scale for Chinese Citizen of SS and SWB. However, the SS positively correlated with SWB (r = −0.36, p < 0.001). It meant that high levels of SS in parents were associated with high levels of SWB. of SS and SWB. However, the SS positively correlated with SWB (r = −0.36, p < 0.001). It meant that high levels of SS in parents were associated with high levels of SWB. The brief subjective well-being scale for Chinese citizen [SWBS- CC (brief)] is a 20-items scale used to measure SWB of Chinese parents with ASD children (Xing, 2003). Sample items are as follows: “Compared with the people around me, I feel satisﬁed.”, “I often get annoyed by some trivial things.”. The items are rated on a 6-point scale from 1 (totally disagree) to 6 (totally agree). The total score can be calculated by averaging all 20 items. The lower score represents the lower levels of SWB. In the present study, parents scoring above 4.5 were considered to have high levels of SWB, and below 2.5 were considered to have low levels of SWB. The Cronbach’s alpha reliability was α = 0.85. In this study, the computed Cronbach’s α was 0.81. Participants Two hundred and four parents of children with ASD participated in the present study. The parents were recruited through seven special education centers and schools in the cities of Guiyang, Zunyi, and Liupanshui in China. All of the children had been diagnosed with ASD and received disability-related services in special education centers and schools. The study was approved by the ethics committee of Anshun university. Frontiers in Psychology \| www.frontiersin.org 2 SWB, Social Support, Perceived Discrimination Ban et al. Procedures With assist from several special education centers and schools, the research team contacted parents with ASD children who showed interest in participating in this study. Parents were informed the study’s objectives and their right (e.g., protect the anonymity of them). All participants gave clear written informed consent before data collection. Then, an individual meeting was held with each parent to ensure that all parents accurately understood the measures and procedures. DISCUSSION The statistical analyses of data were conducted using SPSS software (Version 27.0). At ﬁrst, Pearson correlation analyses were conducted to examine the relationships between PD, SS, and SWB in parents with ASD children. Then, we used PROCESS macro for SPSS which developed by Hayes to examine the mediation role of PD in the relationship between SS and SWB. The PROCESS macro uses a path analysis modeling tool which is based on regression to assess direct and indirect eﬀects of variables (Hayes, 2018). Following the recommendations of Hayes, we performed the mediation analyses using Model 4 with 5000 bias-corrected bootstrap samples. The current study aimed to examine the relationship between SS and SWB and the mediating role of PD in Chinese parents of children with ASD. First, the ﬁndings showed that SS was positively related to SWB. Speciﬁcally, parents with ASD children who perceived high levels of SS reported high levels of SWB. These results conform with previous studies (Lin et al., 2011; Frantz et al., 2018; Khusaifan and Keshky, 2020). From the theoretical point of view, the main eﬀects of SS suggested that the presence of SS may provide important contributions to positive mental health and SWB (Kondrat et al., 2018). Parents with ASD children value SS as an active coping strategy against conﬂicting life events (Davis and Brekke, 2014). As positive social-psychological resources of families and individuals, SS leads parents to accept children’s condition and positively deal with the complex challenges accompanying the rearing of ASD children (Ludlow et al., 2012). For ASD families, an intense and The Mediation Role of Perceived Discrimination in the Relationship Between Social Support and Subjective Well-Being To test whether PD would mediate the relationship between SS and SWB, we ran a mediation analysis with SS as a predictor, PD as a mediator and SWB as the dependent variable. Figure 1 presents the mediation model. The results showed that the eﬀect of SS on PD was −0.24 (SE = 0.07, t = −3.43, p < 0.001, 95% CI: −0.38 to −0.10). Also, the eﬀect of PD on SWB was −0.61 (SE = 0.05, t = −11.43, p < 0.001, 95% CI: −0.71 to −0.50). Importantly, SS has a total eﬀect on SWB of 0.36 (SE = 0.07, t = 5.46, p < 0.001, 95% CI: 0.23 to 0.50). Speciﬁcally, the indirect eﬀect of SS on SWB was 0.15 and statistically diﬀerent from zero (95% CI: 0.04 to 0.25), while the direct eﬀect was 0.22 (SE = 0.05, t = 4.11, p < 0.001, 95% CI: 0.11 to 0.32). This model explained 48.01% of SWB variance (F = 90.02, p < 0.001). Overall, these results suggested that PD partially mediated the relationship between SS and SWB. CONCLUSION active system of SS is one of the most inﬂuential factors that can improve SWB of families. Second, the present study’s ﬁnding suggested a signiﬁcant negative relationship between PD and SWB. High levels of PD were associated with low levels of SWB for Chinese parents of children with ASD. This ﬁnding is consistent with the previous studies (Ji et al., 2019; Schauman et al., 2019). For ASD families, facing stigmatization and discrimination from their environment intensiﬁed the perceived stress and reinforced the feeling of being rejected, self-blame, and angry of individuals (Yasuhiro, 2018). Furthermore, PD might make it diﬃcult for families’ members to seek supportive resources (Benomir et al., 2016). On the other hand, the members of ASD families would employ ineﬀective coping strategies (e.g., avoiding interpersonal contacts) to deal with PD and further exacerbated the severe negative impact on families’ social networks (Kondrat et al., 2018). In this manner, PD harmed the SWB of ASD families (Schmitt et al., 2014). To our knowledge, this is one of few studies to examine the relationships among PD, SS, and SWB in Chinese parents of children with ASD. The results supported the relationship between SS and SWB. Furthermore, PD performed as a mediator of this relationship. These ﬁndings may provide valuable information on how to improve well-being in Chinese parents of children with ASD. Given the mediating role of PD, programs related to PD should be designed to enhance well-being in Chinese parents of children with ASD who report high levels of SS. In addition, consider the direct eﬀect of SS on SWB, interventions aiming at improving SS should be implemented to increase well-being in Chinese parents of children with ASD. To our knowledge, this is one of few studies to examine the relationships among PD, SS, and SWB in Chinese parents of children with ASD. The results supported the relationship between SS and SWB. Furthermore, PD performed as a mediator of this relationship. These ﬁndings may provide valuable information on how to improve well-being in Chinese parents of children with ASD. Given the mediating role of PD, programs related to PD should be designed to enhance well-being in Chinese parents of children with ASD who report high levels of SS. In addition, consider the direct eﬀect of SS on SWB, interventions aiming at improving SS should be implemented to increase well-being in Chinese parents of children with ASD. ETHICS STATEMENT The studies involving human participants were reviewed and approved by School of Education Science, Anshun University. The patients/participants provided their written informed consent to participate in this study. DATA AVAILABILITY STATEMENT Third, and most importantly, the ﬁnding of this study indicated the mediating role of PD. More speciﬁcally, PD mediated the association between SS and SWB. The results supported the view of the PVEST. As PVEST pointed out, the balance between protective and risk factors in individuals’ and families’ development produced speciﬁc psycho-social outcomes. The SS was an important protective factor that buﬀered against the stress of individuals and families. In addition, the self- evaluation process (e.g., PD), which indicated the perception of one’s association with others in social interactions (e.g., the members of hospitals, neighbors), had a substantial eﬀect on mental health outcomes (Spencer et al., 2015). When contacting with other people, inappropriate supports may lead people from ASD families to experiencing negative stereotypes about them and their families, such as the belief that “the burden of disability is unending for the family and they are the most perfect objects of charity” (Block, 2018). Parents and their ASD children may have to bear all the pressure associated with the unsuitable and excessive assistance from fully human people, especially in collectivist societies (e.g., China and Japan) where helping people with disabilities in public places is an important social norm. As Liao et al. (2019) pointed out, the PD related to social interactions had a negative eﬀect on SWB for parents with ASD children. The data presented in this study are available on request from the corresponding author. The data are not publicly available due to the privacy of the participants. FUNDING This research was funded by the National Social Science Foundation of China (18CSH042) and the Department of Education of Guizhou Province (2018520103). AUTHOR CONTRIBUTIONS JL, JS, and YB performed the research, collected data, analysis, and interpretation of data. JS and YB wrote the manuscript and designed the research. All authors contributed to the article and approved the submitted version. Finally, these ﬁndings should be interpreted cautiously, and there are some limitations in the present study. First, all the data of this study were based on the parents’ self-reports about PD, SS, and SWB, and our study lacked the quantitative indexes of these variables. Second, although this study provides evidence of the relationships between SS, PD, and SWB, it is also essential to know how these relationships change and what factors inﬂuence them. In addition, this study was cross-sectional design for parents with ASD children. Still, it remains unclear whether the relationship between SS and SWB and the mediating role of PD could also be obtained in samples of parents with normal children. Future research may examine these issues. Descriptive and Correlational Analyses Descriptive and Correlational Analyses The mean total scores and correlations of three scales are presented in Table 2. The results showed that the PD was negatively correlated with the SS (r = −0.24, p < 0.001) and the SWB (r = −0.66, p < 0.001), indicating that the parents with ASD children reported high levels of PD if they had low levels FIGURE 1 \| The mediation model for the effect of PD in the relationship between SS and SWB. p < 0.001. TABLE 2 \| Descriptive and correlational analyses (n = 198). Mean SD 1 2 3 1 PD 2.24 0.66 — 2 SS 2.29 0.45 −0.24* — 3 SWB 4.01 0.62 −0.66* −0.36* — *p < 0.001. FIGURE 1 \| The mediation model for the effect of PD in the relationship between SS and SWB. p < 0.001. Frontiers in Psychology \| www.frontiersin.org 3 November 2021 \| Volume 12 \| Article 781794 TABLE 2 \| Descriptive and correlational analyses (n = 198). Mean SD 1 2 3 1 PD 2.24 0.66 — 2 SS 2.29 0.45 −0.24 — 3 SWB 4.01 0.62 −0.66* −0.36* — p < 0.001. FIGURE 1 \| The between SS and Frontiers in Psychology \| www.frontiersin.org 3 TABLE 2 \| Descriptive and correlational analyses (n = 198). Mean SD 1 2 3 1 PD 2.24 0.66 — 2 SS 2.29 0.45 −0.24 — 3 SWB 4.01 0.62 −0.66* −0.36* — p < 0.001. FIGURE 1 \| The mediation model for the effect of PD in the relationship between SS and SWB. **p < 0.001. November 2021 \| Volume 12 \| Article 781794 3 SWB, Social Support, Perceived Discrimination Ban et al. REFERENCES Social support, resilience, and quality of life for families with children with intellectual disabilities. Res. Dev. Disabil. 112:103910. doi: 10.1016/j.ridd.2021.103910 Yasuhiro, A. M. (2018). Parents of Children Diagnosed with Autism Spectrum Disorder: Perceptions of Stigma and Quality of Life. Ph.D. thesis. San Francisco, CA: Alliant International University. Hayes, A. (2018). Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approcch, 2nd Edn. New York, NY: Guilford Press. Zhao, Q. J. (2018). Relationship Among Perceived Discrimination, Social Anxiety and Perceived Social Support of Parents of Children with Autism Spectrum Disorder. Master’s thesis. Zhangzhou: Minnan Normal University. Herek, G. (2007). Confronting sexual stigma and prejudice: theory and practice. J. Soc. Issues 63, 905–925. doi: 10.1111/j.1540-4560.2007.00544.x Zhou, T., Wang, Y., and Yi, C. (2018). Aﬃliate stigma and depression in caregivers of children with autism spectrum disorders in China: eﬀects of self-esteem, shame and family functioning. Psychiat. Res. 264, 260–265. doi: 10.1016/j. psychres.2018.03.071 Jasinskaja-Lahti, I., Liebkind, K., Jaakkola, M., and Reuter, A. (2006). Perceived discrimination, social support networks, and psychological well-being among three immigrant groups. J. Cross Cult. Psychol. 37, 293–311. doi: 10.1177/ 0022022106286925 Conﬂict of Interest: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Ji, Y., Rana, C., Shi, C., and Zhong, Y. (2019). Self-esteem mediates the relationships between social support, subjective well-being, and perceived discrimination in Chinese people with physical disability. Front. Psychol. 10:02230. doi: 10.3389/ fpsyg.2019.02230 Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their aﬃliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Khusaifan, S. J., and Keshky, M. (2020). Social support as a protective factor for the well-being of parents of children with autism in Saudi Arabia. J. Pediatr. Nurs. 58, e1–e7. doi: 10.1016/j.pedn.2020.11.014 Kondrat, D. C., Sullivan, W. P., Wilkins, B., Barrett, B. J., and Beerbower, E. (2018). The mediating eﬀect of social support on the relationship between the impact of experienced stigma and mental health. Stigma Health 3, 305–314. doi: 10.1037/sah0000103 Copyright © 2021 Ban, Sun and Liu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). REFERENCES Social support and functional outcome in severe mental illness: the mediating role of proactive coping. Psychiatry Res. 215, 39–45. doi: 10.1016/j.psychres.2013.09.010 Schmitt, M. T., Branscombe, N. R., Postmes, T., and Garcia, A. (2014). The consequences of perceived discrimination for psychological well-being: a meta- analytic review. Psychol. Bull. 140, 921–948. doi: 10.1037/a0035754 Fernández, I., Silván-Ferrero, P., Molero, F., Gaviria, E., and García-Ael, C. (2014). Perceived discrimination and well-being in romanian immigrants: the role of social support. J. Happiness Stud. 16, 857–870. doi: 10.1007/s10902-014- 9537-0 Shorey, R. C., and Lakey, B. (2011). Perceived and capitalization support are substantially similar: implications for social support theory. Pers. Soc. Psychol. B 37, 1068–1079. doi: 10.1177/0146167211406507 Frantz, R., Hansen, S. G., and Machalicek, W. (2018). Interventions to promote well-being in parents of children with autism: a systematic review. Rev. J. Autism. Dev. Dis. 5, 58–77. doi: 10.1007/s40489-017-0123-3 Spencer, M. B., Dupree, D., and Hartmann, T. (1997). A Phenomenological Variant of Ecological Systems Theory (PVEST): a self-organization perspective in context. Dev. Psychopathol. 9, 817–833. doi: 10.1017/s0954579497001454 Garrido, D., Petrova, D., Cokely, E., Carballo, G., and Garcia-Retamero, R. (2021). Parental risk literacy is related to quality of life in spanish families of children with autism spectrum disorder. J. Autism. Dev. Disord. 51, 2475–2484. doi: 10.1007/s10803-020-04733-8 Spencer, M. B., Swanson, D. P., and Harpalani, V. (2015). “Conceptualizing the self: contributions of normative human processes, diverse contexts and social opportunity,” in Handbook of Child Psychology and Developmental Science, eds M. Lamb, C. G. Coll, and R. Lerner (New York, NY: John Wiley & Sons Inc Press), 750–793. Hashemi, N., Marzban, M., Sebar, B., and Harris, N. (2019). Acculturation and psychological well-being among middle eastern migrants in Australia: the mediating role of social support and perceived discrimination. Int. J. Intercult. Relat. 72, 45–60. doi: 10.1016/j.ijintrel.2019.07.002 Wallston, B. S., Alagna, S. W., DeVellis, B. M., and DeVellis, R. F. (1983). Social support and physical health. Health Psychol. 2, 367–391. doi: 10.1037/0278- 6133.2.4.367 Hashemi, N., Marzban, M., Sebar, B., and Harris, N. (2020). Perceived discrimination and subjective well-being among middle eastern migrants in australia: the moderating role of perceived social support. Int. J. Soc. Psychiatr. 67, 110–119. doi: 10.1177/0020764020940740 Xing, Z. J. (2003). Developing the brief subjective well-being scale for Chinese citizen. Chin. Behav. Med. Sci. 12, 703–705. doi: 10.3760/cma.j.issn.1674-6554. 2003.06.054 Hassanein, E., Adawi, T. R., and Johnson, E. S. (2021). REFERENCES social support and coping in family adaptation and maternal well-being. Res. Autism Spect. Dis. 5, 144–156. doi: 10.1016/j.rasd.2010.03.004 American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental Disorders, 5th Edn. Arlington, VA: American Psychiatric Association. Lu, M., Yang, G., Skora, E., Wang, G., Cai, Y., Sun, Q., et al. (2015). Self-esteem, social support, and life satisfaction in Chinese parents of children with autism spectrum disorder. Res. Autism Spect. Dis. 17, 70–77. doi: 10.1016/j.rasd.2015. 05.003 Barrera, M., Sandler, I. N., and Ramsay, T. B. (1981). Preliminary development of a scale of social support: studies on college students. Am. J. Commun. Psychol. 9, 435–447. doi: 10.1007/BF00918174 Ludlow, A., Skelly, C., and Rohleder, P. (2012). Challenges faced by parents of children diagnosed with autism spectrum disorder. J. Health Psychol. 17, 702–711. doi: 10.1177/1359105311422955 Benomir, A. M., Nicolson, R. I., and Beail, N. (2016). Attitudes towards people with intellectual disability in the uk and libya: a cross-cultural comparison. Res. Dev. Disabil. 51–52, 1–9. doi: 10.1016/j.ridd.2015.12.009 Luo, K. N. (2001). A Study of the Eﬀects of Social Support, Personality Traits, and Personal Attributes on Psychological Well-Being Satisfaction of the Elderly. Ph.D. thesis. Taibei: National Chengchi University. Block, L. (2018). Stereotypes About People With Disabilities. Available online at: http://www.disabilitymuseum.org/dhm/edu/essay.html?id=24 (accessed June 15, 2021) Mitter, N., Ali, A., and Scior, K. (2019). Stigma experienced by families of individuals with intellectual disabilities and autism: a systematic review. Res. Dev. Disabil. 89, 10–21. doi: 10.1016/j.ridd.2019.03.001 Centers for Disease Control and Prevention (2020). Prevalence of Autism Spectrum Disorder Among Children Aged 8 Years- Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2016. Available online at: https://www.cdc.gov/mmwr/volumes/69/ss/ss6904a1.htm?s_cid=ss6904a1_ w (accessed June 17, 2021) Recio, P., Molero, F., Silván-Ferrero, P., and Nouvilas, E. (2021). Perceived discrimination and emotional distress among family caregivers of children with physical disabilities: the mediational role of aﬃliate stigma and self-eﬃcacy. Am. J. Orthopsychiatry 91, 367–374. doi: 10.1037/ort0000534 Costa, A. P., Steﬀgen, G., and Ferring, D. (2017). Contributors to well-being and stress in parents of children with autism spectrum disorder. Res. Autism. Spect. Dis. 37, 61–72. doi: 10.1016/j.rasd.2017.01.007 J. Orthopsychiatry 91, 367–374. doi: 10.1037/ort0000534 Schauman, O., Macleod, A. K., Thornicroft, G., and Clement, S. (2019). Mental illness related discrimination: the role of self-devaluation and anticipated discrimination for decreased well-being. Stigma Health 4, 11–18. doi: 10.1037/ sah0000117 Davis, L., and Brekke, J. (2014). ACKNOWLEDGMENTS We would like to thank our participants who devoted their time to support this research. November 2021 \| Volume 12 \| Article 781794 Frontiers in Psychology \| www.frontiersin.org SWB, Social Support, Perceived Discrimination Ban et al. REFERENCES The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Liao, X. L., Lei, X. Y., and Li, Y. M. (2019). Stigma among parents of children with autism: a literature review. Asian J. Psychiatr. 45, 88–94. doi: 10.1016/j.ajp.2019. 09.007 Lin, L. Y., Orsmond, G. I, Coster, W. J., and Cohn, E. S. (2011). Families of adolescents and adults with autism spectrum disorders in taiwan: the role of November 2021 \| Volume 12 \| Article 781794 Frontiers in Psychology \| www.frontiersin.org 5
https://openalex.org/W2965851702	https://escholarship.org/content/qt0n57m95t/qt0n57m95t.pdf?t=qak5vq	English	null	The Trials and Tribulations of Structure Assisted Design of KCa Channel Activators	Frontiers in pharmacology	2,019	cc-by	21,423	UC Davis UC Davis Previously Published Works Title The Trials and Tribulations of Structure Assisted Design of KCa Channel Activators Permalink https://escholarship.org/uc/item/0n57m95t Journal Frontiers in Pharmacology, 10(SEP) ISSN 1663-9812 Authors Shim, Heesung Brown, Brandon M Singh, Latika et al. Publication Date 2019 DOI 10.3389/fphar.2019.00972 Peer reviewed UC Davis UC Davis Previously Published Works Title The Trials and Tribulations of Structure Assisted Design of KCa Channel Activators Permalink https://escholarship.org/uc/item/0n57m95t Journal Frontiers in Pharmacology, 10(SEP) ISSN 1663-9812 Authors Shim, Heesung Brown, Brandon M Singh, Latika et al. Publication Date 2019 DOI 10.3389/fphar.2019.00972 Peer reviewed UC Davis UC Davis UC Davis Previously Published Works Title The Trials and Tribulations of Structure Assiste Permalink https://escholarship.org/uc/item/0n57m95t Journal Frontiers in Pharmacology, 10(SEP) ISSN 1663-9812 Authors Shim, Heesung Brown, Brandon M Singh, Latika et al. Publication Date 2019 DOI 10.3389/fphar.2019.00972 Peer reviewed UC Davis Previously Published Works Title The Trials and Tribulations of Structure Assisted Design of KCa Channel A Permalink https://escholarship.org/uc/item/0n57m95t Journal Frontiers in Pharmacology, 10(SEP) ISSN 1663-9812 Authors Shim, Heesung Brown, Brandon M Singh, Latika et al. Publication Date 2019 DOI 10.3389/fphar.2019.00972 Peer reviewed UC Davis Previously Published Works eScholarship.org Powered by the California Digital Library University of California Powered by the California Digital Library University of California The Trials and Tribulations of Structure Assisted Design of KCa Channel Activators Heesung Shim 1,2†, Brandon M. Brown 1†, Latika Singh 1, Vikrant Singh 1, James C. Fettinger 2, Vladimir Yarov-Yarovoy 3 and Heike Wulff 1* 1 Department of Pharmacology, School of Medicine, University of California, Davis, Davis, CA, United States, 2 Department of Chemistry, University of California, Davis, Davis, CA, United States, 3 Department of Physiology and Membrane Biology, School of Medicine, University of California, Davis, Davis, CA, United States Calcium-activated K+ channels constitute attractive targets for the treatment of neurological and cardiovascular diseases. To explain why certain 2-aminobenzothiazole/oxazole-type KCa activators (SKAs) are KCa3.1 selective we previously generated homology models of the C-terminal calmodulin-binding domain (CaM-BD) of KCa3.1 and KCa2.3 in complex with CaM using Rosetta modeling software. We here attempted to employ this atomistic level understanding of KCa activator binding to switch selectivity around and design KCa2.2 selective activators as potential anticonvulsants. In this structure-based drug design approach we used RosettaLigand docking and carefully compared the binding poses of various SKA compounds in the KCa2.2 and KCa3.1 CaM-BD/CaM interface pocket. Based on differences between residues in the KCa2.2 and KCa.3.1 models we virtually designed 168 new SKA compounds. The compounds that were predicted to be both potent and KCa2.2 selective were synthesized, and their activity and selectivity tested by manual or automated electrophysiology. However, we failed to identify any KCa2.2 selective compounds. Based on the full-length KCa3.1 structure it was recently demonstrated that the C-terminal crystal dimer was an artefact and suggested that the “real” binding pocket for the KCa activators is located at the S4-S5 linker. We here confirmed this structural hypothesis through mutagenesis and now offer a new, corrected binding site model for the SKA-type KCa channel activators. SKA-111 (5-methylnaphtho[1,2-d]thiazol-2-amine) is binding in the interface between the CaM N-lobe and the S4-S5 linker where it makes van der Waals contacts with S181 and L185 in the S45A helix of KCa3.1. Keywords: calcium-activated potassium channels, KCa3.1, KCa2.2, calmodulin binding domain, SKA-111, Rosetta, KCa activators Edited by: Mounir Tarek, Centre National de la Recherche Scientifique (CNRS), France Reviewed by: Lucie Parent, Université de Montréal, Canada Bernard Attali, Tel Aviv University, Israel Luis Gonzalo Cuello, Texas Tech University Health Sciences Center, United States Correspondence: Heike Wulff hwulff@ucdavis.edu †These authors have contributed equally to this work Reviewed by: Lucie Parent, Université de Montréal, Canada Bernard Attali, Tel Aviv University, Israel Luis Gonzalo Cuello, Texas Tech University Health Sciences Center, United States Correspondence: Heike Wulff †These authors have contributed equally to this work Specialty section: This article was submitted to Pharmacology of Ion Channels and Channelopathies, a section of the journal Frontiers in Pharmacology Received: 30 March 2019 Accepted: 29 July 2019 Published: 20 September 2019 Specialty section: This article was submitted to Pharmacology of Ion Channels and Channelopathies, a section of the journal Frontiers in Pharmacology Received: 30 March 2019 Accepted: 29 July 2019 Published: 20 September 2019 Powered by the California Digital Library University of California eScholarship.org eScholarship.org ORIGINAL RESEARCH doi: 10.3389/fphar.2019.00972 published: 20 September 2019 Citation: Specifically, activation of KCa2 channels has been proposed for the treatment of diseases characterized by increased neuronal excitability, like ataxia (Shakkottai et al., 2011; Kasumu et al., 2012) and epilepsy, whereas activation of KCa3.1 has been proposed as a treatment for hypertension (Wulff and Kohler, 2013) and as a possible way to pharmacologically enhance anti-tumor T cell responses (Chandy and Norton, 2016). The idea behind the later hypothesis is that by enhancing K+ efflux and thus lowering intracellular K+ it might be possible to reset the “ionic checkpoint” and boost anti-tumor T cell functions in tumor infiltrating T cells, which have been shown to have increased intracellular K+ concentrations suppressing their ability to activate (Eil et al., 2016). In support of this exciting therapeutic postulate, it has recently been demonstrated that pharmacological KCa3.1 activation can restore the ability of cancer patient derived CD8+ T cells to chemotax (Chimote et al., 2018). we determined that S372 in KCa3.1 (or the corresponding S632 in KCa2.3) is crucial for the activity of all 2-aminobenzothiazole/ oxazole-type KCa activators, which are further stabilized by an extensive hydrogen bond network including E295, N300, R362 in KCa3.1 and M51 and E54 in calmodulin. Based on our findings we suggested that R362, a residue which is at the center of this network in the KCa3.1 but not the KCa2.3 models, is responsible for the 5–10-fold selectivity of these compounds for KCa3.1 over KCa2.3 (Brown et al., 2017). In the current work we set out to use our atomistic level understanding of KCa activator binding to develop a third generation of KCa channel activators; this time with selectivity for KCa2.2 over KCa3.1. In this structure-based drug design approach we again used Rosetta ligand docking and carefully compared the binding poses of 2-aminobenzothiazole/oxazole-type KCa activators in the KCa2.2 and KCa3.1 CaM-BD/CaM interface pocket. Based on differences between residues in the KCa2.2 and KCa.3.1 models we manually designed 168 new virtual SKA compounds and tried to predict whether the compounds would show selectivity for KCa2.2 based on the computational docking models. The most promising compounds were synthesized, and their potency and selectivity for KCa2.2, and KCa3.1 tested by manual or automated electrophysiology. However, despite all these efforts, we failed to identify any KCa2.2 selective compounds and recently learned from the full-length cryo-EM structure of KCa3.1 published by the MacKinnon group (Lee and MacKinnon, 2018), that the C-terminal crystal was an artefact. Citation: Small- and intermediate-conductance calcium-activated potassium channels (KCa) are voltage independent and are gated by the binding of calcium to calmodulin, which functions as their calcium- sensing beta subunit (Xia et al., 1998; Fanger et al., 1999). There are four members in the small- and intermediate-conductance KCa channel subfamily, the small-conductance KCa2.1, KCa2.2, and KCa2.3, collectively known as SK channels, and the intermediate-conductance KCa3.1, also known as IK (Kaczmarek et al., 2017). These tetrameric membrane proteins consist of four six-transmembrane Shim H, Brown BM, Singh L, Singh V, Fettinger JC, Yarov-Yarovoy V and Wulff H (2019) The Trials and Tribulations of Structure Assisted Design of KCa Channel Activators. Front. Pharmacol. 10:972. doi: 10.3389/fphar.2019.00972 Small- and intermediate-conductance calcium-activated potassium channels (KCa) are voltage independent and are gated by the binding of calcium to calmodulin, which functions as their calcium- sensing beta subunit (Xia et al., 1998; Fanger et al., 1999). There are four members in the small- and intermediate-conductance KCa channel subfamily, the small-conductance KCa2.1, KCa2.2, and KCa2.3, collectively known as SK channels, and the intermediate-conductance KCa3.1, also known as IK (Kaczmarek et al., 2017). These tetrameric membrane proteins consist of four six-transmembrane September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org KCa Channel Activator Design Shim et al. domains with their N- and C-termini positioned intracellularly. Calmodulin (CaM) is constitutively bound with its C-lobe to the calmodulin binding domain (CaM-BD) in the C-terminus of each KCa channel subunit and opens the channel when the N-lobe binds calcium following increases in intracellular calcium in the proximity of the channel (Xia et al., 1998; Lee and MacKinnon, 2018). KCa2/3 channels are differentially expressed in the human body with KCa2 channels primarily expressed in, but not limited to, the CNS and KCa3.1 primarily found in peripheral tissues, lymphocytes and red blood cells (Adelman et al., 2012; Wulff and Kohler, 2013). KCa2 channels mediate after hyperpolarization and regulate firing frequency in neurons (Adelman et al., 2012). KCa3.1, on the other hand, is responsible for generating the driving force for calcium influx in immune cells and contributes to the regulation of vascular tone via the vascular endothelium (Wulff and Kohler, 2013). These expression patterns make both KCa2 and KCa3.1 channels attractive pharmacological targets (Wulff and Zhorov, 2008). Citation: Based on the full-length structure it was suggested that the “real” binding pocket for the KCa activators is located between the S4-S5 linker and the CaM N-lobe (Lee and MacKinnon, 2018). We here confirmed this structural hypothesis through mutagenesis and now offer a new, corrected binding site model for the SKA- type KCa channel activators. We hope that our findings provide a cautionary tale for the field in warning against some of the pitfalls in structure-based drug design. ( ) Identifying potent and selective KCa channel activators has been challenging. The first generation, which includes 1-EBIO, NS309 (Strobaek et al., 2004), and SKA-31 (Sankaranarayanan et al., 2009), are relatively unselective and only display a 5–10-fold selectivity for KCa3.1 over KCa2 channels (Christophersen and Wulff, 2015). This lack of selectivity has led to CNS related side effects when trying to use KCa3.1 activation as a new, endothelial targeted antihypertensive approach (Radtke et al., 2013). The second generation of KCa channel activators, as exemplified by SKA-121 and SKA-111, are 40–100-fold selective for KCa3.1 and were efficacious in lowering blood pressure in normotensive and hypertensive mice while avoiding KCa2 channel mediated side effects on the CNS and on heart rate because of the improvement in selectivity (Coleman et al., 2014). For the development of our first and second generation KCa activators (the “SKA” compounds), we used a classical medicinal chemistry approach with no structural input during the structure activity optimization (Sankaranarayanan et al., 2009; Coleman et al., 2014). However, more recently, following the publication (Zhang et al., 2013) of the crystal structure of the KCa2.2 CaM-BD in complex with CaM and containing NS309 (pdb: 4J9Z), we generated homology models of the KCa3.1 and KCa2.3 CaM-BD in complex with CaM using the Rosetta molecular modeling suite and RosettaLigand for compound docking (Brown et al., 2017). Combining structural modeling and site-directed mutagenesis Frontiers in Pharmacology \| www.frontiersin.org Molecular Modeling g We previously described the generation of models (Brown et al., 2017) of the KCa3.1 and KCa2.3 C-terminal CaM-binding domain in complex with CaM with Rosetta computational modeling software (Rohl et al., 2004; Bender et al., 2016; Alford et al., 2017) using the x-ray structure of the KCa2.2 channel CaM-binding domain in complex with CaM and NS309 (Zhang et al., 2013) (pdb id: 4J9Z) as a template. We also previously provided a detailed description (Brown et al., 2017) of the procedure for ligand docking using the RosettaLigand docking application (Meiler and Baker, 2006; Davis and Baker, 2009; Bender et al., 2016) in the Rosetta program suite, version 3.7. Briefly, for this study, ligand conformers of variously substituted benzothiazoles/oxazoles were generated using Open Eye OMEGA software version 2.5.1.4 (OpenEye Scientific Software, Santa Fe, NM; http://www.eyesopen.com) (Hawkins et al., 2010; Hawkins and Nicholls, 2012; OEChem v OpenEye Scientific Software, Inc.), randomly placed within the binding pocket and then taken through the three stages of the RosettaLigand modeling which progresses Frontiers in Pharmacology \| www.frontiersin.org September 2019 \| Volume 10 \| Article 972 2 KCa Channel Activator Design Shim et al. from low-resolution conformational sampling and scoring to full atom optimization using Rosetta’s all-atom energy function. A total of 10,000 models were generated for each virtual compound, the top 1,000 models with the lowest total energy score were selected, and the top 10 models with the lowest binding energy were identified, manually inspected for ligand/channel interactions and convergence between the KCa2.2 and the KCa3.1 model compared. A model was considered converged if the top 10 models overlaid. 6 transfection reagent (Promega, Madison, WI) in OptiMEM reduced-serum medium (Life Technologies, Benicia, CA) for manual patch-clamp experiments or via electroporation (Nucleofector 2b and Lonza Amaxa Cell Line Nucleofector Kit T) according to the manufacturer’s instructions for automated electrophysiology. p y gy Cells transfected using FuGENE 6 were cultured in six-well plates for 24–48 h and then detached by TrypLE Express (Gibco, Grand Island, NY) and plated on coverslips for 30 min to 1 h for whole-cell recordings. For inside-out recordings, cells were plated 2–3 h before the experiments to attach them more firmly. Coverslips were placed in a 15 ml recording chamber mounted on an inverted microscope (Olympus XI-70 equipped with a pE-300Lite LED UV light source and filters; Olympus, Tokyo, Japan), and only clearly green fluorescent cells were patch-clamped. Molecular Modeling CHO cells transfected via electroporation were cultured in T75 flasks until 70% confluency and lifted with TrypLE, spun down, and resuspended. 1X106 cells were used per transfection with 1 µg of plasmid DNA. The setting for the Nucleofector 2b was, cell type: CHO, high efficiency. Cells were afterwards cultured in a T25 flask for 24 h before they were lifted for electrophysiology. Cryo-EM structures of the full length KCa3.1 channel (Lee and MacKinnon, 2018) in two open states and one closed state (pdb id: 6CNN, 6CNO, 6CNM) were refined using the Rosetta cryo-EM refinement protocol (Wang et al., 2016) (Cryo-EM density map, Rosetta version 3.8) and the models with the lowest energy scores were chosen for docking of SKA-111. All molecular graphics of ligand, KCa channel C-terminal CaM-binding domain in complex with CaM, or full length KCa3.1 channels were rendered using the UCSF Chimera software (Resource for Biocomputing, Visualization, and Informatics, San Francisco, CA). Potential druggable sites in the Rosetta refined KCa3.1 open state-1 structure were identified using the SiteMap function in Glide (Schrödinger, LLC, New York, NY, 2018). SiteMap uses a site localization method based on interaction energies between the protein and grid probes, which is analogous to the Goodford’s GRID algorithm. Sites were kept if they comprised at least 50 site points. A restrictive hydrophobicity definition and a standard grid (1.0 Å) were used for identifying potential binding pockets. Protein Data Bank (pdb) format files of the Rosetta models of KCa3.1 open state 1 and open state 2 with SKA-111 docked in the interface between S45A helix and the CaM N-lobe are provided in the Data Supplement1; pdb files of all other models are available upon request. Potential druggable sites in the Rosetta refined KCa3.1 open state-1 structure were identified using the SiteMap function in Glide (Schrödinger, LLC, New York, NY, 2018). SiteMap uses a site localization method based on interaction energies between the protein and grid probes, which is analogous to the Goodford’s GRID algorithm. Sites were kept if they comprised at least 50 site points. A restrictive hydrophobicity definition and a standard grid (1.0 Å) were used for identifying potential binding pockets.i yt p y gy For manual whole-cell experiments, the extracellular solution contained 160 mM NaCl, 4.5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES (pH 7.4, 300 mOsm). 1 https://frontiersin.figshare.com/collections/The_Trials_and_Tribulations_of_ Structure_Assisted_Design_of_KCa_channel_Activators/4641455 Molecular Modeling Solutions on the intracellular side contained 154 mM KCl, 10 mM HEPES, 10 mM EGTA, 1.75 mM MgCl2, and 5.9 mM CaCl2 for a calculated free Ca2+ concentration of 250 nM (pH 7.2, 290 mOsm). For inside- out experiments in symmetrical K+, the extracellular solution contained 154 mM KCl, 10 mM HEPES, 1 mM MgCl2, and 2 mM CaCl2 (pH 7.4, 300 mOsm). Intracellular solutions contained 154 mM KCl, 10 mM HEPES, 10 mM EGTA, 1.75 mM MgCl2, and varying amounts of CaCl2 for calculated free Ca2+ concentrations of 0.05, 0.1, 0.25, 0.5, 1, 10, and 30 mM (pH 7.2, 280–300 mOsm). Free Ca2+ concentrations were calculated using the July 3, 2009 online version of MaxChelator (https://somapp.ucdmc.ucdavis. edu/pharmacology/bers/maxchelator/webmaxc/webmaxcS.htm) assuming a temperature of 25°C, a pH of 7.2, and an ionic strength of 160 mM. Patch pipettes were pulled from soda lime glass (micro- hematocrit tubes; Kimble Chase, Rochester, NY) and had resistances of 1.5–3 MΩ when submerged. Experiments were controlled with a HEKA EPC-10 amplifier and Patchmaster software (HEKA, Lambrecht/Pfalz, Germany). In whole-cell experiments, cells were clamped to a holding potential of -80 mV, and KCa currents were elicited by 200-ms voltage ramps from -120 to +40 mV applied every 10 s. For the inside-out experiments, cells were clamped to a holding potential of -80 mV, macro-patches were pulled and KCa currents were elicited by 200-ms voltage ramps from -80 to +80 mV applied every 5 s. Solutions with different free Ca2+ concentrations were then perfused in rapid successions to avoid run down. Each Ca2+ concentration was normalized to 10 µM of free Ca2+ by perfusing 10 µM Ca2+ before and after each Ca2+ concentration. If the two 10 µM values differed more than 10% from each other, the data point was excluded from the Ca2+ concentration response curve. Protein Data Bank (pdb) format files of the Rosetta models of KCa3.1 open state 1 and open state 2 with SKA-111 docked in the interface between S45A helix and the CaM N-lobe are provided in the Data Supplement1; pdb files of all other models are available upon request. Molecular Biologyh The cloning of human KCa3.1 (#AF033021) has been reported in the late 1990s (Logsdon et al., 1997). The gene was subcloned in-frame downstream to green fluorescent protein in the pEGFP-C1 expression vector (CLONTECH) (Wulff et al., 2001). All clones were verified by sequencing. Mutations were introduced using QuikChange site directed mutagenesis kit (Stratagene, La Jolla, CA) and were verified by fluorescence sequencing. For amino acid numbering for KCa3.1 we used the gene: Homo sapiens potassium channel, calcium activated intermediate/small conductance subfamily N alpha, member 4 (KCNN4). NCBI Reference Sequence: NP_002241.1. September 2019 \| Volume 10 \| Article 972 Electrophysiology All experiments were performed in either the inside-out or the whole- cell configuration of the patch-clamp technique on either transiently transfected CHO cells or Human Embryonic Kidney (HEK) cell lines stably expressing hKCa3.1 or hKCa2.2 (Sankaranarayanan et al., 2009). All cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. Wild-type (WT) and mutant hKCa3.1 channel constructs were transfected using FuGENE The procedure for performing automated whole-cell KCa channel recordings on a QPatch-16 automated electrophysiology platform (Sophion Biosciences) was previously described in detail by our laboratory (Jenkins et al., 2013). For the current study we used the solutions described above with disposable 16-channel September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 3 KCa Channel Activator Design Shim et al. planar patch chip plates (QPlates; patch hole diameter 1 mm, resistance 2.00–0.02MΩ, Sophion Biosciences, Woburn, MA). Cell positioning and sealing parameters were set as follows: positioning pressure -70 mbar, resistance increase for success 750%, minimum seal resistance 0.1 GΩ, holding potential -0 mV, holding pressure -20 mbar. To avoid rejection of cells with large KCa3.1 currents, the minimum seal resistance for whole-cell requirement was lowered to 0.001 GΩ. Access was obtained with the following sequence: 1) suction pulses in 29 mbar increments from -250 mbar to -453 mbar; 2) a suction ramp of an amplitude of -450 mbar; 3) -400 mV voltage zaps of 1 ms duration. Following establishment of the whole-cell configuration, cells were held at -80 mV and KCa currents elicited by a voltage protocol that held at -80 mV for 20 ms, stepped to -120 mV for 20 ms, ramped from -120 to + 40 mV in 200 ms, and then stepped back to -120 mV for 20 ms. This pulse protocol was applied every 10 s. KCa activator dilutions were prepared freshly (within 5 min of the initiation of the QPatch) with Ringer’s solution from 10 mM stock solutions in dry DMSO. Final DMSO concentrations never exceeded 1%. Glass vial inserts (to avoid adsorption) were filled with 400 µL of compound solution and placed into the insert base plate for use in the QPatch assay. Exemplary QPatch raw current traces and plots of slope conductance versus time are shown in Supplementary Figure 3. Calcium Sensitivity Testingi When screening mutants, we always first assessed whether mutations had altered Ca2+ sensitivity in a two-step process. First, mutant channels were patched with an intracellular solution containing 10 µM of free Ca2+. If the mutant did not display currents with amplitudes in the nA range under these conditions, we then tested whether the channel’s control current (at 250 nM free Ca2+) could be increased 5–10-fold in the presence of 100 µM EBIO (this fold increase level is common in the wild type channel). If the mutant did not meet either threshold it was deemed to have altered Ca2+ sensitivity and was excluded from subsequent experiments investigating sensitivity to SKA-111. All mutants shown in Figure 7 or mentioned in the text exhibited current densities at +40 mV that were comparable to the WT channel and produced currents that were large enough to perform experiments. WT: 19.5 ± 23.4 pA/ pF; T212F-V272F: 12.4 ± 10.4 pA/pF; S181A: 11.6 ± 16.2 pA/pF; A184F: 7.2 ± 5.4 pA/pF; L185A: 6.9 ± 6.5 pA/pF; S181A-L185A: 8.8 ± 5.5 pA/pF; S372R: 13.8 ± 12.5 pA/pF (n = 5). General Method I. Preparation of nitro-substituted 2-metylnaphtho[1,2-d]thiazoles or 1-substituted 4-nitronaphthalenes Data analysis, fitting, and plotting were performed with IGOR-Pro (Wavemetrics, Lake Oswego, OR) or Origin 9 (OriginLab Corporation, Northampton, MA). One milliliter of HNO3 was added slowly to 1 g of 2-methyl-β- naphthothiazole or 1-substituted naphthalene in a round bottom flask at room temperature without solvent. The reaction mixture was stirred for 1 h and then 1.5 ml of H2SO4 (95%) was added slowly. The mixture was stirred at room temperature for 30 min until completion of the reaction was indicated by the disappearance of starting material on thin-layer chromatography (TLC). Then the pH of the reaction mixture was adjusted to between 7 and 7.5 with 4N NaOH, the mixture extracted with ethyl acetate and the organic phase washed with water and brine. The extract was dried with anhydrous sodium sulfate and solvent was evaporated under reduced pressure. The crude product was purified via flash- chromatography (petroleum ether: ethyl acetate, 8:2). Chemical Synthesis y Compounds that were not commercially available were synthesized in our laboratory according to the general methods described below. Compounds reported previously were characterized by melting point, proton nuclear magnetic resonance (1H NMR) and 13 carbon (13C) NMR. All NMR spectra were recorded on an 800 MHz Bruker Avance III spectrometer. Data are reported as follows: chemical shift (δ), multiplicity, integration, coupling constant (Hz). Signals are designated as follows: s (singlet), d (doublet), dd (doublet of doublets), ddd (doublet of doublet of doublets), dt (doublet of triplets), t (triplet), quint (quintet), m (multiplet). New chemical entities were additionally characterized by high-resolution mass spectrometry (HRMS). For compounds (SKA-218, SKA-339, SKA- 340, SKA-343 and SKA-347) where it was difficult to unambiguously confirm the structure based on 1H and 13C NMR, we grew crystals and subjected them to X-ray analysis, which allowed us to see the exact position of the substituents in the compounds. All compounds used for electrophysiological experiments were at least 95–98% pure based on NMR and/or mass spectrometry. General Method II. Preparation of amino-substituted 2-metylnaphtho[1,2-d]thiazoles or 1-amino 4-substituted naphthalenes One gram of the previously prepared nitro-substituted 2-metylnaphtho[1,2-d]thiazole or 1-substituted 4-nitronaphthalene was dissolved in 30 mL of 95% ethanol and 5% water, 20 mg of palladium charcoal was added and then 2 mL of hydrazine were dropped in slowly. After the addition was completed, the reaction mixture was stirred at 80°C for 12 h and the progress of the reaction was monitored by TLC with dichloromethane. After completion, the reaction mixture was filtered to remove the palladium charcoal and concentrated to dryness under vacuum. The residue was dissolved in ethyl acetate and was washed with water and brine. The organic Electrophysiology For each compound we typically used 12 consecutive liquid periods Break-in, 3 saline additions to stabilize the current, 2 additions of test compound at 10 µM, 2 saline washes, 2 additions of SKA-31 at 10 µM, followed by 2 more saline washes. In each liquid period 10 pulse protocols were run. Bis-4,4’-(2-amino-4-thiazolyl)biphenyl (SKA-255, CAS 438233- 93-3) was purchased from Oakwood Products, INC (West Columbia, SC). Identity and purity (<95%) of commercial compounds was confirmed by 1H NMR before using compounds for electrophysiological testing. General Method VIII. Preparation of N-substituted naphtho[1,2-d]thiazol-2-amine p [ ] 1-Naphthylamine 500 mg (3.5 mmol) was dissolved in 25 mL chloroform and 0.3 mL of trimethylamine before adding substituted isothiocyanate (3.8 mmol). The resulting mixture was refluxed at 65°C for 48 h. Solvent was then evaporated, and the resulting solid was washed with diethyl ether to obtain N,N’-disubstituted thiourea. The N,N’-disubstituted thiourea (2.2 mmol) obtained in the previous step was suspended in 25 mL chloroform, and a solution of liquid bromine in chloroform (1 eq.) added over a period of 1 h, after which sodium thiosulfate (aq.) was added to the reaction mixture. The reaction mixture was washed with 4N NaOH (aq.) and the organic layer was dried over anhydrous sodium sulfate and evaporated under reduced pressure. General Method IV. Preparation of substituted 2-bromo- 1-(methyl substituted naphthalen-1-yl)ethan-1-one or 2-bromo-1-(methyl substituted cyclohex-1-yl)ethan-1-one Appropriately substituted acetophenones were dissolved in chloroform (20 ml) at room temperature and the reaction mixture was stirred. Liquid bromine (1.2–1.5 equivalent) in chloroform (5 ml) was then added drop-wise. The progress of the reaction was monitored by TLC. The reaction mixture was quenched with a saturated aqueous solution of sodium hydrogen carbonate and the pH was adjusted to between 7 and 7.5. The organic phase was washed with water and brine and the solvent was removed under vacuum. The crude residue was purified via flash chromatography (petroleum ether: ethyl acetate, 7:3). 2-Methylnaphtho[1,2-d]thiazole-5-carbonitrile (SKA-126). To a solution of 5-bromo-2-methylnaphtho[1,2-d]thiazole (SKA-132, 280 mg, 1 mmol) in a mixture of DMF (3 mL) and water (200 uL), kept under nitrogen, were added copper (I) cyanide (CuCN, 121 mg, 1.33 mmol), Pd2(dba)3 (Tris (dibenzylideneacetone) dipalladium (0), 70 mg) and 1,1′-ferrocenediyl-bis(diphenylphosphine) (dppf, 80 mg). The resulting solution was heated at 110°C for 75 h, after which the mixture was filtered and diluted with 10 mL water and was extracted using chloroform (3X10 mL). The organic layer was separated, dried over anhydrous sodium sulfate and evaporated under reduced pressure. Chromatographic purification of the crude product on silica gel using ethyl acetate and hexane (10:90) furnished 120 mg (50%) of the required product as off white solid and 85 mg starting material: m.p = 139–140°C. 1H NMR (800 MHz, CDCl3): δ = 8.85–8.82 (m, 1H), 8.34–8.31 (m, 2H), 7.80–7.74 (m, 2H), 2.99 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 170.8, 152.7, 130.7, 130.6, 128.5, 128.30, 128.0, 126.5, 125.6, 124.7, 117.9, 107.3, 20.6. General Method V. Preparation of substituted 4-phenylthiazole or 4-naphthalenyl thiazoles General Method V. Preparation of substituted 4-phenylthiazole or 4-naphthalenyl thiazoles 4-phenylthiazole or 4-naphthalenyl thiazoles Substituted 2-bromo-1-(methyl substituted naphthalen-1-yl) ethan-1-one or 2-bromo-1-(methyl substituted cyclohex-1-yl) ethan-1-one was added to a solution of a substituted thiourea in 20 ml of absolute ethanol. The mixture was refluxed for 2 h. After completion of the reaction, the ethanol was evaporated under vacuum. The dried reaction mixture was dissolved in dichloromethane and neutralized with saturated NaHCO3. The organic phase was washed with water and brine, and then dried with anhydrous Na2SO4. Finally, the solvent was evaporated under vacuum. The crude product was reconstituted in a methanol, treated with charcoal and recrystallized. General Method VII. Preparation of substituted 2-aminonaphthothiazoles General Method VII. Preparation of substituted 2-aminonaphthothiazoles phase was dried with anhydrous sodium sulfate and the solvent was removed under vacuum. The crude product was purified by flash- chromatography with ethyl acetate/petroleum ether (3:7 v/v) as eluent to give the product. Liquid bromine (150 μL, 3 mmol) and KSCN (485 mg, 5 mmol) were added to a solution of the 1-amino 4-substituted naphthalenes (3 mmol) in 10 ml of acetic acid and the reaction mixture was stirred at room temperature for 30 min. The reaction was quenched by adding 4N NaOH. The reaction mixture was then washed with water and brine and extracted with ethyl acetate. Solvent was evaporated to give the crude product, which was purified by flash-chromatography using ethyl acetate/ petroleum ether (3:7 v/v). General Method III. Preparation of methyl-substituted acetonaphthones Acetylchloride (750 μl, 10 mmol) and AlCl3 (1.33 g, 10 mmol) were added sequentially to 20 ml of chloroform at room temperature. Methyl substituted naphthalenes (1 ml, 7 mmol) in 5 ml of chloroform were then added, the reaction mixture was stirred for 2 h, and the progress of the reaction was monitored by TLC. The reaction mixture was quenched with sodium bicarbonate solution and washed with water and brine. Solvent was removed under reduced pressure to give the crude product. The crude product was purified by flash–chromatography using ethyl acetate/petroleum ether (2:8 v/v). General Method VIII. Preparation of N-substituted naphtho[1,2-d]thiazol-2-amine Chemistry Commercially Available Compounds 4-Biphenyl-4-yl-thiazol-2-ylamine (SKA-232, CAS 2834- 79-9) was purchased from Sigma–Aldrich (St. Louis, MO). September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org KCa Channel Activator Design Shim et al. Frontiers in Pharmacology \| www.frontiersin.org General Method VIII. Preparation of N-substituted naphtho[1,2-d]thiazol-2-amine HRMS (ESI): m/z calculated for C13H8N2S (M+H)+: 225.0485 found: 225.0485. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes To a solution of 2-methylnaphtho[1,2-d]thiazole (1.0 g, 5 mmol) in 25 mL chloroform was added a solution of molecular bromine in chloroform (I eq) over a period of 2 min and the resulting solution was stirred for 30 min at room temperature after which sodium thiosulfate (aq.) was added. This was followed by washing with 4N NaOH (aq.). The organic layer was dried over anhydrous sodium sulfate and evaporated under reduced pressure to obtain 1202 mg (85%) The product was isolated as white solid; m.p = 98–99°C (CAS 125427-01-2).1H NMR (800 MHz, CDCl3): δ = 8.78 (dd, J = 8.1, 1.4 Hz, 1H), 8.33–8.30 (m, 1H), 8.16 (s, 1H), 7.70 (ddd, J = 8.1, 6.8, 1.3 Hz, 1H), 7.67 (ddd, J = 8.3, 6.9, 1.4 Hz, 1H), 2.93 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 166.5, 149.4, 132.0, 130.2, 129.0, 127.8, 127.7, 127.2, 124.4, 122.6, 119.6, 20.3. N-Cyclohexylnaphtho[1,2-d]thiazol-2-amine (SKA-165). 1-Naphthyl isothiocyanate (2 g, 10.8 mmol) was added to 20 mL of ethanol and the mixture was stirred and refluxed at 80°C. Then, 1.24 mL (10.8 mmol) of cyclohyxylaminenaphtho[1,2-d]thiazol-2- ylamine was added and the mixture refluxed for 2 h. The progress of the reaction was monitored by TLC. After completion of the reaction ethanol was removed under vacuum, the residue extracted with ethyl acetate and the organic phase washed with water followed by brine. The organic layer was dried with anhydrous Na2SO4 and the solvent was evaporated. 700 mg (2.46 mmol) of the intermediate product, 1-cyclohexyl-3-(naphthalen- 1-yl)thiourea, was dissolved in 20 mL of chloroform. Then 131.1 uL (2.46 mmol) of liquid bromine was added dropwise. After completion the reaction was quenched by adding 4N NaOH, the resulting mixture was washed with water and brine and extracted with ethyl acetate. Finally, the solvent was evaporated under vacuum. The crude product was reconstituted in a methanol, treated with charcoal and recrystallized. The product was isolated as a brownish crystal. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes 2-Aminonaphtho[1,2-d]oxazole-5-sulfonic acid (SKA-128). To a solution of 4-amino-3-hydroxynaphthalene-1-sulfonic acid (1.0 g, 4.18 mmol) and 3.0 mL Hunig’s base in 50 mL (9:1) ethanol:water was added 850 mg of cyanogen bromide. The resulting mix was stirred at room temperature for 4 d. After completion of the reaction the solvent was evaporated, and the residue was dissolved in water. It was acidified by adding 1N HCl. The precipitate thus obtained was filtered, washed with acetone, and dried to render 800 mg (81%) of pink-brown solid: m.p = 295°C (Decomposes).1H NMR (800 MHz, DMSO-d6 & Et3N): δ 1-Fluoro-4-nitronaphthalene and a secondary or heterocyclic amine (molar ratio 1:3) were dissolved in 10 mL of DMF at 90°C. Potassium carbonate (3 equivalents) was then added. The reaction mixture was stirred and monitored by TLC. After completion, the reaction mixture was washed with brine several times to remove the DMF and extracted with ethyl acetate. Solvent was evaporated to give the crude product. The crude product was purified by flash- chromatography using ethyl acetate/petroleum ether (2:8 v/v). September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 5 KCa Channel Activator Design Shim et al. oxazole (SKA-133) and CuCN following the method described for the synthesis of SKA-126. The product was isolated as a brownish crystal (42%); m.p. = 145–146°C (CAS 60111-00-4). 1H NMR (800 MHz, CDCl3): δ = 8.50–8.46 (m, 1H), 8.31 (dt, J = 8.2, 0.9 Hz, 1H), 8.05 (s, 1H), 7.74 (ddd, J = 8.2, 6.9, 1.3 Hz, 1H), 7.71 (ddd, J = 8.3, 6.9, 1.4 Hz, 1H), 2.78 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 166.4, 146.2, 141.3, 130.5, 128.5, 127.8, 126.0, 125.9, 122.8, 117.9, 117.3, 106.4, 14.9. = 8.92 (dt, J = 8.6, 0.9 Hz, 1H), 8.20 – 8.16 (m, 1H), 8.07 (s, 1H), 7.63 (s, 2H), 7.56 (ddd, J = 8.1, 6.7, 1.2 Hz, 1H), 7.50 (ddd, J = 8.3, 6.7, 1.5 Hz, 1H). 13C NMR (201 MHz, DMSO-d6 & Et3N): δ = 163.6, 141.8, 139.3, 136.9, 128.3, 126.8, 125.3, 124.2, 124.1, 121.5, 108.9. HRMS (ESI): m/z calculated for C11H8N2O4S (M+H)+: 265.0278 found: 265.0284. 5-Chloro-2-methylnaphtho[1,2-d]thiazole (SKA-130). 2- Methyl-β-naphthothiazole (400 mg, 2 mmol), N-chlorosuccinamide (266 mg, 2 mmol), NaCl (117 mg, 2 mmol) and P-TSOH (344 mg, 2 mmol) were dissolved in 30 ml of acetonitrile. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes The reaction mixture was stirred at 40°C for 2 h, and the progress of the reaction was monitored by TLC using an ethyl acetate:petroleum ether mixture (3:7). After completion, the reaction mixture was filtered and washed with water and acetonitrile. The residue was recrystallized from methanol as light yellow crystals (100 mg, 21%); m.p. = 103°C. 1H NMR (800 MHz, CDCl3): δ = 8.81 (dd, J = 8.2, 1.3 Hz, 1H), 8.37 (d, J = 8.3 Hz, 1H), 7.99 (s, 1H), 7.74 (ddd, J = 8.2, 6.8, 1.3 Hz, 1H), 7.70 (ddd, J = 8.3, 6.8, 1.4 Hz, 1H), 2.96 (s, 3H). 13C NMR (201 MHz, CDCl3),: δ = 166.3, 148.6, 131.3, 129.0, 129.0, 128.7, 127.6, 126.8, 124.9, 124.2, 118.9, 20.1. HRMS (ESI): m/z calculated for C12H8ClNS (M+H)+: 243.0138; found: 243.0135. N-Phenylnaphtho[1,2-d]thiazol-2-amine (SKA-146). SKA-146 was prepared from 1-naphthylamine and phenylisothiocyanate according to General Method VIII. The product was isolated as a white solid (320 mg, 42%); m.p = 140–141°C (CAS 21431-44- 7). 1H NMR (800 MHz, DMSO-d6): δ = 10.67 (s, 1H), 8.58 (dt, J = 8.1, 1.0 Hz, 1H), 8.02 (dd, J = 8.1, 1.1 Hz, 1H), 7.99 – 7.94 (m, 3H), 7.75 (d, J = 8.5 Hz, 1H), 7.67 (ddd, J = 8.1, 6.8, 1.2 Hz, 1H), 7.58 (ddd, J = 8.1, 6.8, 1.3 Hz, 1H), 7.49 – 7.44 (m, 2H), 7.09 (tt, J = 7.3, 1.1 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 162.3, 147.3, 140.8, 131.7, 129.1, 128.0, 126.2, 126.1, 125.4, 124.7, 123.4, 122.2, 121.9, 119.1, 117.5. N-Methylnaphtho[1,2-d]thiazol-2-amine (SKA-158). SKA-158 was prepared from1-naphthylamine and methylisothiocyanate according to General Method VIII. The product was isolated as a white solid (42%); m.p = 187–188°C (CAS 876484-23-0).1H NMR (800 MHz, DMSO-d6): δ = 8.45 (dd, J = 8.2, 1.2 Hz, 1H), 8.13 (d, J = 4.8 Hz, 1H), 7.97 – 7.84 (m, 2H), 7.60 (d, J = 8.5 Hz, 1H), 7.57 (ddd, J = 8.2, 6.7, 1.3 Hz, 1H), 7.51 (ddd, J = 8.1, 6.7, 1.3 Hz, 1H), 3.07 (d, J = 4.5 Hz, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 168.1, 147.9, 131.6, 127.9, 125.8, 125.5, 125.1, 124.3, 123.5, 120.5, 119.3, 30.7. 5-Bromo-2-methylnaphtho[1,2-d]thiazole (SKA-132). (SKA-218). 5,6-Dinitronaphtho[1,2-d]thiazol-2-amine (SKA-218). SKA-218 was prepared from 1 g of 2-methyl-β-naphthothiazole according to General Method I. The product was isolated as light brownish crystals (210 mg, 13.3%): m.p. = 172°C. 1H NMR (800 MHz, DMSO-d6): δ = 9.31 (s, 1H), 9.15 (dd, J = 8.2, 1.2 Hz, 1H), 8.50 (dd, J = 7.6, 1.2 Hz, 1H), 8.04 (dd, J = 8.2, 7.6 Hz, 1H), 3.02 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 175.7, 152.1, 145.9, 141.5, 131.8, 130.4, 128.8, 128.6, 126.9, 123.8, 114.7, 20.8. HRMS (ESI): m/z calculated for C12H7N3O4S (M+H)+: 290.02; found: 290.02. 4-(7-Methylnaphthalen-1-yl)thiazol-2-amine (SKA-193). SKA-193 was prepared from 2-methylnaphthalene in 3 steps according to General methods III, IV, and V. The product was isolated as light yellow powder (70 mg, 4%); m.p. = 154–156°C. 1H NMR (800 MHz, CDCl3): δ = 8.06 (d, J = 1.8 Hz, 1H), 7.82 (d, J = 8.2 Hz, 1H), 7.79 (d, J = 8.3 Hz, 1H), 7.60 (dd, J = 7.0, 1.2 Hz, 1H), 7.43 (dd, J = 8.1, 7.0 Hz, 1H), 7.35 (dd, J = 8.3, 1.6 Hz, 1H), 6.64 (s, 1H), 5.41 (d, J = 150.2 Hz, 2H), 2.52 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 168.3, 150.6, 135.4, 133.3, 132.2, 131.4, 128.4, 128.3, 128.1, 127.2, 125.4, 124.9, 105.2, 22.2. HRMS (ESI): m/z calculated for C14H12N2S (M+H)+: 241.0794; found: 241.0793. 5-Bromo-4-(naphthalen-2-yl)thiazol-2-amine (SKA-220). 2-Amino-4-(2-naphthy)thiazole (106 mg, 0.46 mmol) was dissolved in 10 ml of chloroform. Liquid bromine (30 μl, 0.46 mmol) in chloroform (3 ml) was then added drop-wise. The reaction mixture was stirred at room temperature for 30 min. The progress of the reaction was monitored by TLC. The reaction mixture was quenched with 4N NaOH and washed with sodium thiosulfate, water and brine. The organic phase was dried with anhydrous sodium sulfate and the solvent was evaporated under vacuum. The crude product was recrystallized from methanol. The product was isolated as light brown crystals (100 mg, 40%); m.p. = 140–141°C (CAS 99514-91-7). 1H NMR (800 MHz, DMSO-d6): δ = 8.35 (s, 1H), 7.96 (s, 2H), 7.99–7.91 (m, 2H), 7.56–7.52 (m, 2H), 7.38 (s, 2H). 13C NMR (201 MHz, DMSO-d6): δ = 167.4, 147.5, 133.0, 132.7, 131.7, 128.6, 128.0, 127.9, 127.5, 126.96, 126.9, 126.1, 88.0. 4-(6-Methylnaphthalen-2-yl)thiazol-2-amine (SKA-198). SKA-198 was prepared from 2-methylnaphthalene in 3 steps according to General methods III, IV, and V. The product was isolated as light yellow powder (110 mg, 7%); m.p. = 178–179°C. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes 13C NMR (201 MHz, DMSO-d6): δ = 167.3, 146.7, 130.6, 126.5, 126.0, 125.3, 125.0, 124.4, 124.0, 123.98, 119.3, 30.7, 19.0. 2-Methyl-6-nitronaphtho[1,2-d]thiazole (SKA-215). SKA-215 was prepared from 1 g of 2-methyl-β-naphthothiazole according to General Method I. The product was isolated as light-yellow powder (200 mg, 6%); m.p. = 178–179°C. 1H NMR (800 MHz, DMSO-d6): δ = 9.15 (dt, J = 8.2, 1.1 Hz, 1H), 8.49 (dd, J = 9.2, 0.8 Hz, 1H), 8.29 (dd, J = 7.6, 1.3 Hz, 1H), 8.09 (d, J = 9.2 Hz, 1H), 7.74 (t, J = 7.9 Hz, 1H), 2.99 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 167.6, 149.3, 147.0, 133.1, 130.2, 129.3, 125.2, 123.5, 123.5, 122.4, 119.6, 20.2. HRMS (ESI): m/z calculated for C12H8N2O2S (M+H)+: 245.0379; found: 245.0376. methylisothiocyanate according to General Method VIII. The product was isolated as a white solid (35%). m.p = 172–173°C (CAS 1369251-74-0). 1H NMR (800 MHz, DMSO-d6): δ = 8.50– 8.48 (m, 1H), 8.04–8.00 (m, 2H), 7.70 (d, J = 1.2 Hz, 1H), 7.58 (dddd, J = 20.1, 8.2, 6.7, 1.4 Hz, 2H), 3.05 (d, J = 4.7 Hz, 3H), 2.67 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 167.3, 146.7, 130.6, 126.5, 126.0, 125.3, 125.0, 124.4, 124.0, 123.98, 119.3, 30.7, 19.0. 2-Methyl-6-nitronaphtho[1,2-d]thiazole (SKA-215). SKA-215 was prepared from 1 g of 2-methyl-β-naphthothiazole according to General Method I. The product was isolated as light-yellow powder (200 mg, 6%); m.p. = 178–179°C. 1H NMR (800 MHz, DMSO-d6): δ = 9.15 (dt, J = 8.2, 1.1 Hz, 1H), 8.49 (dd, J = 9.2, 0.8 Hz, 1H), 8.29 (dd, J = 7.6, 1.3 Hz, 1H), 8.09 (d, J = 9.2 Hz, 1H), 7.74 (t, J = 7.9 Hz, 1H), 2.99 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 167.6, 149.3, 147.0, 133.1, 130.2, 129.3, 125.2, 123.5, 123.5, 122.4, 119.6, 20.2. HRMS (ESI): m/z calculated for C12H8N2O2S (M+H)+: 245.0379; found: 245.0376. 2-Methylnaphtho[1,2-d]thiazol-6-amine (SKA-172). SKA- 172 was prepared from 2-methyl-6-nitronaphtho[1,2-d]thiazole (SKA-215) in 2 steps according to General methods I and II. The resulting residue was purified by flash-chromatography with ethyl acetate/petroleum ether (3:7 v/v) as eluent to give SKA-172 as reddish crystals (40 mg, 19%); m.p. = 148–150°C. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes 1H NMR (800 MHz, CDCl3): δ = 8.08 (dd, J = 8.9, 0.8 Hz, 1H), 7.94 (d, J = 8.9 Hz, 1H), 7.87 (dt, J = 8.1, 1.0 Hz, 1H), 7.38 (t, J = 7.8 Hz, 1H), 6.81 (dd, J = 7.6, 1.1 Hz, 1H), 5.89 (s, 2H), 2.91 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 165.7, 149.4, 145.8, 131.7, 129.4, 128.2, 120.8, 120.3, 117.1, 111.1, 108.7, 20.2. HRMS (ESI): m/z calculated for C12H10N2S (M+H)+: 215.0637; found: 215.0635. methylisothiocyanate according to General Method VIII. The product was isolated as a white solid (35%). m.p = 172–173°C (CAS 1369251-74-0). 1H NMR (800 MHz, DMSO-d6): δ = 8.50– 8.48 (m, 1H), 8.04–8.00 (m, 2H), 7.70 (d, J = 1.2 Hz, 1H), 7.58 (dddd, J = 20.1, 8.2, 6.7, 1.4 Hz, 2H), 3.05 (d, J = 4.7 Hz, 3H), 2.67 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 167.3, 146.7, 130.6, 126.5, 126.0, 125.3, 125.0, 124.4, 124.0, 123.98, 119.3, 30.7, 19.0. 2-Methyl-6-nitronaphtho[1,2-d]thiazole (SKA-215). SKA-215 was prepared from 1 g of 2-methyl-β-naphthothiazole according to General Method I. The product was isolated as light-yellow powder (200 mg, 6%); m.p. = 178–179°C. 1H NMR (800 MHz, DMSO-d6): δ = 9.15 (dt, J = 8.2, 1.1 Hz, 1H), 8.49 (dd, J = 9.2, 0.8 Hz, 1H), 8.29 (dd, J = 7.6, 1.3 Hz, 1H), 8.09 (d, J = 9.2 Hz, 1H), 7.74 (t, J = 7.9 Hz, 1H), 2.99 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 167.6, 149.3, 147.0, 133.1, 130.2, 129.3, 125.2, 123.5, 123.5, 122.4, 119.6, 20.2. HRMS (ESI): m/z calculated for C12H8N2O2S (M+H)+: 245.0379; found: 245.0376. 2-Methylnaphtho[1,2-d]thiazol-9-amine (SKA-178). SKA- 178 was prepared from 2-methyl-9-nitronaphtho[1,2-d]thiazole (SKA-216) in 2 steps according to General methods I and II as green crystals (39 mg, 17%); m.p = 87-88°C. 1H NMR (800 MHz, CDCl3): δ = 8.01 (d, J = 8.7 Hz, 1H), 7.76 (d, J = 8.7 Hz, 1H), 7.34 (t, J = 7.7 Hz, 1H), 7.22–7.19 (m, 1H), 7.11 (s, 2H), 6.86 (dd, J = 7.6, 1.1 Hz, 1H), 2.97 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 164.9, 148.6, 145.7, 134.1, 130.7, 127.4, 126.4, 119.5, 115.4, 115.1, 109.5, 20.2. HRMS (ESI): m/z calculated for C12H10N2S (M+H)+: 215.0637; found: 215.0635. 2-Methyl-9-nitronaphtho[1,2-d]thiazole (SKA-216). SKA-216 was prepared from 1 g of 2-methyl-β-naphthothiazole according to General Method I. The product was isolated as light-yellow crystals (280 mg, 22%); m.p. = 151–152°C. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes 1H NMR (800 MHz, DMSO-d6): δ = 8.10 (d, J = 8.0 Hz, 1H), 8.01 (d, J = 8.7 Hz, 1H), 7.85 (d, J = 8.7 Hz, 1H), 7.72 (d, J = 7.4 Hz, 1H), 7.60 (t, J = 7.8 Hz, 1H), 2.92 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 167.6, 146.9, 144.3, 135.4, 133.0, 131.9, 125.8, 125.6, 122.2, 121.6, 117.2, 20.4. HRMS (ESI): m/z calculated for C12H8N2O2S (M+H)+: 245.0379; found: 245.0376. 2-Methyl-9-nitronaphtho[1,2-d]thiazole (SKA-216). SKA-216 was prepared from 1 g of 2-methyl-β-naphthothiazole according to General Method I. The product was isolated as light-yellow crystals (280 mg, 22%); m.p. = 151–152°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.10 (d, J = 8.0 Hz, 1H), 8.01 (d, J = 8.7 Hz, 1H), 7.85 (d, J = 8.7 Hz, 1H), 7.72 (d, J = 7.4 Hz, 1H), 7.60 (t, J = 7.8 Hz, 1H), 2.92 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 167.6, 146.9, 144.3, 135.4, 133.0, 131.9, 125.8, 125.6, 122.2, 121.6, 117.2, 20.4. HRMS (ESI): m/z calculated for C12H8N2O2S (M+H)+: 245.0379; found: 245.0376. 4-(4-Methylnaphthalen-1-yl)thiazol-2-amine (SKA-190). SKA-190 was prepared from 1-methylnaphthalene in 3 steps according to General methods III, IV, and V. The product was obtained as golden crystals (355 mg, 21%); m.p = 168–169°C (CAS 332064-25-2). 1H NMR (800 MHz, DMSO-d6): δ = 8.49 (dd, J = 8.4, 1.3 Hz, 1H), 8.09–8.06 (m, 1H), 7.60 (ddd, J = 8.3, 6.7, 1.4 Hz, 1H), 7.57–7.53 (m, 2H), 7.40 (dd, J = 7.2, 1.1 Hz, 1H), 7.12 (s, 2H), 6.74 (s, 1H), 2.70 (d, J = 1.0 Hz, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 168.6, 150.3, 135.5, 132.9, 131.9, 131.8, 128.9, 128.4, 127.6, 126.8, 124.5, 124.3, 102.3, 21.7. 4-(4-Methylnaphthalen-1-yl)thiazol-2-amine (SKA-190). SKA-190 was prepared from 1-methylnaphthalene in 3 steps according to General methods III, IV, and V. The product was obtained as golden crystals (355 mg, 21%); m.p = 168–169°C (CAS 332064-25-2). 1H NMR (800 MHz, DMSO-d6): δ = 8.49 (dd, J = 8.4, 1.3 Hz, 1H), 8.09–8.06 (m, 1H), 7.60 (ddd, J = 8.3, 6.7, 1.4 Hz, 1H), 7.57–7.53 (m, 2H), 7.40 (dd, J = 7.2, 1.1 Hz, 1H), 7.12 (s, 2H), 6.74 (s, 1H), 2.70 (d, J = 1.0 Hz, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 168.6, 150.3, 135.5, 132.9, 131.9, 131.8, 128.9, 128.4, 127.6, 126.8, 124.5, 124.3, 102.3, 21.7. General Method VI. Preparation of heterocyclic 1-substituted 4-nitronaphthalenes (460 mg, 17%); m.p = 115– 116°C (CAS 1368045-48-0).1H NMR (800 MHz, DMSO-d6): δ = 8.38 (d, J = 8.2 Hz, 1H), 8.10 (d, J = 7.4 Hz, 1H), 7.89 (d, J = 8.0 Hz, 1H), 7.79 (d, J = 8.5 Hz, 1H), 7.56 - 7.50 (m, 2H), 7.46 (t, J = 7.4 Hz, 1H), 3.74 (s, 1H), 2.12 - 1.96 (m, 2H), 1.76 (dt, J = 13.1, 4.0 Hz, 2H), 1.61 (dt, J = 13.3, 4.0 Hz, 1H), 1.36 (dddd, J = 37.2, 15.2, 12.0, 5.8 Hz, 4H), 1.22 (dtd, J = 15.5, 11.8, 9.5, 6.0 Hz, 1H).13CNMR(201MHz, DMSO-d6): δ = 166.8, 148.3, 132.1, 128.3, 125.9, 125.52, 124.5, 123.9, 120.9, 119.7, 79.6, 53.7, 32.7, 25.7, 24.9. 5-Bromo-2-methylnaphtho[1,2-d]oxazole (SKA-133). SKA-133 was prepared from 2-methylnaphtho[1,2-d]oxazole (1.0 g, 5.45 mmol) following the method described above for the synthesis of SKA-132. The product was isolated as a white solid (90%); m.p = 101–102°C (CAS 1838658-03-9). 1H NMR (800 MHz, CDCl3): δ = 8.45 (dd, J = 8.2, 1.3 Hz, 1H), 8.33 (dd, J = 8.5, 1.1 Hz, 1H), 7.99 (s, 1H), 7.68 (ddd, J = 8.1, 6.9, 1.1 Hz, 1H), 7.63 (ddd, J = 8.3, 6.9, 1.3 Hz, 1H), 2.73 (s, 3H). 13C NMR (201 MHz, CDCl3): δ = 163.4, 147.6, 136.7, 129.3, 128.1, 127.8, 126.7, 126.6, 122.5, 118.9, 115.1, 14.8. Naphtho[2,3-d]oxazol-2-amine (SKA-134). SKA-134 was prepared from 3-aminonaphthalen-2-ol and cyanogen bromide following the method described for the synthesis of SKA-128. The product was isolated as a white solid (97%); m.p = 247– 248°C (CAS 1820618-76-5). 1H NMR (800 MHz, DMSO-d6): δ = 7.93–7.87 (m, 2H), 7.80 (d, J = 0.8 Hz, 2H), 7.76 (s, 3H), 7.61 (d, J = 0.9 Hz, 2H), 7.39 (dddd, J = 26.3, 8.3, 6.8, 1.5 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 164.3, 148.5, 144.3, 131.6, 129.1, 127.9, 127.5, 124.5, 123.7, 110.8, 104.3. 2-Methylnaphtho[1,2-d]oxazole-5-carbonitrile (SKA-135). SKA-135 was prepared from 5-bromo-2-methylnaphtho[1,2-d] September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 6 KCa Channel Activator Design Shim et al. methylisothiocyanate according to General Method VIII. The product was isolated as a white solid (35%). m.p = 172–173°C (CAS 1369251-74-0). 1H NMR (800 MHz, DMSO-d6): δ = 8.50– 8.48 (m, 1H), 8.04–8.00 (m, 2H), 7.70 (d, J = 1.2 Hz, 1H), 7.58 (dddd, J = 20.1, 8.2, 6.7, 1.4 Hz, 2H), 3.05 (d, J = 4.7 Hz, 3H), 2.67 (s, 3H). (SKA-218). 1H NMR (800 MHz, DMSO-d6): δ = 8.31–8.28 (m, 1H), 7.94 (dd, J = 8.6, 1.7 Hz, 1H), 7.83 (dd, J = 24.8, 8.5 Hz, 2H), 7.69 (s, 1H), 7.38 (dd, J = 8.4, 1.7 Hz, 1H), 7.15 (d, J = 14.1 Hz, 3H), 2.55 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 168.6, 150.3, 135.5, 132.9, 131.9, 131.8, 128.9, 128.4, 127.6, 126.8, 124.5, 124.3, 102.3, 21.7. HRMS (ESI): m/z calculated for C14H12N2S (M+H)+: 241.0794; found: 241.0792. 4,4'-(1,4-Phenylene)bis(thiazol-2-amine) (SKA-230). SKA- 230 was prepared from 1,4-acetylbenzene in 2 steps according to general method IV, V. The product was isolated as light- yellow powder (197 mg, 47%); m.p. = 350°C (CAS 13355-22-1). 1H NMR (800 MHz, DMSO-d6): δ = 8.59 (s, 4H), 7.89 (s, 4H), N,5-Dimethylnaphtho[1,2-d]thiazol-2-amine (SKA-204). SKA-204 was prepared from 2-Methyl-1-naphthylamine and Frontiers in Pharmacology \| www.frontiersin.org September 2019 \| Volume 10 \| Article 972 7 KCa Channel Activator Design Shim et al. 134.7, 129.1, 128.2, 126.8, 126.8, 126.2, 125.8, 124.4, 123.1, 116.9, 104.6. 134.7, 129.1, 128.2, 126.8, 126.8, 126.2, 125.8, 124.4, 123.1, 116.9, 104.6. 7.35 (s, 2H). 13C NMR (201 MHz, DMSO-d6): δ = 170.3, 128.0, 126.6, 119.9, 103.9. 7.35 (s, 2H). 13C NMR (201 MHz, DMSO-d6): δ = 170.3, 128.0, 126.6, 119.9, 103.9. 7.35 (s, 2H). 13C NMR (201 MHz, DMSO-d6): δ = 170.3, 128.0, 126.6, 119.9, 103.9. 4-(Naphthalen-1-yl)-N-(3-(trifluoromethyl)phenyl) thiazol-2-amine (SKA-251). SKA-251 was prepared from 1-acetonaphthone (500 mg, 2.94 mmol) and 3-(trifluoromethyl) phenylthiourea in 2 steps according to General methods IV and V. The product was isolated as white crystals (50 mg, 4.6%); m.p. = 135°C. 1H NMR (800 MHz, DMSO-d6): δ = 10.73 (s, 1H), 8.55 (d, J = 8.5 Hz, 1H), 8.38 (d, J = 2.1 Hz, 1H), 8.01 (m, 2H), 7.77 (ddd, J = 16.4, 7.6, 1.7 Hz, 2H), 7.56 (m, 4H), 7.27 (m, 2H). 13C NMR (201 MHz, DMSO-d6): δ = 162.7, 150.7, 142.2, 134.0, 133.1, 131.1, 130.5, 128.9, 128.7, 127.4, 126.6, 126.4, 126.2, 125.9, 120.7, 117.7, 117.6, 113.0, 113.0, 108.0. HRMS (ESI): m/z calculated for C20H13F3N2S(M+H)+: 371.0825; found: 371.0832. 5-Ethylnaphtho[1,2-d]thiazol-2-amine (SKA-306). SKA-306 was prepared from 1-ethylnaphthalene in 3 steps according to General methods I, II, and VII. The product was isolated as a brownish powder. (154 mg, 22%); m.p = 144–145°C. (SKA-218). 1H NMR (800 MHz, DMSO-d6): δ = 8.39 (d, J = 7.7 Hz, 1H), 8.04 (d, J = 1.7 Hz, 1H), 7.66 (s, 1H), 7.54–7.48 (m, 4H), 3.11–3.01 (m, 2H), 1.31 (t, J = 7.5 Hz, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 167.2, 146.9, 133.2, 130.2, 126.5, 125.7, 125.4, 125.2, 124.5, 124.4, 118.3, 25.7, 15.7. HRMS (ESI): m/z calculated for C13H12N2S (M+H)+: 229.0794; found: 229.0794. 5-Propylnaphtho[1,2-d]thiazol-2-amine (SKA-307). SKA- 307 was prepared from 1-allylnaphthalene in 3 steps according to General Method I, II, and VII. The product was isolated as white crystals (495 mg, 70%); m.p. = 147–150°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.41–8.33 (m, 1H), 8.07–7.98 (m, 1H), 7.65 (s, 1H), 7.56–7.45 (m, 4H), 3.03–2.96 (m, 2H), 1.70 (q, J = 7.5 Hz, 2H), 0.98 (t, J = 7.3 Hz, 3H).13C NMR (201 MHz, DMSO-d6): δ = 167.2, 147.2, 131.5, 130.3, 126.6, 125.6, 125.3, 125.2, 124.6, 124.5, 119.3, 34.9, 24.2, 14.5. HRMS (ESI): m/z calculated for C14H14N2S (M+H)+: 243.0950; found: 243.0950. 5-Propylnaphtho[1,2-d]thiazol-2-amine (SKA-307). SKA- 307 was prepared from 1-allylnaphthalene in 3 steps according to General Method I, II, and VII. The product was isolated as white crystals (495 mg, 70%); m.p. = 147–150°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.41–8.33 (m, 1H), 8.07–7.98 (m, 1H), 7.65 (s, 1H), 7.56–7.45 (m, 4H), 3.03–2.96 (m, 2H), 1.70 (q, J = 7.5 Hz, 2H), 0.98 (t, J = 7.3 Hz, 3H).13C NMR (201 MHz, DMSO-d6): δ = 167.2, 147.2, 131.5, 130.3, 126.6, 125.6, 125.3, 125.2, 124.6, 124.5, 119.3, 34.9, 24.2, 14.5. HRMS (ESI): m/z calculated for C14H14N2S (M+H)+: 243.0950; found: 243.0950. ; 4-(Naphthalen-1-yl)-N-(pyridin-2-yl)thiazol-2-amine (SKA- 258). SKA-258 was prepared from 1-acetonaphthone and 2-pyridylthiourea in 2 steps according to General methods IV and V. The product was isolated as white powder (436 mg, 35%); m.p. = 225°C. 1H NMR (800 MHz, DMSO-d6): δ = 11.51 (s, 1H), 8.53 (m, 1H), 8.39 (ddd, J = 5.0, 1.9, 0.9 Hz, 1H), 7.99 (m, 1H), 7.96 (m, 1H), 7.75 (m, 2H), 7.59 (m, 3H), 7.26 (s, 1H), 7.15 (dt, J = 8.3, 1.0 Hz, 1H), 6.99 (ddd, J = 7.2, 5.0, 1.0 Hz, 1H).13C NMR (201 MHz, DMSO-d6): δ = 159.6, 152.3, 149.2, 147.0, 138.3, 134.0, 133.7, 131.1, 128.6, 128.5, 127.3, 126.5, 126.5, 126.3, 125.9, 116.4, 111.2, 110.1. HRMS (ESI): m/z calculated for C18H13N3S (M+H)+: 304.0903; found: 304.0905. 5-Isopropylnaphtho[1,2-d]thiazol-2-amine (SKA-308). (SKA-218). SKA-308 was prepared from 1-isopropylnaphthalene in 3 steps according to general method I, II, and VII. The product was isolated as white powder (179 mg, 10%); m.p. = 179–181°C.1H NMR (800 MHz, DMSO-d6): δ = 8.47–8.41 (m, 1H), 8.18–8.13 (m, 1H), 7.77 (s, 1H), 7.59–7.52 (m, 4H), 1.38 (d, J = 6.8 Hz, 7H). 13C NMR (201 MHz, DMSO-d6): δ = 167.3, 146.9, 137.6, 130.3, 129.8, 126.6, 125.5, 125.4, 124.7, 124.0, 115.5, 28.4, 24.0. HRMS (ESI): m/z calculated for C14H14N2S (M+H)+: 243.0950; found: 243.0949. 147.0, 138.3, 134.0, 133.7, 131.1, 128.6, 128.5, 127.3, 126.5, 126.5, 126.3, 125.9, 116.4, 111.2, 110.1. HRMS (ESI): m/z calculated for C18H13N3S (M+H)+: 304.0903; found: 304.0905. 4-(p-Tolyl)-N-(3-(trifluoromethyl)phenyl)thiazol-2- amine (SKA-260). SKA-260 was prepared from 2-bromo-4’- methylacetophenone (500 mg, 2.34 mmol) and 3-(trifluoromethyl) phenylthiourea according to General Method V. The product was isolated as orange crystals (744 mg, 95%); m.p. = 108°C (CAS 778566-59-9). 1H NMR (800 MHz, DMSO-d6): δ = 10.68 (s, 1H), 8.42 (t, J = 2.0 Hz, 1H), 7.89 (dd, J = 8.2, 2.2 Hz, 1H), 7.85 (m, 2H), 7.61 (t, J = 7.9 Hz, 1H), 7.39 (s, 1H), 7.33 (m, 1H), 7.29 (m, 2H), 2.37 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 162.9, 150.5, 142.2, 137.5, 132.1, 130.6, 130.2, 130.1, 129.7, 125.9, 125.4, 124.1, 120.6, 117.6, 113.1, 103.3, 21.2. Naphtho[1,2-d:5,6-d’]bis(thiazole)-2,7-diamine (SKA-318). 1,5-Diaminonaphthalene (500 mg, 3.16 mmol) and ammonium thioscyanate (480 mg, 6.32 mmol) were dissolved in 20 mL of anhydrous acetonitrile at room temperature. Benzyl trimethyl ammonium tribromide (2.48 g, 6.36 mmol) was added to the solution, and the reaction mixture was stirred at room temperature for 5 h. After completion of the reaction as determined by TLC, the reaction mixture was concentrated to dryness under vacuum, and the residue was diluted with 100 ml of water. The pH of the solution was adjusted to between 7 to 7.5 with saturated sodium bicarbonate solution. The product was isolated as a red crystal (20 mg, 2.5%); m.p = 194°C (CAS 1621401-78-2). 1H NMR (800 MHz, DMSO-d6): δ = 7.41 (d, J = 9.0 Hz, 2H), 7.38 (d, J = 9.0 Hz, 2H), 5.80 (s, 4H). 13C NMR (201 MHz, DMSO-d6): δ = 168.1,141.8, 129.2, 123.4, 112.3, 102.7. 4-Phenyl-N-(pyridin-2-yl)thiazol-2-amine (SKA-265). SKA-265 was prepared from 2-bromoacetophenone (500 mg, 2.5 mmol) and 2-pyridylthiourea according to General Method V. The product was isolated as white crystals (495 mg, 78%); m.p. = 158°C (CAS 92663-22-4). (SKA-218). 13C NMR (201 MHz, DMSO-d6): δ = 169.4, 169.1, 148.9, 139.0, 138.5, 137.5, 128.4, 127.5, 126.8, 126.8, 126.8, 126.4, 125.8, 124.5, 124.4, 124.4, 122.8, 122.6, 119.9, 119.2, 118.3, 55.3, 14.1. HRMS (ESI): m/z calculated for C15H12N4S (M+H)+: 281.0855; found: 281.0859. 5-(2-Methyl-1H-imidazol-1-yl)naphtho[1,2-d]thiazol- 2-amine (SKA-344). SKA-344 was prepared from 2-methylimidazole in 3 steps according to General methods II, VI, and VII. The product was isolated as light brown powder (300 mg, 41%); m.p = 264–265°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.46 (ddd, J = 8.2, 1.2, 0.7 Hz, 1H), 7.97 (s, 1H), 7.87 (s, 2H), 7.61 (ddd, J = 8.1, 6.8, 1.2 Hz, 1H), 7.51 (ddd, J = 8.2, 6.8, 1.3 Hz, 1H), 7.28 (d, J = 1.3 Hz, 1H), 7.05 (dt, J = 8.4, 0.9 Hz, 1H), 7.02 (d, J = 1.3 Hz, 1H), 2.03 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 169.4, 149.3, 145.8, 128.8, 127.5, 127.2, 127.0, 126.8, 126.0, 124.5, 124.5, 123.1, 122.6, 119.5, 13.2. HRMS (ESI): m/z calculated for C15H12N4S (M+H)+: 281.0855; found: 281 0859 5-(4-Methyl-1H-imidazol-1-yl)naphtho[1,2-d]thiazol- 2-amine (SKA-343). SKA-343 was prepared from 4-methylimidazole in 3 steps according to General methods II, VI, and VII. The product was isolated as brownish crystals (160 mg, 22%); m.p = 269–271°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.45 (d, 1H), 7.95 (d,1H), 7.86 (s, 1H), 7.82 (s, 2H), 7.77 (d, 1H), 7.63–7.58 (m, 1H), 7.52 (dtd, J = 7.6, 6.5, 1.3 Hz, 1H), 7.39 (dt, J = 8.4, 0.9 Hz, 1H), 2.23 (d, J = 1.1 Hz, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 169.4, 169.1, 148.9, 139.0, 138.5, 137.5, 128.4, 127.5, 126.8, 126.8, 126.8, 126.4, 125.8, 124.5, 124.4, 124.4, 122.8, 122.6, 119.9, 119.2, 118.3, 55.3, 14.1. HRMS (ESI): m/z calculated for C15H12N4S (M+H)+: 281.0855; found: 281.0859. 5-(1H-Imidazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA- 334). SKA-334 was prepared from imidazole in 3 steps according to General Method II, VI, and VII. The product was isolated as brownish powder (130 mg, 19%); m.p. = 272–275°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.46 (d, 1H), 7.99 (s, 1H), 7.94 (t, J = 1.1 Hz, 1H), 7.85 (s, 2H), 7.62 (ddd, J = 8.1, 6.8, 1.1 Hz, 1H), 7.55–7.50 (m, 1H), 7.33 (dt, J = 8.4, 0.9 Hz, 2H), 7.18 (t, J = 1.1 Hz,1H). 13C NMR (201 MHz, DMSO-d6): δ = 169.3, 149.1, 139.3, 129.1, 128.4, 127.2, 126.9, 126.8, 125.8, 124.5, 124.4, 123.0, 122.6, 118.6. (SKA-218). = 177–178°C 1H NMR (800 MHz, DMSO-d6): δ = 8.43 (d, 1H), 7.78 (dt, J = 8.5, 0.9 Hz, 1H), 7.74 (s, 1H), 7.68 (s, 2H), 7.57–7.50 (m, 3H), 7.50–7.46 (m, 2H), 7.48– 7.42 (m, 2H).13C NMR (201 MHz, DMSO-d6): δ = 168.1, 148.2, 140.8, 133.0, 130.4, 130.0, 128.8, 127.5, 126.3, 126.1, 126.0, 125.7, 125.3, 124.4, 120.5. HRMS (ESI): m/z calculated for C17H12N2S (M+H)+: 277.0794; found: 277.0798. 5-Phenylnaphtho[1,2-d]thiazol-2-amine (SKA-326). SKA-326 was prepared from 1-phenylnaphthalene in 3 steps according to General Method I, II, and VII. The product was isolated as white crystals (495 mg, 78%); m.p. = 177–178°C 1H NMR (800 MHz, DMSO-d6): δ = 8.43 (d, 1H), 7.78 (dt, J = 8.5, 0.9 Hz, 1H), 7.74 (s, 1H), 7.68 (s, 2H), 7.57–7.50 (m, 3H), 7.50–7.46 (m, 2H), 7.48– 7.42 (m, 2H).13C NMR (201 MHz, DMSO-d6): δ = 168.1, 148.2, 140.8, 133.0, 130.4, 130.0, 128.8, 127.5, 126.3, 126.1, 126.0, 125.7, 125.3, 124.4, 120.5. HRMS (ESI): m/z calculated for C17H12N2S (M+H)+: 277.0794; found: 277.0798. 5-Phenylnaphtho[1,2-d]thiazol-2-amine (SKA-326). SKA-326 was prepared from 1-phenylnaphthalene in 3 steps according to General Method I, II, and VII. The product was isolated as white crystals (495 mg, 78%); m.p. = 177–178°C 1H NMR (800 MHz, DMSO-d6): δ = 8.43 (d, 1H), 7.78 (dt, J = 8.5, 0.9 Hz, 1H), 7.74 (s, 1H), 7.68 (s, 2H), 7.57–7.50 (m, 3H), 7.50–7.46 (m, 2H), 7.48– 7.42 (m, 2H).13C NMR (201 MHz, DMSO-d6): δ = 168.1, 148.2, 140.8, 133.0, 130.4, 130.0, 128.8, 127.5, 126.3, 126.1, 126.0, 125.7, 125.3, 124.4, 120.5. HRMS (ESI): m/z calculated for C17H12N2S (M+H)+: 277.0794; found: 277.0798. 5-(3-Methyl-1H-pyrazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA-340). SKA-340 was prepared from 3-methylpyrazole in 3 steps according to General methods II, VI, and VII. The product was isolated as brownish crystals (140 mg, 19%); m.p = 189–190°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.46–8.41 (m, 1H), 7.97 (d, J = 2.2 Hz, 1H), 7.92 (s, 1H), 7.81 (s, 2H), 7.61–7.57 (m, 2H), 7.50 (ddd, J = 8.4, 7.0, 1.3 Hz, 1H), 6.35 (d, J = 2.2 Hz, 1H), 2.31 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 169.1, 149.0, 148.7, 133.8, 131.0, 128.1, 126.6, 126.4, 125.9, 124.3, 124.2, 123.8, 117.9, 106.4,13.9. HRMS (ESI): m/z calculated for C15H12N4S (M+H)+: 281.0855; found: 281.0858. N5,N5-Diethylnaphtho[1,2-d]thiazole-2,5-diamine (SKA- 330). SKA-330 was prepared from diethylamine hydrochloride in 3 steps according to General Method II, VI, and VII. The product was isolated as yellowish crystals (50 mg, 8%); m.p. (SKA-218). 1H NMR (800 MHz, DMSO-d6),: δ = 11.41 (s, 1H), 8.32 (ddd, J = 5.0, 1.9, 0.9 Hz, 1H), 7.92 (m, 2H), 7.72 (ddd, J = 8.4, 7.1, 1.9 Hz, 1H), 7.43 (m, 3H), 7.31 (td, J = 7.3, 1.3 Hz, 1H), 7.11 (dd, J = 8.3, 1.0 Hz, 1H), 6.94 (ddd, J = 7.2, 5.0, 1.0 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 159.9, 152.3, 149.0, 146.9, 138.3, 135.2, 129.0, 127.9, 126.0, 116.4, 111.2, 106.3, 40.3, 39.7. 5-Nitronaphtho[1,2-d]thiazol-2-amine (SKA-321). SKA-321 was prepared from 1-naphtho[1,2-d]thiazol-2-amine (500 mg, 7 mmol) according to General Method I. The product was isolated as white powder (30 mg, 2%); m.p. = 289–291°C (CAS 320340- 92-9) 1H NMR (800 MHz, DMSO-d6): δ = 8.95 (s, 1H), 8.67 (dt, J = 8.7, 0.9 Hz, 1H), 8.54 (dd, J = 8.3, 1.6 Hz, 1H), 8.40 (s, 2H), 7.76 (ddd, J = 8.6, 6.8, 1.4 Hz, 1H), 7.69 (ddd, J = 8.1, 6.8, 1.1 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 173.2, 155.2, 138.5, 129.3, 127.1, 125.2, 125.1, 124.8, 123.7, 121.4. 4-Phenyl-N-(4-(trifluoromethyl)phenyl)thiazol-2-amine (SKA-268). SKA-268 was prepared from 2-bromoacetophenone (500 mg, 2.5 mmol) and [4-(trifluoromethyl)phenyl]thiourea according to General Method V. The product was isolated as yellow crystals (530 mg, 66%); m.p. = 153°C (CAS 1303995- 45-0). 1H NMR (800 MHz, DMSO-d6): δ = 10.75 (s, 1H), 8.01 (m, 4H), 7.75 (d, J = 8.5 Hz, 2H), 7.49(m, 3H), 7.37 (m, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 162.8, 150.6, 144.8, Frontiers in Pharmacology \| www.frontiersin.org September 2019 \| Volume 10 \| Article 972 8 Shim et al. KCa Channel Activator Design NMR (800 MHz, DMSO-d6): δ = 8.48 (d, Hz,1H), 8.20 (s, 2H), 8.15 (s, 1H), 7.91 (s, 2H), 7.66–7.61 (m, 2H), 7.54 (ddd, J = 8.3, 6.7, 1.3 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 169.8, 149.7, 136.3, 130.2, 126.9, 126.9, 126.8, 125.7, 124.4, 124.2, 123.3, 118.2. HRMS (ESI): m/z calculated for C13H9N5S (M+H)+: 268.0651; found: 268.0654. 5-Phenylnaphtho[1,2-d]thiazol-2-amine (SKA-326). SKA-326 was prepared from 1-phenylnaphthalene in 3 steps according to General Method I, II, and VII. The product was isolated as white crystals (495 mg, 78%); m.p. (SKA-218). = 119–121°C.1H NMR (800 MHz, DMSO-d6): δ = 8.32 (d, 1H), 8.27 (d, 1H), 7.60 (s, 1H), 7.50 (ddd, J = 8.1, 6.7, 1.3 Hz, 1H), 7.45 (ddd, J = 8.2, 6.7, 1.4 Hz, 1H), 7.43 (s, 2H), 3.09 (q, J = 7.1 Hz, 4H), 0.95 (t, J = 7.1 Hz, 6H). 13C NMR (201 MHz, DMSO-d6): δ = 167.0, 145.0, 141.6, 130.3, 126.9, 125.9, 125.4, 124.9, 124.59, 124.1, 112.7, 48.4, 21.5, 12.9. HRMS (ESI): m/z calculated for C15H17N3S (M+H)+: 272.1216; found: 272.1217. 5-(3-(Trifluoromethyl)-1H-pyrazol-1-yl)naphtho[1,2-d] thiazol-2-amine (SKA-341). SKA-341 was prepared from 3-(trifluoromethyl)pyrazole in 3 steps according to General methods II, VI, and VII. The product was isolated as white crystals (150 mg, 17%); m.p = 206–207°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.47 (dt, J = 8.3, 1.0 Hz, 1H), 8.41 (dd, J = 2.4, 1.1 Hz, 1H), 8.11 (s, 1H), 7.91 (s, 2H), 7.63 (ddd, J = 8.2, 6.7, 1.2 Hz, 1H), 7.55 (ddd, J = 8.3, 6.8, 1.3 Hz, 1H), 7.37 (dt, J = 8.5, 0.9 Hz, 1H), 7.07 (d, J = 2.4 Hz, 1H).13C NMR (201 MHz, DMSO-d6): δ = 169.8, 149.8, 142.4, 136.0, 129.4, 127.8, 127.1, 126.9, 125.7, 124.5, 124.1, 122.7, 122.6, 118.9, 105.5. HRMS (ESI): m/z calculated for C15H9F3N4S (M+H)+: 335.0573; found: 335.0575. N5,N5-Dimethylnaphtho[1,2-d]thiazole-2,5-diamine (SKA- 331). SKA-331 was prepared from dimethylamine hydrochloride in 3 steps according to General Method II, VI, and VII. The product was isolated as white crystals (30 mg, 5%); m.p. = 171– 174°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.32 (d, 1H), 8.17 (d, 1H), 7.51 (d,1H), 7.51 (s, 1H), 7.48 (d,1H), 7.42 (s, 2H), 2.78 (s, 6H). 13C NMR (201 MHz, DMSO-d6): δ = 166.7, 145.2, 144.4, 127.5, 127.0, 126.0, 125.4, 124.9, 124.5, 124.3, 108.8, 45.7 HRMS (ESI): m/z calculated for C13H13N3S (M+H)+: 244.0903; found: 244.0902. found: 335.0575. 5-(4-Methyl-1H-imidazol-1-yl)naphtho[1,2-d]thiazol- 2-amine (SKA-343). SKA-343 was prepared from 4-methylimidazole in 3 steps according to General methods II, VI, and VII. The product was isolated as brownish crystals (160 mg, 22%); m.p = 269–271°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.45 (d, 1H), 7.95 (d,1H), 7.86 (s, 1H), 7.82 (s, 2H), 7.77 (d, 1H), 7.63–7.58 (m, 1H), 7.52 (dtd, J = 7.6, 6.5, 1.3 Hz, 1H), 7.39 (dt, J = 8.4, 0.9 Hz, 1H), 2.23 (d, J = 1.1 Hz, 3H). Crystal Structure Determination y The SKA-218, SKA-339, SKA-340, SKA-343 and SKA347 crystals selected for data collection were mounted and optically centered in a nitrogen low temperature stream –183°C (90K), on the Bruker diffractometer with an APEX2 CCD detector or a Bruker D8 Venture diffractometer equipped with a Photon100 CMOS detector (Bruker, Madison, WI). Data were collected with the use of Mo Ka radiation in all cases (λ = 0.71073 Å). The structures were solved by direct methods (SHELXT) and refined by full-matrix least-squares on F2 (SHELXL-2018/3). All non-hydrogen atoms were refined with anisotropic displacement parameters. For a description of the method, see (Sheldrick, 2008). p g Ca ( pp y g ) Using the information from these computational docking models, we here intended to design new KCa2 selective activators by attempting to predict whether the compounds would show selectivity for KCa2 over KCa3.1 channels. However, instead of our previously generated KCa2.3 CaM-BD/CaM homology model, we here used a KCa2.2 CaM-BD/CaM model based on the KCa2.2 CaM-BD/CaM-NS309 crystal structure (pdb:4J9Z). We made this switch because KCa2.2 is the most abundantly expressed KCa2 channel in the mammalian CNS (Adelman et al., 2012) and therefore constitutes an attractive target for the treatment of ataxia and epilepsy. We additionally docked eight more 2-amino-naphthobenzothiazole derivatives (SKA-31, SKA-44, SKA-45, SKA-72, SKA-73, SKA-107, SKA-117, and SKA-120) into the KCa2.2 and KCa3.1 homology models and found that these compounds exhibited the same hydrogen bond network as SKA-121 and SKA-111 in KCa3.1 (Supplementary Figure 1). Based on these docking poses we hypothesized that disruption of the hydrogen bond between the –NH2 group of the benzothiazole ring, and the CaM M51 and E54 residues, which are present in both the KCa3.1 and KCa2.2 models, might be a way to achieve KCa2.2 selectivity. Our goal here was to first “break” this hydrogen bond to ideally achieve KCa2.2 selectivity and then regain potency by adding substituents in other positions to pick up unique contacts in KCa2.2. (SKA-218). HRMS (ESI): m/z calculated for C14H10N4S (M+H)+: 267.0699; found: 267.0700. 15 12 4 ( ) ; 5-(2-Methyl-1H-imidazol-1-yl)naphtho[1,2-d]thiazol- 2-amine (SKA-344). SKA-344 was prepared from 2-methylimidazole in 3 steps according to General methods II, VI, and VII. The product was isolated as light brown powder (300 mg, 41%); m.p = 264–265°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.46 (ddd, J = 8.2, 1.2, 0.7 Hz, 1H), 7.97 (s, 1H), 7.87 (s, 2H), 7.61 (ddd, J = 8.1, 6.8, 1.2 Hz, 1H), 7.51 (ddd, J = 8.2, 6.8, 1.3 Hz, 1H), 7.28 (d, J = 1.3 Hz, 1H), 7.05 (dt, J = 8.4, 0.9 Hz, 1H), 7.02 (d, J = 1.3 Hz, 1H), 2.03 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 169.4, 149.3, 145.8, 128.8, 127.5, 127.2, 127.0, 126.8, 126.0, 124.5, 124.5, 123.1, 122.6, 119.5, 13.2. HRMS (ESI): m/z calculated for C15H12N4S (M+H)+: 281.0855; found: 281.0859. 5-(1H-Pyrazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA- 335). SKA-335 was prepared from pyrazole in 3 steps according to General methods II, VI, and VII. The product was isolated as brown crystal (220 mg, 31%); m.p. = 226–227°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.44 (dt, J = 8.3, 1.1 Hz, 1H), 8.12 (d, J = 2.3 Hz, 1H), 7.97 (s, 1H), 7.85–7.80 (m, 3H), 7.60 (ddd, J = 8.2, 4.7, 3.2 Hz, 1H), 7.52–7.47 (m, 2H), 6.57 (t, J = 2.1 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 169.2, 149.0, 140.6, 133.2, 130.8, 128.2, 126.7, 126.5, 125.9, 124.3, 124.2, 123.6, 118.1, 106.7. HRMS (ESI): m/z calculated for C14H10N4S (M+H)+: 267.0699; found: 267.0700. 130.8, 128.2, 126.7, 126.5, 125.9, 124.3, 124.2, 123.6, 118.1, 106.7. HRMS (ESI): m/z calculated for C14H10N4S (M+H)+: 267.0699; found: 267.0700. 5-(1H-1,2,3-Triazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA-339). SKA-339 was prepared from 1,2,3-triazole in 3 steps according to General methods II, VI, and VII. The product was isolated as purplish crystals (120 mg, 17%); m.p = 208°C. 1H 5-(1H-1,2,3-Triazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA-339). SKA-339 was prepared from 1,2,3-triazole in 3 steps according to General methods II, VI, and VII. The product was isolated as purplish crystals (120 mg, 17%); m.p = 208°C. 1H 5-(4-Methyl-1H-pyrazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA-345). SKA-345 was prepared from 4-methylpyrazole in 3 steps according to General methods II, VI, and VII. The product was 5-(4-Methyl-1H-pyrazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA-345). SKA-345 was prepared from 4-methylpyrazole in 3 steps according to General methods II, VI, and VII. (SKA-218). The product was September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 9 KCa Channel Activator Design Shim et al. monoclinic P21/n a = 3.7801(3) Å, b = 14.7453(12) Å, c = 20.2851(17) Å, β = 94.550(2)°, V = 1127.10(16) Å3, Z = 4, R1 = 0.0305, wR2 = 0.0797, 209 parameters, 0 restraints. isolated as brownish powder (280 mg, 39%); m.p = 249°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.43 (dt, J = 8.3, 0.9 Hz, 1H), 7.97 (s, 1H), 7.92 (s, 1H), 7.87 (t, J = 0.9 Hz, 1H), 7.81 (s, 2H), 7.62–7.54 (m, 2H), 7.49 (ddd, J = 8.3, 6.7, 1.3 Hz, 1H), 2.16 (s, 3H). 13C NMR (201 MHz, DMSO-d6): δ = 169.1, 148.8, 141.2, 131.6, 131.1, 128.1, 126.6, 126.4, 125.9, 124.3, 123.7, 117.8, 116.3, 9.2. HRMS (ESI): m/z calculated for C15H12N4S (M+H)+: 281.0855; found: 281.0857. Structure-Based Drug Design Using the Crystal Structure of the KCa2.2 CaM-BD/ CaM Interface 5-(1H-1,2,4-Triazol-1-yl)naphtho[1,2-d]thiazol-2-amine (SKA-347). SKA-347 was prepared from 1,2,4-triazole in 3 steps according to General methods II, VI, and VII. The product was isolated as light yellowish crystals (200 mg, 28%); m.p = 273°C. 1H NMR (800 MHz, DMSO-d6): δ = 8.97 (s, 1H), 8.47 (dt, J = 8.3, 1.0 Hz, 1H), 8.33 (s, 1H), 8.08 (s, 1H), 7.92 (s, 2H), 7.63 (ddd, J = 8.1, 6.8, 1.1 Hz, 1H), 7.54 (ddd, J = 8.3, 6.8, 1.3 Hz, 1H), 7.42 (dt, J = 8.5, 0.9 Hz, 1H). 13C NMR (201 MHz, DMSO-d6): δ = 169.7, 152.6, 149.8, 146.6, 127.8, 127.0, 126.9, 126.7, 125.84, 124.4, 124.1, 122.9, 118.8. HRMS (ESI): m/z calculated for C13H9N5S (M+H)+: 268.0651; found: 268.0655. We previously generated homology models of KCa3.1 and KCa2.3 (Brown et al., 2017) using the Rosetta membrane method (Rohl et al., 2004; Bender et al., 2016; Alford et al., 2017) and the x-ray crystal structure of the KCa2.2 CaM-BD/CaM (Zhang et al., 2013) as a template. We localized the binding site of the benzothiazoles/oxazoles to the CaM-BD/CaM interface and generate models of the KCa3.1 and KCa2.3 CaM-BD/CaM complexes with SKA-121 and SKA-111 using Rosetta Ligand docking (Meiler and Baker, 2006; Davis and Baker, 2009; Bender et al., 2016). The docking models of KCa3.1 showed that the amino groups of the benzoxazole ring of SKA-121 and of the benzothiazole ring of SKA-111 form hydrogen bonds with M51 and E54 in calmodulin (Brown et al., 2017). Moreover, E54 was further stabilized by an extensive hydrogen bond network with R362, E295 and N300 in the KCa3.1 channel, which we hypothesized to be responsible for the KCa3.1 selectivity of SKA-121 and SKA-111. In the KCa2.3 or KCa2.2 model, however, SKA-111 and SKA-121 formed only hydrogen bonds with M51 and E54, due to the shorter length of the sidechain of S622 in KCa2.3 or N474 in KCa2.2 than that of the corresponding R362 in KCa3.1 (Supplementary Figure 1). Crystal Structure Determination We therefore virtually added various substituents in the C-4,5,6,7,8,9 positions of SKA-74 (Figure 1A), a compound which contains a methyl group in C-2 position instead of an -NH2 group and which we had previously found to activate KCa2 and KCa3.1 channels with a Crystal Data SKA-218, JF2786, APEX2, C12H7N3O4S, F.W = 289.27, purple block, dimensions = 0.472 x 0.400 x 0.232 mm3, monoclinic, P21/n, a = 9.9178(15) Å, b = 15.885(2) Å, c = 7.4528(11) Å, β = 93.4141(19)°, V = 1172.0(3) Å3, Z = 4, 182 parameters, 0 restraints. Crystal Data SKA-339, JF2770, D8 Venture, C13H9N5S, F.W = 267.31, yellow plate, dimensions = 0.136 x 0.071 x 0.023 mm3, monoclinic, P21/n, a = 16.3915(19) Å, b = 3.7972(4) Å, c = 19.437(2) Å, β = 112.065(2)°, V = 1121.2(2) Å3, Z = 4, 208 parameters, 0 restraints. Crystal Data SKA-340, JF2772, D8 Venture, C15H12N4S, F.W = 280.35, yellow rod, dimension = 0.401 x 0.143 x 0.073 mm3, triclinic, P-1, a = 8.7703(10) Å, α = 96.219(4)°, b = 9.3235(11) Å, β = 103.017(4)°, c = 10.7112(13) Å, γ = 110.077(4)°, V = 784.83(16) Å3, Z = 2, 237 parameters, 0 restraints. Crystal Data SKA-343, JF2787, D8 Venture, C15H12N4S F.W = 280.35, colorless rod, dimensions = 0.221 x 0.086 x 0.051 mm3, monoclinic, P21/n, a = 15.9106(10) Å, b = 20.4896(13) Å, c = 7.9634(5) Å, β = 94.589(2)°, V = 2587.8(3) Å3, Z = 8, 458 parameters, 0 restraints. Crystal Data SKA-347, JF2769, D8 Venture, C13H9N5S, F.W. = 267.31, colorless block, dimensions = 205 x 0.150 x 0.088 mm3, September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 10 KCa Channel Activator Design Shim et al. similar ~30 μM potency (Coleman et al., 2014). To improve van der Waals contacts we introduced -CH3, -Br, -Cl, -CF3 groups. The C-5 position was chosen for the first trial since the KCa2.2 docking model of SKA-74 showed that it is adjacent to A484 and in range for new interactions. Crystal Structure Determination We also virtually introduced larger substituents such as cyclohexyl, cyclopentyl, cyclopropyl and phenyl on the 2-position amino group because the KCa2.2 model showed more space in this region of the binding site than 11 September 2019 \| Volume 10 \| A tiers in Pharmacology \| www frontiersin org GURE 1 \| Design scheme of KCa2.2 selective activators and Rosetta models of the top 10 binding poses with the lowest energy of template compounds in erface between CaM (pink) and the CaM-BD (light green) of KCa3.1 and KCa2.2. The docking model of SKA-74 (A) and SKA-76 (B) showed that the ten low nding energy scored models exhibit good structural convergence in KCa2.2 but not in KCa3.1 suggesting selectivity for KCa2.2 over KCa3.1. FIGURE 1 \| Design scheme of KCa2.2 selective activators and Rosetta models of the top 10 binding poses with the lowest energy of template compounds in the interface between CaM (pink) and the CaM-BD (light green) of KCa3.1 and KCa2.2. The docking model of SKA-74 (A) and SKA-76 (B) showed that the ten lowest binding energy scored models exhibit good structural convergence in KCa2.2 but not in KCa3.1 suggesting selectivity for KCa2.2 over KCa3.1. September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 11 Shim et al. KCa Channel Activator Design The virtually proposed 63 SKA-74 derivatives, 87 SKA-75 and SKA-76 derivatives as well as another 18 SKA-31-related compounds (see Supplementary Figure 2 for all structures) were randomly placed into the KCa2.2 and KCa3.1 homology models of the CaM-BD/CaM interface pocket, energy minimized through the three stages of the RosettaLigand method, and the top 10 lowest binding energy scoring models were analyzed. SKA-74 derivatives with –CH3, -CF3, -Br and –Cl in 5-position were predicted to show selectivity for KCa2.2 channels over KCa3.1 (Figures 2A, B) because they converged well in the KCa2.2 model and made van der Waals interactions (dark purple in Figure 2B) with A477, V481 in KCa2.2 and M72, F68, I63 and M51 in CaM. In contrast, the top 10 lowest energy models of all SKA-74 derivatives with bulky substituents in C-2 position did not converge in either KCa2.2 or KCa3.1, whereas all double substituted SKA 74 derivatives were predicted to be KCa2.2 selective with good structural convergence in KCa2.2. Crystal Structure Determination For the SKA-75 and SKA 76 derivatives, the docking model suggested that addition of a methyl group to the 6-position of the naphthalene ring in SKA 76 (SKA-198) created new hydrogen bonds and good structural convergence in the KCa2.2 model but not in KCa3.1. Replacement of the naphthalene ring of SKA-75 with a bi-phenyl ring, a 2-aminophenylthiazole the KCa3.1 model. In addition, we also virtually generated double substituted SKA-74 derivatives (a bulky substituent in the C-2 position and methyl in C-5 position). In parallel we designed a small focused library of 2-aminothiazoles (SKA-75 and SKA-76 derivatives; Figure 1B). Our reasoning for the choice of these two compounds as additional templates was that SKA-75, like SKA- 74, had previously been found to be of similar potency (~30 μM) on both KCa2.3 and KCa3.1, and that SKA-76 was slightly more potent on KCa2.3 (~25 μM) than KCa3.1 (~50 μM) (Coleman et al., 2014). In addition, the molecular docking models of SKA- 75 and SKA-76, which both lack the continuous conjugation between the naphthalene and the 2-aminothiazole ring (Figure 1 right), showed that the ten lowest energy scored models exhibited good structural convergence in KCa2.2 but not in KCa3.1 (Figure 1B). In order to improve the potency and selectivity of SKA-75 and SKA-76 we virtually added a methyl group in the 2,3,4,5,6,7 or 8 positions of the naphthalene ring in SKA-75 and SKA-76. We then replaced the naphthalene ring with a bi-phenyl ring, a 2-aminophenylthiazole or a 2-aminobiphenylthiazole. We further replaced the naphthalene ring with a phenyl ring and added larger groups such as N-trifluorophenyl, N-pyridine to the -NH2 group of the 2-aminothiazole ring. FIGURE 2 \| Rosetta ligand docking models of the lowest energy-binding poses of SKA-130 (A, B) and SKA-230 (C, D) in the interface between CaM (pink) and CaM-BD (light green) of KCa3.1 and KCa2.2. Hydrogen bonds and van der Waals interactions are indicated by purple and black lines. (A, B) Molecular docking suggests that SKA-130 converged well and formed several van der Waals contacts in KCa2.2 but not in KCa3.1. (C, D) Both KCa3.1 and KCa2.2 models exhibited good structural convergence. The molecular docking model showed that while SKA-230 only formed 2 hydrogen bonds in KCa3.1, it formed four hydrogen bonds in KCa2.2 suggesting selectivity for KCa2.2. Crystal Structure Determination FIGURE 2 \| Rosetta ligand docking models of the lowest energy-binding poses of SKA-130 (A, B) and SKA-230 (C, D) in the interface between CaM (pink) and CaM-BD (light green) of KCa3.1 and KCa2.2. Hydrogen bonds and van der Waals interactions are indicated by purple and black lines. (A, B) Molecular docking suggests that SKA-130 converged well and formed several van der Waals contacts in KCa2.2 but not in KCa3.1. (C, D) Both KCa3.1 and KCa2.2 models exhibited good structural convergence. The molecular docking model showed that while SKA-230 only formed 2 hydrogen bonds in KCa3.1, it formed four hydrogen bonds in KCa2.2 suggesting selectivity for KCa2.2. FIGURE 2 \| Rosetta ligand docking models of the lowest energy-binding poses of SKA-130 (A, B) and SKA-230 (C, D) in the interface between CaM (pink) and CaM-BD (light green) of KCa3.1 and KCa2.2. Hydrogen bonds and van der Waals interactions are indicated by purple and black lines. (A, B) Molecular docking suggests that SKA-130 converged well and formed several van der Waals contacts in KCa2.2 but not in KCa3.1. (C, D) Both KCa3.1 and KCa2.2 models exhibited good structural convergence. The molecular docking model showed that while SKA-230 only formed 2 hydrogen bonds in KCa3.1, it formed four hydrogen bonds in KCa2.2 suggesting selectivity for KCa2.2. September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 12 KCa Channel Activator Design Shim et al. for synthesis including some that were not predicted to be selective in order to verify that the predicted selectivity in the model is consistent with experiments (see Supplementary Figure 2 where the chosen structures are highlighted in color). A general scheme of the compound synthesis is given in Figure 3. To obtain 5-CH3 substituted SKA-74 derivatives the commercially available starting materials, 2-methylnaphtho[1,2-d]thiazole (SKA-74) and 2-methylnaphtho[1,2-d]oxazole (SKA-103), were first brominated using liquid bromine to obtain SKA-132 and SKA- 133, which were then reacted with CuCN in a cross coupling reaction with a palladium catalyst to obtain SKA-126 and SKA-135. SKA-130 was synthesized by electrophilic aromatic or a 2-aminobiphenylthiazole (SKA-232, SKA-230 and SKA-255) created new hydrogen bonds and good structural convergence in both KCa2.2 and KCa3.1. Crystal Structure Determination The molecular docking model showed that while these SKA-75 derivatives (SKA-232, SKA-230 and SKA-255) only formed 2 hydrogen bonds in KCa3.1, they formed four hydrogen bonds in KCa2.2 (see Figures 2C, D for SKA-230) suggesting selectivity for KCa2.2 as well as relatively high potency. Synthesis and Activity Testing of the Newly Designed KCa Channel Activators Based on the docking models we chose 26 (16 SKA-74 and 10 SKA-75/76 derivatives) of the 168 virtual compounds Synthesis and Activity Testing of the Synthesis and Activity Testing of the Newly Designed KCa Channel Activators Based on the docking models we chose 26 (16 SKA-74 and 10 SKA-75/76 derivatives) of the 168 virtual compounds GURE 3 \| General scheme for the synthesis of SKA-74, SKA-75, SKA-76 (A) and SKA-111 derivatives (B). FIGURE 3 \| General scheme for the synthesis of SKA-74, SKA-75, SKA-76 (A) and SKA-111 derivatives (B). FIGURE 3 \| General scheme for the synthesis of SKA-74, SKA-75, SKA-76 (A) and SKA-111 derivatives (B). FIGURE 3 \| General scheme for the synthesis of SKA-74, SKA-75, SKA-76 (A) and SKA-111 derivatives (B). Frontiers in Pharmacology \| www.frontiersin.org 13 September 2019 \| Volume 10 \| Article 972 September 2019 \| Volume 10 \| Article 972 Shim et al. KCa Channel Activator Design be KCa2.2 selective and potent based on multiple predicted hydrogen bonds (see Figure 2D for SKA-230 in KCa2.2). substitution with N-chlorosuccinimide. N-substituted SKA- 74 derivatives were prepared from 1-aminonaphthalene, which was reacted with isothiocyanates to the N,N’- disubstituted thioureas, which were then cyclized using bromine. SKA 75 and SKA-76 derivatives were synthesized starting from methyl naphthalene, 1,4 acetylbenzene, 1-acetonaphthone or 2-bromo substituted acetophenone. A classic Fridel-Craft acylation was used to synthesize monoacylated intermediates, which were then brominated on the alpha carbon followed by cyclization of the compounds with substituted thiourea and liquid bromine.h Return to “Classical Medicinal” Chemistry At this stage of our work we became very skeptical about our approach of using RosettaLigand docking for structure based KCa2.2 activator design and decided to return to a purely activity driven compound design approach. We therefore went back to the 2-aminonaphothiazole system of SKA-31 and opted to further exploit our previous observation that introduction of a single -CH3 group in 5-position could achieve a 100-fold gain in selectivity for KCa3.1 over KCa2 channels (Coleman et al., 2015). In order to more thoroughly explore the structure-activity relationship (SAR) in the 5-position (Figure 5A) we replaced the methyl group of SKA-111 with longer alkyl chains such as ethyl (SKA-306), propyl (SKA-307), isopropyl (SKA-308) or other functional groups, which were not tried in our previous study such sulfonyl (SKA-128), nitro (SKA-321), dimethyl (SKA-330) or diethylamino (SKA-331). Synthesis and Activity Testing of the We further decided to generate a small library of compounds with aromatic substituents in 5-position such as phenyl, imidazole, pyrazole, and triazole in order to provide ample possibilities to pick up new van der Waals, π–π, cation–π or hydrogen bond interactions (Figure 5B). In total we synthesized 19 SKA-111 derivatives according to the schemes shown in Figure 3B. Alkyl or phenyl substituted compounds were synthesized starting from differently substituted naphthalenes, which we first nitrated using nitric acid and then reduced with hydrazine. Finally, to obtain the benzothizole ring- system the resulting amino-substituted naphthalenes were subjected to a classic Hugerschoff benzothiazole synthesis with potassium thiocyanate and liquid bromine. Secondary amines or heterocyclic ring substituted compounds were prepared from 1-fluoro-4-nitronaphthalen, which was reacted with differently substituted amines through amine arylation. This nucleophilic aromatic substitution allowed us to introduce the heterocycle ring system or alkylamine into the 5-position of the naphthalene ring. Other 2-aminobenzothiazoles were generated through nitration and Hugerschoff benzothiazole synthesis. q The newly synthesized compounds were tested by automated whole-cell patch-clamp for their ability to activate KCa2.2 and KCa3.1 channels stably expressed in HEK cells. In order to quickly identify promising compounds, we devised a screening system in which we measured the response of each compound at a concentration of 10 µM and normalized that response to the response elicited by 10 µM of SKA-31, which at this concentration maximally activates both KCa3.1 and KCa2.2 (Sankaranarayanan et al., 2009). This screen was performed at a free intracellular Ca2+ concentration of 250 nM, which is ideal for activator testing (Jenkins et al., 2013), and therefore allowed us to determine two things; (1) if each compound could elicit a maximal effect, and (2) if a compound displayed any selectivity towards KCa2.2. A promising compound would ideally do both. Supplementary Figure 3 shows representative raw current traces to illustrate the sequence and timing of compound additions and washes. We started by comparing the effect of the SKA-74 derivatives to SKA-31 (Figure 4A). Addition of a cyanide group to the C5-position (SKA-135) or of a nitro group to the 6-position (SKA-215), rendered the compounds inactive on both channels. Introduction of lipophilic chloro- or bromo- groups (SKA-130, SKA-132 and SKA-133), decreased activity on KCa2.2 but preserved selectivity for KCa3.1. Synthesis and Activity Testing of the Similarly, adding – NH2 or –NO2 groups in the 6- or 9- or both the 5- and 6-position preserved the selectivity for KCa3.1 (SKA-172, SKA-178, SKA- 216 and SKA-218). Substitution of the 2-position -NH2 group also did not result in any KCa2.2 selective compounds. While introduction of a -CH3 group (SKA-158) abolished activity on KCa2.2 but preserved it again on KCa3.1, introduction of an aromatic phenyl (SKA-146) or a cyclopropyl (SKA-169) ring rendered the resulting compounds less potent than SKA-31. Introduction of a larger, aliphatic cyclohexyl group (SKA-165) completely eliminated activity on both channels. gf y We next proceeded to screen these SKA-111 derivatives (Figure 5) and found that replacement of the 5-position methyl group with ethyl (SKA-306), propyl (SKA-307), isopropyl (SKA-308) and dimethylamine (SKA-331) resulted in compounds that were roughly as potent and as KCa3.1 selective as SKA-111. Addition of a diethylamine group (SKA-330) somewhat reduced potency, while large, polar substituents like a nitro-group (SKA-312) or a sulfonic acid (SKA-128) drastically reduced potency on both KCa3.1 and KCa2.2. Unfortunately, introduction of aromatic substituents in 5-positions also did not provide any KCa2.2 selective compounds and overall again reduced potency on both channels. Taken together, we failed to generate any KCa2.2 selective SKA compounds using either structure based or “classical” approaches. We next tested the linear SKA-75 and SKA-76 derivatives and found that, compared to their undecorated templates, introduction of a methyl or a bromo group to SKA-75 and SKA 76 (SKA-190, SKA-193, SKA-198 and SKA- 220) somewhat increased KCa channel activating potency (Figure 4B). However, when compared to SKA-31, none of these modifications produced the desired KCa2.2 selectivity or improved potency. Additional introduction of a large substituent on the -NH2 group completely abolished activity (SKA-251 and SKA-258). Similarly, various linearized compounds (SKA-230, SKA-232, SKA-255, SKA-260, SKA-265 and SKA-268) were completely devoid of activity even though RosettaLigand docking predicted these linear compounds to September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 14 KCa Channel Activator Design Shim et al. FIGURE 4 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were performed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline additions, and then SKA-31 was applied as positive control for normalization of the response. Synthesis and Activity Testing of the (A) Screening of SKA-74 derivatives. (B) Screening of SKA-75 and SKA-76 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical structures are shown below the bar graphs. KCa Channel Activator Design Shim et al. FIGURE 4 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were performed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline additions, and then SKA-31 was applied as positive control for normalization of the response. (A) Screening of SKA-74 derivatives. (B) Screening of SKA-75 and SKA-76 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical structures are shown below the bar graphs. FIGURE 4 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were performed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline additions, and then SKA-31 was applied as positive control for normalization of the response. (A) Screening of SKA-74 derivatives. (B) Screening of SKA-75 and SKA-76 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical structures are shown below the bar graphs. FIGURE 4 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were performed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline additions, and then SKA-31 was applied as positive control for normalization of the response. (A) Screening of SKA-74 derivatives. (B) Screening of SKA-75 and SKA-76 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical structures are shown below the bar graphs. Frontiers in Pharmacology \| www.frontiersin.org September 2019 \| Volume 10 \| Article 972 15 KCa Channel Activator Design Shim et al. The Full-Length KCa3.1 Cryo-EM Structure Reveals That the C-Terminal CaM-BD/ CaM Dimer Crystal Is an Artefact y In our structure-based drug design attempt described above, we had used the C-terminal CaM-BD/CaM crystal structure, which consist of two vertically orientated CaM molecules and two horizontal KCa2.2 C-terminal fragments in an antiparallel arrangement (Zhang et al., 2012). In this structure, the CaM N-lobe interacts with the C-terminal region of the CaM-BD, whereas the C-lobe is bound to N-terminal region of the CaM-BD. Several KCa channel activators, EBIO (Zhang et al., 2012), NS309 (Zhang et al., 2013) and, most recently in a publication from the Structural Biology group at Pfizer (Cho et al., 2018), CyPPA and riluzole were shown by X-ray crystallography and solid-state NMR to be located at the interface between the CaM N-lobe and the C-terminal region of the CaM-BD, where we docked our compounds. For clarity, only half of this so-called “dimer of dimers” complex, one CaM and one C-terminal fragment, was used for our modeling (see Supplementary Figure 1). In order to probe this “new” binding site we decided to mutate S181, L185 and the neighboring A184 to smaller, bulkier or charged residues to either disrupt contacts with SKA-111 or disturb the overall shape and size of the S45A helix/CaM N-lobe interface pocket. As described in the Materials and Methods, mutations were first tested for expression and Ca2+ sensitivity and then used for evaluation of SKA-111 sensitivity. Mutating serine 181 to a shorter alanine, reduced SKA-111 potency 7-fold by removing the van der Waals contact to the 5-position –CH3 group of SKA-111 (Figures 6B and 7). Mutating S181 to a charged arginine or large phenylalanine also reduced SKA-111 potency presumably by “pushing” SKA-111 forward. Introducing even larger tryptophan or tyrosine residues had an interesting effect on KCa3.1 gating and resulted in channels that already produced large nA currents in the presence of only 250 nM free Ca2+, suggesting that these mutants are more sensitive to Ca2+ then the wild-type channel. We therefore did not use these two mutants for testing SKA-111 sensitivity. Mutating the other S45A helix residue, L185, that is in direct van der Waals contact with SKA-111 had a similar effect as mutating position 181. Replacing leucine 185 with a smaller alanine reduced SKA-111 potency by 7-fold, while substitution of a larger phenylalanine in 185-position produced an 8-fold reduction in potency (Figures 6B and 7). Synthesis and Activity Testing of the IGURE 5 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were erformed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline dditions, and then SKA-31 was applied as positive control for normalization of the response. (A) Screening of SKA-111 derivatives. (B) Screening of aromatically ubstituted SKA-111 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical tructures are shown below the bar graphs. FIGURE 5 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were performed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline additions, and then SKA-31 was applied as positive control for normalization of the response. (A) Screening of SKA-111 derivatives. (B) Screening of aromatically substituted SKA-111 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical structures are shown below the bar graphs. FIGURE 5 \| Whole-cell KCa3.1 and KCa2.2 responses elicited by 10 µM of activator and normalized to 10 µM of the mixed activator SKA-31. Experiments were performed by automated electrophysiology with 250 nM of free internal Ca2+. In each experiment the new activator was tested first, washed out with 2 saline additions, and then SKA-31 was applied as positive control for normalization of the response. (A) Screening of SKA-111 derivatives. (B) Screening of aromatically substituted SKA-111 derivatives. The bar graphs show means ± SD of the slope conductance measured between －85 and －65 mV (n = 2–7 cells). Chemical structures are shown below the bar graphs. September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 16 16 KCa Channel Activator Design Shim et al. interface between the S45A helix and the CaM N-lobe. Synthesis and Activity Testing of the Similar to our previous docking model, SKA-111 is stabilized in open state 1 by hydrogen bonds involving M51 and E54 and multiple CaM N-lobe residues are located within a 5 Å radius sphere around SKA-111 (F19, I27, L32, M51, I52, E54 V55, I63, F68, M71, M72, R74, K75) (Figure 6A, open state 1). However, on the channel side SKA-111 now makes van der Waals interactions with S181 and L185 in the S45A helix (Figure 6A). Interestingly, in open state 2 the lowest energy docking pose of SKA-111 is “flipped” 180 degrees around the hydrogen bond with CaM M51 and the molecule now makes van der Waals contacts with S181 and A184 in the S45A helix and again multiple residues in the CaM N-lobe (Supplementary Figure 4). The parent compound SKA-31 and the benzoxazole SKA-121 take up similar low energy docking poses in which they make contacts with M51 and E54 in CAM and S181 and/or L185 in KCa3.1 (Supplementary Figure 5). The Full-Length KCa3.1 Cryo-EM Structure Reveals That the C-Terminal CaM-BD/ CaM Dimer Crystal Is an Artefact Introduction of even larger (W, Y) or charged residues unfortunately resulted in non-functional KCa3.1 channels that no longer respond to free Ca2+ concentrations as high as 10 µM (Figure 6B). Although it is not in direct contact with SKA-111 we also mutated the neighboring A184 position to F because our model suggested that a phenylalanine in this position would push the CaM-K75 residue, that is contacting SKA-111 upwards. In keeping with this hypothesis, the A184F mutant was 9-fold less sensitive to SKA-111 than the WT KCa3.1 channel (Figures 6B and 7). g ( pp y g ) Recently, the MacKinnon group (Lee and MacKinnon, 2018) determined the full-length cryo-EM structures of KCa3.1 in the closed and in two activated states (pdb: 6cnm, 6cnn, and 6cno) and revealed that the C-terminal CaM-BD/CaM dimer crystal is an artefact. The full-length structure showed four CaMs per channel tetramer, with the CaM C-lobe of each CaM tightly bound to the CaM-BD of each subunit in the closed and the two activated states (Figure 6A). However, the N-lobes were only clearly visible in the open, Ca2+-bound states and poorly resolved in the closed, Ca2+-free structure suggesting that they are flexible in the absence of Ca2+. When Ca2+ binds to the N-lobe it moves from the bottom of the S2 segment to the bottom of the S4-S5 linker (which in KCa3.1 consists of two helices), while the C-lobe maintains its interaction with the HA and HB helices in the C-terminus. The N-lobe then pulls part of the S4–S5 linker, namely the S45A helix downward and this displacement expands the S6 helices and opens the pore (Lee and MacKinnon, 2018). In their study the MacKinnon group also proposed a new binding pocket for the KCa activator EBIO formed by the S45A helix and the CaM N-lobe, in which EBIO binds to L185 in the S45A linker (Lee and MacKinnon, 2018) instead of L480 in the C-terminal crystal complex (Zhang et al., 2012). However, this very plausible alternative binding site hypothesis was not experimentally tested. Probing the “New” SKA Compound Binding Site at the Interface Between the S45A Helix and the CaM N-Lobe Probing the “New” SKA Compound Binding Site at the Interface Between the S45A Helix and the CaM N-Lobe 45 Based on this new binding site hypothesis we here docked and energy minimized SKA-111 in the interface pocket between the S45A helix and the CaM N-lobe (Figure 6A) using the open state of the full length cryo-EM structure of KCa3.1. The two open cryo-EM structures (6cnn and 6cno) were refined using the Rosetta cryo-EM refinement protocol with cryo-EM density map (Wang et al., 2016). A total of 10,000 models were generated for each state and the model with the lowest energy among the largest clusters of the top 1,000 models was used for RosettaLigand docking of SKA-111 starting from random ligand positions at the In order to more dramatically reduce SKA-111 sensitivity we next generated the double S181A-L185A mutant reasoning that it would disrupt both the van der Waals interactions SKA-111 makes with the S45A helix. In keeping with this idea, the S181A- L185A mutant produced a roughly additive effect when compare with the two single A mutations and reduced SKA-111 potency by 18-fold (Figures 6B and 7). In order to determine that this change in SKA-111 sensitivity is not caused by a reduced Ca2+ September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 17 KCa Channel Activator Design Shim et al. KCa Channel Activator Des him et al. ensitivity we performed in-side out recordings comparing he Ca2+ sensitivity of the S181A-L185A mutant with the WT sensitivity (Figure 7C). As a further control experiment, w tested the sensitivity of a double “fenestration” mutation, T212 FIGURE 6 \| Docking model of SKA-111 in the full-length KCa3.1 structure and mutational strategy. (A) Bottom and side view of the full-length KCa3.1 cryo-EM structure following Rosetta refinement in the Ca2+ free closed state (pdb: 6cnm) and Open state I (pdb: 6cnn). The channel is shown in gray, the CaM C-lobe in purple and the CaM N-lobe in yellow. Next to open state 1 we show a zoom out of the lowest energy docking pose of SKA-111 in the interface between the CaM N-lobe and the S45A helix interface. Hydrogen bonds are shown in black, van der Waals interactions are visualized in purple. For clarity, not all side chains of CaM residues within contact range of SKA-111 are explicitly shown. Please note the channel residues S181, A184 and L185. (B) Alignment of the S45A helix sequence in KCa2.1, KCa2.2, KCa2.3 and KCa3.1. Probing the “New” SKA Compound Binding Site at the Interface Between the S45A Helix and the CaM N-Lobe Whole-cell KCa3.1 currents were elicited by voltage-ramps from －120 to + 40 mV with an intracellular free calcium concentration of 250 nM. Data points are mean ± S.D. from 3–5 independent cells/recordings. The reported P values are for an extra sum-of-squares F test (GraphPad Prism5; GraphPad Software, La Jolla, CA) to compare the curves of KCa3.1 mutants to WT. (C) Inside-out calcium concentration- response curves for WT KCa3.1 (EC50 = 437 nM, 95% CI: 353–521 nM, nH = 1.98) and the S181A-L185A double mutant (EC50 = 392 nM, 95% CI: 287-497 nM, nH = 1.39, P = 0.4951). Data points are the mean ± S.D. from 3–5 independent recordings. The calcium-sensitivity of the mutant is statistically not different from the WT KCa3.1 channel (P = 0.4951 in extra sum-of-squares F test). FIGURE 7 \| Mutations of S181 and L185 in the S45A helix disturb SKA-111 activity but not calcium gating. (A) Representative whole-cell WT and mutant KCa3.1 currents with an intracellular free calcium concentration of 250 nM in the presence and absence of SKA-111. (B) Concentration–response for SKA-111 induced current activation: WT (EC50 = 146 nM, 95% CI: 99–193 nM), T212F-V272F (EC50 = 153 nM, 95% CI: 114–182 nM, P = 0.1017), S181A (EC50 = 1.012 µM, 95% CI: 0.780–1.244 µM, P < 0.0001), A184F (EC50 = 1.326 µM, 95% CI: 1.205–1.447 µM, P < 0.0001), L185A (EC50 = 0.993 µM, 95% CI: 0.903–1.083 µM, P < 0.0001), S181A-L185A (EC50 = 2.654 µM, 95% CI: 2.619–2.624 µM, P < 0.0001). Whole-cell KCa3.1 currents were elicited by voltage-ramps from －120 to + 40 mV with an intracellular free calcium concentration of 250 nM. Data points are mean ± S.D. from 3–5 independent cells/recordings. The reported P values are for an extra sum-of-squares F test (GraphPad Prism5; GraphPad Software, La Jolla, CA) to compare the curves of KCa3.1 mutants to WT. (C) Inside-out calcium concentration- response curves for WT KCa3.1 (EC50 = 437 nM, 95% CI: 353–521 nM, nH = 1.98) and the S181A-L185A double mutant (EC50 = 392 nM, 95% CI: 287-497 nM, nH = 1.39, P = 0.4951). Data points are the mean ± S.D. from 3–5 independent recordings. The calcium-sensitivity of the mutant is statistically not different from the WT KCa3.1 channel (P = 0.4951 in extra sum-of-squares F test). C-terminal HC helix, which forms a coiled coil located at the center of the channel. Probing the “New” SKA Compound Binding Site at the Interface Between the S45A Helix and the CaM N-Lobe Residues that were mutated are highlighted in yellow and the EC50 values for SKA-111 shown next to each mutant (for confidence intervals see Figures 7 and 8). FIGURE 6 \| Docking model of SKA-111 in the full-length KCa3.1 structure and mutational strategy. (A) Bottom and side view of the full-length KCa3.1 cryo-EM structure following Rosetta refinement in the Ca2+ free closed state (pdb: 6cnm) and Open state I (pdb: 6cnn). The channel is shown in gray, the CaM C-lobe in purple and the CaM N-lobe in yellow. Next to open state 1 we show a zoom out of the lowest energy docking pose of SKA-111 in the interface between the CaM N-lobe and the S45A helix interface. Hydrogen bonds are shown in black, van der Waals interactions are visualized in purple. For clarity, not all side chains of CaM residues within contact range of SKA-111 are explicitly shown. Please note the channel residues S181, A184 and L185. (B) Alignment of the S45A helix sequence in KCa2.1, KCa2.2, KCa2.3 and KCa3.1. Residues that were mutated are highlighted in yellow and the EC50 values for SKA-111 shown next to each mutant (for confidence intervals see Figures 7 and 8). sensitivity we performed in-side out recordings comparing the Ca2+ sensitivity of the S181A-L185A mutant with the WT KCa3.1 channel and found that mutant did not differ in its Ca2+ sensitivity (Figure 7C). As a further control experiment, we tested the sensitivity of a double “fenestration” mutation, T212F- V272F, to SKA-111. We had previously generated this mutation, September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 18 KCa Channel Activator Design Shim et al. FIGURE 7 \| Mutations of S181 and L185 in the S45A helix disturb SKA-111 activity but not calcium gating. (A) Representative whole-cell WT and mutant KCa3.1 currents with an intracellular free calcium concentration of 250 nM in the presence and absence of SKA-111. (B) Concentration–response for SKA-111 induced current activation: WT (EC50 = 146 nM, 95% CI: 99–193 nM), T212F-V272F (EC50 = 153 nM, 95% CI: 114–182 nM, P = 0.1017), S181A (EC50 = 1.012 µM, 95% CI: 0.780–1.244 µM, P < 0.0001), A184F (EC50 = 1.326 µM, 95% CI: 1.205–1.447 µM, P < 0.0001), L185A (EC50 = 0.993 µM, 95% CI: 0.903–1.083 µM, P < 0.0001), S181A-L185A (EC50 = 2.654 µM, 95% CI: 2.619–2.624 µM, P < 0.0001). Probing the “New” SKA Compound Binding Site at the Interface Between the S45A Helix and the CaM N-Lobe As shown in Figure 8A, S372 is facing outward from this coil towards N42 in the CaM N-lobe and accepting a hydrogen bond from its NH2 group. Mutating S372 to a larger, charged arginine (Figure 8B) in our Rosetta model of open state I “pushes” the outer edge of the CaM N-lobe 6.4 Å downwards, now resulting in a hydrogen bond with L39 in the CaM N-lobe and changing the shape and the volume of the interface pocket (see Figure 8C for an overlay of the WT and mutant channel). In keeping with this “distortion” of the interface pocket SKA-111 does not converge in the S372R mutant, while all top 50 lowest energy-binding poses virtually overlay in the WT channel pocket (Supplementary Figure 6). Interestingly, introducing the double AA mutation (S181A-L185A) into the S372R mutant, resulted only in small, but not statistically significant additional shift in the concentration-response curve of SKA-111 (Figure 8D). which closes the KCa3.1 fenestration between S5 and S6 with two bulky aromatic residues without changing the biophysical properties of the channel when identifying the binding site of the dihydropyridine nifedipine in KCa3.1 (Nguyen et al., 2017). As expected, the double “fenestration” mutant was as sensitive to SKA-111 as the WT channel (Figure 7). Why Did the S372R Mutation in Our Previous Study Have Such a Large Effect of KCa Activator Potency? Taken together the above presented mutagenesis data suggest that the KCa channel activator SKA-111 is indeed binding in the interface between the S45A helix and the CaM N-lobe as hypothesized by Lee et al. (Lee and MacKinnon, 2018). However, we were still puzzled by the fact that we had previously seen such a strong reduction in potency for multiple KCa channel activators including SKA-31, SKA-111, EBIO and NS309, when mutating S372 in KCa3.1 or the corresponding S632 residue in KCa2.3 to arginine (Brown et al., 2017). Based on the C-terminal crystal fragment we had believed this residue to be located at the back of the interfacial binding site pocket. In the full-length KCa3.1 structure S372 is located at the top of the DISCUSSION We here attempted to perform structure assisted design of KCa2.2 selective small molecule activators using the KCa2.2 CaM-BD/CaM September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 19 Shim et al. KCa Channel Activator Design FIGURE 8 \| Rosetta Model of the wild type KCa3.1 channel in open state 1 with serine in position 372 (A) and of the S372R mutant (B). Hydrogen bonds are visualized in black. (C) Overlay of the Rosetta models of the S45A helix/CaM N-lobe interface pocket in the wild type (yellow) and the S372R mutant shown in space fill. (D) Concentration-response curve for SKA-111 induced current activation: WT (EC50 = 146 nM, 95% CI: 99–193 nM), S372R (EC50 = 6.860 µM, 95% CI: 6.788–6.932 µM, P < 0.0001), S181A-L185A-S372R (EC50 = 8.876 µM, 95% CI: 8.652–9.100 µM, P < 0.0001). Data points are mean ± S.D. from 3–5 independent cells/recordings. FIGURE 8 \| Rosetta Model of the wild type KCa3.1 channel in open state 1 with serine in position 372 (A) and of the S372R mutant (B). Hydrogen bonds are visualized in black. (C) Overlay of the Rosetta models of the S45A helix/CaM N-lobe interface pocket in the wild type (yellow) and the S372R mutant shown in space fill. (D) Concentration-response curve for SKA-111 induced current activation: WT (EC50 = 146 nM, 95% CI: 99–193 nM), S372R (EC50 = 6.860 µM, 95% CI: 6.788–6.932 µM, P < 0.0001), S181A-L185A-S372R (EC50 = 8.876 µM, 95% CI: 8.652–9.100 µM, P < 0.0001). Data points are mean ± S.D. from 3–5 independent cells/recordings. FIGURE 8 \| Rosetta Model of the wild type KCa3.1 channel in open state 1 with serine in position 372 (A) and of the S372R mutant (B). Hydrogen bonds are visualized in black. (C) Overlay of the Rosetta models of the S45A helix/CaM N-lobe interface pocket in the wild type (yellow) and the S372R mutant shown in space fill. (D) Concentration-response curve for SKA-111 induced current activation: WT (EC50 = 146 nM, 95% CI: 99–193 nM), S372R (EC50 = 6.860 µM, 95% CI: 6.788–6.932 µM, P < 0.0001), S181A-L185A-S372R (EC50 = 8.876 µM, 95% CI: 8.652–9.100 µM, P < 0.0001). Data points are mean ± S.D. from 3–5 independent cells/recordings. Frontiers in Pharmacology \| www.frontiersin.org DISCUSSION was afterwards repeatedly observed in structural studies addressing the mechanism of action of small molecule KCa channel activators and of PIP2 on KCa2.2 channel function (Zhang et al., 2012; Zhang et al., 2013; Zhang et al., 2014). Zhang et al. showed in several high- resolution X-ray structures that the KCa activator EBIO, the more potent NS309, as well as several NS309 derivatives bind in the interface between the CaM N-lobe and the KCa2.2 CaM-BD (Zhang et al., 2012; Zhang et al., 2013; Nam et al., 2017). Independently of this work a group at Pfizer also crystallized the KCa2.2 CaM-BD/ CaM complex and showed the presence of two other KCa activators, riluzole and a CyPPA derivatives, in the interface between the CaM N-lobe and the channel CaM-BD (Cho et al., 2018). This interaction crystal structure. This structure seemed extremely attractive for a structure based approach because it had been repeatedly crystalized by multiple groups. The first structure of the KCa2.2 channel C-terminal calmodulin binding domain (CaM-BD) in complex with CaM was reported in 2001 at 1.6 Å resolution and showed an elongated, anti-parallel dimer of two KCa2.2 C-terminal 76 amino acid long fragments. On each of these channel pieces CaM was tightly bound with its C-lobe to two alpha helices connected by a turn from the same channel subunit and “grabbed” the free end of the fragment from the other subunit in the dimer with its N-lobe suggesting that CaM-BD dimerization might gate KCa2 channels (Schumacher et al., 2001). The same dimeric crystal orientation September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 20 KCa Channel Activator Design Shim et al. was further confirmed by solution state NMR experiments demonstrating that riluzole and the CyPPA analog perturb the chemical shifts for both 15N-labeled CaM and 15N-labeled KCa2.2 CaM-BD fragments (Cho et al., 2018). Moreover, our own previous molecular modeling studies using the KCa2.2 CaM-BD/CaM crystal structure to generate a homology model of the analogous binding pocket for the related KCa3.1 channel in order to explain the KCa3.1 selectivity of our KCa activators SKA-111 and SKA-121 (Coleman et al., 2014), seemed in good agreement with experimental mutagenesis and structure-activity relationship observations for the SKA-type activators (Brown et al., 2017). DISCUSSION As described above we were therefore very astonished when our computationally designed SKA molecules, which based on the Rosetta modeling work should have been both KCa2.2 selective and potent, showed such disappointing activity in electrophysiological experiments. By demonstrating that the C-terminal CaM-BD/CaM crystal is an artefact the full-length KCa3.1 structure (Lee and MacKinnon, 2018) explained this conundrum for us. The study also provided us with a very testable structural hypothesis for an alternative KCa activator binding site in between the highly flexible CaM N-lobe and the S45A helix in the unusually long S4-S5 linker of KCa3.1. We here tested this hypothesis from the MacKinnon laboratory and found that at least our KCa channel activator SKA-111 is indeed binding in the interface pocket between the S45A helix and the CaM N-lobe. Interestingly, while the channel residues implicated in SKA-111 binding, namely S181, A184 and L185 (Figure 9A), are now different, the residues in the CaM N-lobe (F19, I27, L32, M51, E54, M71, and K75) are the same residues as in our previous FIGURE 9 \| Interactions between SKA-111 and residues in the KCa3.1 S45A helix (A) and the CaM N-lobe (B) interface pocket. Hydrogen bonds are indicated by black lines, van der Waals interactions are indicated by purple lines. (C) The KCa3.1-CaM channel complex seen from the intracellular side. The S45A helix/CaM N-lobe interface colored dark blue. Two other, potential binding sites for small molecules are highlighted with orange or sky-blue circles. The channel is shown in gray, the CaM C-lobe in purple and the CaM N-lobe in yellow. FIGURE 9 \| Interactions between SKA-111 and residues in the KCa3.1 S45A helix (A) and the CaM N-lobe (B) interface pocket. Hydrogen bonds are indicated by black lines, van der Waals interactions are indicated by purple lines. (C) The KCa3.1-CaM channel complex seen from the intracellular side. The S45A helix/CaM N-lobe interface colored dark blue. Two other, potential binding sites for small molecules are highlighted with orange or sky-blue circles. The channel is shown in gray, the CaM C-lobe in purple and the CaM N-lobe in yellow. FIGURE 9 \| Interactions between SKA-111 and residues in the KCa3.1 S45A helix (A) and the CaM N-lobe (B) interface pocket. Hydrogen bonds are indicated by black lines, van der Waals interactions are indicated by purple lines. (C) The KCa3.1-CaM channel complex seen from the intracellular side. The S45A helix/CaM N-lobe interface colored dark blue. DATA AVAILABILITY The physical data of all synthesized compounds are provided in the Method section of this article. NMR spectra are available on request. Protein Data Bank (pdb) format files of the Rosetta models of KCa3.1 open state 1 and open state 2 with SKA-111 docked in the interface between S45A helix and the CaM N-lobe are provided in footnote 1; pdb files of all other models are available upon request. SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar.2019.00972/ full#supplementary-material modeling and design. J. Chem. Theory Comput. 13, 3031–3048. doi: 10.1021/ acs.jctc.7b00125 Bender, B. J., Cisneros, A., 3rd, Duran, A. M., Finn, J. A., Fu, D., Lokits, A. D., et al. (2016). Protocols for molecular modeling with rosetta3 and rosettascripts. Biochemistry 55, 4748–4763. doi: 10.1021/acs.biochem.6b00444 Brown, B. M., Shim, H., Zhang, M., Yarov-Yarovoy, V., and Wulff, H. (2017). Structural dDeterminants for the selectivity of the positive KCa3.1 gating modulator ACKNOWLEDGMENTS We gratefully acknowledge use of OMEGA version 2.5.1.4 OpenEye Scientific Software, Santa Fe, NM. http://www. eyesopen.com FUNDING This work was supported by the CounterACT Program, National Institutes of Health Office of the Director [U54NS079202], and the National Institute of Neurological Disorders and Stroke [R21NS 101876]. BB was supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through grant number UL1 TR001860 and linked award TL1 TR001861. This work was supported by the CounterACT Program, National Institutes of Health Office of the Director [U54NS079202], and the National Institute of Neurological Disorders and Stroke [R21NS 101876]. BB was supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through grant number UL1 TR001860 and linked award TL1 TR001861. i y g g Based on the KCa3.1 structural model we would further like to hypothesize here that the cytoplasmic facing surface of KCa3.1 and the closely related KCa2 channels offer multiple binding pockets for positive and potentially also negative gating modulators. As visualized in Figure 9C in dark blue, the S45A helix/CaM N-lobe interface is present four times in the complex consisting of four KCa3.1 α-subunits and four CaM molecules. How many of these pockets can be occupied by a small molecule activator is currently unknown and we would like to posit that it is possible that the occupancy of these four pockets might differ between different activators molecules and different KCa channel subtypes. In addition to the S45A helix/CaM N-lobe interface, the SiteMap function of the Schrödinger Glide software identifies two other potential druggable sites in the cytoplasmic surface of KCa3.1 that AUTHOR CONTRIBUTIONS Participated in research design: HS, BB, VY-Y, HW. Performed chemical synthesis and compound analysis: HS, LS, VS. Conducted patch-clamp recordings and data analysis: BB. Performed molecular modeling and related data analysis: HS, VY-Y. Determined crystal structures: JF. Wrote or contributed to the writing of the manuscript: BB, HS, VY-Y, HW. Participated in research design: HS, BB, VY-Y, HW. Performed chemical synthesis and compound analysis: HS, LS, VS. Conducted patch-clamp recordings and data analysis: BB. Performed molecular modeling and related data analysis: HS, VY-Y. Determined crystal structures: JF. Wrote or contributed to the writing of the manuscript: BB, HS, VY-Y, HW. We believe we have now identified the “correct” binding site for SKA-111 and related SKA compounds. However, since we only used mutagenesis in combination with functional electrophysiological recordings to confirm the binding site, there remains the caveat that we could have observed allosteric effects. It will therefore be important to perform additional experiments in future that demonstrate binding to the S45A helix for example with a photo- affinity probe and to directly measure ligand binding affinities using surface plasma resonance, isothermal titration calorimetry or differential scanning fluorimetry. These experiments should be performed with physiological ligand concentrations and not with saturating concentrations as was the case in the above described crystallography and solution state NMR studies in order to avoid the identification of “non-physiological relevant” binding sites. DISCUSSION Two other, potential binding sites for small molecules are highlighted with orange or sky-blue circles. The channel is shown in gray, the CaM C-lobe in purple and the CaM N-lobe in yellow. Frontiers in Pharmacology \| www.frontiersin.org September 2019 \| Volume 10 \| Article 972 21 KCa Channel Activator Design Shim et al. could accommodate small molecules: one adjacent site located between the S45B helix and the HA helix in the C-terminus (orange circle in Figure 9C) and another site in the space between the CaM C-lobe and the HB helix in the C-terminus (sky blue circle in Figure 9C). Additional experimental and structural work will be necessary to determine if any, and how many, of these sites are targeted by positive or negative KCa channel gating modulators. binding pocket (Figure 9B and Supplementary Figure 1). In both the “old” and the new binding pocket the amino group of the benzothiazole ring of SKA-111 forms two hydrogen bonds with M51 and E54 in the CaM N-lobe and makes nine van der Waals contacts with additional CaM residues. We suspect that these multiple, strong interactions with CaM are responsible for the fact that the benzothiazole riluzole and the structurally related EBIO and NS309 could be soaked into the C-terminal crystal where the highly flexible N-lobe of CaM had “grabbed” the only available part of the C-terminal helix that was not already occupied by the C-lobe. And as chance would have it the sequence in this part of the KCa2.2 C-terminus also contains a serine and leucin, similarly spaced apart, creating a very similar environment on the channel side for additional van der Waals contacts. The fact, that the majority of the contacts that SKA-111 is making in the interface pocket between the S45A helix and the CaM N-lobe are with CaM are also in line with our findings that we could significantly alter SKA-111 potency by mutating KCa3.1 residues in the S45A helix but not render KCa3.1 completely insensitive to its activating activity without rendering the channel non-functional through more drastic mutations.i Adelman, J. P., Maylie, J., and Sah, P. (2012). Small-conductance Ca2+-activated K+ channels: form and function. Ann. Rev. Physiol. 74, 245–269. doi: 10.1146/ annurev-physiol-020911-153336 Alford, R. F., Leaver-Fay, A., Jeliazkov, J. R., O’Meara, M. J., DiMaio, F. P., Park, H., et al. (2017). The rosetta all-atom energy function for macromolecular REFERENCES Structure of the gating domain of a Ca2+-activated K+ channel complexed with Ca2+/calmodulin. Nature 410, 1120–1124. doi: 10.1038/35074145 Coleman, N., Brown, B. M., Olivan-Viguera, A., Singh, V., Olmstead, M. M., Valero, M. S., et al. (2014). New positive Ca2+-activated K+ channel gating modulators with selectivity for KCa3.1. Mol. Pharmacol. 86, 342–357. doi: 10.1124/mol.114.093286 Shakkottai, V. G., do Carmo Costa, M., Dell’Orco, J. M., Sankaranarayanan, A., Wulff, H., and Paulson, H. L. (2011). Early changes in cerebellar physiology accompany motor dysfunction in the polyglutamine disease spinocerebellar ataxia type 3. J. Neurosci. 31, 13002–13014. doi: 10.1523/JNEUROSCI.2789-11.2011 Coleman, N., Nguyen, H. M., Cao, Z., Brown, B. M., Jenkins, D. P., Zolkowska, D., et al. (2015). The riluzole derivative 2-amino-6-trifluoromethylthio- benzothiazole (SKA-19), a mixed KCa2 activator and NaV blocker, is a potent novel anticonvulsant. Neurotherapeutics 12, 234–249. doi: 10.1007/ s13311-014-0305-y Sheldrick, G. M. (2008). A short history of SHELX. Acta Crystallogr., Section A 64, 112–121. doi: 10.1107/S0108767307043930 Strobaek, D., Teuber, L., Jorgensen, T. D., Ahring, P. K., Kjaer, K., Hansen, R. S., et al. (2004). Activation of human IK and SK Ca2+-activated K+ channels by NS309 (6,7-dichloro-1H-indole-2,3-dione 3-oxime). Biochim. Biophys. Acta 1665, 1–5. doi: 10.1016/j.bbamem.2004.07.006 Davis, I. W., and Baker, D. (2009). RosettaLigand docking with full ligand and receptor flexibility. J. Mol. Biol. 385, 381–392. doi: 10.1016/j. jmb.2008.11.010 Eil, R., Vodnala, S. K., Clever, D., Klebanoff, C. A., Sukumar, M., Pan, J. H., et al. (2016). Ionic immune suppression within the tumour microenvironment limits T cell effector function. Nature 537, 539–543. doi: 10.1038/ nature19364 Wang, R. Y., Song, Y., Barad, B. A., Cheng, Y., Fraser, J. S., and DiMaio, F. (2016). Automated structure refinement of macromolecular assemblies from cryo-EM maps using Rosetta. Elife 5, e17219. doi: 10.7554/eLife.17219f Wulff, H., Gutman, G. A., Cahalan, M. D., and Chandy, K. G. (2001). Delination of the clotrimazole/TRAM-34 binding site on the intermediate conductance calcium-activated potassium channel IKCa1. J. Biol. Chem. 276, 32040–32045. doi: 10.1074/jbc.M105231200 Fanger, C. M., Ghanshani, S., Logsdon, N. J., Rauer, H., Kalman, K., Zhou, J., et al. (1999). Calmodulin mediates calcium-dependent activation of the intermediate conductance KCa channel, IKCa1. J. Biol. Chem. 274, 5746–5754. doi: 10.1074/jbc.274.9.5746 Wulff, H., and Kohler, R. (2013). Endothelial small-conductance and intermediate- conductance KCa channels: an update on their pharmacology and usefulness as cardiovascular targets. J. Cariovasc. Pharmacol. 61, 102–112. doi: 10.1097/ FJC.0b013e318279ba20f Hawkins, P. C., and Nicholls, A. (2012). REFERENCES Adelman, J. P., Maylie, J., and Sah, P. (2012). Small-conductance Ca2+-activated K+ channels: form and function. Ann. Rev. Physiol. 74, 245–269. doi: 10.1146/ annurev-physiol-020911-153336 Alford, R. F., Leaver-Fay, A., Jeliazkov, J. R., O’Meara, M. J., DiMaio, F. P., Park, H., et al. (2017). The rosetta all-atom energy function for macromolecular Adelman, J. P., Maylie, J., and Sah, P. (2012). Small-conductance Ca2+-activated K+ channels: form and function. Ann. Rev. Physiol. 74, 245–269. doi: 10.1146/ annurev-physiol-020911-153336 Alford, R. F., Leaver-Fay, A., Jeliazkov, J. R., O’Meara, M. J., DiMaio, F. P., Park, H., et al. (2017). The rosetta all-atom energy function for macromolecular September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 22 Shim et al. KCa Channel Activator Design 5-methylnaphtho[2,1-d]oxazol-2-amine (SKA-121). Mol. Pharmacol. 92, 469–480. doi: 10.1124/mol.117.109421 OEChem v OpenEye Scientific Software, Inc. http://wwweyesopencom Santa Fe, NM, USA. f Chandy, K. G., and Norton, R. S. (2016). Immunology: channelling potassium to fight cancer. Nature 537, 497–499. doi: 10.1038/nature19467 Radtke, J., Schmidt, K., Wulff, H., Kohler, R., and de Wit, C. (2013). Activation of KCa3.1 by SKA-31 induces arteriolar dilation and lowers blood pressure in normo- and hypertensive connexin40-deficient mice. Br. J. Pharmacol. 170, 293–303. doi: 10.1111/bph.12267 Chimote, A. A., Balajthy, A., Arnold, M. J., Newton, H. S., Hajdu, P., Qualtieri, J., et al. (2018). A defect in KCa3.1 channel activity limits the ability of CD8+ T cells from cancer patients to infiltrate an adenosine-rich microenvironment. Sci. Signal 11: (527), eaaq1616. doi: 10.1126/scisignal.aaq1616 Rohl, C. A., Strauss, C. E., Misura, K. M., and Baker, D. (2004). Protein structure prediction using Rosetta. Methods Enzymol. 383, 66–93. doi: 10.1016/ S0076-6879(04)83004-0 i Sci. Signal 11: (527), eaaq1616. doi: 10.1126/scisignal.aaq1616 Cho, L. T., Alexandrou, A. J., Torella, R., Knafels, J., Hobbs, J., Taylor, T., et al. (2018). An intracellular allosteric modulator binding pocket in SK2 ion channels is shared by multiple chemotypes. Structure 26, 533–544 e533. doi: 10.1016/j.str.2018.02.017 Sankaranarayanan, A., Raman, G., Busch, C., Schultz, T., Zimin, P. I., Hoyer, J., et al. (2009). Naphtho[1,2-d]thiazol-2-ylamine (SKA-31), a new activator of KCa2 and KCa3.1 potassium channels, potentiates the endothelium-derived hyperpolarizing factor response and lowers blood pressure. Mol. Pharmacol. 75, 281–295. doi: 10.1124/mol.108.051425 Christophersen, P., and Wulff, H. (2015). Pharmacological gating modulation of small- and intermediate-conductance Ca2+-activated K+ channels (KCa2.x and KCa3.1). Channels (Austin) 9, 336–343. doi: 10.1080/19336950.2015.1071748 Schumacher, M. A., Rivard, A. F., Bachinger, H. P., and Adelman, J. P. (2001). REFERENCES Conformer generation with OMEGA: learning from the data set and the analysis of failures. J. Chem. Inf. Model 52, 2919–2936. doi: 10.1021/ci300314k Hawkins, P. C., Skillman, A. G., Warren, G. L., Ellingson, B. A., and Stahl, M. T. (2010). Conformer generation with OMEGA: algorithm and validation using high quality structures from the Protein databank and cambridge structural database. J. Chem. Inf. Model 50, 572–584. doi: 10.1021/ci100031x Wulff, H., and Zhorov, B. S. (2008). K+ channel modulators for the treatment of neurological disorders and autoimmune diseases. Chem. Rev. 108, 1744–1773. doi: 10.1021/cr078234p Xia, X. M., Fakler, B., Rivard, A., Wayman, G., Johnson-Pais, T., Keen, J. E., et al. (1998). Mechanism of calcium gating in small-conductance calcium-activated potassium channels. Nature 395, 503–507. doi: 10.1038/26758 Jenkins, D. P., Yu, W., Brown, B. M., Lojkner, L. D., and Wulff, H. (2013). Development of a QPatch automated electrophysiology assay for identifying KCa3.1 inhibitors and activators. ASSAY Drug Dev. Technol. 11, 551–560. doi: 10.1089/adt.2013.543 Zhang, M., Meng, X. Y., Cui, M., Pascal, J. M., Logothetis, D. E., and Zhang, J. F. (2014). Selective phosphorylation modulates the PIP2 sensitivity of the CaM-SK channel complex. Nat. Chem. Biol. 10, 753–759. doi: 10.1038/ nchembio.1592 Kaczmarek, L. K., Aldrich, R. W., Chandy, K. G., Grissmer, S., Wei, A. D., and Wulff, H. (2017). International union of basic and clinical pharmacology. c. nomenclature and properties of calcium-activated and sodium-activated potassium channels. Pharmacol. Rev. 69, 1–11. doi: 10.1124/pr.116.012864 Zhang, M., Pascal, J. M., Schumann, M., Armen, R. S., and Zhang, J. F. (2012). Identification of the functional binding pocket for compounds targeting small- conductance Ca2+-activated potassium channels. Nat. Commun. 3, 1021. doi: 10.1038/ncomms2017 Kasumu, A. W., Hougaard, C., Rode, F., Jacobsen, T. A., Sabatier, J. M., Eriksen, B. L., et al. (2012). Selective positive modulator of calcium-activated potassium channels exerts beneficial effects in a mouse model of spinocerebellar ataxia type 2. Chem. Biol. 19, 1340–1353. doi: 10.1016/j.chembiol.2012.07.013 Zhang, M., Pascal, J. M., and Zhang, J. F. (2013). Unstructured to structured transition of an intrinsically disordered protein peptide in coupling Ca2+- sensing and SK channel activation. Proc. Natl. Acad. Sci. U.S.A. 110, 4828–4833 doi: 10.1073/pnas.1220253110. Lee, C. H., and MacKinnon, R. (2018). Activation mechanism of a human SK-calmodulin channel complex elucidated by cryo-EM structures. Science 360, 508–513. doi: 10.1126/science.aas9466 Logsdon, N. J., Kang, J., Togo, J. A., Christian, E. P., and Aiyar, J. (1997). REFERENCES A novel gene, hKCa4, encodes the calcium-activated potassium channel in human T lymphocytes. J. Biol. Chem. 272, 32723–32726. doi: 10.1074/jbc.272.52.32723 Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationship that could be constructed as a potential conflict of interest. Meiler, J., and Baker, D. (2006). Rosettaligand: protein-small molecule docking with full side-chain flexibility. Proteins 65, 538–548. doi: 10.1002/prot.21086f Copyright © 2019 Shim, Brown, Singh, Singh, Fettinger, Yarov-Yarovoy and Wulff. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. l Nam, Y. W., Orfali, R., Liu, T., Yu, K., Cui, M., Wulff, H., et al. (2017). Structural insights into the potency of SK channel positive modulators. Sci. Rep. 7, 17178. doi: 10.1038/s41598-017-16607-8 Nguyen, H. M., Singh, V., Pressly, B., Jenkins, D. P., Wulff, H., and Yarov-Yarovoy, V. (2017). Structural insights into the atomistic mechanisms of action of small molecule inhibitors targeting the KCa3.1 Channel Pore. Mol. Pharmacol. 91, 392–402. doi: 10.1124/mol.116.108068 September 2019 \| Volume 10 \| Article 972 Frontiers in Pharmacology \| www.frontiersin.org 23
https://openalex.org/W1598277916	https://redined.educacion.gob.es/xmlui/bitstream/11162/98384/1/Edutec-e_n44-Rodriguez-Alvarez.pdf	es		Análisis didáctico de las aulas virtuales. Una investigación en un contexto de educación superior	Edutec	2,013	cc-by	5,270	EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 ANÁLISIS DIDÁCTICO DE LAS AULAS VIRTUALES. UNA INVESTIGACIÓN EN UN CONTEXTO DE EDUCACIÓN SUPERIOR DIDACTIC ANALYSIS OF VIRTUAL CLASSROOM. AN INVESTIGATION IN A CONTEXT OF HIGHER EDUCATION Carlos Rodríguez-Hoyos; rodriguezhc@unican.es Universidad de Cantabria Mª José Álvarez Álvarez; mj.alvarez@unileon.es Universidad de León RESUMEN En este artículo se exponen los principales resultados de una investigación destinada a analizar la utilización didáctica de las aulas virtuales utilizadas en las asignaturas de los diferentes campos científicos de la Universidad de León. El estudio constata que, en estos momentos, la incorporación de la plataforma virtual de esta institución no ha supuesto una modificación cualitativa de las estrategias metodológicas empleadas ni de las prácticas docentes. PALABRAS CLAVE: E-learning, plataformas virtuales, docencia virtual en la educación superior. ABSTRACT This article presents the main results of an investigation to analyze the didactic use of virtual classrooms used in the subjects of different fields of knowledge at the University of León. The study finds that, at present, the incorporation of institutional Learning Management System has not been a qualitative change of the approaches employed or teaching practices. KEY WORDS: E-learning, Learning Management Systems, virtual teaching in higher education. Página 1 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 1. INTRODUCCIÓN La introducción de las Tecnologías de la Información y Comunicación (TIC) en los procesos de educación ha estado presidida por discursos que, de forma mayoritaria, han expuesto las ventajas que trae aparejadas la utilización de estos recursos para la mejora del rendimiento académico, el incremento de la motivación del alumnado, la accesibilidad, la interacción o el abaratamiento de los costes destinados a la formación (Khan, 2001; Morán, 2012), entre otros. Esas mismas expectativas de innovación y cambio se generaron con la aparición de la modalidad de e-learning. Si bien el origen de esta modalidad estuvo muy ligado a las experiencias de formación continua, en estos momentos también se está implantando en los diferentes niveles de los sistemas educativos formales (Pérez-Juárez y Verdú, 2000) y en otros espacios educativos que Dondi (2007) ha calificado como “territorios del e-learning”, en los que se incluyen nuevos campos de educación no formal e informal como, por ejemplo, las comunidades de práctica profesionales. 2.MARCO TEÓRICO Muchas universidades presenciales europeas han incorporado plataformas de educación virtual con la intención de diversificar y ampliar su oferta formativa. Las universidades españolas no se han quedado al margen de esta tendencia, tal y como se deduce de los datos aportados en el informe publicado por la CRUE (2010) relativo al período 2006-2010, en el que se señala que el 97,6% de ellas dispone de una plataforma de docencia virtual institucional. También resulta destacable el alto volumen de utilización de estos entornos por parte del profesorado y del alumnado, ya que el 72,6% del personal docente y el 90,5% de los alumnos hacen uso de estas herramientas educativas telemáticas. En la actualidad ya se han desarrollado varias investigaciones en diferentes países que tratan de analizar los factores clave y los desafíos de los procesos de implementación del elearning en las instituciones de educación superior. En un trabajo de revisión de la literatura sobre los procesos de implementación del e-learning en las universidades del Reino Unido, O’Neill, Singh, y O’Donoghue (2004) evidenciaron que algunos de los factores de éxito de esos procesos estaban relacionados con la dotación en infraestructuras tecnológicas, la experiencia en la utilización de tecnología o la preparación del profesorado para hacer frente a los retos que plantea esa modalidad. Jones y O´Shea (2004) señalaron que para que el e-learning tuviera un mayor impacto en las universidades del Reino Unido sería necesario que éstas reformularan sus estrategias en diferentes áreas como los recursos humanos, la calidad, la financiación, la administración o en su dimensión pedagógica. El análisis desarrollado por Tuparova y Tuparov (2007) sobre el estado del arte del elearning en la educación superior en Bulgaria puso de manifiesto que si bien esta modalidad había crecido durante los últimos años, seguía presentando algunos problemas y desafíos. Entre ellos señalaban aspectos como la escasez de contenidos en algunas de las áreas de conocimiento, la falta de preparación didáctica y tecnológica del profesorado universitario Página 2 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 para hacer frente a los desafíos de esta modalidad o la inexistencia de documentación en los centros educativos que estimulara la utilización de esta modalidad. Chengfeng (2003) analizó las barreras de la implantación del e-learning en China señalando que algunos de los factores que impedían su desarrollo estaban relacionados con aspectos que afectaban al alumnado, al profesorado, a la configuración de los planes de estudio y a otras variables estructurales u organizativas. Por su parte, Aoki (2010) ha reflejado que en la actualidad las instituciones de educación superior japonesas ya han comenzado a implementar lentamente la modalidad de e-learning. Este autor pone de manifiesto que, en la mayoría de los casos, esa integración tecnológica no está suponiendo una verdadera transformación de las prácticas docentes, dado que se reproducen estrategias metodológicas tradicionales, cuya única ventaja es la posibilidad de acceder a los recursos siempre que el alumnado disponga de una conexión a Internet. En España se han desarrollado varias investigaciones que han tratado de analizar cómo se ha desarrollado la oferta formativa a través del e-learning en diferentes universidades españolas. Area (2002) dirigió un trabajo en el que se trató de ofrecer una visión de conjunto sobre su estado en las universidades españolas. Esa investigación puso de manifiesto que las universidades del territorio nacional se encontraban en un proceso de creación de los campus virtuales que favorecería el aumento, a corto y medio plazo, de las acciones formativas desarrolladas en línea por estas instituciones. El estudio dirigido por Infante (2004) evidenció que la principal actividad formativa de las universidades era aquella que se dirigía a apoyar la docencia presencial a través de sistemas virtuales. Entre sus resultados se ponía de manifiesto que alrededor del 71% de los tutores de las experiencias de e-learning revelaban tener poca formación sobre la modalidad en cuestión. Valverde (2004) centró su análisis en la evolución de la oferta formativa de las acciones en línea de las universidades públicas españolas entre los años 2001 y 2004. Ese informe revelaba que ya existía una importante oferta formativa desarrollada por las universidades a través de esta modalidad, en las que se trataba una gran variedad de contenidos. El informe señalaba que junto a la dotación en infraestructuras tecnológicas era necesario elaborar un proyecto pedagógico que pudiera dar coherencia a las prácticas docentes y planes estratégicos. Otras investigaciones que se han orientado a analizar el tipo de prácticas docentes desarrolladas por el profesorado universitario a través de plataformas reflejaron que la utilización de esas tecnologías tiende a reproducir modelos educativos clásicos que se orientan a la transmisión de información más que a la potenciación de la participación del alumnado (Rodríguez Malmierca, 2006); a desarrollar la mayor parte de las asignaturas de forma presencial y usar del e-learning como un complemento para distribuir materiales y realizar algunas actividades (Salinas, 2007); a utilizar los campus virtuales para desarrollar las mismas metodologías y estrategias pedagógicas que en las clases presenciales tradicionales (Cabero, 2010) o a emplear una concepción logocéntrica de la enseñanza (Area, 2008). Página 3 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 Los resultados de estas investigaciones ponen de manifiesto que el e-learning y el blendedlearning se han convertido en los últimos años en modalidades ampliamente utilizadas en las universidades. Estos estudios constatan que muchas de las prácticas docentes desarrolladas a través de esas modalidades siguen reproduciendo modelos pedagógicos tradicionales propios de la formación presencial. La investigación en este ámbito ha señalado que el profesorado tiende a aceptar y utilizar esta modalidad únicamente si es capaz de controlar desde una perspectiva académica aquello que sucede en esos espacios (Hardaker y Singh, 2011). 3. CONTEXTUALIZACIÓN DE LA INVESTIGACIÓN La investigación que presentamos se realizó durante el curso académico 2009/2010 y se contextualiza en la Universidad de León. Esta institución universitaria pública dispone, desde el curso 2006/2007, de una plataforma de educación virtual denominada “aul@unileon” que sirve como soporte tecnológico a la docencia presencial de todas las asignaturas de las titulaciones. Nuestra investigación se dirigió a realizar un análisis, desde una perspectiva didáctica, de una muestra representativa de las aulas virtuales presentes en esa plataforma de los diferentes campos científicos con la intención de conocer cómo se estaban organizando, los tipos de actividades predominantes, los materiales empleados por el profesorado, los usos de las herramientas de comunicación o la propia percepción que el profesorado y el alumnado tenían sobre el aula virtual. Tratamos de investigar si la utilización de las aulas virtuales de la Universidad de León, como complemento a la docencia presencial, había supuesto una innovación únicamente a nivel tecnológico o si verdaderamente había supuesto un replanteamiento general del diseño y desarrollo de los procesos de enseñanza-aprendizaje. El interés de esta investigación se justifica por la inexistencia de trabajos anteriores que hayan profundizado en el análisis de la utilización de este espacio virtual por parte del profesorado y del alumnado de esta universidad. 4. METODOLOGÍA Y OBJETO DE ESTUDIO El diseño metodológico se asentó en un paradigma mixto de investigación, en el que combinamos herramientas de recogida de datos que tradicionalmente se han incluido en el paradigma cuantitativo y otros instrumentos propios del cualitativo. Consideramos que este abordaje combinado nos permitiría realizar un acercamiento complejo a la realidad educativa que pretendimos estudiar al utilizar diferentes fuentes metodológicas (Corbetta, 2007). La diferencia fundamental que se ha establecido entre estos enfoques se refiere a la finalidad última que los guía. Así, hemos pretendido mantener un equilibrio entre la búsqueda de la explicación de las causas del objeto de estudio (propio del paradigma cuantitativo) y la comprensión de la realidad analizada (cualitativo), utilizando las siguientes técnicas de recogida de datos: Página 4 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250  Núm. 44 / Junio 2013 Observación no participante: Durante la primera fase del trabajo de campo se realizó una observación no participante de una muestra de las aulas virtuales de la Universidad de León con la intención de recoger evidencias que nos permitiera definir sus principales características técnicas, organizativas, comunicativas y didácticas, valorar el enfoque curricular predominante en las mismas e identificar el papel desempeñado por los profesores y los estudiantes en ellas. Para registrar los datos obtenidos elaboramos un instrumento que facilitara su recogida sistemática. La parrilla de observación provisional fue revisada y validada por tres profesores expertos en la incorporación de las TIC a la educación de la Universidad de León. Entre las dimensiones recogidas se incluyeron datos sobre la identificación del aula virtual, contenidos, recursos, actividades, organización didáctica, etc.  Cuestionario: Se construyeron dos modelos de cuestionario ad hoc dirigidos a conocer la opinión de los actores implicados, uno para el grupo de estudiantes, inspirado en el empleado por Álvarez y González (2009), y otro para el colectivo de docentes, apoyado en los diseñados por Pavón (2008) y Sánchez et al. (2008). El cuestionario dirigido al alumnado incluyó variables como la experiencia previa en el uso del aula virtual, los lugar/es de acceso, la estructura de los contenidos, los recursos ofertados, etc. El instrumento destinado a los docentes recogía cuestiones como la frecuencia de uso del aula virtual, estrategias de enseñanza-aprendizaje, objetivos del uso de los recursos, etc. Se enviaron telemáticamente a una muestra de 269 profesores y 363 alumnos, seleccionados a través de un muestro aleatorio entre el universo de docentes (897) y de estudiantes (12348) de la Universidad de León (curso 2009/2010). Se recibieron 169 cuestionarios de los 363 enviados al colectivo de estudiantes y 79 cuestionarios, de los 269 remitidos a los docentes. En ambos casos el tamaño se calculó para un nivel de confianza del 95% con un error muestral del 0,5%.  Entrevistas estructuradas: Optamos por la modalidad de pregunta estandarizada y respuesta abierta y fue realizada a una muestra de diez informantes clave, cinco de cada uno de los colectivos, con representación de todos los campos científicos de la citada Universidad (Ciencias Experimentales y de la Salud, Ciencias Sociales y Jurídicas, Humanidades e Ingeniería).  Grupos de discusión: Se realizaron dos grupos de discusión, uno por cada colectivo implicado. El grupo de discusión de los estudiantes estuvo constituido por 5 participantes (1 alumno de Ciencias Sociales, 1 de Ciencias Experimentales, 1 de Ciencias de la Salud, 1 de Ingeniería y 1 de Humanidades) y en el grupo del colectivo docente participaron 7 personas (2 profesores de Ciencias Sociales, 2 de Ciencias Experimentales, 1 de Ciencias de la Salud, 1 de Ingeniería y 1 de Humanidades). Cada grupo de discusión se constituyó mediante un muestreo casual, con sujetos seleccionados como informadores clave basándonos en diferentes aspectos como la representación de los distintos campos científicos, las necesidades específicas de información detectadas en las fases previas del trabajo de campo o que fueran usuarios de la plataforma virtual institucional y tuvieran experiencia suficiente en su uso. Página 5 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 5. RESULTADOS 5.1 Configuración didáctica de las aulas virtuales La observación no participante se realizó en un total de 80 aulas virtuales, seleccionadas mediante un muestreo aleatorio estratificado por campo científico, de entre el universo total de aulas virtuales de la Universidad de León (N=2602). Nº total aulas virtuales Nº aulas virtuales muestra 534 19 982 28 Humanidades 452 13 Ingenierías 634 20 Total: 2602 Total: 80 Campo científico Ciencias Experimentales y de la Salud Ciencias Sociales y Jurídicas Tabla 1. Número de aulas virtuales de la muestra Por lo que respecta al tipo de materiales empleados hay que señalar que en el 77,5% de las aulas virtuales analizadas se identificaron documentos textuales y en un 55% de las mismas existían recursos multimedia. Este porcentaje descendió al 15% en el caso de los recursos audiovisuales y a un 6,3% para los esquemas, gráficos y mapas conceptuales. En el 76,3% de las aulas virtuales pudo identificarse un listado con la bibliografía de la asignatura mientras que en el 51,3%, se localizaron enlaces a páginas web y/o a otros recursos de interés. Las actividades más predominantes fueron aquellas que solicitaban al alumnado la búsqueda de información sobre un tema específico (30,4%) así como las destinadas a la redacción de ensayos u otro tipo de documentos, como portafolios o propuestas didácticas (21,5%). La resolución de ejercicios (17,5%) y la planificación y desarrollo de proyectos (16,3%) fueron actividades que se identificaron con menor frecuencia en el contexto de las aulas virtuales. En un 10% de los espacios analizados también se registraron diferentes debates organizados en torno a un área temática determinada. Con relación a las herramientas de comunicación pudimos constatar que el correo electrónico institucional fue el recurso que se empleó con mayor frecuencia para interactuar en el entorno de las aulas virtuales de la Universidad de León (77,2%). En un 42,5% de las aulas virtuales el tablón de anuncios se empleó por los docentes para informar sobre cambios horarios de las sesiones presenciales de la asignatura, para establecer plazos de entrega de actividades virtuales o para informar de eventos relacionados con la materia. El foro estuvo activado únicamente en el 20% de los entornos virtuales analizados. A pesar de que se habían creado entradas en 16 foros de las aulas virtuales analizadas, solamente en 8 de ellos se observó una participación continuada entre el profesorado y el alumnado. Página 6 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 En el resto, solamente aparecían los mensajes de presentación del profesor, sin que se registraran intervenciones por parte de los estudiantes. Gráfico 1. Frecuencia de uso de los recursos de comunicación Con relación al tipo de actividad planteada, pudimos constatar que el profesorado optó mayoritariamente por realizar trabajos desarrollados individualmente (75,8%), aunque también identificamos algunas prácticas colaborativas organizadas en pequeños grupos compuestos por 3 ó 4 alumnos (21%). Solamente dos asignaturas (una del campo científico de las Ciencias Sociales y otra de las Ciencias de la Salud) organizaron a los estudiantes en grupos grandes para el desarrollo del trabajo telemático. De las 80 aulas virtuales observadas, en 59 de ellas (73,8%) se pudo analizar el tipo de instrumento empleado para proceder a la evaluación del aprendizaje del alumnado. En las 21 aulas restantes, no fue posible conocer el tipo de evaluación, por no aparecer, explícita ni implícitamente, en ninguno de los apartados del aula virtual. En la mayor parte de las aulas virtuales (70%) en las que se pudo identificar la evaluación se optó por un procedimiento de evaluación mediante un examen teórico y/o práctico. En las 24 aulas virtuales restantes, se desarrolló una estrategia de evaluación continua, en la que se contemplaron aspectos tales como la participación en los foros, la entrega de tareas a lo largo del curso, la asistencia a las sesiones presenciales, etc. 5.2 Percepciones de los usuarios La tasa de respuesta a los cuestionarios por parte del alumnado fue del 46, 5% y del 29,4% por parte del colectivo docente. En el caso de las respuestas procedentes del profesorado, el campo científico de Ciencias Sociales fue el más representado (29,3%), seguido de Ciencias Experimentales (25,3%), Ciencias de la Salud (20%), Ingenierías (14,7%) y Humanidades (8%). El ámbito con menor representación fue el de Ciencias Jurídicas (2,7%). Página 7 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 Gráfico 2. Distribución de la tasa de respuesta del profesorado por campo científico En los cuestionarios remitidos por el colectivo de estudiantes, el campo científico de Ciencias Sociales supuso el 33% de las repuestas, mientras que los ámbitos con menor representación fueron el de Ciencias Jurídicas (6%) e Ingenierías (7%). La tasa de respuesta de las titulaciones de Humanidades fue del 9%, mientras que la del ámbito de Ciencias Experimentales y de la Salud tuvieron una tasa de respuesta del 21% y del 24%, respectivamente. Gráfico 3. Distribución de la tasa de respuesta del alumnado por campo científico Por lo que respecta a la valoración general del aula virtual, el 41,3% de los profesores de la Universidad de León que participó en el estudio manifestó que la plataforma institucional le había posibilitado modificar sus estrategias de enseñanza mucho o muchísimo, siendo este aspecto el mejor puntuado dentro de la valoración global del aula virtual. Las estrategias didácticas desarrolladas con mayor frecuencia fueron los trabajos individuales (el 29,9% de los docentes manifestó haberla utilizado siempre o casi siempre) y el estudio de casos (22,7% de uso frecuente –siempre o casi siempre-). Los profesores consideraron que, en términos generales, el aprendizaje de los estudiantes no había mejorado sustancialmente debido al uso de este entorno de educativo (un 10,7% de los docentes estima que este aspecto no se ha beneficiado nada y un 31% considera que ha mejorado sólo algo). Los datos obtenidos mediante las diferentes herramientas Página 8 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 evidenciaron que las experiencias educativas de carácter colaborativo son excepcionales, siendo dominante la utilización de estrategias individuales. En el grupo de discusión en el que participó el alumnado se puso de manifiesto que este colectivo no contemplaba el trabajo colaborativo entre las estrategias metodológicas que pueden emplearse en el aula virtual institucional y que, tal vez debido a su falta de experiencia, mostraran una concepción limitada y negativa de ese tipo de actividades. A la hora de valorar la contribución del aula virtual al seguimiento de las asignaturas y al desarrollo de los procesos de enseñanza-aprendizaje, el alumnado de la Universidad de León señaló que su mayor contribución estaba relacionada con la mejora en la organización y los procesos de comunicación. Un 45,2% de los estudiantes valoraron el aula virtual como un recurso muy útil para organizarse mejor y un 36,9% destacaron su utilidad para incrementar la comunicación con el profesorado implicado en la docencia de la asignatura. Aunque esa opinión recogida en los cuestionarios contrasta con las críticas vertidas en el grupo de discusión del alumnado, dado que las personas participantes manifestaron que uno de los mayores problemas a la hora de comunicarse con el profesorado a través del aula virtual fue el retraso a la hora de remitir las respuestas o la ausencia de las mismas. El 66,2% de los estudiantes consideraron que la plataforma es poco o nada útil para asistir menos a clases mientras que el 47,1% manifestó que le había resultado poco o nada útil para reducir el número de tutorías. Otro de los aspectos valorados de forma mayoritariamente negativa por el alumnado encuestado fue la motivación para seguir las asignaturas desarrolladas a través del aula virtual. El 51,6% manifestó que la plataforma virtual era poco o nada útil para estar más motivado a la hora de proceder al seguimiento de una determinada materia. Por lo que respecta a las percepciones en torno a las utilidades del aula virtual, cabe señalar que ambos colectivos apuntaron que las aplicaciones que se empleaban con mayor frecuencia eran aquellas que facilitaban la distribución de contenidos. En ese sentido, el 81,3% de los docentes manifestó que utilizaba siempre o casi siempre el gestor de contenidos, mientras que el 86,6% de los estudiantes aseveró que “aul@unileon” era una plataforma buena o muy buena para organizar los contenidos de las asignaturas. Estos datos están consonancia con los obtenidos a través de las entrevistas y los grupos de discusión. En las respuestas de los profesores y estudiantes entrevistados aparecieron constantes referencias a que el elemento más importante de los procesos de enseñanzaaprendizaje virtuales desarrollados a través de “aula@unileon” eran los contenidos. Página 9 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 Gráfico 4. Valoración del profesorado de la utilidad de los recursos del aula Con relación a las herramientas de comunicación de la plataforma, los foros fueron el instrumento que el profesorado utilizaba con menor frecuencia. Cabe destacar que un 68% de los docentes manifestó no haberlos utilizado nunca, mientras que sólo el 5,3% hizo uso de ellos siempre o casi siempre durante el curso. Por su parte, el 48% de los docentes encuestados manifestaron que utilizaron el correo electrónico siempre o casi siempre durante el curso para comunicarse con su alumnado. Si analizamos los datos referentes a la utilidad percibida por los estudiantes y a la valoración global de las herramientas de comunicación virtuales, estratificándolos según la frecuencia de uso, podemos observar que el 93,3% de los alumnos que nunca utilizó el foro consideró que era un recurso poco útil para la comunicación y el 98,6% valoró, en términos generales, muy mal esta herramienta. Sin embargo, de entre los discentes de la Universidad de León que manifestaron emplear con bastante frecuencia el foro de discusión, el 66,7% valoró como muy bueno este recurso y el 40% lo consideró bastante útil. En las entrevistas y el grupo de discusión, los docentes que reconocieron haber usado los foros virtuales defendieron que una de las mayores potencialidades didácticas de estas herramientas, si se comparan con los debates presenciales, es que exigen una mayor reflexión por parte de los estudiantes y que facilitan el seguimiento y la evaluación continua del alumnado. Por su parte, el alumnado que participó en el grupo de discusión señaló que el escaso empleo de los foros se debe a las pocas actividades didácticas desarrolladas a través de estos espacios virtuales y al carácter asincrónico de esta herramienta. Página 10 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 6. DISCUSIÓN DE RESULTADOS Tal y como hemos podido constatar en este estudio, en el contexto de “aul@unileon” existe una clara primacía del trabajo individual en detrimento de las actividades de tipo colaborativo. Esta tendencia a realizar actividades individuales en los procesos de enseñanza-aprendizaje virtuales también ha sido puesta de manifiesto en otros trabajos previos destinados a evaluar la docencia universitaria en los entornos virtuales, tanto a nivel nacional (Area, 2008; Sánchez et al., 2008) como internacional (Mandizadeh, Biemans y Mulder, 2008). De hecho, algunos autores han llegado a constatar que la percepción de los docentes indica que las actividades en entornos virtuales exigen una mayor carga de trabajo individual por parte del estudiantes (Peña y Avendaño, 2006). Otra de las conclusiones que pueden extraerse en este trabajo es que a pesar de encontrarnos inmersos en el proceso de convergencia al Espacio Europeo de la Educación Superior, en las aulas virtuales de la Universidad de León que fueron analizadas se observó un claro predominio de la evaluación final en detrimento de estrategias más procesuales. Esta tendencia centrada en los resultados ya ha sido identificada en la mayor parte de los estudios realizados sobre los entornos virtuales de otras Universidades españolas (Pavón, 2008; Sánchez, 2008; Cabero, 2010). Al mismo tiempo, en ninguno de los espacios virtuales analizados existían mecanismos para valorar el grado de satisfacción de los estudiantes con el proceso de enseñanza-aprendizaje. Esta investigación ha puesto de manifiesto la necesidad de evaluar otros de los elementos básicos de cualquier proceso de enseñanzaaprendizaje virtual como los materiales, actividades, contenidos o la propia evaluación. Tomando en consideración los resultados obtenidos en este trabajo no podemos concluir que las herramientas de comunicación presentes en el entorno virtual hayan incrementado o mejorado la comunicación entre las personas que participaron en estas experiencias. A pesar de que se habilitaron diferentes foros y se abrieron algunas líneas de debate en los que proceder al intercambio de ideas, opiniones, etc., no se llegaron a establecer líneas de diálogo entre las personas implicadas en el proceso. Esto nos llevaría a concluir que el uso que se realice de estos instrumentos de comunicación es más determinante que las propias potencialidades pedagógicas que, a priori, presentan y su utilización dependerá, en gran medida, de las estrategias metodológicas diseñadas por los docentes o la forma en la que se favorecen los procesos de comunicación (Pérez i Garcias, 2006). En el contexto de esta investigación hay algunos docentes que manifestaron inquietudes por modificar sus propuestas metodológicas hacia planteamientos curriculares más dialógicos y flexibles. En ese sentido, la plataforma virtual institucional se utiliza en este momento, de forma mayoritaria, como un repositorio para que los docentes pongan a disposición del alumnado los materiales de las diferentes materias, pero no está transformando sustancialmente las prácticas docentes. Tras el análisis e interpretación de los diferentes resultados obtenidos podemos llegar a la conclusión de que la utilización del “aul@unileon” por parte del profesorado tiende a reproducir en el aula virtual una Página 11 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 metodología transmisiva y modelos de comunicación unidireccionales, similares a los empleados tradicionalmente en los procesos de enseñanza-aprendizaje presenciales. Estos aspectos ya han sido apuntados en otros estudios previos sobre docencia virtual en la educación superior (Sahzad, y Khan, 2010; Tirado-Moureta, Pérez-Rodríguez y Aguaded, 2011). 7. BIBLIOGRAFÍA ÁLVAREZ, P. R. y GONZÁLEZ, M. (2009). El aula virtual como recurso para la docencia y tutorización académica del alumnado universitario. En A. García-Valcárcel (Ed.), Experiencias de innovación docente universitaria (pp. 205-211). Salamanca: Ediciones Universidad de Salamanca. AOKI, K. (2010). The Use of ICT and e-Learning in Higher Education in Japan. World Academy of Science, Engineering and Technology, 66, 868-872. AREA, M. (2008). Evaluación del campus virtual de la Universidad de La Laguna. Análisis de las aulas virtuales (periodo 2005-07). Informe final. Recuperado el 03/04/2012 de http://webpages.ull.es/users/manarea/informeudv.pdf AREA, M (Dir.) (2002). Los Campus Virtuales Universitarios en España. Análisis del estado actual. Trabajo presentado en el II Congreso Europeo TIEC, Junio, Barcelona. CABERO, J. (Dir.) (2010). Usos del e-lerarning en las Universidades Andaluzas: estado de la situación y análisis de buenas prácticas (resumen ejecutivo). Sevilla: Grupo de investigación Didáctica de la Universidad de Sevilla. CORBETTA, P. (2007). Los paradigmas de la investigación social. En P. Corbetta, Metodología y Técnicas de Investigación Social (2ª ed.) (pp. 3-31). Madrid: Mc GrawHill. CHENGFENG, J. (2009). Barriers to e-Learning: Literature Review and Analysis. China Distance Education, 11, pp. 45-56. CRUE (2010). Evolución de las TIC en el Sistema Universitario Español: UNIVERSITIC 2010. Madrid: CRUE. Recuperado el 13/04/12 de http://www.crue.org/export/sites/Crue/Publicaciones/Documentos/Universitic/UNI VERSITIC 2010.pdf DONDI, C. (2007). The underground rivers of innovative e-Learning: a preview from the HELIOS Yearly Report 2006/07. e-learning Papers, 4. Recuperado el 11/04/2008 desde http://www.e-learningeuropa.info/files/media/media 12720.pdf HARDAKER, G. y SINGH, G. (2011). The adoption and diffusion of e-learning in UK universities: A comparative case study using Giddens's Theory of Structuration. Campus-Wide Information Systems, 28(4), 221–233. Página 12 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 INFANTE, A. (Dir.) (2004). La Enseñanza Virtual en España ante el Nuevo Espacio Europeo de la Educación Superior. Proyecto EA2004-0090. Recuperado el 24/03/12 de http://www.uhu.es/alfonso.infante/Enlaces/ librodefinitivo.pdf JONES, N. y O’SHEA, J. (2004). Challenging hierarchies: The impact of e-learning. Higher Education, 48, 379-395. Khan, B. (2001): Web-Based Training. New Jersey: Englewood Cliffs. MANDIZADEH, H., BIEMANS, H., y MULDER, M. (2008). Determining factors of the use of elearning environments by university teachers. Computers & Education, 52(3), 142154. MORÁN, L. (2012). Blended-learning. Desafío y oportunidad para la educación actual. Edutec, Revista Electrónica de Tecnología Educativa, 39. Recuperado el 04/04/12 de http://edutec.rediris.es/Revelec2/Revelec39/ blended_learning_desafio_oportunidad_educacion_actual.html O’NEILL, K, SINGH, G. y O’DONOGHUE, J. (2004). Implementing e-learning Programmes for Higher Education: A Review of the Literature. Journal of Information Technology Education, 3, 313-323. PAVÓN, F. (2008). Aulas virtuales para la docencia en la Universidad de Cádiz. RELATEC Revista Latinoamericana de Tecnología Educativa, 7(2), 119-134. Recuperado el 02/02/12 de http://campusvirtual.unex.es/cala/editio/ index.php?journal=relatec&page=article&op=view&path[]=421&path[]=359 PEÑA, M. R. y AVENDAÑO. B.L. (2006) Evaluación de la implementación del aula virtual en una institución de Educación Superior. Suma Psicológica, 13(2),173-192. PÉREZ JUÁREZ, M. A. y VERDÚ, M. J. (2000). Teleformación: primaria, secundaria, universitaria y permanente. Valladolid: Secretariado de publicaciones e intercambio editorial. Universidad de Valladolid. PÉREZ i GARCIAS, A. (2007): Internet aplicado a la educación: aspectos técnicos y comunicativos. Las plataformas. En Cabero, J. (2007) (coord.) Nuevas Tecnologías Aplicadas a la Educación (pp.189-203). Madrid: Mc Graw-Hill. RODRÍGUEZ MALMIERCA, M. J. (dir.) (2006). El estado del e-learning en Galicia. Análisis en la Universidad y empresa. Recuperado el 21/03/12 de http://observatorioel.cesga.es/resultados/DO_ELE_InformeFinalObsEL_ca.pdf SALINAS, J. (2007). Modelos didácticos en los campus virtuales universitarios: perfiles metodológicos de los profesores en procesos de enseñanza-aprendizaje en entornos virtuales. Recuperado el 21/03/12 de http://www.virtualeduca.info/ponencias/252/virtual_salinas.doc SÁNCHEZ, J.A., et al. (2008). El Campus Virtual de la Universidad de Barcelona. Modelos de enseñanza y aprendizaje emergentes. RELATEC Revista Latinoamericana de Página 13 / 14 EDUTEC. Revista Electrónica de Tecnología Educativa. ISSN 1135-9250 Núm. 44 / Junio 2013 Tecnología Educativa, 7(2), 33-44. Recuperado http://campusvirtual.unex.es/cala/editio/index.php?journal= relatec&page=article&op=view&path=413&path=341 el 23/03/12 de SAHZAD, A.H., KHAN, A. (2010). Virtual learning and students perception, a research study. Procedia Social and Behavioral Sciences, 2, 5463–5467. TIRADO-MOURETA, R., PÉREZ-RODRÍGUEZ, M.A., AGUADED, J.I. (2011). Blended e-learning en universidades andaluzas. Aula Abierta, 39(2), 47-58. Recuperado el 12/04/12 de http://www.uniovi.net/ICE/publicaciones/ Aula_Abierta/numeros_anteriores/i16/05_Aula_Abierta_vol39_n2_mayo_2011 TUPAROVA, D. y TUPAROV, G. (2007). e-Learning in Bulgaria – the State of the Art. elearning Papers, 4, 1-20. VALVERDE, J. (dir.) (2004). Evolución de la oferta formativa on-line en las universidades públicas españolas (2001-2004) y elaboración de protocolo de buenas prácticas sobre las competencias del tutor universitario on-line. Recuperado el 06/03/12 de http://www.nodoeducativo.org/indice.htm RODRÍGUEZ, C. & ÁLVAREZ, M. J. (2013). Análisis didáctico de las aulas virtuales. Una investigación en un contexto de educación superior. EDUTEC, Revista Electrónica de Tecnología Educativa, 44. Recuperado el dd/mm/aa de http://edutec.rediris.es/Revelec2/Revelec44/analisis_didactico_aulas_virtuales_educacion_ superior.html Fecha de recepción: 2012-07-20 Fecha de aceptación: 2013-05-29 Fecha de publicación: 2013-06-26 Página 14 / 14
https://openalex.org/W3117106800	https://iris.unitn.it/bitstream/11572/285837/1/Integration%20of%20Ecosystem%20Services%20in%20Strategic%20Environmental%20assessment%20of%20a%20peri-urban%20development%20plan.pdf	English	null	Integration of Ecosystem Services in Strategic Environmental Assessment of a Peri-Urban Development Plan	Sustainability	2,020	cc-by	37,002	Teodoro Semeraro 1,* , Benedetta Radicchio 2, Pietro Medagli 3, Stefano Arzeni 3, Alessio Tur and Davide Geneletti 4,* 1 Department of Biological and Environmental Sciences and Technologies (DiSTeBA), University of Salento, 73100 Lecce, Italy 1 Department of Biological and Environmental Sciences and Technologies (DiSTeBA), University of Salento, l y 2 Regional Agency for Prevention and Environmental Protection, Puglia, 70126 Bari, Italy; b.radicchio@arpa.puglia.it 2 Regional Agency for Prevention and Environmental Protection, Puglia, 70126 Bari, Italy; b.radicchio@arpa.puglia.it 3 DHITECH S.c.a r.l.Prov. le Lecce Monteroni, 73100 Lecce, Italy; pietro.medagli@unisalento.it (P.M.); stefano.arzeni@unisalento.it (S.A.); alessio.turco@unisalento.it (A.T.) 4 Department of Civil, Environmental and Mechanical Engineering, University of Trento, 38123 Trento, Italy * Correspondence: teodoro.semeraro@unisalento.it (T.S.); davide.geneletti@unitn.it (D.G.) p p g 3 DHITECH S.c.a r.l.Prov. le Lecce Monteroni, 73100 Lecce, Italy; pietro.medagli@unisalento.it (P.M.); stefano.arzeni@unisalento.it (S.A.); alessio.turco@unisalento.it (A.T.) 4 Department of Civil, Environmental and Mechanical Engineering, University of Trento, 38123 Trento, Italy * Correspondence: teodoro semeraro@unisalento it (T S ); davide geneletti@unitn it (D G ) 3 DHITECH S.c.a r.l.Prov. le Lecce Monteroni, 73100 Lecce, Italy; pietro.medagli@unisalento.it (P.M.); stefano.arzeni@unisalento.it (S.A.); alessio.turco@unisalento.it (A.T.) , , y; p g ( ); stefano.arzeni@unisalento.it (S.A.); alessio.turco@unisalento.it (A.T.) 4 Department of Civil, Environmental and Mechanical Engineering, University of Trento, 38123 Trento, Italy * Correspondence: teodoro.semeraro@unisalento.it (T.S.); davide.geneletti@unitn.it (D.G.) 4 Department of Civil, Environmental and Mechanical Engineering, University of Trento, 38123 Trento, Italy * Correspondence: teodoro.semeraro@unisalento.it (T.S.); davide.geneletti@unitn.it (D.G.) p g g y y * Correspondence: teodoro.semeraro@unisalento.it (T.S.); davide.geneletti@unitn.it (D.G.) Abstract: Strategic environmental assessment (SEA) can support decision-makers in constructing more sustainable plans, programs, and policies (PPPs). To be more coherent with new frontiers of sustainable cities, PPPs need to include conservation objectives and to increase ecosystem service (ES) strategies. The ES concept is not intrinsic to the SEA process; therefore, it is necessary to develop an approach and methodology to include it. In this paper, we propose a methodology to integrate the concept of ecosystem services in all phases of the SEA process for a sub-urban plan, including the design of mitigation measures. The case study is represented by a peri-urban development plan in the municipality of Gallipoli in South Italy, characterized by a strong tourism economy and valuable agro-ecosystems. The analysis shows the priority ecosystem services that are selected considering the sustainable development and environmental goals, the context of referment, and the aims of the peri-urban plan. After, we highlight the potential ecosystem services developed considering the design of mitigation actions like green infrastructure, which could be implemented in the peri-urban plan. Citation: Semeraro, T.; Radicchio, B.; Medagli, P.; Arzeni, S.; Turco, A.; Geneletti, D. Integration of Ecosystem Services in Strategic Environmental Assessment of a Peri-Urban Development Plan. Sustainability 2021, 13, 122. https://dx.doi.org/ 10.3390/su13010122 Keywords: environmental assessment; environmental goals; ecosystem services; mitigation actions; green infrastructure; urban plan; landscape planning Received: 3 November 2020 Accepted: 22 December 2020 Published: 24 December 2020 Received: 3 November 2020 Accepted: 22 December 2020 Published: 24 December 2020 Teodoro Semeraro 1,* , Benedetta Radicchio 2, Pietro Medagli 3, Stefano Arzeni 3, Alessio Tur and Davide Geneletti 4,* The capacity to develop green infrastructure in SEA processes can conﬁgure the SEA as a tool for ecological urban design that is integrated with urban planning. This requires the ability to transfer ecological and planning theories into practical actions and the capacity of different disciplines to work in a transdisciplinary approach. Citation: Semeraro, T.; Radicchio, B.; Medagli, P.; Arzeni, S.; Turco, A.; Geneletti, D. Integration of Ecosystem Services in Strategic Environmental Assessment of a Peri-Urban Development Plan. Sustainability 2021, 13, 122. https://dx.doi.org/ 10.3390/su13010122 Integration of Ecosystem Services in Strategic Environmental Assessment of a Peri-Urban Development Plan Teodoro Semeraro 1,* , Benedetta Radicchio 2, Pietro Medagli 3, Stefano Arzeni 3, Alessio Turco 3 and Davide Geneletti 4,* o 1,* , Benedetta Radicchio 2, Pietro Medagli 3, Stefano Arzeni 3, Alessio Turco 3 l i 4 * sustainability sustainability sustainability sustainability 1. Introduction Ecosystem services (ESs) represent the human beneﬁts derived from a combination of biophysical structure and ecological functions that characterize the landscape and can be conditioned from the method and purpose for which humans use the land [1,2]. Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional claims in published maps and institutional afﬁliations. Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional claims in published maps and institutional afﬁliations. The concept of ecosystem services should be included in decision-support procedures, such as strategic environmental assessments (SEAs) for public plans or programs and environmental impact assessments (EIAs) for individual projects because the conservation of ecosystem services is essential to safeguarding people’s well-being [3,4]. SEA refers to a “range of analytical and participatory approaches that aim to integrate environmental considerations into Policies, Plans and Programmes (PPPs) and evaluate the inter-linkages with economic and social considerations” [5]. Copyright: © 2020 by the authors. Li- censee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/). Currently, interest is growing in the potential role of the ecosystem service (ES) con- cept within SEA [6–8]. Integrating ecosystem services into SEA is aimed at reducing the environmental impacts of PPPs and the potential impacts on ecosystem service provision. https://www.mdpi.com/journal/sustainability Sustainability 2021, 13, 122. https://dx.doi.org/10.3390/su13010122 Sustainability 2021, 13, 122 2 of 25 However, the inclusion of ecosystem services should be contextualized considering the objectives and the administrative scale of PPPs. Particularly, neither ESs can be addressed with a standard approach for all PPPs nor do all ESs linked with landscape transformations have the same importance [9]. Hence, it is important to determine priority ESs that could be positively and negatively affected by policy- and decision-maker choices [3,6]. The application of ES analysis in SEA is still in the exploration phase, although several appli- cations have been developed [6,7,10,11]. Many of these applications concern plans at the regional or urban scales, but few studies are related to peri-urban plans at the sub-urban scale. These plans are important because they determine the features of the future urban space and, hence, of the well-being of the people living therein. Peri-urban landscapes play an important role in ES provision, but they are often characterized by ES trade-offs [12]. 1. Introduction p p y y The difﬁculty of adapting regional- and urban-scale approaches at the sub-urban scale is in the nature of the planning. In the ﬁrst case, the planning exercise is a public initiative for the public interest of a vast range of stakeholders and affects private interests. This is the case of top-down planning processes. On the sub-urban scale, planning can be promoted for private interest. Therefore, it is important to match private economic interest with the social-ecological effects of the transformation plan. This is, in turn, the case of bottom-up processes, which, however, are conditioned by top-down plans [13]. Currently, an emerging concept for urban planners to foster sustainable development is represented by green infrastructure (GI), which is promoted by the European Union (EU) policy for both rural and urban areas [14,15]. It is possible to consider GI as part of nature-based solutions (NbSs) integrated with grey infrastructure (e.g., roads, sewer systems, and buildings) able to provide multiple ecological functions supporting the city’s activities and human well-being, involving both public and private spaces [16–19]. For these guidelines to consider ecosystem services in SEA processes, the priority ecosystem services considering other strategic actions need to be deﬁned and the variation in ecosystem services in different scenarios must be assessed considering the evolution approach developed [3,20,21]. Currently, the decision support processes in SEA do not consider GI/NbS-based ecosystem service assessment. This is an important gap requiring action to clarify the feedback between the built environment, GI, and human well-being during planning to ﬁnd the best solutions to achieve urban sustainability. The aim of this work was to construct a methodology to integrate ES in all phases of the SEA process for a sub-urban plan, including the design of mitigation measures, such as green infrastructure. Mainly, we wanted to include GI in the analysis of alternative scenarios to reinforce the impact assessment on ecosystem services. A central element of the approach is the identiﬁcation of priority ESs based on analysis of the objectives, environmental goals, and strategic actions of the plan. The case study is represented by a peri-urban development plan in the municipality of Gallipoli, in Southern Italy, which is characterized by a strong tourist economy and valuable agro-ecosystems. 2.2. Methods 2.2. Methods The methodology was inspired by the UNEP guidelines [3] and the work of Geneletti [11] and Vasquez et al. [8], which is based on establishing the ecosystem service context (stage 1), determining and assessing priority ecosystem services (stage 2), identifying al- ternatives and assessing impacts on ecosystem services (stage 3), and following up on ecosystem services (stage 4). The methodology was inspired by the UNEP guidelines [3] and the work of Geneletti [11] and Vasquez et al. [8], which is based on establishing the ecosystem service context (stage 1), determining and assessing priority ecosystem services (stage 2), identifying alternatives and assessing impacts on ecosystem services (stage 3), and following up on ecosystem services (stage 4). The methodology was inspired by the UNEP guidelines [3] and the work of Geneletti [11] and Vasquez et al. [8], which is based on establishing the ecosystem service context (stage 1), determining and assessing priority ecosystem services (stage 2), identifying al- ternatives and assessing impacts on ecosystem services (stage 3), and following up on ecosystem services (stage 4). The methodology was inspired by the UNEP guidelines [3] and the work of Geneletti [11] and Vasquez et al. [8], which is based on establishing the ecosystem service context (stage 1), determining and assessing priority ecosystem services (stage 2), identifying alternatives and assessing impacts on ecosystem services (stage 3), and following up on ecosystem services (stage 4). y g The definition of priority ecosystem services together with environmental goals can be useful for incorporating the ecosystem service concept in all phases of the SEA process, creating an interconnected workflow between the SEA process and the inclusion of eco- system services in the peri-urban plan. The main difference compared to the reference approach is the phase in which the mitigation measures are considered. This is applied as the last action in stage 3. We instead implement this phase as a specific action to foresee mitigation figures to be developed as GI. The effect of GI in ecosystem service production is assessed during the analysis of plan alternatives (Figure 2). The deﬁnition of priority ecosystem services together with environmental goals can be useful for incorporating the ecosystem service concept in all phases of the SEA process, cre- ating an interconnected workﬂow between the SEA process and the inclusion of ecosystem services in the peri-urban plan. 2. Materials and Methods 2.1. Study Area The study area (Figure 1) is a relatively small urban plot characterized by buildings scattered within a mosaic of arable lands, uncultivated areas, and olive groves. Olive groves represent the most critical element because they are characterized by monumental olive trees, which are protected under regional legislation by virtue of their production, ecological, and hydrogeological functions as well as their cultural and historical value. However, in recent years, almost all olive groves in Gallipoli have been severely damaged by the spread of Xylella fastidiosa [22–24]. 3 of 25 3 of 26 Sustainability 2021, 13, 122 Sustainability 2021 13 x FO Figure 1. Study area. Figure 1. Study area. Figure 1. Study area. Figure 1. Study area. 2.2.1. Step 1 Thi 2.2.1. Step 1 This step is focused on establishing the ecosystem service context and on determining and assessing priority ecosystem services. Priority ESs are those that contribute the most to social welfare on the basis of the objectives, environmental goals, and strategic actions of the urban plan. Environmental goals were selected considering international and na- tional directives. The definitions of the environmental goals started with consideration of the UN’s Sustainable Development Goals (SDGs) [26]. Afterwards, we analyzed the rela- tionships with other relevant plans (at different spatial tiers) by comparing the objec- tives/actions of the plan and the guidelines/forecasts of other plans (external consistency analysis) [27]. This is important because the SEA process needs to analyze the coherence between the peri-urban plan with PPP strategies of different administrative levels and scales, as stated by international and national directives and laws. Once goals were iden- tified, priority ESs that can support the achievement of these goals in the study area were selected. ESs were classified according to the Economics of Ecosystems and Biodiversity: Ecological and Economic Foundations, Earthscan [28]. In this phase, it is necessary to analyze the land use that characterizes the biophysical This step is focused on establishing the ecosystem service context and on determining and assessing priority ecosystem services. Priority ESs are those that contribute the most to social welfare on the basis of the objectives, environmental goals, and strategic actions of the urban plan. Environmental goals were selected considering international and national directives. The deﬁnitions of the environmental goals started with consideration of the UN’s Sustainable Development Goals (SDGs) [26]. Afterwards, we analyzed the relationships with other relevant plans (at different spatial tiers) by comparing the objectives/actions of the plan and the guidelines/forecasts of other plans (external consistency analysis) [27]. This is important because the SEA process needs to analyze the coherence between the peri-urban plan with PPP strategies of different administrative levels and scales, as stated by international and national directives and laws. Once goals were identiﬁed, priority ESs that can support the achievement of these goals in the study area were selected. ESs were classiﬁed according to the Economics of Ecosystems and Biodiversity: Ecological and Economic Foundations, Earthscan [28]. 2.2. Methods 2.2. Methods I thi t d did t d l ti f f ll t i ( ti 4) Experts, knowledge and scientific references Identify priority ecosystem services STEP 1 •Characterization of the study area • Identification of the main environmental issues and connections with ecosystem services Reference to international, national and local plans directives and indications Environmental goals and urban plan objectives Land use and ecosystem services characterization Experts, knowledge and scientific references Experts, knowledge and scientific references Nature-based Solutions Scenarios analysed: (A) Current land use; (B) Land use generated by the urban plan (C) Land use generated by the urban plan with the implementation of green infrastructure Identification of land use scenario to evaluate different Plan’s alternatives Assessment of ecosystem services changes: - Between A and B - Between B and C - Between A and C Analysis of the potential changes in the provision of ecosystem services between the scenarios considered Figure 2 Scheme of the methodology modified considering Therivel UNEP Geneletti and Figure 2. Scheme of the methodology modiﬁed considering Therivel, UNEP, Geneletti, and Vasquez et al. [3,8,11,25]. g gy g asquez et al. [3,8,11,25]. In this study, we did not develop actions for follow-up on ecosystem services (action 4). g gy g Vasquez et al. [3,8,11,25]. In this study, we did not develop actions for follow-up on ecosystem services (action 4). g gy g asquez et al. [3,8,11,25]. In this study, we did not develop actions for follow-up on ecosystem services (action 4). 2.2. Methods 2.2. Methods The main difference compared to the reference approach is the phase in which the mitigation measures are considered. This is applied as the last action in stage 3. We instead implement this phase as a speciﬁc action to foresee mitigation ﬁgures to be developed as GI. The effect of GI in ecosystem service production is assessed during the analysis of plan alternatives (Figure 2). 4 of 25 of 26 Sustainability 2021, 13, 122 S t i bilit 2021 13 F Figure 2. Scheme of the methodology modified considering Therivel, UNEP, Geneletti, and Vasquez et al. [3,8,11,25]. Environmental goals and urban plan objectives Identify priority ecosystem services Nature-based Solutions Identification of land use scenario to evaluate different Plan’s alternatives Scenarios analysed: (A) Current land use; (B) Land use generated by the urban plan (C) Land use generated by the urban plan with the implementation of green infrastructure Reference to international, national and local plans directives and indications Experts, knowledge and scientific references Experts, knowledge and scientific references Assessment of ecosystem services changes: - Between A and B - Between B and C - Between A and C Land use and ecosystem services characterization Experts, knowledge and scientific references STEP 1 •Characterization of the study area • Identification of the main environmental issues and connections with ecosystem services STEP 2 • Identification of mitigation actions as green infrastructures (Design ecosystem services) STEP 3 • Identification of land-use scenarios linked with urban development considering green infrastructures Analysis of the potential changes in the provision of ecosystem services between the scenarios considered STEP 4 • Ecosystem Services Impact Assessment Figure 2. Scheme of the methodology modiﬁed considering Therivel, UNEP, Geneletti, and Vasquez et al. [3,8,11,25]. 2.2.3. Step 3 In this step, we analyzed the ecosystem service evolution of land use considering the current conditions of the study area, potential land use decisions of the peri-urban plan, and the scenario of a peri-urban plan that integrates the planned GI. Land use scenarios were then generated and described. 2.2.1. Step 1 Thi 2.2.1. Step 1 In this phase, it is necessary to analyze the land use that characterizes the biophysical structure and the ecological functions that support priority ecosystem services, such as the cultural value of the landscape elements that have to be preserved or regenerated. It is nec- essary to consider their evolution in time to evaluate the loss to social welfare with the loss or gain of ES vis-à-vis the benefit of development. This is important for defining how to develop the peri-urban plan to reduce the negative effects on ecosystem services. Moreover, In this phase, it is necessary to analyze the land use that characterizes the biophysical structure and the ecological functions that support priority ecosystem services, such as the cultural value of the landscape elements that have to be preserved or regenerated. It is necessary to consider their evolution in time to evaluate the loss to social welfare with the loss or gain of ES vis-à-vis the beneﬁt of development. This is important for deﬁning how to develop the peri-urban plan to reduce the negative effects on ecosystem services. Moreover, it can represent the starting point in the evaluation of the interaction between Sustainability 2021, 13, 122 5 of 25 the built environment and the existing vegetation, which supports ecosystem services important for human well-being. the built environment and the existing vegetation, which supports ecosystem services important for human well-being. 2.2.2. Step 2 The SEA process can introduce mitigation actions to reduce negative environmental impacts and to simultaneously improve the quality of human life. In an urban context, mitigation actions should be identiﬁed to develop urban elements harmonizing economic income with social-ecological needs and without compromising the ecological functions and ecosystem services’ provision [17,29]. By linking scientiﬁc knowledge and planned activities, practical solutions were identiﬁed that can meet ecological needs with economic and social needs. These solutions consist of green infrastructure (GI) proposals in coherence with priority ecosystem services, land use characterizations, and structural elements that characterize the landscape. GI provides an alternative to grey infrastructure, having many co-beneﬁts for planners to foster sustainable urban land-use, to compensate for the negative effects of urban density, and to enhance human wellbeing [30,31]. GI has been deﬁned as “a strategically planned network of natural and semi-natural areas with other environmental features designed and managed to deliver a wide range of ecosystem services” [15]. g g y Following the deﬁnition of GI, it is possible to divide it into green (e.g., parks, play- ing ﬁelds, and private gardens) or blue (e.g., rivers, wetlands, and canals) areas, inter- connected with technological built forms and grey infrastructure (e.g., roads and sewer systems) [17,32–35]. Urban GI can partially replace speciﬁc urban ecosystem services like gas regulation; carbon sequestration; micro-climate regulation; rainwater drainage, and sewage and waste treatment; biodiversity conservation, preservation of traditional cultural needs, and historical characteristics of the landscape; recreational activities; and educa- tional values [35–37]. Other services that are lost to development, such as food production, usually have lesser importance within urban boundaries but may become relevant when examining more extended contexts (i.e., metropolitan or regional areas), especially under the eventual pressures potentially exerted by climate change on urbanized areas [35,38]. Mainly, we adopted the solutions reported in the literature [34,37,39–42] but focused on the study area. Therefore, each speciﬁc solution needs to follow speciﬁc conditions, such as the type of vegetation, functional use, and localizations. yp g The design of ecosystem service actions is also considered in the next phase, where plan scenarios are identiﬁed and subsequently analyzed to value the best alternative. 2.2.4. Step 4 A qualitative analysis was conducted by identifying losses and gains in ES provi- sion [3,43] using a seven-point Likert scale: from low, medium, high (for losses and for gains), and null values, where there was no type of variation [24]. The ecosystem service evaluation included methods identiﬁed in the literature, which link ecosystem service production with different classes of land use and integrated ﬁeld surveys to characterize vegetation types. The ES valuation, therefore, links biophysical structures indicated in the literature important in determining ecosystem services with speciﬁc ecological functions supported by the status and type of the vegetation currently present and proposed in different scenarios, also including GI (Table 1). We used the Economics of Ecosystems and Biodiversity Ecological and Economic Foundations ecosystem services classiﬁcation as a reference for this study [44]. Sustainability 2021, 13, 122 6 of 25 Table 1. Actions applied to assess ecosystem services in peri-urban area and in different scenarios analyzed. ES, ecosystem services; SEA, strategic environmental assessment. Elements of ES Assessment Actions References Scientiﬁc Literature Valuations and mapping of ecosystem services considering land-cover analysis Reference manuscripts [24,43,45–47] Method to assess the symptoms of X. fastidiosa related to leaf scorching and wilting of the canopy Reference manuscripts [23,24,48] Regional law of monumental olive groves Parameters indicated by the regional legislation for the classiﬁcation of monumental olive trees and monumental olive groves Regional law 14/2007 [49] Field activities Floristic and vegetation relief to better calibrate the ecosystem services linked with land use, considering species that characterize them and the local context Ecologist, botanist, agronomist, faunist Visual inspection to assess the symptoms of X. fastidiosa related to leaf scorching and wilting of the canopy Classiﬁcation of monumental olive trees and monumental olive groves Skills and experience of the expert involved Ecosystem service design to improve beneﬁt Expert in endemic vegetation; ecological role of vegetation, biodiversity restoration; ecosystem service analysis, planner, and SEA processes; references of local vegetation [50–52] Regional law of Xylella fastidiosa. Actions implemented in olive groves affected by Xylella fastidiosa Apulian Regulation guideline 02/2019 [53] Table 1. Actions applied to assess ecosystem services in peri-urban area and in different scenarios analyzed. ES, ecosystem services; SEA, strategic environmental assessment. Table 1. Actions applied to assess ecosystem services in peri-urban area and in different scenarios analyzed. ES, ecosystem services; SEA, strategic environmental assessment. Ecosystem service design to improve beneﬁt 3.1. Step 1 3.1.1. Priority Ecosystem Services 3. Results Next, we applied the conceptual framework to the case study. The structure of the sections and subsections follow the methodologies reported in Figure 2. 3.1. Step 1 3 1 1 P i it E t S i 3.1.3. Land Use and Ecosystem Services Characterization 3.1.3. Land Use and Ecosystem Services Characterization The land use of the area shown in Figure 1 is mainly characterized by arable land (33.3% of the total surface) and olive groves (28.7% of the total surface) (Table 3 and Figure 3). There are also artiﬁcial surfaces (roads and buildings) covering 14% of the total study area. The presence of natural vegetation is very low (shrub, covering 0.3 ha). The herbaceous vegetation is nitrous and ruderal ﬂora with a low ecological value and its development has been conditioned by present and past anthropic activities. Although 62% of the area is characterized by agricultural use, its productive role is minimal. The arable land is characterized by strong owners’ fragmentation and, in many cases, represents small plots used as vegetable gardens for self-production. These garden plots might not have strong economic value for landowners, as 97% of these plots have joined the development of the peri-urban plan, thus renouncing future use as vegetable gardens. The olive groves were almost all affected by Xylella fastidiosa. Symptoms of Xylella fastidiosa related to leaf scorching and wilting of the canopy were assessed by visual inspection, estimating disease severity using the following scale: 0 = symptomless, 1 = leaf scorching on few branches or few desiccated branches affecting the portion of the canopy, 2 = leaf scorching on several branches or desiccation affecting a large part of the portion of the canopy, and 3 = canopy with desiccated branches uniformly distributed [23,24,48]. The olive trees were classiﬁable as 2 and 3; in this case, almost all olive trees were dry or potentially dry because they were severely infected by Xylella fastidiosa (Figure 4). Therefore, it was possible to consider these plants dead [48]. p p Xylella fastidiosa produced a strong ecological impact because the olive groves lost their ecological function in the landscape, as some of them were monumental trees with important cultural and aesthetic values in the landscape. Many olive groves are character- ized by historical and monumental trees. The cultural value of monumental olive groves was then analyzed through objective measures concerning the diameter of the trunk and through a visual analysis of the shape of the trunk of the plants, as established by regional law 14/2007 [23,49]. The elimination of infected monumental olive trees will cause the loss of an identity element of the landscape of Gallipoli (Table 4). 3.1.2. Environmental Goals Table 2 reports the overall environmental sustainability objectives relevant to urban development plans drawn from international, national, and regional environmental legisla- tion (internal consistency analysis). The study area is mainly characterized by a peri-urban landscape, and provisioning services are not relevant because the value of food and raw material production is low. However, regulating and cultural services have strong relevance considering the area as a socio-ecological system [24,35,36]. g g y It is difﬁcult to establish a connection between ecosystem services and all environ- mental objectives. For example, aspects connected with mobility and electromagnetism are mainly related to the project techniques, so they can indirectly concern the ecosystem services. For this reason, in this phase, we only report the environmental goals that are directly linked to ecosystem services and that satisfy the internal and external consistency analysis, which characterizes the SEA process. 3.1.1. Priority Ecosystem Services 3.1.1. Priority Ecosystem Services The plan was developed on a local scale; therefore, it was possible to deﬁne priority ecosystem services that can be developed considering the speciﬁc environmental issues that characterize the urban area, such as air-quality microclimate issues associated with built environmental and human activities. For these reasons, air-quality regulation, climate regulation, and moderation of extreme events were selected as priority ecosystem services that need to be enhanced in the area through mitigation actions [26,54–57]. In addition, we considered cultural and amenity services as priority ecosystem services, which are important due to the presence of many monumental olive trees, currently threatened by Xylella fastidiosa. y f Focusing on priority ecosystem services does not exclude the possibility to enhance additional services. For example, in the case study, habitat services are considered co- beneﬁts since the aim of the sub-urban plan is the development of residential areas and Sustainability 2021, 13, 122 7 of 25 since green areas will be planned to increase human well-being. Speciﬁc design choices can improve the capacity of green areas to support biodiversity, but this is not the primary scope and we cannot predict with certainty those birds and animal species that will be able to settle in the spaces. 3.1.3. Land Use and Ecosystem Services Characterization Moreover, the olive groves were important carbon sinks and therefore played an important role in mitigating the local microclimate and air quality [24]. 8 of 25 8 of 25 Sustainability 2021, 13, 122 Table 2. Indications of priority ecosystem services connected with environmental goals for each sector indicated by the international, national, and regional directives, policies, programs, and plans: ecosystem services from 1 to 6 are provisioning services; from 7 to 15 are biological services; from 16 to 17 are habitat services; and from 18 to 22 area cultural and amenity services [28,44]. P, priority ecosystem services that we want to directly enhance by integrating it into the peri-urban plan’s green infrastructure related to the mainly environmental problems. C, other ecosystem services that can be linked indirectly with the green infrastructure. Sector Environmental Goals National and International Directives Ecosystem Services 1.Population and human health 1-1. Protecting the population from health risks originating from situations of environmental degradation 1-2. Preventing and reducing sources of urban pollution and the risk of major accidents 1-4. Improving the quality of life in urban context 1-5. Promotion of shared territorial choices 1-6. Human health 2001/42/CE 2014/52/EU National environmental law: D.Lgs. 152/2006 and modiﬁcations 1 Food (C) 2. Water (C) 7. Air quality regulation (P) 8. Climate regulation (P) 9. Moderation of extreme events (P) 10. Regulation of water ﬂows(C) 11.Water treatment (C) 15. Biological control (C) 17. Maintenance of genetic diversity (C) 19. Opportunities for recreation and tourism (C) 20. Inspiration for culture, art and design (P) 21. Spiritual experience (P) 22. Information for cognitive development (P) 2. Soil 2-1. Protection, prevention, and mitigation of hydrogeological and hydraulic risks 2-2. Preventing and mitigating current and potential risks associated with forest ﬁres 2-3. Ensuring the permeability of the soils and quality National environmental law: D.Lgs. 152/2006 and modiﬁcations Regional Territorial Landscape Plan (Apulian Regiona); Regional Document of General Asset 9. Moderation of extreme events (P) 10. Regulation of water ﬂows (C) 3. Water 3-1. Guaranteeing the quantitative/qualitative protection of groundwater bodies 3-2. Reducing water consumption 91/271/EC; 2006/118/EC; 2007/60/CE; 2008/105/CE; National environmental law: D.Lgs. 152/2006 and modiﬁcations Regional Water Protection Plan; Regional Hydrogeological Plan. (Apulian Regiona) 2. Water (C) 10. Regulation of water ﬂows (C) 11. Waste treatment (C) 4.Climate, microclimate and air quality 4-1. Reduction of climate altering and polluted atmospheric pollutant gas emissions 4-2. Enhancement of carbon sink vegetation cover 4-3. 3.1.3. Land Use and Ecosystem Services Characterization Reduction of local climatic changes 2018/841/EU 2016/2284 7. Air quality regulation (P) 8. Climate regulation (P) 2008/50/CE National environmental law: D.Lgs. 152/2006 and modiﬁcations Environmental Goals 7. Air quality regulation (P) 8. Climate regulation (P) 7. Air quality regulation (P) 8. Climate regulation (P) 9 of 25 Sustainability 2021, 13, 122 Table 2. Cont. Sector Environmental Goals National and International Directives Ecosystem Services 5. Biodiversity 5-1 Enhancement of agro-ecosystems to support fauna and ﬂora (especially endemic) 5-2. Development of widespread ecological connectivity at the provincial level 5-3. Increase in endemic and local biodiversity 79/409/CEE 92/43/CEE; Nature 2000. Biodiversity strategies 2020 Biodiversity strategies 2030 16. Maintenance of life cycles of migratory species (C) 17. Maintenance of genetic diversity (C) 14. Pollination (C) 15. Biological control (C) 6. Landscape and cultural heritage 6-1. Conservation or enhancement of the elements of the local landscape 6-2. Implementing the integration of the urban sprawl in the landscape context of reference 6-3. Policies and actions of discovery and rapprochement of residents to cultural capital 6-4. Development of strategies for sustainable use of landscape heritage Regional Territorial Landscape Plan (Apulian Region) 18. Aesthetic information (P) 19. Opportunities for recreation and tourism (C) 20. Inspiration for culture, art and design (P); 21. Spiritual experience (P) 22. Information for cognitive development (P) 7. Energy 7-1. Promoting energy savings 7-2. Incentivize energy efﬁciency in buildings 2009/28/CE 2001/77/CE 7. Air quality regulation(P) 8. Climate regulation (P) 8. Waste 8-1. Efﬁciency of natural resources National environmental law: D.Lgs. 152/2006 and modiﬁcations 7. Air quality regulation (P) 11. Waste treatment (C) Table 2. Cont. National and International Directives 16. Maintenance of life cycles of migratory species (C) 17. Maintenance of genetic diversity (C) 14. Pollination (C) 15. Biological control (C) Sustainability 2021, 13, 122 10 of 25 ility 2021, 13, 122 10 of 25 Table 3. Composition of the main land use of the study area. Land Use % Fruit trees 0.6 Herbaceous vegetation 23.4 Shrubs 1.1 Arable land 45.1 Artiﬁcial surfaces (building) 19.0 Olive groves 38.9 Green urban areas 4.2 Vineyards 3.0 bility 2021, 13, x FOR PEER REVIEW 3 of 26 3.1.3. Land Use and Ecosystem Services Characterization The land use of the area shown in Figure 1 is mainly characterized by arable land (33.3% of the total surface) and olive groves (28.7% of the total surface) (Table 3 and Figure 3). 3.1.3. Land Use and Ecosystem Services Characterization There are also artificial surfaces (roads and buildings) covering 14% of the total study area. The presence of natural vegetation is very low (shrub, covering 0.3 ha). The herba- ceous vegetation is nitrous and ruderal flora with a low ecological value and its develop- ment has been conditioned by present and past anthropic activities. Figure 3. Spatial configuration of the main land use of the study area. Table 3. Composition of the main land use of the study area. Figure 3. Spatial conﬁguration of the main land use of the study area. Table 3. Composition of the main land use of the study area. Land Use % Fruit trees 0.6 Herbaceous vegetation 23.4 Shrubs 1.1 Arable land 45.1 Artiﬁcial surfaces (building) 19.0 Olive groves 38.9 Green urban areas 4.2 Vineyards 3.0 REVIEW 3 of 26 3.1.3. Land Use and Ecosystem Services Characterization The land use of the area shown in Figure 1 is mainly characterized by arable land (33.3% of the total surface) and olive groves (28.7% of the total surface) (Table 3 and Figure 3). There are also artificial surfaces (roads and buildings) covering 14% of the total study area. The presence of natural vegetation is very low (shrub, covering 0.3 ha). The herba- ceous vegetation is nitrous and ruderal flora with a low ecological value and its develop- ment has been conditioned by present and past anthropic acti ities Table 3. Composition of the main land use of the study area. R REVIEW Table 3. Composition of the main land use of the study area. REVIEW Table 3. Composition of the main land use of the study area. REVIEW Table 3. Composition of the main land use of the study area. Land Use % Fruit trees 0.6 Herbaceous vegetation 23.4 Shrubs 1.1 Arable land 45.1 Artiﬁcial surfaces (building) 19.0 Olive groves 38.9 Green urban areas 4.2 Vineyards 3.0 REVIEW 3 of 26 3.1.3. Land Use and Ecosystem Services Characterization The land use of the area shown in Figure 1 is mainly characterized by arable land (33.3% of the total surface) and olive groves (28.7% of the total surface) (Table 3 and Figure 3). There are also artificial surfaces (roads and buildings) covering 14% of the total study area. The presence of natural vegetation is very low (shrub, covering 0.3 ha). 3.1.3. Land Use and Ecosystem Services Characterization The herba- ceous vegetation is nitrous and ruderal flora with a low ecological value and its develop- ment has been conditioned by present and past anthropic activities Figure 3. Spatial configuration of the main land use of the study area. Figure 3. Spatial conﬁguration of the main land use of the study area. 11 of 25 Sustainability 2021, 13, 122 p Figure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of the monumental olive trees. Xylella fastidiosa produced a strong ecological impact because the olive groves lost Figure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of the monumental olive trees. gure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of monumental olive trees. Figure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the ure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of monumental olive trees. Xylella fastidiosa produced a strong ecological impact because the olive groves lost Figure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of the monumental olive trees. Figure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of the monumental olive trees. Xylella fastidiosa produced a strong ecological impact because the olive groves lost Figure 4. Example of the olive trees affected by Xylella fastidiosa in the study area, highlighting the state of desiccation of the monumental olive trees. Xylella fastidiosa produced a strong ecological impact because the olive groves lost their ecological function in the landscape, as some of them were monumental trees with important cultural and aesthetic values in the landscape. Many olive groves are charac- terized by historical and monumental trees. The cultural value of monumental olive groves was then analyzed through objective measures concerning the diameter of the trunk and through a visual analysis of the shape of the trunk of the plants, as established by regional law 14/2007 [23,49]. Table 4. Qualitative characterization of the ecosystem service provision linked with the main land use in the study area. -, not relevant for ecosystem services [13,34,44]. planned considering the typologies of the space t building, private green areas, public green areas, a 3.2. Step 2: Identiﬁcation of Natural-Based Solutions g p g p g p g lutions were conceived and planned considering the typologies of vegetation species that can be used to develop priority ecosystem services. NbS considered the typologies of the area that could be developed and potentially used (Table 5). For example, we propose two types of green roofs considering the type of buildings. Mainly, we previously suggested the use of extensive green roofs for houses and intensive green roofs for schools, markets, and public buildings [30,58]. The use of scientific literature was important for setting human benefit and ecosystem services connected with the GI. If food production could not be considered a priority ecosystem service in the urban context, urban community gardens can be important strategies to promote social activities and integrations between different cultures in the newly created neighborhood. It is im- portant to plan urban parks integrated with a water body and wetland vegetation to give it This phase was developed considering the GI that can be developed in urban areas to support priority ecosystem services and its associated co-beneﬁts. The type of solution was planned considering the typologies of the space that can be used: roofs and walls of the building, private green areas, public green areas, and parking areas. The natural-based solutions were conceived and planned considering the typologies of vegetation species that can be used to develop priority ecosystem services. NbS considered the typologies of the area that could be developed and potentially used (Table 5). For example, we propose two types of green roofs considering the type of buildings. Mainly, we previously suggested the use of extensive green roofs for houses and intensive green roofs for schools, markets, and public buildings [30,58]. The use of scientiﬁc literature was important for setting human beneﬁt and ecosystem services connected with the GI. portant to plan urban parks integrated with a water body and wetland vegetation to give it multifunctional social uses and to contribute to important ecological functions [59]. These pieces of green infrastructure, although relatively small in size, can improve specific ecolog- ical functions, increase human health, and enhance local biodiversity. In a highly urbanized context where ecological land use is in competition with social and economic land use, these may represent interventions. 3.1.3. Land Use and Ecosystem Services Characterization Panoramic view of dead olive trees in the study, highlighting the state of desiccation of the olive grove of dead olive trees in the study, highlighting the state of desiccation of the olive groves. of dead olive trees in the study, highlighting the state of desiccation of the olive groves. 3.2. Step 2: Identification of Natural-Based Solutions This phase was developed considering the GI that can be developed in urban areas to support priority ecosystem services and its associated co-benefits. The type of solution was The species detected for the herbaceous vegetation belong to the ruderal type, which characterize land exploited for human activities of low ecological value and not relevant for food or medicinal uses. 3.1.3. Land Use and Ecosystem Services Characterization Main Services Type Olive Groves Olive Groves Affected by X. fastidiosa Arable Land Herbaceous Vegetation 1-Food medium - medium - 2-Water - - - - 3-Raw Materials medium - low - 4-Genetic resources medium medium low medium 5-Medicinal resources low - - - 6-Ornamental resources - - - - 7-Air quality regulation strong low low low 8-Climate regulation strong low low low 9-Moderation of extreme events medium low low low 10-Regulation of water ﬂows low low low low 11-Waste treatment low low low low 12-Erosion prevention medium low - low 13-Maintenance of soil fertility medium low low low 14-Pollination low low low low 15-Biological control low low low low 16-Maintenance of life cycles of migratory species medium low low low 17-Maintenance of genetic diversity low medium medium medium 18-Aesthetic information strong low low low 19-Opportunities for recreation and tourism strong low low low 20-Inspiration for culture, art and design strong - - - 21-Spiritual experience strong - - - 22-Information for cognitive development strong - - - As a paradox, the olive trees that were previously a valuable element of the landscape now represent an element of degradation of the landscape (Figure 5). The provision of ecosystem services of the olive trees affected by Xylella fastidiosa is now more comparable to uncultivated areas characterized by herbaceous ruderal vegetation [23,24]. Xylella fastidiosa produced a strong ecological impact because the olive groves lost their ecological function in the landscape, as some of them were monumental trees with important cultural and aesthetic values in the landscape. Many olive groves are charac- terized by historical and monumental trees. The cultural value of monumental olive Table 4. Qualitative characterization of the ecosystem service provision linked with the main land use in the study area. -, not relevant for ecosystem services [13,34,44]. 12 of 25 6 of 26 12 of 25 6 of 26 Sustainability 2021, 13, 122 Sustainability 2021 13 x FO Figure 5. Panoramic view of dead olive trees in the study, highlighting the state of desiccation of the olive groves. Figure 5. Panoramic view of dead olive trees in the study, highlighting the state of desiccation of the olive groves. e 5. Panoramic view of dead olive trees in the study, highlighting the state of desiccation of the olive grove re 5. planned considering the typologies of the space t building, private green areas, public green areas, a 3.2. Step 2: Identiﬁcation of Natural-Based Solutions These actions can be framed as ecological urban design tools, which need to combine scientific knowledge and applied knowledge from different disciplines, such as ecology, en- gineering, and planning. Moreover, it needs to involve private and public bodies to guaran- tee the development and the capacity to maintain these solutions in the long term If food production could not be considered a priority ecosystem service in the urban context, urban community gardens can be important strategies to promote social activities and integrations between different cultures in the newly created neighborhood. It is important to plan urban parks integrated with a water body and wetland vegetation to give it multifunctional social uses and to contribute to important ecological functions [59]. These pieces of green infrastructure, although relatively small in size, can improve speciﬁc ecological functions, increase human health, and enhance local biodiversity. In a highly urbanized context where ecological land use is in competition with social and economic land use, these may represent interventions. 13 of 25 Sustainability 2021, 13, 122 13 of 25 onnections between beneﬁt and biophysical structures of mitigation action design, like green infrastructure (GI) useful for increasing priority ecosystem services and satisfying the ntal goals identiﬁed in Table 2 [27]. Mitigation Actions and GI Ecosystem Services Beneﬁt Reference Environmental Goals (Table 2) n in the impermeable surface with o the indications of the ality urban plan development and extensive green roofs for ings all tion: mixed shrub, arboreal ral uses, and natural ﬂora of endemic vegetation with m derivate from the local context eds and cuttings collected in the context) of microhabitats and structures upport local insects and animal of melliferous, aromatic ﬂora of a wetland and ﬂooding area of forest habitat mmunity gardens streets, and parking lots with oors ra for environmental monitoring er collection and use systems 1-Food (C) Fruits, small-scale subsistence; Food security; Raising awareness of the inhabitants; Food production and processing; Energy consumption and production. [42,60–62] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6 2-Water (C) Water availability [42,58,59] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6; 3-1; 3-2 7-Air quality regulation (P) Remove air toxins and nursery gasses like particulate matter, carbon dioxide, nitrogen dioxide, carbon monoxide, and sulfur dioxide; Carbon sinks; Reduction of carbon footprint. 3.3. Step 3: Analysis of Different Scenarios In this phase, we considered four different land-use scenarios: Past, with olive groves in good status; Scenario 0, representing the current situation with olive groves affected by Xylella fastidiosa; Scenario 1, with the foreseen actions by the peri-urban plan; and Scenario 2, with the peri-urban plan integrated with GI. We included the scenario with Xylella fastidiosa (Past scenario) because introducing the concept of ecosystem services in the urbanization process has to drive land use trans- formation considering pre-disturbance conditions. This increases the level of comparison between the scenarios to produce solutions that improve ecosystem service production. As such, the urbanization processes can develop the actions needed to regenerate the de- graded landscape. The ante-Xylella fastidiosa scenario (Past scenario) was included because it shows an agro-ecosystem in which there was not this speciﬁc disturbance. Consider- ing Scenario 0 (olive groves affected by Xylella fastidiosa), we applied a likely scenario (from olive groves to uncultivated soil) following implementation of the regulation of the Apulia region of 5 February 2019 [53], which requires the felling of trees infected with Xylella fastidiosa [23,24]. In the current state, the low value of the area for agricultural production can push landowners to abandon land [23], which negatively affects ecosystem services or landscape restoration. Moreover, in the medium term, considering the interest of landowners to develop a peri-urban plan, they would not be interested in planting new olive groves while waiting for approval of the plan. g g We hypothesized an urban development plan (Scenario 1) that reduces the area destined for residential and commercial structures, increasing the areas assigned to urban standards, mainly green areas. This scenario was developed by reducing the interaction of the built environment with monumental olive groves. The few monumental olive trees that interfered with viability of the municipality plan at the urban scale have to be relocated to an area of interest in the peri-urban plan [49]. This is important to preserve the agricultural element that characterizes the area. In these scenarios, vegetation was designed only to keep monumental olive groves but without considering the priority ecosystem services. Therefore, no indications of the type of biophysical structure and ecological functions to develop in speciﬁc locations of the sub-urban area were provided. However, the spread of Xylella fastidiosa in the area results in the abatement of infected olive trees regardless of the urban plan. planned considering the typologies of the space t building, private green areas, public green areas, a 3.2. Step 2: Identiﬁcation of Natural-Based Solutions [42,58,59,63–68] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6 4-1; 4-2; 4-3 7-1; 7-2 8-1 8-Climate regulation (P) Reduced heat ﬂux into the building Reduced energy demand for space climate conditioning; Reduction of urban heat island effect; Increase in thermal comfort; Reduced urban energy consumption; Reduction of carbon footprints; Decreasing cooling and heating loads. [42,58,59,64,66–71] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6 4-1; 4-2; 4-3 9-Moderation of extreme events (P) Reduction in stormwater volume; Decrease the burden of the water treatment facilities in an area; Stormwater retention to reduce peak ﬂow and runoff; Improve the use of rainwater. [42,55,58,59,66–72] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6 10-Regulation of water ﬂows (C) 2-1; 2-2; 2-3 11-Waste treatment (C) 3-1; 3-2 8-1 14 of 25 Sustainability 2021, 13, 122 Table 5. Cont. Mitigation Actions and GI Ecosystem Services Beneﬁt Reference Environmental Goals (Table 2) 14-Pollination (C) Provide habitat for spiders, mites, beetles, grasshoppers, butterﬂies, and birds; Implementation of vegetation biodiversity and improve landscape value. [42,58,59,62,66–68] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6 15-Biological control (C) 16. Maintenance of life cycles of migratory species (C) 5-1; 5-2; 5-3 17-Maintenance of genetic diversity (C) 18-Aesthetic information (P) Promote green lifestyles; Increase community engagement; Provide recreational green spaces; Reduced anxiety and tension; Positive effect on mental processes and behavior; Attentional restoration Reduced mental fatigue/fatigue; Improved cognitive function; Improved productivity/ability to perform tasks/positive workplace attitude; Aesthetic appreciation; Increased inspiration; Increased recreational satisfaction Positive effect on physical function and/or physical health; Relaxation and recreation, release stress, and relax the mind; Provide recreational space; Horticultural practices; Community support; Social activities; Social integrations between people of different cultures. [35,42,58,60,62,66,69, 73–75] 1-1; 1-2; 1-3, 1-4; 1-5; 1-6 19-Opportunities for recreation & tourism (C) 20-Inspiration for culture, art and design (P) 21-Spiritual experience (P) 6-1; 6-2; 6-3; 6-4 22-Information for cognitive development (P) Table 5. Cont. [35,42,58,60,62,66,69, 73–75] 6-1; 6-2; 6-3; 6-4 Sustainability 2021, 13, 122 15 of 25 15 of 25 These actions can be framed as ecological urban design tools, which need to combine scientiﬁc knowledge and applied knowledge from different disciplines, such as ecology, engineering, and planning. Moreover, it needs to involve private and public bodies to guarantee the development and the capacity to maintain these solutions in the long-term. 3.3. Step 3: Analysis of Different Scenarios 3.3. Step 3: Analysis of Different Scenarios Table 6 Qualitative variation in ecosystem services from the scenarios analyzed 😊indicates positive variation ☹indi and design; spiritual experience; and information for cog be considered irreversible transformations. Table 6 Qualitative variation in ecosystem services from the scenarios analyzed 😊indicat p y p The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p The most important impacts, however, affect the foll thetic information; opportunities for recreation and touri and design; spiritual experience; and information for cog be considered irreversible transformations. 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air quality regulation climate regulation and moderation o 3.4. Step 4: Ecosystem Service Impact Assessment Sustainability 2021, 13, x FOR PEER REVIEW Sustainability 2021, 13, x FOR PEER REVIEW The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Fi 2 d T bl 6) Sustainability 2021, 13, x FOR PEER REVIEW Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even 3.4. Step 4: Ecosystem Service Impact Assessment Th i i f h P i S i 0 d l f 3.4. Step 4: Ecosystem Service Impact Assessment Th i i f h P i S i 0 d l f 3.4. Step 4: Ecosystem Service Impact Assessment Th i i f h P i S i 0 d l f Sustainability 2021, 13, x FOR PEER REVIEW 10 of Sustainability 2021, 13, x FOR PEER REVIEW 10 Sustainability 2021, 13, x FOR PEER REVIEW Sustainability 2021, 13, x FOR PEER REVIEW 10 of Sustainability 2021, 13, x FOR PEER REVIEW 10 Sustainability 2021, 13, x FOR PEER REVIEW Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce- nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 21-Spiritual experience ☹☹☹ = 😊😊 22-Information for cognitive develop- ment ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pres- ence of green urban areas structured with shrub and arboreal vegetation. Dead olive trees will be replaced with new trees, promoting biophysical structure able to support pollution absorbance such as CO2. The climate regulation service was not evaluated because, even if the green urban areas can improve this service, 39% of urbanization of the total area (considering building construction, roads, and parking areas) can increase local tempera- ture. Therefore, there is no direct information about the relation between vegetation and environment building that indicates a changing value However, without specific mitiga- indicates positive variation, 7-Air quality regulation ☹☹ 😊 8-Climate regulation ☹☹ 9-Moderation of extreme events ☹ ☹ 10-Regulation of water flows = ☹ 11-Waste treatment = 14-Pollination = 15-Biological control ☹ 17-Maintenance of genetic diversity ☹ 18-Aesthetic information ☹☹☹ 19-Opportunities for recreation and tourism ☹☹☹ 20-Inspiration for culture, art and de- sign ☹☹☹ 21-Spiritual experience ☹☹☹ 22-Information for cognitive develop- ment ☹☹☹ Moving from Scenario 0 to Scenar ence of green urban areas structured w will be replaced with new trees, promo absorbance such as CO2. The climate r if the green urban areas can improve (considering building construction, ro ture. 3.3. Step 3: Analysis of Different Scenarios Therefore, there is no direct infor environment building that indicates a tion actions, the urbanization area red tion of water flow services because th addition, the presence of dry olive tre of a degraded and abandoned landsca During the SEA process, specific ecosystem services (Table 5 and Figu were planned for the realization of new agroecosystem matrix and natural hab l l fl b h indicates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no signiﬁcant variation. “Not evaluated” indicates the inability to form a signiﬁcant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Scenario 1 From Scenario 1 to Scenario 2 7-Air quality regulation Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 21-Spiritual experience ☹☹☹ = 😊😊 22-Information for cognitive develop- ment ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tree Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 21-Spiritual experience ☹☹☹ = 😊😊 22-Information for cognitive develop- ment ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the p ence of green urban areas structured with shrub and arboreal vegetation. Dead olive t p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: aes- thetic information; opportunities for recreation and tourism; inspiration for culture, art, and design; spiritual experience; and information for cognitive development. These can be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce- nario 2 Air quality regulation ☹☹ 😊 😊😊 Climate regulation ☹☹ Not evaluated 😊 Moderation of extreme events ☹ ☹ 😊 0-Regulation of water flows = ☹ 😊 1-Waste treatment = Not evaluated 😊 4-Pollination = Not evaluated 😊😊 5-Biological control ☹ Not evaluated 😊😊 7-Maintenance of genetic diversity ☹ Not evaluated 😊😊 8-Aesthetic information ☹☹☹ = 😊😊 9-Opportunities for recreation and urism ☹☹☹ = 😊😊 0-Inspiration for culture, art and de- gn ☹☹☹ = 😊😊 p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 8-Climate regulation be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 21-Spiritual experience ☹☹☹ = 😊😊 22-Information for cognitive develop- ment ☹☹☹ = 😊😊😊 M i f S i 0 S i 1 i li l i i d h be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 21-Spiritual experience ☹☹☹ = 😊😊 22-Information for cognitive develop- ment ☹☹☹ = 😊😊😊 M i f S i 0 S i 1 i li l i i d h Not evaluated vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 9-Moderation of extreme events and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and tourism ☹☹☹ = 😊😊 20-Inspiration for culture, art and de- sign ☹☹☹ = 😊😊 21-Spiritual experience ☹☹☹ = 😊😊 22-Information for cognitive develop- ment ☹☹☹ = 😊😊😊 and design; spiritual experience; and information for cog be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicat cates negative variation, and the number of emoticons indicates the potential intensity of the medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to nario 1 7-Air quality regulation ☹☹ 😊 8-Climate regulation ☹☹ Not evaluated 9-Moderation of extreme events ☹ ☹ 10-Regulation of water flows = ☹ 11-Waste treatment = Not evaluated 14-Pollination = Not evaluated 15-Biological control ☹ Not evaluated 17-Maintenance of genetic diversity ☹ Not evaluated 18-Aesthetic information ☹☹☹ = 19-Opportunities for recreation and tourism ☹☹☹ = 20-Inspiration for culture, art and de- sign ☹☹☹ = 21-Spiritual experience ☹☹☹ = 22-Information for cognitive develop- ment ☹☹☹ = The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and ☹☹☹ 😊😊 10-Regulation of water ﬂows = The most important impacts, however, affect the foll thetic information; opportunities for recreation and touri and design; spiritual experience; and information for cog be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicat cates negative variation, and the number of emoticons indicates the potential intensity of the medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to nario 1 7-Air quality regulation ☹☹ 😊 8-Climate regulation ☹☹ Not evaluated 9-Moderation of extreme events ☹ ☹ 10-Regulation of water flows = ☹ 11-Waste treatment = Not evaluated 14-Pollination = Not evaluated 15-Biological control ☹ Not evaluated 17-Maintenance of genetic diversity ☹ Not evaluated 18-Aesthetic information ☹☹☹ = 19-Opportunities for recreation and tourism ☹☹☹ = 20-Inspiration for culture, art and de- sign ☹☹☹ = 21-Spiritual experience ☹☹☹ = I f f d l 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18 A th ti i f ti ☹☹☹ 😊😊 11-Waste treatment = Not evaluated 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 14-Pollination = Not evaluated Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 19-Opportunities for recreation and ☹☹☹ = 😊😊 Not evaluated Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 17-Maintenance of genetic diversity 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. 3.3. Step 3: Analysis of Different Scenarios Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 17-Maintenance of genetic diversity ☹ Not evaluated 😊😊 18-Aesthetic information ☹☹☹ = 😊😊 Not evaluated Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. 3.3. Step 3: Analysis of Different Scenarios Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 18-Aesthetic information 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem vices such as air-quality regulation, climate regulation, and moderation of extreme ev related to the loss of the thick crown of the dead olive trees that are still present in the (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. These be considered irreversible transformations. Table 6. 3.3. Step 3: Analysis of Different Scenarios Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosyste vices such as air-quality regulation, climate regulation, and moderation of extreme related to the loss of the thick crown of the dead olive trees that are still present in t (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem servic thetic information; opportunities for recreation and tourism; inspiration for cultu and design; spiritual experience; and information for cognitive development. Th be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ in cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signi cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 t nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 15-Biological control ☹ Not evaluated 😊😊 = Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc 7-Air quality regulation ☹☹ 😊 😊😊 8 Cli t l ti ☹☹ N t l t d 😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From 7-Air quality regulation ☹☹ 😊 😊😊 8 Cli t l ti ☹☹ N t l t d 😊 19-Opportunities for recreation and tourism Sustainability 2021, 13, x FOR PEER REVIEW 10 of 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. 3.3. Step 3: Analysis of Different Scenarios Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 Sustainability 2021, 13, x FOR PEER REVIEW 10 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem vices such as air-quality regulation, climate regulation, and moderation of extreme ev related to the loss of the thick crown of the dead olive trees that are still present in the (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. These be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosyste vices such as air-quality regulation, climate regulation, and moderation of extreme related to the loss of the thick crown of the dead olive trees that are still present in t (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosystem servic thetic information; opportunities for recreation and tourism; inspiration for cultu and design; spiritual experience; and information for cognitive development. Th be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ in cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signi cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 t nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11-Waste treatment = Not evaluated 😊 14-Pollination = Not evaluated 😊😊 = Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Sc Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. 3.3. Step 3: Analysis of Different Scenarios Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From 20-Inspiration for culture, art and design Sustainability 2021, 13, x FOR PEER REVIEW 10 of 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11 W t t t t N t l t d 😊 Sustainability 2021, 13, x FOR PEER REVIEW 10 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem vices such as air-quality regulation, climate regulation, and moderation of extreme ev related to the loss of the thick crown of the dead olive trees that are still present in the (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. These be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11 W t t t t N t l t d 😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosyste vices such as air-quality regulation, climate regulation, and moderation of extreme related to the loss of the thick crown of the dead olive trees that are still present in t (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem servic thetic information; opportunities for recreation and tourism; inspiration for cultu and design; spiritual experience; and information for cognitive development. Th be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ in cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signi cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 t nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 10-Regulation of water flows = ☹ 😊 11 W t t t t N t l t d 😊 = Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment. From Past Scenario to From Scenario 0 to Sce From Sc Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment. From Past Scenario to From Scenario 0 to Sce From 21-Spiritual experience Sustainability 2021, 13, x FOR PEER REVIEW 10 of 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 Sustainability 2021, 13, x FOR PEER REVIEW 10 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem vices such as air-quality regulation, climate regulation, and moderation of extreme ev related to the loss of the thick crown of the dead olive trees that are still present in the (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. These be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosyste vices such as air-quality regulation, climate regulation, and moderation of extreme related to the loss of the thick crown of the dead olive trees that are still present in t (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem servic thetic information; opportunities for recreation and tourism; inspiration for cultu and design; spiritual experience; and information for cognitive development. Th be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ in cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signi cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 t nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8-Climate regulation ☹☹ Not evaluated 😊 9-Moderation of extreme events ☹ ☹ 😊 = Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to f cant judgment Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability cant judgment 22-Information for cognitive development Sustainability 2021, 13, x FOR PEER REVIEW 10 of 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the are (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosystem services: ae thetic information; opportunities for recreation and tourism; inspiration for culture, ar and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to Sce nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8 Climate regulation ☹☹ Not evaluated 😊 Sustainability 2021, 13, x FOR PEER REVIEW 10 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem vices such as air-quality regulation, climate regulation, and moderation of extreme ev related to the loss of the thick crown of the dead olive trees that are still present in the (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. These be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 to S nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8 Climate regulation ☹☹ Not evaluated 😊 Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosyste vices such as air-quality regulation, climate regulation, and moderation of extreme related to the loss of the thick crown of the dead olive trees that are still present in t (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosystem servic thetic information; opportunities for recreation and tourism; inspiration for cultu and design; spiritual experience; and information for cognitive development. Th be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ in cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signi cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From Scenario 1 t nario 2 7-Air quality regulation ☹☹ 😊 😊😊 8 Climate regulation ☹☹ Not evaluated 😊 = Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration and design; spiritual experience; and information for cognitive develop be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive varia cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderatio related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspirat and design; spiritual experience; and information for cognitive deve be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive va cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, Sustainability 2021, 13, x FOR PEER REVIEW 3.4. 3.3. Step 3: Analysis of Different Scenarios , no significant variation. Not evaluated indicates the inability to udgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From S dium variation; 3, strong variation. , no significant variation. Not evaluated indicates the inability nt judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From medium variation; 3, strong variation. , no significant variation. Not evaluated indicates the inabil cant judgment. Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 Fr The most important impacts, however, affect the following ecosystem services: aes- thetic information; opportunities for recreation and tourism; inspiration for culture, art, and design; spiritual experience; and information for cognitive development. These can be considered irreversible transformations. y ( g ) g , were planned for the realization of new land cover characterized by urban elements in an agroecosystem matrix and natural habitat. Therefore, the vegetation is not designed like simple ornamental flora but how new ecosystems are able to create structures and func- tions similar to the land cover characterized by olive groves before the spread of Xylella ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu tion of water flow services because there is an increase in the impermeable surfaces. addition, the presence of dry olive trees and uncultivated trees can create the percepti of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific prior ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, Nb were planned for the realization of new land cover characterized by urban elements in agroecosystem matrix and natural habitat. Therefore, the vegetation is not designed li simple ornamental flora but how new ecosystems are able to create structures and fun tions similar to the land cover characterized by olive groves before the spread of Xyle ture. Therefore, there is no direct information about the relation between vegetation environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios However, without specific m tion actions, the urbanization area reduces the moderation of extreme events and re tion of water flow services because there is an increase in the impermeable surface addition, the presence of dry olive trees and uncultivated trees can create the perce of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific pri ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, were planned for the realization of new land cover characterized by urban elements agroecosystem matrix and natural habitat. Therefore, the vegetation is not designed simple ornamental flora but how new ecosystems are able to create structures and tions similar to the land cover characterized by olive groves before the spread of X Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pres- e of green urban areas structured with shrub and arboreal vegetation. Dead olive trees l be replaced with new trees, promoting biophysical structure able to support pollution orbance such as CO2. The climate regulation service was not evaluated because, even he green urban areas can improve this service, 39% of urbanization of the total area nsidering building construction, roads, and parking areas) can increase local tempera- e. Therefore, there is no direct information about the relation between vegetation and vironment building that indicates a changing value. However, without specific mitiga- n actions, the urbanization area reduces the moderation of extreme events and regula- n of water flow services because there is an increase in the impermeable surfaces. In dition, the presence of dry olive trees and uncultivated trees can create the perception a degraded and abandoned landscape. Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ture. Therefore, there is no direct information about the relation betwe environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios However, witho tion actions, the urbanization area reduces the moderation of extreme e tion of water flow services because there is an increase in the imperm addition, the presence of dry olive trees and uncultivated trees can cre of a degraded and abandoned landscape. Moving from Scenario 0 to Scenario 1, air-quality regulation incre ence of green urban areas structured with shrub and arboreal vegetat will be replaced with new trees, promoting biophysical structure able absorbance such as CO2. The climate regulation service was not eval if the green urban areas can improve this service, 39% of urbanizat (considering building construction, roads, and parking areas) can inc ture. Therefore, there is no direct information about the relation betw environment building that indicates a changing value. However, wit tion actions, the urbanization area reduces the moderation of extrem tion of water flow services because there is an increase in the impe addition, the presence of dry olive trees and uncultivated trees can c of a degraded and abandoned landscape. (considering building construction, roads, and parking areas) can increase local temper ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu tion of water flow services because there is an increase in the impermeable surfaces. addition, the presence of dry olive trees and uncultivated trees can create the percepti of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific prior ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, Nb were planned for the realization of new land cover characterized by urban elements in agroecosystem matrix and natural habitat. Therefore, the vegetation is not designed li simple ornamental flora but how new ecosystems are able to create structures and fun (considering building construction, roads, and parking areas) can increase local tem ture. Therefore, there is no direct information about the relation between vegetation environment building that indicates a changing value. However, without specific m tion actions, the urbanization area reduces the moderation of extreme events and re tion of water flow services because there is an increase in the impermeable surface addition, the presence of dry olive trees and uncultivated trees can create the perce of a degraded and abandoned landscape. 3.3. Step 3: Analysis of Different Scenarios During the SEA process, specific solutions are provided to increase specific pri ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, were planned for the realization of new land cover characterized by urban elements agroecosystem matrix and natural habitat. Therefore, the vegetation is not designed simple ornamental flora but how new ecosystems are able to create structures and Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ture. Therefore, there is no direct information about the relation betwe environment building that indicates a changing value. However, witho tion actions, the urbanization area reduces the moderation of extreme e tion of water flow services because there is an increase in the imperm addition the presence of dry olive trees and uncultivated trees can cre absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu tion of water flow services because there is an increase in the impermeable surfaces. addition, the presence of dry olive trees and uncultivated trees can create the percepti of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific prior ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, Nb were planned for the realization of new land cover characterized by urban elements in t t i d t l h bit t Th f th t ti i t d i d li absorbance such as CO2. The climate regulation service w if the green urban areas can improve this service, 39% o (considering building construction, roads, and parking ar ture. Therefore, there is no direct information about the r environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios H tion actions, the urbanization area reduces the moderatio tion of water flow services because there is an increase addition, the presence of dry olive trees and uncultivated of a degraded and abandoned landscape. During the SEA process, specific solutions are provi ecosystem services (Table 5 and Figure 5). Moving from were planned for the realization of new land cover charac t t i d t l h bit t Th f th ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ture. Therefore, there is no direct information about the relation betwe environment building that indicates a changing value. However, witho tion actions, the urbanization area reduces the moderation of extreme e will be replaced with new trees, promoting biophysical str absorbance such as CO2. The climate regulation service w if the green urban areas can improve this service, 39% o (considering building construction, roads, and parking ar ture. Therefore, there is no direct information about the r environment building that indicates a changing value. H tion actions, the urbanization area reduces the moderatio tion of water flow services because there is an increase addition, the presence of dry olive trees and uncultivated of a degraded and abandoned landscape. During the SEA process, specific solutions are provi ecosystem services (Table 5 and Figure 5). Moving from al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ture. 3.3. Step 3: Analysis of Different Scenarios Therefore, there is no direct information about the relation betwe i t b ildi th t i di t h i l H ith ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if th b i thi i 39% f b i ti m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation incre ence of green urban areas structured with shrub and arboreal vegetat will be replaced with new trees, promoting biophysical structure able absorbance such as CO2. The climate regulation service was not eval if th b i thi i 39% f b i t ive develop ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu i f fl i b h i i i h i bl f tic information ☹☹☹ 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to thetic information ☹☹☹ 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation incre ence of green urban areas structured with shrub and arboreal vegetat will be replaced with new trees, promoting biophysical structure able ☹☹☹ = 😊😊 ive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios However, without specific mitig enance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas f b d i h h b d b l i ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation incre f b d i h h b d b l , art and de ☹☹☹ = 😊😊 ☹☹☹ = 😊😊 ive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ure, art and de ☹☹☹ = 😊😊 e ☹☹☹ = 😊😊 nitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the ence of green urban areas structured with shrub and arboreal vegetation. Dead olive will be replaced with new trees, promoting biophysical structure able to support poll absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service, 39% of urbanization of the total (considering building construction, roads, and parking areas) can increase local tem ulture, art and de ☹☹☹ = 😊😊 nce ☹☹☹ = 😊😊 ognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to t ence of green urban areas structured with shrub and arboreal vegetation. Dead oli will be replaced with new trees, promoting biophysical structure able to support p absorbance such as CO2. 3.3. Step 3: Analysis of Different Scenarios The climate regulation service was not evaluated becau if the green urban areas can improve this service, 39% of urbanization of the to (considering building construction, roads, and parking areas) can increase local te 😊😊 ical control ☹ Not evaluated 😊😊 enance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 ☹☹☹ 😊😊 , art and de- ☹☹☹ = 😊😊 ☹☹☹ = 😊😊 ive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service 39% of urbanization of the total ar ☹☹☹ 😊😊 ure, art and de- ☹☹☹ = 😊😊 e ☹☹☹ = 😊😊 nitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the ence of green urban areas structured with shrub and arboreal vegetation. Dead olive will be replaced with new trees, promoting biophysical structure able to support poll absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service 39% of urbanization of the total ☹☹☹ 😊😊 ulture, art and de- ☹☹☹ = 😊😊 nce ☹☹☹ = 😊😊 ognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to t ence of green urban areas structured with shrub and arboreal vegetation. Dead oli will be replaced with new trees, promoting biophysical structure able to support p absorbance such as CO2. 3.3. Step 3: Analysis of Different Scenarios The climate regulation service was not evaluated becau if the green urban areas can improve this service 39% of urbanization of the to treatment = Not evaluated 😊 tion = Not evaluated 😊😊 ical control ☹ Not evaluated 😊😊 enance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 te treatment = Not evaluated 😊 ination = Not evaluated 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 ☹☹☹ 😊😊 eation and ☹☹☹ = 😊😊 , art and de- ☹☹☹ = 😊😊 ☹☹☹ = 😊😊 ive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti ☹☹☹ 😊😊 ecreation and ☹☹☹ = 😊😊 ure, art and de- ☹☹☹ = 😊😊 e ☹☹☹ = 😊😊 nitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the ence of green urban areas structured with shrub and arboreal vegetation. Dead olive will be replaced with new trees, promoting biophysical structure able to support poll ☹☹☹ 😊😊 r recreation and ☹☹☹ = 😊😊 ulture, art and de- ☹☹☹ = 😊😊 nce ☹☹☹ = 😊😊 ognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to t ence of green urban areas structured with shrub and arboreal vegetation. 3.3. Step 3: Analysis of Different Scenarios Dead oli will be replaced with new trees, promoting biophysical structure able to support p tion of water flows = ☹ 😊 treatment = Not evaluated 😊 tion = Not evaluated 😊😊 ical control ☹ Not evaluated 😊😊 enance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ulation of water flows = ☹ 😊 te treatment = Not evaluated 😊 ination = Not evaluated 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 c diversity ☹ Not evaluated 😊😊 ☹☹☹ = 😊😊 eation and ☹☹☹ = 😊😊 , art and de- ☹☹☹ = 😊😊 ☹☹☹ = 😊😊 ive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation Dead olive tre etic diversity ☹ Not evaluated 😊😊 on ☹☹☹ = 😊😊 ecreation and ☹☹☹ = 😊😊 ure, art and de- ☹☹☹ = 😊😊 e ☹☹☹ = 😊😊 nitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the ence of green urban areas structured with shrub and arboreal vegetation Dead olive genetic diversity ☹ Not evaluated 😊😊 ation ☹☹☹ = 😊😊 r recreation and ☹☹☹ = 😊😊 ulture, art and de- ☹☹☹ = 😊😊 nce ☹☹☹ = 😊😊 ognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to t ence of green urban areas structured with shrub and arboreal vegetation Dead oli regulation ☹☹ Not evaluated 😊 tion of extreme events ☹ ☹ 😊 tion of water flows = ☹ 😊 treatment = Not evaluated 😊 tion = Not evaluated 😊😊 ical control ☹ Not evaluated 😊😊 enance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 ate regulation ☹☹ Not evaluated 😊 eration of extreme events ☹ ☹ 😊 ulation of water flows = ☹ 😊 te treatment = Not evaluated 😊 ination = Not evaluated 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 ☹ Not evaluated 😊😊 c diversity ☹ Not evaluated 😊😊 ☹☹☹ = 😊😊 eation and ☹☹☹ = 😊😊 , art and de- ☹☹☹ = 😊😊 ☹☹☹ = 😊😊 ive develop- ☹☹☹ = 😊😊😊 ☹ Not evaluated 😊😊 etic diversity ☹ Not evaluated 😊😊 on ☹☹☹ = 😊😊 ecreation and ☹☹☹ = 😊😊 ure, art and de- ☹☹☹ = 😊😊 e ☹☹☹ = 😊😊 nitive develop- ☹☹☹ = 😊😊😊 l ☹ Not evaluated 😊😊 genetic diversity ☹ Not evaluated 😊😊 ation ☹☹☹ = 😊😊 r recreation and ☹☹☹ = 😊😊 ulture, art and de- ☹☹☹ = 😊😊 nce ☹☹☹ = 😊😊 ognitive develop- ☹☹☹ = 😊😊😊 lity regulation ☹☹ 😊 😊😊 regulation ☹☹ Not evaluated 😊 tion of extreme events ☹ ☹ 😊 tion of water flows = ☹ 😊 treatment = Not evaluated 😊 tion = Not evaluated 😊😊 ical control ☹ Not evaluated 😊😊 enance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 uality regulation ☹☹ 😊 😊😊 ate regulation ☹☹ Not evaluated 😊 eration of extreme events ☹ ☹ 😊 ulation of water flows = ☹ 😊 te treatment = Not evaluated 😊 ination = Not evaluated 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and m ☹☹☹ = 😊😊 r quality regulation ☹☹ 😊 😊 mate regulation ☹☹ Not evaluated 😊 oderation of extreme events ☹ ☹ 😊 egulation of water flows = ☹ 😊 Waste treatment = Not evaluated 😊 ollination = Not evaluated 😊 ological control ☹ Not evaluated 😊 Maintenance of genetic diversity ☹ Not evaluated 😊 esthetic information ☹☹☹ = 😊 pportunities for recreation and sm ☹☹☹ = 😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the presence of green urban areas structured with shrub and arboreal vegetation. 3.3. Step 3: Analysis of Different Scenarios Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a vices such as air-quality regulation, climate regulation, and modera related to the loss of the thick crown of the dead olive trees that are s (Figure 2 and Table 6). The most important impacts, however, affect the following ec thetic information; opportunities for recreation and tourism; inspi and design; spiritual experience; and information for cognitive de be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive cates negative variation, and the number of emoticons indicates the potential intensity of the variation io a io , e u a i a io a ea e u e e o e a io o e e e e e a egu a tion of water flow services because there is an increase in the impermeable surfaces. In addition, the presence of dry olive trees and uncultivated trees can create the perception of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific priority ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, NbSs tions similar to the land cover charac will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture Therefore there is no direct information about the relation between vegetation an will be replaced with new trees, promoting biophysical structure able to support pollu absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service, 39% of urbanization of the total (considering building construction, roads, and parking areas) can increase local tem ture Therefore there is no direct information about the relation between vegetation ☹☹☹ = 😊😊 elop- ☹☹☹ = 😊😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. 3.3. Step 3: Analysis of Different Scenarios The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar ence of green urban areas structured with shrub and arboreal vegetation. Dead olive will be replaced with new trees, promoting biophysical structure able to support pollu absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service, 39% of urbanization of the total ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. 3.3. Step 3: Analysis of Different Scenarios Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2 The climate regulation service was not evaluated because ev Moving from Scenario 0 to Scenario 1, air-quality regu ence of green urban areas structured with shrub and arbo will be replaced with new trees, promoting biophysical str absorbance such as CO2 The climate regulation service w ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 ment ☹☹☹ = Moving from Scenario 0 to Scenario 1, air-quality regu ence of green urban areas structured with shrub and arbo tunities for recreation and ☹☹☹ = 😊😊 ation for culture, art and de- ☹☹☹ = 😊😊 nance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ti f lt t d d ical control ☹ Not evaluated 😊😊 nance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 tunities for recreation and ☹☹☹ = 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 ortunities for recreation and ☹☹☹ = 😊😊 , art and de- ☹☹☹ = 😊😊 ☹☹☹ = 😊😊 ive develop tion = Not evaluated 😊😊 ical control ☹ Not evaluated 😊😊 nance of genetic diversity ☹ Not evaluated 😊😊 tic information ☹☹☹ = 😊😊 ination = Not evaluated 😊😊 ogical control ☹ Not evaluated 😊😊 ntenance of genetic diversity ☹ Not evaluated 😊😊 thetic information ☹☹☹ = 😊😊 eation and ☹☹☹ = 😊😊 , art and de- ☹☹☹ = 😊😊 tion of water flows = ☹ 😊 treatment = Not evaluated 😊 tion = Not evaluated 😊😊 ical control ☹ Not evaluated 😊😊 ulation of water flows = ☹ 😊 te treatment = Not evaluated 😊 ination = Not evaluated 😊😊 ogical control ☹ Not evaluated 😊😊 ☹☹☹ = 😊😊 eation and ☹☹☹ = 😊😊 , art and de- on ☹☹☹ = 😊😊 ecreation and ☹☹☹ = 😊😊 ure, art and de- ation ☹☹☹ = 😊😊 r recreation and ☹☹☹ = 😊😊 ulture, art and de- tion of extreme events ☹ ☹ 😊 tion of water flows = ☹ 😊 treatment = Not evaluated 😊 tion = Not evaluated 😊😊 eration of extreme events ☹ ☹ 😊 ulation of water flows = ☹ 😊 te treatment = Not evaluated 😊 ination = Not evaluated 😊😊 ☹ Not evaluated 😊😊 c diversity ☹ Not evaluated 😊😊 ☹☹☹ = 😊😊 eation and ☹☹☹ = 😊😊 ☹ Not evaluated 😊😊 etic diversity ☹ Not evaluated 😊😊 on ☹☹☹ = 😊😊 ecreation and ☹☹☹ = 😊😊 l ☹ Not evaluated 😊😊 genetic diversity ☹ Not evaluated 😊😊 ation ☹☹☹ = 😊😊 r recreation and ☹☹☹ = 😊😊 lity regulation ☹☹ 😊 😊😊 regulation ☹☹ Not evaluated 😊 tion of extreme events ☹ ☹ 😊 tion of water flows = ☹ 😊 t t t N t l t d 😊 uality regulation ☹☹ 😊 😊😊 ate regulation ☹☹ Not evaluated 😊 eration of extreme events ☹ ☹ 😊 ulation of water flows = ☹ 😊 t t t t N t l t d 😊 = Not evaluated 😊😊 ☹ Not evaluated 😊😊 c diversity ☹ Not evaluated 😊😊 ☹☹☹ = 😊😊 = Not evaluated 😊😊 ☹ Not evaluated 😊😊 etic diversity ☹ Not evaluated 😊😊 on ☹☹☹ = 😊😊 = Not evaluated 😊😊 l ☹ Not evaluated 😊😊 genetic diversity ☹ Not evaluated 😊😊 ation ☹☹☹ = 😊😊 yp Scenario 0 nario 1 lity regulation ☹☹ 😊 😊😊 regulation ☹☹ Not evaluated 😊 tion of extreme events ☹ ☹ 😊 yp Scenario 0 nario 1 uality regulation ☹☹ 😊 😊😊 ate regulation ☹☹ Not evaluated 😊 eration of extreme events ☹ ☹ 😊 o ☹ 😊 = Not evaluated 😊 = Not evaluated 😊😊 ☹ Not evaluated 😊😊 di it ☹ N t l t d 😊😊 o ☹ 😊 = Not evaluated 😊 = Not evaluated 😊😊 ☹ Not evaluated 😊😊 ti di it ☹ N t l t d 😊😊 e o ☹ 😊 = Not evaluated 😊 = Not evaluated 😊😊 l ☹ Not evaluated 😊😊 ti di it ☹ N t l t d 😊😊 Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From S lity regulation ☹☹ 😊 😊😊 regulation ☹☹ Not evaluated 😊 Main Services Type From Past Scenario to Scenario 0 From Scenario 0 to Sce- nario 1 From uality regulation ☹☹ 😊 😊😊 ate regulation ☹☹ Not evaluated 😊 events ☹ ☹ 😊 ows = ☹ 😊 = Not evaluated 😊 = Not evaluated 😊😊 me events ☹ ☹ 😊 r flows = ☹ 😊 = Not evaluated 😊 = Not evaluated 😊😊 reme events ☹ ☹ 😊 ter flows = ☹ 😊 = Not evaluated 😊 = Not evaluated 😊😊 um variation; 3, strong variation. 3.3. Step 3: Analysis of Different Scenarios Dead olive trees will be replaced with new trees, promoting biophysical structure able to support pollution absorbance such as CO2. The climate regulation service was not evaluated because, even if the green urban areas can improve this service, 39% of urbanization of the total area (considering building construction, roads, and parking areas) can increase local temperature. Therefore, there is no direct information about the relation between vegetation and environment building that indicates a changing value. However, without speciﬁc mitigation actions, the urbanization area reduces the moderation of extreme events and regulation of water ﬂow services because there is an increase in the impermeable surfaces. In addition, the presence of dry olive trees and uncultivated trees can create the perception of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific priority system services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, NbSs re planned for the realization of new land cover characterized by urban elements in an oecosystem matrix and natural habitat. Therefore, the vegetation is not designed like mple ornamental flora but how new ecosystems are able to create structures and func- During the SEA process, specific solutions are provided to increas ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to were planned for the realization of new land cover characterized by urb agroecosystem matrix and natural habitat. Therefore, the vegetation is simple ornamental flora but how new ecosystems are able to create str During the SEA process, specific solutions are provided to incre ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 were planned for the realization of new land cover characterized by u agroecosystem matrix and natural habitat. Therefore, the vegetation simple ornamental flora but how new ecosystems are able to create p y tions similar to the land cover characterized by olive groves before the spread of Xyle p y tions similar to the land cover characterized by olive groves before the spread of X addition, the presence of dry olive trees and uncultivated trees can crea of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increas ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to were planned for the realization of new land cover characterized by urb agroecosystem matrix and natural habitat Therefore the vegetation is agroecosystem matrix and natural habitat. 3.3. Step 3: Analysis of Different Scenarios Therefore, the vegetation is not designed li simple ornamental flora but how new ecosystems are able to create structures and fun tions similar to the land cover characterized by olive groves before the spread of Xyle agroecosystem matrix and natural habitat. Therefore, the simple ornamental flora but how new ecosystems are ab tions similar to the land cover characterized by olive gro tion of water flow services because there is an increase in the imperm addition, the presence of dry olive trees and uncultivated trees can crea of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increas ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to y ( g ) g were planned for the realization of new land cover charac agroecosystem matrix and natural habitat. Therefore, the simple ornamental flora but how new ecosystems are ab tions similar to the land cover characterized by olive gro environment building that indicates a changing value. However, witho tion actions, the urbanization area reduces the moderation of extreme e tion of water flow services because there is an increase in the imperm addition, the presence of dry olive trees and uncultivated trees can crea of a degraded and abandoned landscape. During the SEA process specific solutions are provided to increas ( g g , , p g ) ture. Therefore, there is no direct information about the relation betwe environment building that indicates a changing value. However, witho tion actions, the urbanization area reduces the moderation of extreme e tion of water flow services because there is an increase in the imperm addition, the presence of dry olive trees and uncultivated trees can crea if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ture. Therefore, there is no direct information about the relation betwe environment building that indicates a changing value. However, witho tion actions, the urbanization area reduces the moderation of extreme e tion of water flow services because there is an increase in the imperm if the green urban areas can improve this service, 39% of urbanizat (considering building construction, roads, and parking areas) can inc ture. Therefore, there is no direct information about the relation betw environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios However, wit tion actions, the urbanization area reduces the moderation of extrem tion of water flow services because there is an increase in the impe tion of water flow services because there is an increase in the impermeable surfaces. addition, the presence of dry olive trees and uncultivated trees can create the percepti of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific prior ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, Nb l d f th li ti f l d h t i d b b l t i p , p g p y absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre ture. Therefore, there is no direct information about the relation betwe environment building that indicates a changing value. However, witho p , p g p y absorbance such as CO2. The climate regulation service was not eval if the green urban areas can improve this service, 39% of urbanizat (considering building construction, roads, and parking areas) can inc ture. Therefore, there is no direct information about the relation betw environment building that indicates a changing value. However, wit environment building that indicates a changing value. However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu tion of water flow services because there is an increase in the impermeable surfaces. addition, the presence of dry olive trees and uncultivated trees can create the percepti of a degraded and abandoned landscape. During the SEA process, specific solutions are provided to increase specific prior ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization (considering building construction, roads, and parking areas) can incre t Th f th i di t i f ti b t th l ti b t ence of green urban areas structured with shrub and arboreal vegetat will be replaced with new trees, promoting biophysical structure able absorbance such as CO2. 3.3. Step 3: Analysis of Different Scenarios The climate regulation service was not eval if the green urban areas can improve this service, 39% of urbanizat (considering building construction, roads, and parking areas) can inc t Th f th i di t i f ti b t th l ti b t ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu tion of water flow services because there is an increase in the impermeable surfaces. addition, the presence of dry olive trees and uncultivated trees can create the percepti f d d d d b d d l d ture. Therefore, there is no direct information about the relation between vegetation environment building that indicates a changing value. However, without specific m tion actions, the urbanization area reduces the moderation of extreme events and re tion of water flow services because there is an increase in the impermeable surface addition, the presence of dry olive trees and uncultivated trees can create the percep f d d d d b d d l d ture. Therefore, there is no direct information about the relation between vegetat environment building that indicates a changing value. However, without specific tion actions, the urbanization area reduces the moderation of extreme events and tion of water flow services because there is an increase in the impermeable surf addition, the presence of dry olive trees and uncultivated trees can create the per f d d d d b d d l d Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to absorbance such as CO2. The climate regulation service was not evalua if the green urban areas can improve this service, 39% of urbanization Moving from Scenario 0 to Scenario 1, air-quality regulation incre ence of green urban areas structured with shrub and arboreal vegetat will be replaced with new trees, promoting biophysical structure able absorbance such as CO2. 3.3. Step 3: Analysis of Different Scenarios The climate regulation service was not eval if the green urban areas can improve this service, 39% of urbanizat if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. However, without specific mitig tion actions, the urbanization area reduces the moderation of extreme events and regu tion of water flow services because there is an increase in the impermeable surfaces. if the green urban areas can improve this service, 39% of urbanization of the total (considering building construction, roads, and parking areas) can increase local tem ture. Therefore, there is no direct information about the relation between vegetation environment building that indicates a changing value. However, without specific m tion actions, the urbanization area reduces the moderation of extreme events and re tion of water flow services because there is an increase in the impermeable surface if the green urban areas can improve this service, 39% of urbanization of the to (considering building construction, roads, and parking areas) can increase local te ture. Therefore, there is no direct information about the relation between vegetat environment building that indicates a changing value. However, without specific tion actions, the urbanization area reduces the moderation of extreme events and tion of water flow services because there is an increase in the impermeable surf ☹☹☹ 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation will be replaced with new trees, promoting biophysical structure able to b b h CO Th li l i i l ☹☹☹ 😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation incre ence of green urban areas structured with shrub and arboreal vegetat will be replaced with new trees, promoting biophysical structure able b b h CO Th li l i i l absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture. Therefore, there is no direct information about the relation between vegetation an environment building that indicates a changing value. 3.3. Step 3: Analysis of Different Scenarios However, without specific mitig absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service, 39% of urbanization of the total (considering building construction, roads, and parking areas) can increase local tem ture. Therefore, there is no direct information about the relation between vegetation environment building that indicates a changing value. However, without specific m absorbance such as CO2. The climate regulation service was not evaluated becaus if the green urban areas can improve this service, 39% of urbanization of the to (considering building construction, roads, and parking areas) can increase local te ture. Therefore, there is no direct information about the relation between vegetat environment building that indicates a changing value. However, without specific p ☹☹☹ 😊😊 mation for cognitive develop- ☹☹☹ = 😊😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increas ence of green urban areas structured with shrub and arboreal vegetation p ☹☹☹ 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation incre ence of green urban areas structured with shrub and arboreal vegetat ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar (considering building construction, roads, and parking areas) can increase local temper ture Therefore there is no direct information about the relation between vegetation an ence of green urban areas structured with shrub and arboreal vegetation. Dead olive will be replaced with new trees, promoting biophysical structure able to support pollu absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service, 39% of urbanization of the total (considering building construction, roads, and parking areas) can increase local tem ture Therefore there is no direct information about the relation between vegetation ence of green urban areas structured with shrub and arboreal vegetation. Dead oli will be replaced with new trees, promoting biophysical structure able to support p absorbance such as CO2. 3.3. Step 3: Analysis of Different Scenarios The climate regulation service was not evaluated becaus if the green urban areas can improve this service, 39% of urbanization of the to (considering building construction, roads, and parking areas) can increase local te ture Therefore there is no direct information about the relation between vegetat ☹☹☹ 😊😊 al experience ☹☹☹ = 😊😊 mation for cognitive develop- ☹☹☹ = 😊😊😊 ☹☹☹ 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the pre ence of green urban areas structured with shrub and arboreal vegetation. Dead olive tre will be replaced with new trees, promoting biophysical structure able to support polluti absorbance such as CO2. The climate regulation service was not evaluated because, ev if the green urban areas can improve this service, 39% of urbanization of the total ar Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to the ence of green urban areas structured with shrub and arboreal vegetation. Dead olive will be replaced with new trees, promoting biophysical structure able to support pollu absorbance such as CO2. The climate regulation service was not evaluated because, if the green urban areas can improve this service, 39% of urbanization of the total Moving from Scenario 0 to Scenario 1, air-quality regulation increases due to t ence of green urban areas structured with shrub and arboreal vegetation. Dead oli will be replaced with new trees, promoting biophysical structure able to support p absorbance such as CO2. The climate regulation service was not evaluated becaus if the green urban areas can improve this service, 39% of urbanization of the to ation for culture, art and de- ☹☹☹ = 😊😊 al experience ☹☹☹ = 😊😊 mation for cognitive develop- ☹☹☹ = 😊😊😊 iration for culture, art and de- ☹☹☹ = 😊😊 itual experience ☹☹☹ = 😊😊 rmation for cognitive develop- ☹☹☹ = 😊😊 nspiration for culture, art and de- ☹☹☹ = 😊 piritual experience ☹☹☹ = 😊 nformation for cognitive develop- t ☹☹☹ = 😊 During the SEA process, speciﬁc solutions are provided to increase speciﬁc priority ecosystem services (Table 5 and Figure 5). Moving from Scenario 1 to Scenario 2, NbSs were planned for the realization of new land cover characterized by urban elements in an agroecosystem matrix and natural habitat. 3.3. Step 3: Analysis of Different Scenarios The most important impacts however affect the following ecosystem vices such as air quality regulation, climate regulation, and moderation of extre related to the loss of the thick crown of the dead olive trees that are still present vices such as air quality regulation, climate regulation, and moderation of ex related to the loss of the thick crown of the dead olive trees that are still prese y 2021, 13, x FOR PEER REVIEW The transition from the Past scenario to Scenario 0 produces a loss of ecos vices such as air quality regulation climate regulation and moderation of extre The transition from the Past scenario to Scen vices such as air quality regulation climate regul 3.4. Step 4: Ecosystem Service Impact Assessment 3.4. Step 4: Ecosystem Service Impact Assessment vices such as air-quality regulation, climate regulation related to the loss of the thick crown of the dead olive t (Figure 2 and Table 6). The most important impacts however (Figure 2 and Table 6). The most important impacts howev vices such as air quality regulation, climate related to the loss of the thick crown of the d vices such as air quality regulation, clima related to the loss of the thick crown of th ty 2021, 13, x FOR PEER REVIEW The transition from the Past scenario to vices such as air quality regulation climate The trans vices such as a 3.4. Step 4: Eco The transition from the Past scenario to Scenario 0 produces a loss of ecosystem services such as air-quality regulation, climate regulation, and moderation of extreme events related to the loss of the thick crown of the dead olive trees that are still present in the area (Figure 2 and Table 6). e a e o e o o e i o o e ea o i e ee a a e i p e e i e a ea (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: aes- thetic information; opportunities for recreation and tourism; inspiration for culture, art, and design; spiritual experience; and information for cognitive development. These can be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenario cates negative variation, and the number of emoticons indicates the p medium variation; 3, strong variation. =, no significant variation. “No cant judgment. 3.3. Step 3: Analysis of Different Scenarios Main Services Type From Past Scenario to S i 0 The most important impacts, however, affect the following ecosystem services: a thetic information; opportunities for recreation and tourism; inspiration for culture, a and design; spiritual experience; and information for cognitive development. These c be considered irreversible transformations. The most important impacts, however, affect the following ecosystem services thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. Thes be considered irreversible transformations. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem ser- es such as air-quality regulation, climate regulation, and moderation of extreme events ated to the loss of the thick crown of the dead olive trees that are still present in the area 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation related to the loss of the thick crown of the dead olive trees that are still p 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a lo vices such as air-quality regulation, climate regulation, and moderati related to the loss of the thick crown of the dead olive trees that are sti (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: a thetic information; opportunities for recreation and tourism; inspiration for culture, a and design; spiritual experience; and information for cognitive development. These c b id d i ibl t f ti (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services thetic information; opportunities for recreation and tourism; inspiration for culture and design; spiritual experience; and information for cognitive development. Thes b id d i ibl t f ti 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the ar (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss vices such as air quality regulation climate regulation and moderation 3.4. Step 4: Ecosystem Service Impact Assessment Sustainability 2021, 13, x FOR PEER REVIEW Sustainability 2021, 13, x FOR PEER REVIEW The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even related to the loss of the thick crown of the dead olive trees that are still present in the ar (Fi 2 d T bl 6) Sustainability 2021, 13, x FOR PEER REVIEW Sustainability 2021, 13, x FOR PEER REVIEW 3.4. Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 produces a loss of ecosystem se vices such as air-quality regulation, climate regulation, and moderation of extreme even 3.4. Step 4: Ecosystem Service Impact Assessment Th f h P S 0 d l f 3.4. Step 4: Ecosystem Service Impact Assessment Th f h P S 0 d l f 3.4. Step 4: Ecosystem Service Impact Assessment Th f h P S 0 d l f Sustainability 2021, 13, x FOR PEER REVIEW 10 of Sustainability 2021, 13, x FOR PEER REVIEW 10 Sustainability 2021, 13, x FOR PEER REVIEW Sustainability 2021, 13, x FOR PEER REVIEW 10 of Sustainability 2021, 13, x FOR PEER REVIEW 10 Sustainability 2021, 13, x FOR PEER REVIEW Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. indicates positive variation, Scenario 0 7-Air quality regulation ☹☹ 😊 8-Climate regulation ☹☹ 9-Moderation of extreme events ☹ ☹ 10-Regulation of water flows = ☹ indicates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no signiﬁcant variation. “Not evaluated” indicates the inability to form a signiﬁcant judgment. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 3.3. Step 3: Analysis of Different Scenarios 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. related to the loss of the thick crown of the dead olive trees that are still present in the area (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: aes- thetic information; opportunities for recreation and tourism; inspiration for culture, art, related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosyst thetic information; opportunities for recreation and tourism; inspiration related to the loss of the thick crown of the dead olive trees that are stil (Figure 2 and Table 6). The most important impacts, however, affect the following ecosy thetic information; opportunities for recreation and tourism; inspirat be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, d f “ l d” d h b l f f be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, d f “ l d” d h b l f f vices such as air-quality regulation, climate regulation, and moderation o related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts however affect the following ecosyst and design; spiritual experience; and information for cognitive development. These ca be considered irreversible transformations. 3.3. Step 3: Analysis of Different Scenarios The most important impacts, however, affect the following ecosystem services: a thetic information; opportunities for recreation and tourism; inspiration for culture, a vices such as air-quality regulation, climate regulation, an related to the loss of the thick crown of the dead olive tree (Figure 2 and Table 6). The most important impacts, however, affect the fo thetic information; opportunities for recreation and tour y 2021, 13, x FOR PEER REVIEW 3 4 Step 4: Ecosystem Service Impact Assessment The transition from the Past scenario to Scenario 0 p vices such as air-quality regulation, climate regulation, an related to the loss of the thick crown of the dead olive tree (Figure 2 and Table 6). Th i i h ff h f y 2021, 13, x FOR PEER REVIEW y 2021, 13, x FOR PEER REVIEW EER REVIEW 10 of 3.4. Step 4: Ecosystem Service Impact Assessment EER REVIEW 10 of EER REVIEW 10 of R PEER REVIEW 10 FOR PEER REVIEW Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. indicates positive variation, Scenario 0 7-Air quality regulation ☹☹ 😊 8-Climate regulation ☹☹ 9-Moderation of extreme events ☹ ☹ 10-Regulation of water flows = ☹ indicates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no signiﬁcant variation. “Not evaluated” indicates the inability to form a signiﬁcant judgment. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. “Not evaluated” indicates the inability to form a signifi- cant judgment. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, medium variation; 3, strong variation. =, no significant variation. 3.3. Step 3: Analysis of Different Scenarios “Not evaluated” indicates the inability to form a signifi- cant judgment. related to the loss of the thick crown of the dead olive trees that are still present in the area (Figure 2 and Table 6). The most important impacts, however, affect the following ecosystem services: aes- thetic information; opportunities for recreation and tourism; inspiration for culture, art, related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts, however, affect the following ecosys thetic information; opportunities for recreation and tourism; inspiratio related to the loss of the thick crown of the dead olive trees that are sti (Figure 2 and Table 6). The most important impacts, however, affect the following ecos thetic information; opportunities for recreation and tourism; inspira be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, d f l d d h b l f f be considered irreversible transformations. Table 6. Qualitative variation in ecosystem services from the scenarios analyzed. 😊 indicates positive variation, ☹ indi- cates negative variation, and the number of emoticons indicates the potential intensity of the variation: 1, low variation; 2, d f l d d h b l f f vices such as air-quality regulation, climate regulation, and moderation related to the loss of the thick crown of the dead olive trees that are still p (Figure 2 and Table 6). The most important impacts however affect the following ecosys and design; spiritual experience; and information for cognitive development. These c be considered irreversible transformations. Table 6 Qualitative variation in ecosystem services from the scenarios analyzed 😊indicates positive variation ☹indi and design; spiritual experience; and information for cog be considered irreversible transformations. Table 6 Qualitative variation in ecosystem services from the scenarios analyzed 😊indicat p y p The transition from the Past scenario to Scenario 0 produces a loss vices such as air-quality regulation, climate regulation, and moderation related to the loss of the thick crown of the dead olive trees that are still p The most important impacts, however, affect the fol thetic information; opportunities for recreation and tour and design; spiritual experience; and information for cog be considered irreversible transformations. 3.4. 3.3. Step 3: Analysis of Different Scenarios This was foreseen by national and regional law [53]. Therefore, in the future, it will be necessary to replace olive trees that will be placed in public or private green space. In Scenario 2, we integrated the sub-urban plan with speciﬁc natural-based solutions to increase ecological functions able to support priority ecosystem services. The urban plan envisages replacing the olive trees with other trees to retain the agricultural value of the area. Mainly, these areas can be destined for the development of community urban gardens and agriculture park. The adoption of the strategies shown in Table 4 could reduce the impact of urban land-use change. In the case of building construction, mitigation solutions include green roofs and green walls. Considering the original land use (arable land), these strategies can increase the provision of priority ecosystem services. For the parking area, we foresaw the use of lawn ﬂooring, and for private and public space, we foresaw the design of urban community gardens to try to recover the loss of cultural value due to drying of the olive groves. Moreover, the mitigation solutions can enhance green infrastructure by creating speciﬁc habitats capable of increasing priority ecosystem services identiﬁed according to environmental objectives. In this case, the amount of green space, the typologies of the green spaces, and the vegetation used are important. This is the main difference between Scenario 1 and Scenario 2. Sustainability 2021, 13, 122 16 of 25 ick crown o system s cosystem 16 of 25 ick crown o system s cosystem 3.4. Step 4: Ecosystem Service Impact Assessment vices such as air-quality regulation, climate regulation, and moderation of extreme events related to the loss of the thick crown of the dead olive trees that are still present in the area g p p be considered irreversible (Figure 2 and Table 6). The most important impacts however affect the following ecosystem ser (Figure 2 and Table 6). 3.3. Step 3: Analysis of Different Scenarios Therefore, the vegetation is not designed like simple ornamental ﬂora but how new ecosystems are able to create structures and functions Sustainability 2021, 13, 122 17 of 25 17 of 25 similar to the land cover characterized by olive groves before the spread of Xylella fastidiosa. Mainly, the new land-cover was designed to implement priority ecosystem services. similar to the land cover characterized by olive groves before the spread of Xylella fastidiosa. Mainly, the new land-cover was designed to implement priority ecosystem services. In particular, NbSs increase the permeable surface of the area thanks to the imple- mentation of green roofs. This can improve ecosystem services such as climate regulation, moderation of extreme events, and regulation of water ﬂows. The presence of green areas integrated with ﬂood areas can help implement these ecosystem services and can improve the biodiversity of the area. In particular, these two areas were designed to try to create a continuous habitat between private green urban areas and public green urban areas. The forecasts for urban agricultural parks and urban gardens can improve the cultural aspect of the area linked to the agricultural landscape. These cannot introduce the same cultural aspects as the monumental olive groves, but they can contribute to providing new value and cultural functionality of the area. Therefore, the transition from Scenario 1 to Scenario 2 is favorable and should be supported. However, the presence of community gardens can improve social aspects like psycho- logical well-being, reinforcing the sense of community with respect the actual status [76]. In a social-ecological system, it is important to have not only these cultural elements but also the accessibility and possibility of stakeholders using them [27]. GI supports main- taining the cultural agricultural value. In this case, it is important to choose agricultural vegetation to implement social and educational values. Table 7 shows the historical fruit, aromatics, and medicinal ﬂora identiﬁed for these community gardens. The peri-urban area is located in the context compromised by Xylella fastidiosa; therefore, the species that are hosts of Xylella fastidiosa (subsp. pacua) can be reproduced directly from seed, such as for Lavandula ofﬁcinalis L. and Rosmarinus ofﬁcinalis L. [77]. Table 7. Examples of the vegetation proposed for community gardens [50–52]. Historical Fruit Aromatics and Medicinal Flora Ceratonia siliqua L Hypericum perforatum L. Ficus carica L. Lavandula ofﬁcinalis L. Malus domestica L. Melissa ofﬁcinalis L. Mespilus germanica L. Menta spicata L. subsp. 3.3. Step 3: Analysis of Different Scenarios Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: ( atonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenera w, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the rms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bi 4. Discussion Figure 6. Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: (2) monumental Ceratonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenerate the esthetical view, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the medium and long terms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bike tour. The peri-urban plan results in the conversion of arable land to construc networks, but this is an intrinsic aspect of the urbanization processes. There cept of ESA in the SEA must contribute to reducing the negative effects of la formation by providing indications to develop the best solutions, reducing th d b h b l d In conclusion, the enhancement of identity can undergo the rediscovery of social and cultural values that have faded over time and are currently reinterpretable only through the testimonies of those who have experienced those times, free books, and information on the Internet. 3.3. Step 3: Analysis of Different Scenarios glabrata (Lej & Court) Lebeau Opuntia ﬁcus- indica L. Rosmarinus ofﬁcinalis L. Punica granatum L. Salvia ofﬁcinalis L. Zyziphus sativus L. Satureja cuneifolia Ten. Cydonia oblonga L. Thymus capitata (L.) Cav. Morus alba L. Ruta graveolens L. Morus nigra L. Capparis spinosa L. Origanum vulgare L. Table 7. Examples of the vegetation proposed for community gardens [50–52]. These urban gardens can therefore represent a point of reference for educational and social activities, which can also strengthen the neighborhood community [76]. In addition, these gardens can constitute a constant safeguard by the population over time, allowing increasing urban resilience to abandonment and degradation phenomena when socioeconomic conditions are no longer favorable. In addition, they avoid undesirable abandoned lots [76]. Some proposed plants have a cultural and landscape value similar to the monumen- tal olive tree but that is less known. Ceratonia siliqua L-Carob could be compared with monumental olive trees both for its beauty and for its high cultural value in the landscape because its fruit was used in Southern Italy when food was scarce (Figure 6). The carob tree can provide the characteristics of monumentality and beauty over time, comparable to the olive trees (Figure 6). Finally, the carob tree has edaphic and water needs similar to the Sustainability 2021, 13, 122 18 of 25 18 of 25 olive tree. Therefore, this species ﬁts perfectly into the Salento landscape to help restore the biophysical structure that characterizes the agricultural landscape. cultural values that have faded over time and are currently reinterpretable the testimonies of those who have experienced those times, free books, and on the Internet. olive tree. Therefore, this species ﬁts perfectly into the Salento landscape to help restore the biophysical structure that characterizes the agricultural landscape. cultural values that have faded over time and are currently reinterpretable the testimonies of those who have experienced those times, free books, and on the Internet. Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: (2) eratonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenerate iew, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the m terms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bike Figure 6. 3.3. Step 3: Analysis of Different Scenarios Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: (2) monumental Ceratonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenerate the esthetical view, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the medium and long terms; (3) a scene of a group of i h h f i f C t i ili l h i i f bik Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: ratonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenera ew, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the erms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bi 4. Discussion Figure 6. Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: (2) monumental Ceratonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenerate the esthetical view, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the medium and long terms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bike tour. omparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: ( atonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenera w, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the rms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bi 4. Discussion Figure 6. Comparison between monumental Ceratonia siliqua L and monumental olive trees in the landscape: (2) monumental Ceratonia siliqua L; (1) monumental olive trees, highlighting the capacity of new vegetation to regenerate the esthetical view, the opportunities for recreation and tourism, and the spiritual experience of the landscape in the medium and long terms; (3) a scene of a group of tourists who taste the fruit of Ceratonia siliqua along the itinerary of a bike tour. 4. Discussion ing both public and private spaces. The enhancement of priority ecosystem services assessed for this plan achieve the environmental objectives set in the first part of the SEA process; sub-urban plan can be considered largely consistent with the hierarchical PP inspired the environmental objectives. This phase is important because it can The peri-urban plan results in the conversion of arable land to construction and road networks, but this is an intrinsic aspect of the urbanization processes. Therefore, the concept of ESA in the SEA must contribute to reducing the negative effects of land use transformation by providing indications to develop the best solutions, reducing the loss of ecosystem services generated by the built environment and integrating natural- Sustainability 2021, 13, 122 19 of 25 19 of 25 based solutions to increase ecosystem service provision. These actions should be selected considering both public and private spaces. based solutions to increase ecosystem service provision. These actions should be selected considering both public and private spaces. The enhancement of priority ecosystem services assessed for this plan can help to achieve the environmental objectives set in the ﬁrst part of the SEA process; therefore, the sub-urban plan can be considered largely consistent with the hierarchical PPPs that have inspired the environmental objectives. This phase is important because it can inﬂuence all phases of the SEA process, mainly the design of mitigation actions like green infrastructure, the analysis of different scenarios, and the choices that will determine future transforma- tions of the area. This increases the coherence between bottom-up planning and top-down planning, reinforcing the sustainability of human chose and actions. p g g y However, the analysis of ecosystem services is not a substitute for evaluations of the effects caused by waste production, mobility, quality of the area, use of water resources, etc. ESAs should be considered as a tool to better understand the relationships between biodiversity, ecological functions, and land use transformation and to avoid unintended negative consequences on the provision of ecosystem services and human well-being. Therefore, the analysis of ecosystem services in the SEA process must be considered complementary to other analysis and methodologies and not preponderant or exclusive. For example, in this case, the analysis of ecosystem services should be weighed with other social and economic factors, such as the needs of houses for the population, especially those with economic difﬁculties. 4. Discussion For example, if the residences that will be built are destined to be public housing, they can support homeless families. The use of ecosystem services within the SEA can be considered an additional tool to bring designers closer to ecological and social issues not yet widespread in sub-urban planning. Above all, it helps to perceive actions aimed at sustainability as important for obtaining a quality and more attractive plan and not as an additional cost for planners. g q y p p The difference between the proposed method of using the ES concept in the SEA process on the sub-urban scale and the existing methods is the accuracy of the evaluation and the design of the ecosystem services. The many difﬁculties facing current methods of evaluation of ecosystem services develop on the regional or local scale in speciﬁc peri- urban areas. Many landscape-level ecosystem service assessments in SEA are performed considering the biophysical structure and ecological functions [46,47] without a focus on the speciﬁc characteristics of vegetation and human intentions. The assessment approach on a ﬁner scale needs to incorporate more complexity in the analysis [2,78]. For example, the olive groves, on an urban or regional scale, can be considered homogeneous. However, on a sub-urban scale, they can be distinguished on the basis of age, structure, health status, understory vegetation, productive or nonproductive, and type of variety. All this information can be used to discriminate the ability to support ecosystem services. This example also extends to other land use classes. p In this study on the sub-urban scale, the ecosystem services approach is inﬂuenced by the skills needed to assign value and priority to speciﬁc ecosystem services considering spe- ciﬁc use of the area. We had to adapt the existing ecosystem service assessment approach to vegetation knowledge (type, specify ecological processes that can be supported, ecological role, and human intentions in future use). Mainly, some ecosystem services highlighted in the literature for land cover like olive groves were corrected considering the status of Xylella fastidiosa and the characteristic of the olive trees. For example, concerning herba- ceous vegetation and shrubs, we considered the capacity of the detected species to support regulating services and others. Therefore, an important contribution to the assessment process was to directly involve different experts such as botanists, agronomists, faunists, ecologists, planners, decision-makers, and different public institutions that participate in SEA processes. 4. Discussion p Peri-urban plans are more inﬂuenced by human interests in land use than urban or regional plans. Therefore, the SEA process needs to assess the effect of speciﬁc land use changes, providing more information on the types of building, localization, speciﬁc interaction with the vegetation, and ecosystem services, which is not possible on the urban Sustainability 2021, 13, 122 20 of 25 20 of 25 and regional scales. These plans provide general information about potential landscape or urban transformation driving and conditioning the type of transformation at the sub-urban scale, but plans at the sub-urban scale need to be actuated. Therefore, we have more detailed information to consider and the ﬁnal decision has a direct effect on landscape transformation. Hence, in this study, the applied ecosystem service approach is not limited to simple assessment but can be extended to designing priority ecosystem services by planning speciﬁc green infrastructures to increase social beneﬁts. This is important because ecosystem service design is directly included in peri-urban plan formation. Differently, mitigation actions are often speciﬁc sections of SEA processes which follow the scenarios’ analyses, indirectly linked with peri-urban plan formation. As such, ecosystem service enhancement is an active part of the plan’s development process and not just a part of SEA processes, including them in the analysis of different scenarios to help ﬁnd the best solution for human well-being. In this paper, we provided ideas about how the ecosystem service concept can be integrated into the workﬂow of the SEA process and the positive effect of integration of vegetation in the urban infrastructure with the possibility of developing speciﬁc green infrastructure. However, in the speciﬁc urban contexts and plan, it is necessary to consider the potential ecosystem disservices that can be produced by vegetation, like the production of allergens [79]. For example, green roofs and urban community gardens may have negative ecological effects or may create ecosystem disservices. Nitrogen, phosphorus, and potassium, for example, may accumulate in high concentrations in garden soil due to the indiscriminate application of fertilizers or compost, polluting urban stormwater runoff or groundwater [37,78]. An important element is the temporal projection of the green infrastructure that is developed. In this study, the cultural amenity services were considered priority ecosystem services given the past characteristics of the study area, where beautiful monumental olive trees were present, which have now been destroyed by Xylella fastidiosa. 4. Discussion The urban plan has to regenerate the cultural value, potentially by developing speciﬁc strategies, and the SEA has to value this with a strategic vision in the future. The sustainable concept upon which ESA is based focuses on future generations; therefore, sustainable cultural strategies have to promote the transformation developed from the sub-urban plan over time. Therefore, the cultural strategies have to be developed by considering how the area can evolve in the future. In this study, for example, it is impossible to quickly restore the cultural value of monumental olive trees However, considering the mitigation actions adopted, within a 50-year timeframe, the landscape structure can partially recover the cultural value of olive trees. In a peri-urban plan, mitigation actions must be programmed during the development phase of individual activities to identify adequate technical solutions from both a structural and vegetational point of view. Furthermore, if these actions are programmed as green infrastructure during the development phase of the plan, they allow a new vision in the planning of both public and private spaces. Private spaces can also contribute to ecological and social well-being. This opens an interactive dialogue between designers of the plan and experts who develop the SEA process by considering the active role of GI in the peri-urban plan. This creates increased openness about the real beneﬁts of these actions and therefore about accepting some design solutions that could otherwise be seen as simple additional costs. The winning aspect of this methodology is the incorporation of the ecosystem service concept and GI in the technical rules for implementing the plan, which thereby become elements of the plan, being binding for its execution. 5. Conclusions This paper presented a methodology that combines the assessment of ES with the possibility of designing suitable planning interventions and mitigation measures. This is essential for developing feedback between evaluation and design by increasing the ability Sustainability 2021, 13, 122 21 of 25 21 of 25 to ﬁnd more sustainable solutions in urban development, which can range from the scale of the individual building to that of the neighborhood. This can support the construction of more accountable peri-urban plans, providing transparency and replicability to decision- makers about the proposed transformations. to ﬁnd more sustainable solutions in urban development, which can range from the scale of the individual building to that of the neighborhood. This can support the construction of more accountable peri-urban plans, providing transparency and replicability to decision- makers about the proposed transformations. The integration of ecosystem services analysis into SEA requires a strong vision of the expected added value. Therefore, before starting the analysis, a workﬂow must be devel- oped that clariﬁes how this analysis can be integrated with the various assessment stages of the SEA and with other elements of the analysis. The consideration of green infrastructure as mitigation measures to reduce the negative effects of plans on the environment can arguably evolve the role of the ecosystem service concept from a simple assessment phase to a design phase of ecosystem services. Therefore, the capacity to develop green infras- tructure in SEA processes can conﬁgure the SEA as a tool of ecological urban design that is integrated into an urban plan design. This requires the ability to transfer the ecological and planning theories into practical actions and the capacity of different disciplines to work in a transdisciplinary approach, where different contributions act complementarily to develop the best solution. Through this research, we provided a pilot case study application that can also be adapted to other similar plans, programs, and policies or can provide stimuli and sugges- tions to improve the integration of ecosystem services into environmental assessments. The ES assessment methodology adopted in this work can be considered independent of the conceptual model developed. However, the methodology chosen must be able to functionally assess the variation in ecosystem services deemed to be priorities or to focus on some of them deemed most important. 5. Conclusions The integration of ecosystem services in environmental assessments can produce many beneﬁts contributing to better designed PPPs, but there are also critical issues, such as the complexity of appropriately identifying priority ecosystem services and the lack of well-established methods. From this perspective, an important issue is the consultation of stakeholders because they are affected by the new PPPs and can provide useful information, such as their needs, values, and dependencies on priority ecosystem services. Ensuring the participatory approach of stakeholders in the SEA process is crucial because it guarantees transparency and the effectiveness and efﬁciency of the strategic actions proposed [27]. The concept of ecosystem services should be used within the SEA to increase the performance of the human well-being aspect of urban plans, creating a stronger connection between ecological urban design choices and human beneﬁts. These beneﬁts have to be monitored after realization of the urban areas to check if the environmental objectives of the plan have been met and if the actions developed were effective in achieving them. This is an important aspect of the SEA process: the monitoring program of the plan has to be developed. Therefore, another point that can be expanded in the future is the follow up on ecosystem services [3], linked to the development of well-being performance indicators to value the achievement of environmental goals. This is particularly challenging for plans on the sub-urban scale due to the lack of site-speciﬁc well-being data and indicators. Author Contributions: Conceptualization, T.S. and D.G.; methodology, T.S.; validation, T.S., P.M., S.A. and A.T.; formal analysis, T.S.; investigation, T.S., P.M., S.A. and A.T.; data curation, T.S.; writing—original draft preparation, T.S.; writing—review and editing, T.S., P.M., S.A., A.T. and D.G.; visualization, D.G.; supervision, B.R. and D.G. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable for studies not involving human or animals. Institutional Review Board Statement: Not applicable for studies not involving human or animals Informed Consent Statement: Not applicable for studies not involving human. Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available due to the original language in italian and organized in different technical reports. References 1. De Groot, R.S.; Alkemade, R.; Braat, L.; Hein, L.; Willemen, L. Challenges in integrating the concept of ecosystem services and values in landscape planning, management and decision making. Ecol. Complex. 2010, 7, 260–272. [CrossRef] p p g g g p 2. Turner, B.L.; Skole, D.; Sanderson, S.; Fischer, G.; Fresco, L.; Leemans, R. Land-Use and Land-Cover Change Science/Research Plan. Joint publication of the International Geosphere-Biosphere Programme (Report No. 35) and the Human Dimensions of Global Environmental Change Programme (Report No. 7); Royal Swedish Academy of Sciences: Stockholm, Sweden, 1995. g g p y y 3. United Nations Environment Programme. Integrating Ecosystem Services in Strategic Environmental Assessment: A Guide for Practitioners. UNON, Publishing Services Section, Nairobi, ISO 14001:2004 Certiﬁed. 2014. Available online: http://www.ing. unitn.it/~{}genelab/documents/GuidelineESintoSEA.pdf (accessed on 13 June 2020). g p 4. Schneider, F.; Feurer, M.; Lundsgaard-Hansen, L.M.; Myint, W.; Don Nuam, C.; Nydegger, K.; Oberlack, C.; Tun, N.N.; Zahringer, J.G.; Tun, A.M.; et al. Sustainable Development Under Competing Claims on Land: Three Pathways Between Land Use Changes, Ecosystem Services and Human Well Being. Eur. J. Dev. Res. 2020, 32, 316–337. [CrossRef] g ategic Environmental Assessment. Good Practice Guidance for Development Co- Operation. DAC Guidelines 5. OECD. Applying Strategic Environmental Assessment. Good Practice Guidance for Development Co- Operation. DAC Guidelines and Reference Series. Organisation for Economic Cooperation and Development. 2006. Available online: https://www.oecd.org/ environment/environment-development/37353858.pdf (accessed on 19 May 2020). p p y 6. Baker, J.; Sheate, W.R.; Philips, P.; Eals, R. Ecosystem services in environmental assessment—Help or hindrance? Environ. Impact Assess. Rev. 2013, 40, 3–13. [CrossRef] 7. Partidario, M.; Gomes, R. Ecosystem Services Inclusive strategic Environmental Assessment. Environ. Impact Assess. Rev. 2013, 40, 36–46. [CrossRef] 8. Va’squez, D.R.; Furst, C.; Geneletti, D.; Almendra, O. Integration of ecosystem services in strategic environmental assessment across spatial planning scales. Land Use Policy 2017, 71, 303–310. [CrossRef] p p g y 9. Watson, K.; Galford, G.; Sonter, L.J.; Koh, I.; Ricketts, T.H. Effects of human demand on conservation and ecosystem services. Conserv. Biol. 2019, 33, 942–952. [CrossRef] y 10. Honrado, J.P.; Vieira, C.; Soares, C.; Monteiro, M.B.; Marcos, B.; Pereira, M.H.; Partidario, M.R. Can we infer about ecosystem services from EIA and SEA practice? A framework for analysis and examples from Portugal. Environ. Impact Assess. Rev. 2013, 40, 14–24. [CrossRef] 11. Geneletti, D. Chapter 3: Ecosystem Services Analysis for Strategic Environmental Assessment: Concepts and Examples. 2016. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. References In Research Handbooks on Impact Assessment Series; Geneletti, D., Ed.; Edward Elgar Publishing: Cheltenham, UK, 2016; pp. 41–61. ISBN 9781783478989. 12. Spyra, M.; La Rosa, D.; Zasada, I.; Sylla, M.; Shkaruba, A. Governance of ecosystem services trade-offs in peri-urban landscapes. Land Use Policy 2020, 95, 104617. [CrossRef] 13. Cabiddu. Territorial Government Law; Giappichel 14. Tzoulas, K.; Korpela, K.; Venn, S.; Yli-Pelkonen, V.; Kazmierczak, A.; Niemela, J.; James, P. Promoting ecosystem and human health in urban areas using green infrastructure: A literature review. Landsc. Urban Plan 2007, 81, 167–178. [CrossRef] 15. European Commission (EC). Green Infrastructure (GI)—Enhancing Europe’s Natural Capital. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Commit- tee of the Regions, COM 249 Final Brussels. 2013. Available online: https://eur-lex.europa.eu/resource.html?uri=cellar: d41348f2-01d5-4abe-b817-4c73e6f1b2df.0014.03/DOC_1&format=PDF (accessed on 9 April 2020). p ( ) ( ) g p p Commission to the European Parliament, the Council, the European Economic and Social Committee and the Commit- tee of the Regions, COM 249 Final Brussels. 2013. Available online: https://eur-lex.europa.eu/resource.html?uri=cellar: d f d b b f b df / f ( d l ) g p p d41348f2-01d5-4abe-b817-4c73e6f1b2df.0014.03/DOC_1&format=PDF (accessed on 9 April 2020). p ; Niemela, L. The concept of ecosystem services in adaptive urban planning and design: A framework for on. Landsc. Urban Plan. 2014, 125, 254–259. [CrossRef] p 16. Ahern, J.; Ciliers, S.; Niemela, L. The concept of ecosystem services in adaptive urban planning and supporting innovation. Landsc. Urban Plan. 2014, 125, 254–259. [CrossRef] supporting innovation. Landsc. Urban Plan. 2014, 125, 254–259. [CrossRef] 17. Pulighe, G.; Fava, F.; Lupia, F. Insights and opportunities from mapping ecosystem services of urban green spaces and potentials in planning. Ecosyst. Serv. 2016, 22, 1–10. [CrossRef] pp g 17. Pulighe, G.; Fava, F.; Lupia, F. Insights and opportunities from mapping ecosystem services of urban green spaces and potentials in planning. Ecosyst. Serv. 2016, 22, 1–10. [CrossRef] p g y [ ] 18. Semeraro, T.; Aretano, R.; Pomes, A.; Del Giudice, C.; Nigro, D. Planning ground based utility scale solar energy as Green Infrastructure to enhance ecosystem services. Energy Policy 2018, 117, 218–227. [CrossRef] y gy y 19. Keesstra, S.; Nunes, L.; Novara, A.; Finger, D.; Avelar, D.; Kalantari, Z.; Cerdà, A. The superior effect of nature based solutions in land management for enhancing ecosystem services. Sci. Total Environ. 2018, 610–611, 997–1009. 5. Conclusions Sustainability 2021, 13, 122 22 of 25 22 of 25 Acknowledgments: Many thanks to Aurelia Scarano for editing the manuscript. Many thank to Daniele Pomes for creating image 6 in Figure 6. This study is not part of the SEA processes, but it represents an independent scientiﬁc activity. Therefore, how the manuscript was structured and the content of this paper are not connected to and not binding for any decisions or analysis related to current and SEA processes in the study area. Many thanks to Aurelia Scarano for editing the manuscript. Conﬂicts of Interest: The authors declare no conﬂict of interest. References [CrossRef] [PubMed] 20 OECD Strategic Environmental in Assessment Ecosystem Services 2010 Available online: http://www oecd org/dataoecd/24 y gy y 19. Keesstra, S.; Nunes, L.; Novara, A.; Finger, D.; Avelar, D.; Kalantari, Z.; Cerdà, A. The superior effect of land management for enhancing ecosystem services. Sci. Total Environ. 2018, 610–611, 997–1009. [Cros 19. Keesstra, S.; Nunes, L.; Novara, A.; Finger, D.; Avelar, D.; Kalantari, Z.; Cerdà, A. The superior effect of nature based solutions in land management for enhancing ecosystem services. Sci. Total Environ. 2018, 610–611, 997–1009. [CrossRef] [PubMed] 20. OECD. Strategic Environmental in Assessment Ecosystem Services. 2010. Available online: http://www.oecd.org/dataoecd/24 /54/41882953.pdf (accessed on 28 May 2020). 20. OECD. Strategic Environmental in Assessment Ecosystem Services. 2010. Available online: http://www.oecd.org/dataoecd/24 /54/41882953.pdf (accessed on 28 May 2020). p y 21. WRI. Ecosystem Services Review for Impact Assessment. 2010. Available online: http://www.google.com/url?sa=t&rct=j&q= &esrc=s&source=web&cd=2&ved=0CDEQFjAB&url=http%3A%2F%2Fcontent.undp.org%2Fgo%2Fcmsservice%2Fdownload% 23 of 25 Sustainability 2021, 13, 122 2Fpublication%2F%3Fversion%3Dlive%26id%3D2078079&ei=8PyHT-TfCuiG0AXX5qnhCQ&usg=AFQjCNEOVxepE2vsY7ZRN8 nTZwllXE0vVQ (accessed on 10 April 2020). p 22. Saponari, M.; Boscia, D.; Altamura, G.; Loconsole, G.; Zicca, S.; D’Attoma, G.; Morelli, M.; Palmisano, F.; Saponari, A.; Tavano, D.; et al. Isolation and pathogenicity of Xylella fastidiosa associated to the olive quick decline syndrome in Southern Italy. Sci. Rep. 2017, 7, 17723. [CrossRef] [ ] 23. Maggiore, G.; Semeraro, T.; Aretano, R.; De Bellis, L.; Luvisi, A. GIS Analysis of Land-Use Change in Threatened Landscapes by Xylella fastidiosa. Sustainability 2019, 11, 253. [CrossRef] y y 24. Semeraro, T.; Gatto, E.; Buccolieri, R.; Vergine, M.; Gao, Z.; De Bellis, L.; Luvisi, A. Changes in Olive Urban Forests Infected by Xylella fastidiosa: Impact on Microclimate and Social Health in urban areas. Int. J. Environ. Res. Public Health 2019, 16, 2642. [CrossRef] 25. Therivel, R. Strategic Environmental Assessment in Action; Earthscan: London, UK, 2004. g 26. UN. Resolution 70/1 in 2015: Transforming Our World: The 2030 Agenda for Sustainable Development. United Nations General Assembly, 2015. Available online: https://www.un.org/ga/search/view_doc.asp?symbol=A/RES/70/1&Lang=E (accessed on 17 September 2020). p ) 27. Semeraro, T.; Zaccarelli, N.; Lara, A.; Sergi Cucinelli, F.; Aretano, R. A Bottom-Up and Top-Down Participatory Approach to Planning and Designing Local Urban Development: Evidence from an Urban University Center. Land 2020, 9, 98. [CrossRef] 28 K P Th E i f E t d Bi di it E l i l d E i F d ti E th L d UK W hi t p 27. Semeraro, T.; Zaccarelli, N.; Lara, A.; Sergi Cucinelli, F.; Aretano, R. References Urban form, biodiversity potential an Urban Plan. 2007, 83, 308–317. [CrossRef] 37. Taylor, S.L.; Taylor, J.R. Supplying urban ecosystem services through multifunctional green infrastructure in the United States. Landsc. Ecol. 2013, 28, 1447–1463. [CrossRef] 38. Sieber, J.; Pons, M. Assessment of Urban Ecosystem Services using Ecosystem Services Reviews and GIS-based Tools. Procedia Eng. 2015, 115, 53–60. [CrossRef] 39. United States General Services Administration. The Beneﬁts and Challenges of Green Roofs on Public and Commercial Buildings. Available online: https://www.gsa.gov/about-us/organization/ofﬁce-ofgovernmentwide-policy/ofﬁce-of-federal- highperformance-buildings/projects-andresearch/green-roofs (accessed on 28 July 2020). g p g p j g y 40. U.S. National Climate Assessment. Global Change Research Program. Available online: https://nca2014.globalchange.gov/ highlights/report-ﬁndings/water-supply (accessed on 22 August 2020). g g p g pp y g 41. Wang, J.; Banzhaf, E. Towards a better understanding of Green Infrastructure: A critical review. Ecol. Ind. 2018, 85, 758–772. [CrossRef] 42. Brzoska, P.; Spage, A. From City- to Site-Dimension: Assessing the Urban Ecosystem Services of Different Types of Green Infrastructure Land 2020 9 150 [CrossRef] 41. Wang, J.; Banzhaf, E. Towards a better understanding of Green Infrastructure: A critical review. Ecol. Ind. 20 g, J ; , g , , [ ] 42. Brzoska, P.; Spage, A. From City- to Site-Dimension: Assessing the Urban Ecosystem Services of Different Types of Green Infrastructure. Land 2020, 9, 150. [CrossRef] 42. Brzoska, P.; Spage, A. From City- to Site-Dimension: Assessing the Urban Ecosystem Services of D Infrastructure. Land 2020, 9, 150. [CrossRef] 43. Aretano, R.; Parlagreco, L.; Semeraro, T.; Zurlini, G.; Petrosillo, I. Coastal dynamics vs beach users attitudes and perceptions to enhance environmental conservation and management effectiveness. Mar. Pollut. Bull. 2017, 123, 142–155. [CrossRef] 44. De Groot, R.; Fisher, B.; Christie, M.; Aronson, J.; Braat, L.; Gowdy, J.; Haines-Young, R.; Maltby, E.; Neuville, A.; Polasky, S.; et al. Integrating the ecological and economic dimensions in biodiversity and ecosystem service valuation. In The Economics of Ecosystems and Biodiversity Ecological and Economic Foundations; Kumar, P., Ed.; Routledge Press: London, UK; Earthscan: London, UK, 2010; pp. 9–40. 45. Strohbacha, M.W.; Haaseb, D. Above-ground carbon storage by urban trees in Leipzig, Germany: Analysis of patterns in a European city. Landsc. Urban Plan. 2012, 104, 95–104. [CrossRef] 46. Burkhard, B.; Kroll, F.; Nedkov, S.; Müller, F. Mapping ecosystem service supply, demand and budgets 29. [CrossRef] 47. Burkhard, B.; Crossman, N.; Nedkov, S.; Petz, K.; Alkemade, R. Mapping and Modelling Ecosystem Services for Science, Policy and Practice. Special Issue. Ecosyst. Serv. References A Bottom-Up and Top-Down Participatory Approach to Planning and Designing Local Urban Development: Evidence from an Urban University Center. Land 2020, 9, 98. [CrossRef] 28. Kumar, P. The Economics of Ecosystems and Biodiversity: Ecological and Economic Foundations; Earthscan: London, UK; Washington, DC, USA, 2010. 28. Kumar, P. The Economics of Ecosystems and Biodiversity: Ecological and Economic Foundations; Earthsca DC, USA, 2010. 29. Semeraro, T.; Pomes, A.; Aretano, R. Green Roof Technology as a Sustainable Strategy to Improve Water Urban Availability. IOP Conf. Series: Materials Science and Engineering. IOP Conf. Ser. Mater. Sci. Eng. 2019, 471, 092065. [CrossRef] g g f g 30. Tzoulas, K.; Galan, J.; Venn, S.; Dennis, M.; Pedroli, B.; Mishra, H.; Haase, D.; Pauleit, S.; Niemelä, J.; James, P. A conceptual model of the social–ecological system of nature-based solutions in urban environments. Ambio 2020. [CrossRef] 30. Tzoulas, K.; Galan, J.; Venn, S.; Dennis, M.; Pedroli, B.; Mishra, H.; Haase, D.; Pauleit, S.; Niemelä, J.; James, P. A conceptual d l f h i l l i l f b d l i i b i A bi [C R f] as, K.; Galan, J.; Venn, S.; Dennis, M.; Pedroli, B.; Mishra, H.; Haase, D.; Pauleit, S.; Niemelä, J.; James, P l of the social–ecological system of nature-based solutions in urban environments. Ambio 2020. [CrossRef] 31. Mell, I.C. Can green infrastructure promote urban sustainability? Proceedings of the Institution of Civil Engineers. Eng. Sustain. 2009, 1, 23–34. [CrossRef] 32. Semeraro, T.; Aretano, R.; Pomes, A. Green Infrastructure to Improve Ecosystem Services in the Landscape Urban Regeneration. IOP Conf. Ser. Mater. Sci. Eng. 2017, 245, 082044. [CrossRef] f g 33. Semeraro, T.; Aretano, R.; Barca, A.; Pomes, A.; Del Giudice, C.; Gatto, E.; Lenucci, M.; Buccolieri, R.; Emmanuel, R.; Gao, Z.; et al. A Conceptual Framework to Design Green Infrastructure: Ecosystem Services as an Opportunity for Creating Shared Value in Ground Photovoltaic Systems. Land 2020, 9, 238. [CrossRef] 34. Majekodunmi, M.; Emmanuel, R.; Jafry, T. A spatial exploration of deprivation and green infrastructure ecosystem services within Glasgow city. Urban For. Urban Green. 2020, 52, 126698. [CrossRef] 35. La Rosa, D.; Spyra, M.; Inostroza, L. Indicators of Cultural Ecosystem Services for urban planning: A review. Ecol. Ind. 2016, 61, 74–89. [CrossRef] ; Warren, P.H.; Davies, R.G.; Gaston, K.J. Urban form, biodiversity potential and ecosystem services. Landsc , 308–317. [CrossRef] 36. Tratalos, J.; Fuller, R.; Warren, P.H.; Davies, R.G.; Gaston, K.J. References 2013, 4, 1–146. [CrossRef] Sustainability 2021, 13, 122 24 of 25 24 of 25 48. Luvisi, A.; Aprile, A.; Sabella, E.; Vergine, M.; Nicolì, F.; Nutricati, E.; Miceli, A.; Negro, C.; De Bellis, L. Xylella fastidiosa subsp. pauca (CoDiRO strain) infection in four olive (Olea europaea L.) cultivars: Proﬁle of phenolic compounds in leaves and progression of leaf scorch symptoms. Phytopathol. Mediterr. 2017, 56, 259–273. [CrossRef] p g y p y p 49. Apulian Law. Protection and Enhancement of the Landscape of the Monumental Olive Trees of Apulia. Ofﬁcial Bulletin of the Apulian Region—n. 83 Suppl. of 7-6-2007. 2007. Available online: http://www.geologipuglia.it/doc/downloads/457-lr_4_06_2 007_n14-legge-regionale-4-giugno-2007-n-14.pdf (accessed on 20 April 2020). gg g g g p p 50. Medagli, P.; Accogli, R.; Turco, A.; Zuccarello, V.; Albano, A. Spontaneous Flowers of Salento; Grifo: Modugno, Italy, 2016. 51. Accogli, E.; Medagli, P. Spontaneous Salento Herbs; Grifo: Modugno, Italy, 2014. g , ; g , ; , ; , ; , p f ; g , y, 51. Accogli, E.; Medagli, P. Spontaneous Salento Herbs; Grifo: Modugno, Italy, 2014. 51. Accogli, E.; Medagli, P. Spontaneous Salento Herbs; Grifo: Modugno, g g g 52. Pignatti, S. Flora of Italy, 2nd ed.; Edagricole: Bologna, Italy, 1982. S. Flora of Italy, 2nd ed.; Edagricole: Bologna, Italy, 19 53. Apulian Law. Operational Guidelines for the Coordinated Application of the Legislation on the Felling of Olive Trees (Law no. 144/51) in the Infected Delimited Areas of the Puglia Region to Counter the Spread of Xylella Fastidiosa. Amendments to the Resolution of the Regional Council of 14 December 1989 no. 7310. Ofﬁcial Bulletin of the Apulian Region—n. 24 of 26-2-2019. 2019. Available online: http://cartograﬁa.sit.puglia.it/DOC/xylella/DEL_1780_2019.pdf (accessed on 27 June 2020). 54. Hirano, Y.; Yoshida, Y. Assessing the effects of CO2 reduction strategies on heat islands in urban areas. Sustain. Cities Soc. 2016, 26, 383–392. [CrossRef] 55. Kaspersen, P.S.; Høegh Ravn, N.; Arnbjerg-Nielsen, K.; Madsen, H.; Drews, M. Comparison of the impacts of urban development and climate change on exposing European cities to pluvial ﬂooding. Hydrol. Earth Syst. Sci. 2017, 21, 4131–4147. [CrossRef] 56. Krüger, E.; Drach, P.; Emmanuel, R. Effects of atmospheric stability and urban morphology on daytime intra-urban temperature variability for Glasgow, UK. Sci. Total Environ. 2018, 627, 782–791. [CrossRef] g p g p p g y y 56. Krüger, E.; Drach, P.; Emmanuel, R. Effects of atmospheric stability and urban morphology on daytime intra-urban temperature variability for Glasgow, UK. Sci. References Harlan, S.L.; Ruddel, D. Climate change and health in cities: Impacts of heat and air pollution and p mitigation and adaptation. Curr. Opin. Environ. Sustain. 2011, 3, 126–134. [CrossRef] g p p 70. Sohn, W.; Kim, H.W.; Kim, J.H.; Li, M.H. The capitalized amenity of green infrastructure in single-family housing values: An application of the spatial hedonic pricing method. Urban. Urban Green. 2020, 49, 126643. [CrossRef] 71. Witzling, L.M.; Wander, E.; Phillips, E. Testing and educating on urban soil lead: A case of Chicago community gardens. J. Agric. Food Syst. Commun. Dev. 2011, 1, 167–185. [CrossRef] 72. Hunt, A.; Watkiss, P. Climate change impacts and adaptation in cities: A review of the literature. Clim. Chang. 2011, 104, 13–49. [CrossRef] 73. Park, B.-J.; Furuya, K.; Kasetani, T.; Takayama, N.; Kagawa, T.; Miyazaki, Y. Relationship between psychological responses and physical environments in forest settings. Landsc. Urb. Plan. 2011, 102, 24–32. [CrossRef] 74. Lindemann-Matthies, P.; Junge, X.; Matthies, D. The inﬂuence of plant diversity on people’s perception and esthetic appreciation 72. Hunt, A.; Watkiss, P. Climate change impacts and adaptation in cities: A review of the literature. Clim. Chang. 2011, 104, 13–49. [CrossRef] 73. Park, B.-J.; Furuya, K.; Kasetani, T.; Takayama, N.; Kagawa, T.; Miyazaki, Y. Relationship between psychological responses and physical environments in forest settings. Landsc. Urb. Plan. 2011, 102, 24–32. [CrossRef] 74. Lindemann-Matthies, P.; Junge, X.; Matthies, D. The inﬂuence of plant diversity on people’s perception and esthetic appreciation of grassland vegetation. Biol. Conserv. 2010, 143, 195–202. [CrossRef] p y g , , [ ] 74. Lindemann-Matthies, P.; Junge, X.; Matthies, D. The inﬂuence of plant diversity on people’s perception and esthetic appreciation of grassland vegetation. Biol. Conserv. 2010, 143, 195–202. [CrossRef] 75. Thompson, C.W.; Roe, J.; Aspinall, P.; Mitchell, R.; Clow, A.; Miller, D. More green space is linked to less stress in deprived communities: Evidence from salivary cortisol patterns. Landsc. Urban Plan. 2012, 105, 221–229. [CrossRef] y p 76. Gray, L.; Guzman, P.; Glowa, K.M.; Drevno, A.G. Can home gardens scale up into movements for social change? The role of home gardens in providing food security and community change in San Jose, California. Local Environ. 2013, 19, 187–203. [CrossRef] 76. Gray, L.; Guzman, P.; Glowa, K.M.; Drevno, A.G. Can home gardens scale up into movements for social change? The role of home gardens in providing food security and community change in San Jose, California. Local Environ. 2013, 19, 187–203. References [CrossRef] 61 M i G Clif R Chi i I U b C l i i d I C ib i S i bili Nibbl f F d b O dl f S i l 60. Taylor, J.R.; Lovell, S.T. Urban home food gardens in the Global North: Research traditions and future directions. Agric. Hum. Values 2014, 31, 285–305. [CrossRef] 61. Martin, G.; Clift, R.; Chistie, I. Urban Cultivation and Its Contributions to Sustainability: Nibbles of Food but Oodles of Social Values 2014, 31, 285–305. [CrossRef] 61. Martin, G.; Clift, R.; Chistie, I. Urban Cultivation and Its Contributions to Sustainability: Nibbles of Food but Oodles of Social Capital Sustainability 2016 8 409 [CrossRef] 61. Martin, G.; Clift, R.; Chistie, I. Urban Cultivation and Its Contributions to Sustainability: Nibbles Capital. Sustainability 2016, 8, 409. [CrossRef] Chistie, I. Urban Cultivation and Its Contributions to Sustainability: Nibbles of Food but Oodles of Social ty 2016, 8, 409. [CrossRef] 62. Lin, B.L.; Egerer, M.H. Global social and environmental change drives the management and delivery of ecosystem services from urban gardens: A case study from Central Coast, California. Glob. Environ. Chang. 2020, 60, 102006. [CrossRef] g y g 63. Setälä, H.; Viippola, V.; Rantalainen, A.L.; Pennanen, A.; Yli-Pelkonen, V. Doesurban vegetation mitigate air pollution in northern conditions? Environ. Pollut. 2013, 183, 104–112. [CrossRef] [PubMed] 64. Santamouris, M. Regulating the damaged thermostat of the cities—Status, impacts andmitigation challenges. Energy Build. 2015, 91, 43–56. [CrossRef] 65. Pascal, M.; Corso, M.; Chanel, O.; Declercq, C.; Badaloni, G.; Henschel, S.; Meister, K.; Haluza, D.; Olmedo, P.M.; Medina, S. Assessing the public health impacts of urban air pollution in 25 European cities: Results of the Aphekom project. Sci. Total Environ. 2013, 449, 390–400. [CrossRef] 66. Farahani, L.M.; Maller, C. Perceptions and Preferences of Urban Greenspaces: A Literature Review and Framework for Policy and Practice. Landsc. Online 2018, 61, 1–22. [CrossRef] 67. Langemeyer, J.; Connolly, J.J.T. Weaving notions of justice into urban ecosystem services research and practice. Environ. Sci. Policy 2020, 109, 1–14. [CrossRef] 68. Phoomirat, R.; Disyatat, N.R.; Park, T.Y.; Lee, D.K.; Dumrongrojwatthana, P. Rapid assessment check services in Bangkok. Ecol. Process. 2020, 9, 19. [CrossRef] services in Bangkok. Ecol. Process. 2020, 9, 19. [CrossRef] 69. Harlan, S.L.; Ruddel, D. Climate change and health in cities: Impacts of heat and air pollution and potential co-beneﬁts from mitigation and adaptation. Curr. Opin. Environ. Sustain. 2011, 3, 126–134. [CrossRef] 69. References Total Environ. 2018, 627, 782–791. [CrossRef] y g 57. Huang, Q.; Yin, D.; He, C.; Yan, J.; Liu, Z.; Meng, S.; Ren, Q.; Zhao, R.; Inostroza, L. Linking ecosystem services and subjective well-being in rapidly urbanizing watersheds: Insights from a multilevel linear model. Ecosyst. Serv. 2020, 43, 101106. [CrossRef] 57. Huang, Q.; Yin, D.; He, C.; Yan, J.; Liu, Z.; Meng, S.; Ren, Q.; Zhao, R.; Inostroza, L. Linking ecosystem services and subjective well-being in rapidly urbanizing watersheds: Insights from a multilevel linear model. Ecosyst. Serv. 2020, 43, 101106. [CrossRef] 58. Sandifer, P.; Sutton-Grier, A.E.; Ward, B.P. Exploring connections among nature, biodiversity, ecosystem services, and human health and well-being: Opportunities to enhance health and biodiversity conservation. Ecosyst. Serv. 2015, 12, 1–15. [CrossRef] 59. Jiangyi, L.J.; Shiquan, D.; Hmeimar, E.H. Cost-effectiveness analysis of different types of payments for ecosystem services: A case in the urban wetland ecosystem. J. Clean. Prod. 2020, 249, 119325. [CrossRef] well being in rapidly urbanizing watersheds: Insights from a multilevel linear model. Ecosyst. Serv. 2020, 43, 101106. [CrossRef] 58. Sandifer, P.; Sutton-Grier, A.E.; Ward, B.P. Exploring connections among nature, biodiversity, ecosystem services, and human health and well-being: Opportunities to enhance health and biodiversity conservation. Ecosyst. Serv. 2015, 12, 1–15. [CrossRef] 59. Jiangyi, L.J.; Shiquan, D.; Hmeimar, E.H. Cost-effectiveness analysis of different types of payments for ecosystem services: A case i th b tl d t J Cl P d 2020 249 119325 [C R f] 58. Sandifer, P.; Sutton-Grier, A.E.; Ward, B.P. Exploring connections among nature, biodiversity, ecosystem services, and human health and well-being: Opportunities to enhance health and biodiversity conservation. Ecosyst. Serv. 2015, 12, 1–15. [CrossRef] health and well-being: Opportunities to enhance health and biodiversity conservation. Ecosyst. Serv. 2015, 12, 1–15. [CrossRef] 59. Jiangyi, L.J.; Shiquan, D.; Hmeimar, E.H. Cost-effectiveness analysis of different types of payments for ecosystem services: A case in the urban wetland ecosystem. J. Clean. Prod. 2020, 249, 119325. [CrossRef] 59. Jiangyi, L.J.; Shiquan, D.; Hmeimar, E.H. Cost-effectiveness analysis of different types of payments for ecosystem services: A case in the urban wetland ecosystem. J. Clean. Prod. 2020, 249, 119325. [CrossRef] 60. Taylor, J.R.; Lovell, S.T. Urban home food gardens in the Global North: Research traditions and future directions. Agric. Hum. Values 2014, 31, 285–305. 77. SIT Apulian Region. Host Plants Identiﬁed as Susceptible to Xylella fastidiosa Subsp. Pauca—ST53. Available online: http: //www.emergenzaxylella.it/portal/portale_gestione_agricoltura/Documenti/Specie (accessed on 30 November 2020). g y p p g g p 78. McGarigal, K.; Marks, B.J. Fragstats: Spatial Pattern Analysis Program for Quantifying Landscape Structure; PNW-GTR-351, USDA Forest Service; Paciﬁc Northwet Research Station: Portland, OR, USA, 1995. 79. von Döhren, P.; Haase, D. Ecosystem disservices research: A review of the state of the art with a focus on cities. Ecol. Ind. 2015, 52, 490–497. [CrossRef] se, D. Ecosystem disservices research: A review of the state of the art with a focus on cities. Ecol. Ind. 2015, 52 77. SIT Apulian Region. Host Plants Identiﬁed as Susceptible to Xylella fastidiosa Subsp. Pauca—ST53. Available online: http: //www.emergenzaxylella.it/portal/portale_gestione_agricoltura/Documenti/Specie (accessed on 30 November 2020). 78. McGarigal, K.; Marks, B.J. Fragstats: Spatial Pattern Analysis Program for Quantifying Landscape Structure; General Technical Report PNW-GTR-351, USDA Forest Service; Paciﬁc Northwet Research Station: Portland, OR, USA, 1995. 79. von Döhren, P.; Haase, D. Ecosystem disservices research: A review of the state of the art with a focus on cities. Ecol. Ind. 2015, 52, 490–497. [CrossRef] K.; Marks, B.J. Fragstats: Spatial Pattern Analysis Program for Quantifying Landscape Structure; General Technical 51, USDA Forest Service; Paciﬁc Northwet Research Station: Portland, OR, USA, 1995. P.; Haase, D. Ecosystem disservices research: A review of the state of the art with a focus on cities. Ecol. Ind. 2 ossRef] p g p y f p p //www.emergenzaxylella.it/portal/portale_gestione_agricoltura/Documenti/Specie (accessed on 30 November 2020). 78. McGarigal, K.; Marks, B.J. Fragstats: Spatial Pattern Analysis Program for Quantifying Landscape Structure; General Technical Report PNW-GTR-351, USDA Forest Service; Paciﬁc Northwet Research Station: Portland, OR, USA, 1995. 77. SIT Apulian Region. Host Plants Identiﬁed as Susceptible to Xylella fastidiosa Subsp. Pauca—ST53. Available online: http: //www.emergenzaxylella.it/portal/portale_gestione_agricoltura/Documenti/Specie (accessed on 30 November 2020). 78. McGarigal, K.; Marks, B.J. Fragstats: Spatial Pattern Analysis Program for Quantifying Landscape Structure; General Technical Report PNW-GTR-351, USDA Forest Service; Paciﬁc Northwet Research Station: Portland, OR, USA, 1995. 79 von Döhren P; Haase D Ecosystem disservices research: A review of the state of the art with a focus on cities Ecol Ind 2015 52 References [CrossRef] 25 of 25 Sustainability 2021, 13, 122
https://openalex.org/W2130903395	https://www.scielo.br/j/ss/a/FQpCGsc9tvfcqK8bYJVqH6H/?lang=pt&format=pdf	Portuguese	null	Uma falsa lição de anatomia ou de um simples caso de impregnação teórica dos fatos	Scientia Agrícola	2,004	cc-by	9,049	1 Limito-me aqui à análise das interpretações a respeito da Capela Sistina, mas a crítica aqui feita também é perti- nente para as duas esculturas de Michelangelo analisadas no livro em questão. resenhas resenhas Uma falsa lição de anatomia... scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Uma falsa lição de anatomia... scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Uma falsa lição de anatomia ou de um simples caso de impregnação teórica dos fatos Eduardo Henrique Peiruque Kickhöfel g Gilson Barreto e Marcelo G. de Oliveira Editora ARX São Paulo, 2004, 230 págs. 2 Por exemplo, os autores (p. 37) interpretam literalmente um texto de Leonardo da Vinci a respeito das diferenças entre o escultor e o pintor, um texto que é evidentemente de feição retórica. 3 Os estudos para a Capela Sistina estão catalogados sob os números 119-174 do catálogo de Tonlay (1975). Os dese- nhos anatômicos de Michelangelo estão sob os números 105-116 do mesmo catálogo e compreendem um conjunto muito heterogêneo de desenhos espalhado em coleções do Ashmolean Museum (Oxford), da Casa Buonarroti (Flo- rença), do Louvre (Paris), do Teyler Museum (Haarlem), da Royal Library (Windsor) e da Stanza del Borgo (Milão). Com exceção do no. 110, considerado uma cópia, todos esses desenhos estão citados como originais de Michelangelo e suas datações oscilam entre ca. 1505-06 e ca. 1530, mas Tonlay não fornece justificativas para suas datações. Perrig (1991, p. 67), um revisionista radical e irreverente, em um livro brilhante a respeito de atribuição de desenhos de Michelangelo, considera todos os desenhos de anatomia de Michelangelo como cópias. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 I Dado o grau de complexidade e a quantidade hoje disponível de conhecimentos acadê- micos, para o bem e para o mal, as formações especializadas fazem parte do cotidiano universitário. Estuda-se meticulosamente assuntos muito específicos na expectativa da realização de um trabalho original. O mundo acadêmico é pródigo de exemplos de que uma formação especializada não garante uma contribuição original em uma área específica, mas sua importância não pode ser subestimada. Após uma idéia que se apre- senta inicialmente dessa forma, uma formação é importante e mesmo necessária para que uma idéia seja desenvolvida e fundamentada. Trata-se de proporcionar à investi- gação contornos objetivos, no sentido de que seus resultados e conclusões possam ser avaliados por outros estudiosos. O livro recentemente lançado A arte secreta de Michelangelo, de autoria do médico Gilson Barreto e do químico Marcelo G. de Oliveira, apresenta uma tese inovadora e ousada: imbuído de estudos sistemáticos de anatomia tal qual muitos médicos anatomistas de sua época, Michelangelo pintou estruturas anatômicas camufladas em muitas figuras da Capela Sistina, de modo a fornecer uma lição de anatomia um tanto original apenas agora decifrada por ambos pesquisadores.1 Entretanto, sem formação na área em que se propuseram a trabalhar, essa tese não resiste a um exame crítico. 427 Eduardo Henrique Peiruque Kickhöfel Figura 1. Michelangelo. Estudos de figura para a Ca- pela Sistina e estudo da mão da Sibila Líbica (sangüí- nea); estudos de escravos para a tumba de Júlio II (pena e nanquim). Oxford, Ashmolean Museum. Além disso, como premissa de sua argu- mentação existe a idéia da união efetiva entre as artes e a ciência no Renascimen- to italiano, idéia que também não resiste a um exame crítico. II Em um texto que freqüentemente apon- ta para a falta de familiaridade com a cul- tura e o estilo do período em questão,2 as menções aos estudos de anatomia de Michelangelo, enfatizadas pelos autores, são necessárias para seu argumento, mas não são suficientes para sustentá-lo. Não há o menor indício das interpretações propostas em seus desenhos para os afrescos da Capela Sistina (fig. 1 e 2) e nem em seus raros desenhos de anato- mia.3 Os textos contemporâneos que, de uma forma ou de outra, refletem as pala- vras de Michelangelo, como as biografi- as de Ascanio Condivi (publicada em 1553) e Giorgio Vasari (publicada em 1550, revi- sada em 1568), também não sugerem isso. Além do mais, não há uma caracterização do período em questão que sustente as interpretações propostas, no sentido de uma apro- ximação da cultura do Renascimento a partir de seus documentos, como textos e dese- nhos de artistas contemporâneos de Michelangelo. Figura 1. Michelangelo. Estudos de figura para a Ca- pela Sistina e estudo da mão da Sibila Líbica (sangüí- nea); estudos de escravos para a tumba de Júlio II (pena e nanquim). Oxford, Ashmolean Museum. Figura 1. Michelangelo. Estudos de figura para a Ca- pela Sistina e estudo da mão da Sibila Líbica (sangüí- nea); estudos de escravos para a tumba de Júlio II (pena e nanquim). Oxford, Ashmolean Museum. 428 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Figura 2. Michelangelo. Estudos para a Sibila Líbica (sangüínea). Nova Iorque, Metropolitan Museum. Figura 2. Michelangelo. Estudos para a Sibila Líbica (sangüínea). Nova Iorque, Metropolitan Museum. Uma falsa lição de anatomia... Numa tentativa de contornar essa falta, os autores usam as chamadas “pis- tas”, com as quais eles pensam ser possí- vel decifrar os estudos anatômicos camu- flados. Por exemplo, segundo os autores (p. 94), a árvore seca que está à esquerda da cena do Dilúvio representa uma rami- ficação brônquica, sendo a “pista” para essa interpretação a atitude de várias fi- guras “que repousam as mãos sobre o tó- rax” e assim chamam “a atenção para essa parte do corpo”. Baxandall (1991), em um livro exemplar a respeito de como inter- pretar pinturas da Itália do século xv, faz uma contextualização dos gestos da épo- ca a partir de textos e gravuras e, a partir disso, propõe interpretações de algumas pinturas. Dor, por exemplo, era repre- sentada ao se pressionar o peito com a palma da mão, e vergonha, ao se cobrir os olhos com os dedos. A partir disso, Figura 2. Michelangelo. Estudos para a Sibila Líbica (sangüínea). Nova Iorque, Metropolitan Museum. é possível interpretar o afresco A expulsão do Paraíso, de Masaccio, de forma plausível (fig. 3): Adão, com os olhos cobertos, exprime vergonha, e Eva, com suas mãos sobre o peito, apenas dor. Baxandall cita outro gesto em uso na ordem dos Beneditinos que diz respeito ao exemplo do Dilúvio citado acima: “Demonstração: indica-se uma coisa que se viu abrindo a palma da mão em sua direção” (Baxandall, 1991, p. 240). Esse gesto, presente em diver- sas pinturas do século xv, não é a “pista” sugerida acima. Usando palavras de Baxandall, uma investigação nesses termos está relacionada a encontrar o estilo cognitivo de uma época. De fato, os autores evitam uma contextua- lização histórica, a qual eles pensam ser posterior às suas é possível interpretar o afresco A expulsão do Paraíso, de Masaccio, de forma plausível (fig. 3): Adão, com os olhos cobertos, exprime vergonha, e Eva, com suas mãos sobre o peito, apenas dor. Baxandall cita outro gesto em uso na ordem dos Beneditinos que diz respeito ao exemplo do Dilúvio citado acima: “Demonstração: indica-se uma coisa que se viu abrindo a palma da mão em sua direção” (Baxandall, 1991, p. 240). Esse gesto, presente em diver- sas pinturas do século xv, não é a “pista” sugerida acima. Usando palavras de Baxandall, uma investigação nesses termos está relacionada a encontrar o estilo cognitivo de uma época. Figura 3. Masaccio. A expulsão do Paraíso (afresco). Capella Brancacci, Santa Maria del Carmine, Florença. Uma falsa lição de anatomia... De fato, os autores evitam uma contextua- lização histórica, a qual eles pensam ser posterior às suas 429 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Eduardo Henrique Peiruque Kickhöfel interpretações. Após sugerirem a existência de um “código secreto”, eles escrevem (p. 27) que “suas possíveis implicações históricas e filosóficas estão além de nossa capaci- dade de análise, que se limitará à descrição da anatomia na arte”. Assim, o “código decifrado” (p. 70-1) é apenas uma lista de indicações a respeito de como observar as interpretações arbitrárias e sem fundamento dos próprios autores uma a uma, mas não uma estrutura de sentidos consistente que permita interpretar as formas anatômicas relacionadas entre si e dentro do programa iconográfico da Capela Sistina, o que configuraria uma lição de anatomia, conforme proposto não apenas no título do livro e diversas vezes ao longo de seu texto, mas enfatizado quando os autores mencio- nam a existência de um “código didático” (p. 214). Tanto quanto os documentos de época permitem supor, não havia um código de época que possibilitava um artista a pintar afrescos conforme os autores sugerem. Eles, a seu modo, admitem isso ao es- crever que apenas agora o código (p. 27) “foi quebrado”. Por isso, pode-se supor que tal código existiu secretamente na mente do divino Michelangelo e que, por algum motivo não cognoscível a comuns mortais, surgiu de súbito no cérebro de alguns raros escolhidos. Talvez, mas sendo as formas anatômicas tridimensionais, muitas, de acor- do com o ângulo de que são vistas, servem tanto às interpretações dos autores quanto a muitas outras, de acordo com a boa vontade e a criatividade de quem as observa, as quais em alguns casos têm de ser extraordinariamente fortes, como para ver, no afresco que narra a história da judia Judite e do general assírio Holofernes, o corpo deste últi- mo como uma representação científica da segunda vértebra do pescoço, chamada áxis, com seu processo odontóide em evidência. Isto posto, não há como pensar por que motivo, em um conjunto de afrescos com um plano nitidamente teológico, Michelangelo representou formas anatômicas. Os au- tores admitem isso (p. 28), ao escrever que as informações que eles apresentam a res- peito de Michelangelo não são suficientes para que se proponha um motivo pelo qual Michelangelo fez o que fez. Uma falsa lição de anatomia... Entretanto, eles seguem com suas interpretações, restando apenas argumentos historicamente pouco convincentes (p. 216): “Considerando-se a sua extensa experiência em dissecação e seu grande interesse pela anatomia dos ór- gãos internos, não há razão para que formas anatômicas (além das formas da mus- culatura externa) não estivessem também presentes no restante de sua arte plástica”. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 III A referência acima à “extensa experiência em dissecação de Michelangelo” e a “seu grande interesse pela anatomia dos órgãos internos” está inteiramente baseada na idéia de uma união entre as artes e a ciência no Renascimento, um topos entre os historiado- 430 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Uma falsa lição de anatomia... res da arte, premissa de toda a argumentação do livro em questão. Os autores destacam isso de diversas formas. Por exemplo, comentando a obra de Andreas Vesalius (p. 48), eles mencionam “aquele encontro histórico entre arte e ciência”, e enfatizam isso pouco após (p. 70), ao mencionar que Michelangelo, “no coração da Igreja, imortalizou a co- munhão da arte com o conhecimento científico”. Entretanto, eles não fornecem defi- nições mínimas de seus termos centrais a partir de textos de época, de modo que seus argumentos também são desprovidos de sentido histórico objetivo. O termo “arte” no Renascimento esteve ligado à definição de Aristóteles, qual seja, produzir a partir de princípios, mas não necessariamente os primeiros. De forma sucinta, Aristóteles diz na Ética a Nicômaco (VI, 1140a 20-22) que arte é “uma disposi- ção que se ocupa de produzir, envolvendo o reto raciocínio”.4 Em outras palavras, o termo arte significava manualidade ou a habilidade voltada a realizar algo material. Um memorando do representante do duque de Milão em Florença, escrito circa 1490, ilustra esse conceito de forma clara: “Filippino, filho do grande pintor Fra Filippo Lippi: aluno do acima mencionado Botticelli e filho do mais ilustre pintor de seu tempo. Suas coisas têm um ar mais doce que as de Botticelli; não creio que tenha[m] tanta arte” (Baxandall, 1991, p. 33; grifo no original). Também derivado do sentido aristotélico, o termo “ciência” no Renascimento, por sua vez, definiu o resultado abstrato da opera- ção de separação entre o que em um determinado conjunto de experiências era comum e o que fazia parte do variável, sendo o que era variável apenas um acidente em relação ao que era comum, sua essência. Publicado em Colônia em 1574, o livro de Franciscus Toletus, Commentaria una cum quaestionibus in VIII libros De physica auscultatione, elucida o sentido desse conceito no período em questão. Para Toletus, as três principais divi- sões da philosophia foram a speculativa, a practica e a factiva, e destas, a primeira era subdividida em metaphysica, mathematica e physica ou philosophia naturalis. 4 É importante evitar confundir esse sentido específico do termo “arte” com seu uso relacionado aos estudos teóri- cos chamados Artes Liberais, compreendidos pelo Trivium (grammatica, logica e rhetorica) e o Quadrivium (arithmetica, geometria, musica e astronomia). Esses termos característicos da escolástica têm sua origem nos sofistas gregos, que lançaram as bases para a elaboração do trivium e do quadrivium ocorrida no final da Antigüidade; a língua grega da época dos sofistas não possuía uma palavra específica para separar a noção de saber e de poder vindos da ação, de modo que a palavra “arte” significava tanto as “artes profissionais” quanto as “artes educativas”, estas últimas volta- das à vida política do cidadão ateniense. Sobre isso, cf. Jaeger, 1979, p. 311-57. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 III philosophia ou scientia naturalis, a partir da qual foram elaboradas as ciências moder- nas, era um corpo de conhecimentos orga- nizado segundo princípios definidos, sen- do seu objetivo principal o conhecimento das primeiras causas. Nesses termos, a dis- tinção entre os estudiosos da filosofia na- tural e os artistas5 do período em questão é simples e clara. Um artista utilizava seus conhecimentos empíricos – eventualmen- te formalizados sob a forma de desenhos, mas raramente sob a forma de um discur- so sistemático6 – para realizar sua arte específica, como a pintura ou a escultura, entre diversas outras (fig. 4). Um filósofo natural formalizava sob a forma de discurso seus conhecimentos teóricos relacionados ao mundo físico. Entre ambos, existiram dife- renças de objetivos, métodos e formas de realização de seus respectivos conhecimentos. Assim, é correto pensar que no século xv houve uma aproximação entre as artes e a filosofia natural. Alguns conhecimentos pertencentes até então ao âmbito da filoso- fia natural foram aplicados nas atividades até então eminentemente práticas. Por exem- plo, o estudo de perspectiva dos artistas do Renascimento foi apenas uma parte de um corpus de conhecimento óptico muito amplo, cujos textos principais eram os de Alhazen, Peckham, Roger Bacon e Pelacano, porção que tinha por objetivo apenas estabelecer corretamente as dimensões dos objetos aos olhos dos homens sobre uma superfície pictórica plana (fig. 5). Antonio di Tuccio Manetti, em sua biografia de Filippo Brunel- leschi escrita circa 1480, deixa isso claro: Figura 4. Michelangelo. A punição de Tityos (giz negro).Windsor Castle, Royal Library. Figura 4. Michelangelo. A punição de Tityos (giz negro).Windsor Castle, Royal Library. so sistemático6 – para realizar sua arte específica, como a pintura ou a escultura, entre diversas outras (fig. 4). Um filósofo natural formalizava sob a forma de discurso seus conhecimentos teóricos relacionados ao mundo físico. Entre ambos, existiram dife- renças de objetivos, métodos e formas de realização de seus respectivos conhecimentos. Assim, é correto pensar que no século xv houve uma aproximação entre as artes e a filosofia natural. Alguns conhecimentos pertencentes até então ao âmbito da filoso- fia natural foram aplicados nas atividades até então eminentemente práticas. 5 Por facilidade de exposição, o termo “artista” nesta resenha está utilizado em seu sentido atual e refere-se a pesso- as como Sandro Boticelli, por exemplo, assim como a Michelangelo Buonarrotti. 6 Cita-se, por exemplo, aqui o pequeno tratado de Cennino Cennini di Valdelsa, do início do século xv, o qual é um manual para a prática de ateliê, sem elaborações teóricas sofisticadas. O terceiro Comentário de Lorenzo Ghiberti (1378-1455), escrito por volta de 1447, é um texto voltado a determinar as bases teóricas das artes, mas resta inacabado. Leon Battista Alberti (1404-1472), por sua vez, escreveu diversos livros que sistematizam as práticas artísticas da época, como a pintura (De pictura) e a escultura (De statua). Entretanto, Alberti pertenceu a um âmbito de cultura muito diverso dos artistas que trabalhavam em ateliês. A maior parte de sua produção escrita não está destinada às artes, e dele não se conhece obras de pintura e escultura, mas apenas de arquitetura. III A meta- physica tratava dos princípios mais comuns de todo ser; a mathematica, por sua vez, era subdividida em mathematica pura, que estudava as entidades que não dependiam do movimento e eram abstraídas desse, e a mathematica media, que estudava as entidades abstraídas mas ainda encontradas em movimento; e, por fim, à physica cabia estudar tudo o que dizia respeito aos sentidos (cf. Wallace, 1988, p. 201). Ciência, physica, 431 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Figura 4. Michelangelo. A punição de Tityos (giz negro).Windsor Castle, Royal Library. Eduardo Henrique Peiruque Kickhöfel philosophia ou scientia naturalis, a partir da qual foram elaboradas as ciências moder- nas, era um corpo de conhecimentos orga- nizado segundo princípios definidos, sen- do seu objetivo principal o conhecimento das primeiras causas. Nesses termos, a dis- tinção entre os estudiosos da filosofia na- tural e os artistas5 do período em questão é simples e clara. Um artista utilizava seus conhecimentos empíricos – eventualmen- te formalizados sob a forma de desenhos, mas raramente sob a forma de um discur- so sistemático6 – para realizar sua arte específica, como a pintura ou a escultura, entre diversas outras (fig. 4). Um filósofo natural formalizava sob a forma de discurso seus conhecimentos teóricos relacionados ao mundo físico. Entre ambos, existiram dife- renças de objetivos, métodos e formas de realização de seus respectivos conhecimentos. Figura 4. Michelangelo. A punição de Tityos (giz negro).Windsor Castle, Royal Library. Eduardo Henrique Peiruque Kickhöfel philosophia ou scientia naturalis, a partir da qual foram elaboradas as ciências moder- nas, era um corpo de conhecimentos orga- nizado segundo princípios definidos, sen- do seu objetivo principal o conhecimento das primeiras causas. Nesses termos, a dis- tinção entre os estudiosos da filosofia na- tural e os artistas5 do período em questão é simples e clara. Um artista utilizava seus conhecimentos empíricos – eventualmen- te formalizados sob a forma de desenhos, mas raramente sob a forma de um discur- so sistemático6 – para realizar sua arte específica, como a pintura ou a escultura, entre diversas outras (fig. 4). Um filósofo natural formalizava sob a forma de discurso seus conhecimentos teóricos relacionados ao mundo físico. Entre ambos, existiram dife- renças de objetivos, métodos e formas de realização de seus respectivos conhecimentos. Figura 4. Michelangelo. A punição de Tityos (giz negro).Windsor Castle, Royal Library. 5 Por facilidade de exposição, o termo “artista” nesta resenha está utilizado em seu sentido atual e refere-se a pesso- as como Sandro Boticelli, por exemplo, assim como a Michelangelo Buonarrotti. III Por exem- plo, o estudo de perspectiva dos artistas do Renascimento foi apenas uma parte de um corpus de conhecimento óptico muito amplo, cujos textos principais eram os de Alhazen, Peckham, Roger Bacon e Pelacano, porção que tinha por objetivo apenas estabelecer corretamente as dimensões dos objetos aos olhos dos homens sobre uma superfície pictórica plana (fig. 5). Antonio di Tuccio Manetti, em sua biografia de Filippo Brunel- leschi escrita circa 1480, deixa isso claro: Aquilo que os pintores atualmente chamam de perspectiva é aquela parte da ci- ência da perspectiva que se volta praticamente a reduzir ou ampliar sistematica- mente, conforme aquilo que o olho percebe, os objetos que estão respectivamente 5 Por facilidade de exposição, o termo “artista” nesta resenha está utilizado em seu sentido atual e refere-se a pesso- as como Sandro Boticelli, por exemplo, assim como a Michelangelo Buonarrotti. as como Sandro Boticelli, por exemplo, assim como a Michelangelo Buonarrotti. 6 Cita-se, por exemplo, aqui o pequeno tratado de Cennino Cennini di Valdelsa, do início do século xv, o qual é um manual para a prática de ateliê, sem elaborações teóricas sofisticadas. O terceiro Comentário de Lorenzo Ghiberti (1378-1455), escrito por volta de 1447, é um texto voltado a determinar as bases teóricas das artes, mas resta inacabado. Leon Battista Alberti (1404-1472), por sua vez, escreveu diversos livros que sistematizam as práticas artísticas da época, como a pintura (De pictura) e a escultura (De statua). Entretanto, Alberti pertenceu a um âmbito de cultura muito diverso dos artistas que trabalhavam em ateliês. A maior parte de sua produção escrita não está destinada às artes, e dele não se conhece obras de pintura e escultura, mas apenas de arquitetura. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 432 Figura 5. Piero della Francesca. Representação pers- pectiva do tratado De prospectiva pingendi (pena e nanquim). Bibioteca Palatina, Parma, MS. Parm. 1576, fol. 29r. Uma falsa lição de anatomia... afastados ou próximos – que se trata de construções, planícies, montanhas e paisagens de todo tipo – e de figuras e outros objetos em todos os lugares pelo tamanho que parecem ter de uma certa distância, correspondendo ao seu maior ou menor distanciamento (apud Baxandall, 1991, p. 200). scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 III No caso da anatomia, muitos artis- tas do século xv possivelmente assistiram (e talvez realizaram) dissecções para re- alizar suas pinturas e esculturas, entre outras obras, e é prudente pensar que eles estudaram muito mais a anatomia da su- perfície do corpo do que as dissecções anatômicas propriamente ditas, como as feitas nas escolas de medicina, e que eles pouco se interessaram pela relação en- tre as formas e os usos (no sentido de funções) do corpo humano. Em Florença, segundo o Statuta universitatis et studii florentini de 1387, os pintores e escultores foram aceitos na Consorteria dei Medici e degli Speziali no ano de 1303. Conforme o documento, aos artistas eram entregues à vontade corpos de pessoas que morriam de causas naturais no hospital de Santa Maria Novella ou em outras enfermarias da cidade, enquanto aos médicos apenas dois cadáveres de criminosos enforcados de ambos os sexos eram for- necidos por ano (excepcionalmente três, dois homens e uma mulher) para o exercício anatômico. O motivo desta desigual distribuição foi que enquanto os artistas apenas se preocupavam com a superfície dos cadáveres, os médicos os dissecavam e destruíam todas as suas partes (Statuta universitatis et studii florentini, 6 Rubriche, 62a até 72a. Cf. Bernabeo, 1984, p. 31). Figura 5. Piero della Francesca. Representação pers- pectiva do tratado De prospectiva pingendi (pena e nanquim). Bibioteca Palatina, Parma, MS. Parm. 1576, fol. 29r. Figura 5. Piero della Francesca. Representação pers- pectiva do tratado De prospectiva pingendi (pena e nanquim). Bibioteca Palatina, Parma, MS. Parm. 1576, fol. 29r. Em conformidade com isso, no sentido de mostrar a diferença de abordagem entre médicos e artistas sobre o corpo humano, segundo o tratado Anothomia de Hie- ronymo Manfredi, escrito em Bolonha circa 1490, a anatomia superficial ainda era de pouco interesse para os médicos. Apenas os últimos quatro fólios do manuscrito desse 433 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Eduardo Henrique Peiruque Kickhöfel tratado – a partir do final do fólio 44r até o fólio 47v – estão dedicados à anatomia das extremidades (cf. Singer, 1975, p. 130-64). Considerando que o texto de Manfredi, uma ampliação do conhecido texto de Mondino de Luzzi, é um guia de dissecção, pode-se perguntar que conhecimentos um artista podia obter em uma demonstração pública de anatomia para seus fins artísticos. III Em suma, não há documentos que sustentem a idéia de que era necessário um artista se familiarizar com a estrutura e os usos do cor- po humano para representá-lo em suas obras, excetuando o estranho e singular caso de Leonardo da Vinci, que operou uma aproximação entre as artes e a ciência do perío- do de uma forma inédita e profunda. Leonardo foi o primeiro a sistematizar os conhe- cimentos técnicos de sua época, de modo a proporcionar à prática uma sistematização teórica elaborada. Uma simples comparação dos manuscritos técnicos de Mariano Taccola e Francesco di Giorgio Martini com os de Leonardo, em especial o Codice Ma- drid I, evidencia isso. Leonardo também aplicou seus métodos de estudar mecânica (sua “scientia della machine”) e máquinas simples a seus estudos de anatomia. Galluzzi (1996) defende essa tese com muita propriedade. Além disso, é importante frisar que a própria disposição de desenhar de Leonardo da Vinci serviu de base para seu modo de elaborar e formalizar suas idéias. Em poucas palavras, pode-se pensar que as artes foram informadas por conhe- cimentos da filosofia natural, e se das aplicações desses conhecimentos problemas novos surgiram, esses problemas foram resolvidos em seu próprio âmbito produtivo, pois os artistas que trabalhavam sobre essas aplicações não tinham intenções além de produzir pinturas, esculturas etc., e por esse motivo, suas aplicações permaneceram e permanecem apenas como aplicações. Nesse sentido, não há meio de sustentar que (p. 32) “na busca pelo ideal de bele- za da Antigüidade Clássica, a maioria dos grandes pintores e escultores renascentistas estudou anatomia tanto quanto os médicos”, que eles (p. 47) “eram freqüentadores assíduos das demonstrações públicas de dissecção” e que “encontrar um escalpelo no ateliê de um pintor ou escultor era algo natural”. Nesse sentido, pergunta-se surpreso a que se refere (p. 47) a “Academia Florentina de Arte”, a qual “tomara a atitude inédita de instituir o curso de anatomia em seu currículo”. Talvez a referência seja à Accademia del Disegno fundada em 1561 por Giorgio Vasari, mas a parte do texto em que está essa informação é confusa e sugere a existência de uma instituição acadêmica antes da me- tade do século xvi. Tampouco há como sustentar o interesse de Michelangelo por anatomia médica. As citações de Condivi presentes no livro em questão apontam para uma direção opos- ta à proposta por seus autores (p. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 III 51): “[Michelangelo] havia aprendido tanto dessa ciência [dissecção] que pensava em escrever um tratado, como um serviço a quem ti- vesse interesse em esculpir e pintar os movimentos do corpo e a estrutura dos ossos de 434 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Uma falsa lição de anatomia... todas as maneiras”. Seus interesses artísticos também ficam muito claros nas seguin- tes palavras de Condivi citadas pelos autores (p. 51), que expressam críticas de Miche- langelo às teorias artísticas de Albrecht Dürer: Verdade seja dita, Albrecht aborda apenas as proporções e diversidades do corpo, para as quais nenhuma regra pode ser dada, e apresenta suas figuras de pé, como estacas. Daquilo que é mais importante, os movimentos e gestos do corpo huma- no, ele não diz uma palavra. Essa menção, de fato, opõe-se à observação dos autores (p. 40) de que “o tema predileto de Michelangelo era a anatomia do homem”, uma interpretação feita especi- almente para justificar sua tese não fundamentada. Por fim, eles citam Condivi mais uma vez (p. 54-55): Agora que ele já está com idade avançada e não espera mais revelar suas idéias no mundo da escrita, tem-me revelado tudo nos mínimos detalhes. Também come- çou a falar a respeito com o mestre Realdo Colombo, um anatomista de primeira linha e amigo pessoal meu e de Michelangelo que, por conta disso, mandou-nos o corpo de um mouro (...). Nesse cadáver, Michelangelo me mostrou coisas raras e escondidas, provavelmente desconhecidas ainda, as quais eu anotei na espe- rança de um dia poder publicá-las, com a ajuda de um estudioso, para a conveni- ência e utilização de todos os que desejam lidar com a pintura e a escultura. Pode-se pensar plausivelmente que Michelangelo não teve formação para es- crever um tratado de anatomia médica, como seu planejado trabalho com Realdo Colombo sugere, da mesmo forma que o próprio Condivi que, com sua formação artís- tica, esperava a ajuda de um estudioso para ordenar e publicar suas próprias anota- ções. Os autores admitem isso ao escrever que (p. 57) “se Michelangelo tivesse realiza- do o seu [tratado de anatomia], teria existido uma obra que, apesar de concebida para estudantes de arte, certamente haveria ultrapassado as fronteiras dos ateliês para ocu- par um lugar de destaque nas bibliotecas de Medicina”. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 III Especular sobre a importância de um tratado de anatomia que não existe é um exercício sem sentido para um histori- ador, e atribuir a uma especulação como essa alguma certeza, um absurdo. Assim, é errôneo chamar Michelangelo (p. 28) de “artista e cientista” e escrever (p. 47) a res- peito de sua “formação científica”, de suas (p. 50) “investigações científicas” e (p. 57) de seu “conhecimento científico”. Na mesma linha de raciocínio, considerar Michelangelo humanista é um erro grave (p. 216). O termo humanismo, oriundo de escritores romanos (em especial Cícero scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 435 Eduardo Henrique Peiruque Kickhöfel e Gellius), foi revivido por Francesco Petrarca, Coluccio Salutati e outros durante o século xiv, e por volta da metade do século xv circunscreveu um bem definido ciclo de estudos, os chamados studia humanitatis, o qual incluía a grammatica, a rhetorica, a poetica, a historia e a philosophia moralis. Diferentemente das Artes Liberais do final da Idade Média, o humanismo não incluiu a logica ou o quadrivium (arithmetica, geome- tria, astronomia e musica), assim como as disciplinas que foram importantes para as universidades durante essa mesma época e o Renascimento, como a teologia, a juris- prudência (direito) e a medicina, e as disciplinas filosóficas que não a ética, como a lógica, a filosofia natural e a metafísica. Entretanto, além de suas especulações filológicas ou filosóficas como literatos, juristas ou filósofos de profissão, muitos es- tudaram outras disciplinas que se estenderam para além do programa humanista, po- dendo ser encontrados – a partir de meados do século xv – humanistas que também se exerceram na philosophia naturalis: médicos e professores de medicina, juristas e mes- mo teólogos (cf. Kristeller, 1988). Mesmo sendo culto a ponto de ser poeta de reco- nhecido valor, cujos versos fazem parte de muitas antologias de literatura italiana, Michelangelo evidentemente não se enquadrou nesse âmbito da cultura renascentista. Além disso, escrever que (p. 45) “os registros de seus estudos influenciaram mais as ciências do que os comentários dos médicos gregos e medievais” e que (p. 216) “a produção artística italiana dos séculos xv e xvi passou a representar um meio de inves- tigação, um instrumento dotado de bases teóricas, que despontou entre as atividade intelectuais mais elaboradas” consiste na consideração do desenvolvimento das ciên- cias naturais modernas principalmente a partir do trabalho dos artistas. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Figura 7. Ugo da Carpi (?). Ilustração do livro Isagogae breves, de Berengario da Carpi (gravura em madeira). III Segundo esta perspectiva, arte e ciência estavam (p. 45) “inexoravelmente unidas”. Não há como negar seu importante papel no desenvolvimento de ciências naturais como a anato- mia, por exemplo. Entretanto, apenas ilustrações não configuram uma ciência, nem mesmo uma ciência descritiva. É errôneo desconsiderar o trabalho de sistematização feito pelos filósofos naturais que estudavam a anatomia do corpo humano e que ti- nham formação específica para isso, como Andreas Vesalius. Não há documentos que sustentem ambas opiniões. Muito pelo contrário, considerando as habilidades de de- senho de Antonio del Pollaiolo e os recursos gráficos então existentes, ilustrações como as do livro de Vesalius podiam ser feitas por volta de 1480 na Itália, mas isso não ocor- reu. De fato, o tratado de Hieronymo Manfredi citado acima, escrito em um ambiente de corte propício às artes, não contém ilustrações, e os livros impressos a respeito de anatomia até meados do século xvi contêm ilustrações evidentemente toscas e muito distantes das habilidades dos artistas da época, como mostra a baixa qualidade das ilus- trações anatômicas dos livros de Magnus Hundt, Antropologium, de hominis dignitate natura et proprietalibus (Leipzig, 1501) e Gregorius Reisch, Margarita philosophica (Leipzig, 1503) (fig. 6), e mesmo os textos especificamente voltados para a anatomia 436 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Uma falsa lição de anatomia... Figura 6. Artista anônimo do norte eu- ropeu. Ilustração do livro Margarita philosophica, de Gregorius Reisch (gra- vura em madeira). de Johannes de Kethan, Fasciculus medicinae (Vene- za, 1491) e Berengario da Carpi, Carpi anatomia Mun- dini noviter impressa ac per Carpum castigata, (Bolo- nha, 1514), Carpi Commentaria cum amplissimus additionibus super anatomia Mundini una cum textu e iusdem in pristinum et verum nitorem redacto (Bolonha, 1521) e Isagogae breves (Bolonha, 1522) (fig. 7). Assim, não é plausível pensar que os artistas foram responsáveis pelo desenvolvimento da anato- mia devido a suas técnicas ilustrativas. Os artistas foram importantes quando trabalharam para os ana- tomistas das universidades, o que ocorreu principal- mente a partir da obra de Andreas Vesalius, De huma- nis corporis fabrica, feita em solo italiano e publicada na Basiléia em 1543. Nesse sentido, pensar a produ- ção artística como um “meio” ou “instrumento” im- plica pensar um fim diferente de um fim artístico, e um fim presumivelmente científico, uma opinião muito discutível, conforme exposto acima. Além de interesses específico de um artista no âmb restrita. III Leonardo da cia ao fazer isso, conf Atlantico 117v (confor Assim, não é plausível pensar que os artistas foram responsáveis pelo desenvolvimento da anato- mia devido a suas técnicas ilustrativas. Os artistas foram importantes quando trabalharam para os ana- tomistas das universidades, o que ocorreu principal- mente a partir da obra de Andreas Vesalius, De huma- nis corporis fabrica, feita em solo italiano e publicada na Basiléia em 1543. Nesse sentido, pensar a produ- ção artística como um “meio” ou “instrumento” im- plica pensar um fim diferente de um fim artístico, e um fim presumivelmente científico, uma opinião muito discutível, conforme exposto acima. Além de interesses específicos e falta de formação, a inserção de um artista no âmbito da ciência natural era muito restrita. Leonardo da Vinci enfrentou muita resistên- cia ao fazer isso, conforme mostra um texto do Codice Atlantico 117v (conforme sua numeração antiga): Figura 6. Artista anônimo do norte eu- ropeu. Ilustração do livro Margarita philosophica, de Gregorius Reisch (gra- vura em madeira). Figura 6. Artista anônimo do norte eu- ropeu. Ilustração do livro Margarita philosophica, de Gregorius Reisch (gra- vura em madeira). Sei bem que por não ser letrado que algum presun- çoso poderá razoavelmente me censurar, alegando justamente que eu sou um homem sem letras. Pes- soas estúpidas ! [...] Dirão que por não ter eu letras, não posso dizer aquilo de que quero tratar, e não sa- bem elas que as minhas coisas são mais de serem tiradas da experiência que das palavras de outros, a qual foi mestra de quem bem escreveu, e assim por ser mestra e em todos casos eu a alegarei. Figura 7. Ugo da Carpi (?). Ilustração do livro Isagogae breves, de Berengario da Carpi (gravura em madeira). 437 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Eduardo Henrique Peiruque Kickhöfel Sua tentativa de fazer da pintura uma ciência era inviável mesmo em sua época, de modo que Leonardo não deixou uma escola e nem mesmo um único discípulo se- guidor de suas idéias muito pouco ortodoxas, como o erudito Paolo Giovio soube já circa 1524 (cf. Giovio apud Barocchi, 1971, p. 7-9). Por fim, se as artes e a ciência esta- vam “inexoravelmente unidas”, aguarda-se ansiosamente uma explicação para o sur- gimento da ciência matematizada no século xvii. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 IV Surpreendentemente, opiniões como essas persistem firmes entre historiado- res da arte. Entre diversos exemplos, Chastel (1974, p. 219), grande historiador da arte francês, menciona que no século xv os ateliês florentinos foram transformados em “pequenos centros de estudo científico”: Os ateliês haviam se transformado em pequenos centros de estudo científico: a anatomia e a perspectiva não eram ensinadas no Estúdio, mas nos ateliês de Verrocchio e Pollaiuolo. Os conhecimentos empíricos e tecnológicos precisavam ser sistematizados, e um enquadramento conceitual e formas lingüísticas apro- priadas era requerido, mas os textos científicos da Antigüidade e mesmo da Ida- de Média apenas parcialmente proporcionavam isso. Os ateliês haviam se transformado em pequenos centros de estudo científico: a anatomia e a perspectiva não eram ensinadas no Estúdio, mas nos ateliês de Verrocchio e Pollaiuolo. Os conhecimentos empíricos e tecnológicos precisavam ser sistematizados, e um enquadramento conceitual e formas lingüísticas apro- priadas era requerido, mas os textos científicos da Antigüidade e mesmo da Ida- de Média apenas parcialmente proporcionavam isso. Entretanto, esquecendo o anacronismo de “conhecimentos tecnológicos”, os artistas estavam interessados na produção de pinturas, esculturas etc. e tinham méto- dos específicos para isso, mas não na elaboração sistemática de conceitos e teorias, de modo que a expressão utilizada por Chastel é sem propriedade alguma. Brizio (1974, p. 22-3), vinciana de renome, considerando os artistas florentinos do século xv, escreve: Ao lado da perspectiva, o outro grande conceito que dominou a arte do Renasci- mento foi o da imitação da natureza. Isso não deve ser interpretado no sentido restrito do século xix relativo à reprodução externa e epidérmica dos fenômenos da natureza. Para os artistas florentinos dos século xv, a imitação da natureza foi uma ‘sutil especulação’: eles seguiram a ordem de investigar as leis da natureza e, de acordo com essas mesmas leis, de criar, reproduzindo os processos criativos da própria natureza. Desconsiderando o anacronismo de “leis da natureza”, essa opinião pode ser colocada em questão a partir da simples observação de que não existem escritos de 438 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Uma falsa lição de anatomia... artistas – excetuando aqueles a respeito de perspectiva, no sentido exposto brevemen- te acima – como os de Leonardo da Vinci no século xv ou mesmo xvi. Uma rápida con- sulta ao grande livro de Schlosser (1994) mostra isso de forma clara. 7 Um exemplo claro disso está em um livro recente de White (2002). Sua “história de reavaliação” da obra de Leo- nardo da Vinci nada mais é que uma tese muito em voga na primeira metade do século passado, mas que desde então não tem o menor respaldo dos estudiosos vincianos que merecem tal nome. De fato, na introdução de seu livro, sem muita surpresa intitulada “O homem total”, White (2002, p. 21) mostra seu desconhecimento grosseiro da histo- riografia vinciana ao mencionar que “quase nada foi escrito das importantes idéias científicas” de Leonardo da Vin- ci. Uma breve consulta aos acervos on line da Biblioteca Leonardiana de Vinci <www.reanet.comune.empoli.fi.it/vinci/ home.htm> e do Istituto e museo di storia della scienza de Florença <www.imss.fi.it/biblio/indice.htm> mostra como White, um jornalista inglês de divulgação científica, não sabia minimamente a respeito do assunto que se propôs a escrever. IV Idéias como essas têm sua origem em um conhecido estudo de Panofsky (1924), segundo o qual artistas como Filippo Brunelleschi e Albrecht Dürer descobriram os princípios da geometria descritiva antes dos matemáticos. Em um ensaio posterior, Panofsky (1962, p. 133) escreve que os esquemas de perspectiva dos artistas renascentistas “colocaram os fun- damentos do que foi desenvolvido após como geometria projetiva e analítica” e, de- fendendo a tese da “descompartimentalização” dos saberes no Renascimento, ele afir- ma que nesse período (p. 136) “os maiores avanços nas ciências naturais foram feitos por engenheiros, construtores de instrumentos e artistas, mas não os acadêmicos” e que (p. 140) “muito do que mais tarde foi isolado como ciência natural veio a existir nos ateliês dos artistas”, idéias dure a morire nas mãos de jornalistas e outros escritores diletantes, tanto quanto a de Leonardo da Vinci como gênio incompreendido e ante- cipador da ciência e de muitas invenções modernas.7 Talvez seja interessante especular brevemente a respeito dos motivos da exis- tência de idéias como essas, pois afinal, em um meio de pessoas tão cultas e eruditas, isso tem de ter alguma explicação plausível. Definir arte é uma tarefa notoriamente difícil, talvez porque arte está hoje associada a uma qualidade presente em certos obje- tos, justamente os objetos de arte, de difícil apreensão discursiva. Entretanto, para os fins aqui propostos, basta pensar que quando se diz hoje “a arte do século xv” pensa-se a respeito das esculturas, pinturas e outras obras do século xv produzidas, ou seja, os produtos de disposições de produzir. Isso é evidente quando se lê um livro cujo título é “História da arte do Renascimento”, o qual descreve a história das obras de arte do Renascimento, assim como no título do livro aqui em questão. Arte como uma disposi- ção talvez seja uma noção muito abstrata para o senso comum, especialmente em uma época em que o alto valor monetário de uma obra de arte freqüentemente é notícia de jornal. A ciência que aqui interessa mantém seus contornos gerais desde a época dos gregos, ao menos entre pessoas educadas, no sentido de expressar a busca de relações 7 Um exemplo claro disso está em um livro recente de White (2002). IV Sua “história de reavaliação” da obra de Leo- nardo da Vinci nada mais é que uma tese muito em voga na primeira metade do século passado, mas que desde então não tem o menor respaldo dos estudiosos vincianos que merecem tal nome. De fato, na introdução de seu livro, sem muita surpresa intitulada “O homem total”, White (2002, p. 21) mostra seu desconhecimento grosseiro da histo- riografia vinciana ao mencionar que “quase nada foi escrito das importantes idéias científicas” de Leonardo da Vin- ci. Uma breve consulta aos acervos on line da Biblioteca Leonardiana de Vinci <www.reanet.comune.empoli.fi.it/vinci/ home.htm> e do Istituto e museo di storia della scienza de Florença <www.imss.fi.it/biblio/indice.htm> mostra como White, um jornalista inglês de divulgação científica, não sabia minimamente a respeito do assunto que se propôs a escrever. 439 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Eduardo Henrique Peiruque Kickhöfel causais entre os fenômenos do mundo físico. Entretanto, desde o século xvii a ciência está vinculada ao modelo das ciências físico-matemáticas baseadas sobre experimen- tos e expressões matemáticas, características que ao senso comum configuram conhe- cimento rigoroso e seguro a ser seguido por todas as disciplinas. Além disso, e eis um ponto importante, desde o século xix a ciência está fortemente vinculada à tecnologia. Pensa-se assim que qualquer disciplina que faça uso de algum método que se asseme- lhe minimamente ao chamado “método científico” merece o estatuto de ciência, e que os objetivos das ciências são os produtos tecnológicos delas derivados. Ciência como teoria talvez seja uma noção muito abstrata para o senso comum, principalmente em uma sociedade que se denomina orgulhosamente “sociedade do conhecimento”, mas que efetivamente valoriza o lado utilitário e imediatista das ciências, na qual qualquer saber de algum modo científico está voltado à tecnologia e a outras “demandas da socie- dade”. Isto posto, de modo geral os textos de história da arte do Renascimento não apresentam definições mínimas de “arte” e “ciência” do período e, sem isso, eles apli- cam as vagas noções atuais expressas por essas palavras. Assim, pensa-se facilmente que os artistas do Renascimento, investigando anatomia, faziam investigações cientí- ficas, e que em suas obras existem conhecimentos científicos, sendo esse efetivamen- te um de seus objetivos. 8 Seu tratado De prospectiva pingendi (A perspectiva dos pintores), da segunda metade do século xv, tem bases evidente- mente euclidianas e é um texto voltado à prática dos pintores, como seu título indica, mas não uma contribuição à ciência matemática. IV Segundo esse raciocínio, uma pintura de Piero della Francesca feita nos moldes definidos por Leon Battista Alberti expressa uma concepção de espa- ço que antecipa o espaço abstrato da ciência do século xvii,8 e uma figura de Antonio del Pollaiolo contém conhecimentos de anatomia médica. Segue-se disso que suas prá- ticas artísticas e suas obras influenciaram as ciências naturais no período. Nesse âm- bito repleto de muita beleza e pouca sofisticação conceitual, “arte” e “ciência” são ter- mos tão vagos e mal aplicados a ponto de serem intercambiáveis. No caso do livro em questão, um livro muitíssimo longe da imensa cultura e erudição dos livros de Panofsky, só faltou os autores escreverem A ciência secreta de Michelangelo. V A obra de Michelangelo e sua relação com a cultura da Renascença são evidentemente complexas, como atesta não só a perenidade de sua obra, mas também a imensa for- tuna crítica a seu respeito. No caso da Capela Sistina, Michelangelo se dirigiu a um público específico com uma educação específica. Passados quase cinco séculos de sua 8 Seu tratado De prospectiva pingendi (A perspectiva dos pintores), da segunda metade do século xv, tem bases evidente- mente euclidianas e é um texto voltado à prática dos pintores, como seu título indica, mas não uma contribuição à ciência matemática. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 440 Uma falsa lição de anatomia... realização, seu público mudou drasticamente; não mais existem seu poderoso comitente e poucos fiéis afortunados, mas observadores eruditos e muitos turistas afoitos, even- tualmente inventivos e curiosos. A tarefa dos historiadores da arte e da cultura é recupe- rar a partir de documentos de época parte do contexto de sua realização para que algum entendimento objetivo seja possível, uma tarefa que não raro ocupa toda uma vida. Segundo Baxandall (1991, p. 38), cada pessoa tem seu estilo cognitivo individual. Na vida cotidiana, isso é praticamente insignificante, mas “em algumas circunstânci- as, a diferença entre um indivíduo e outro, em alguns casos marginais, pode adquirir uma curiosa importância”. Na introdução do livro, um dos autores expressa inocente- mente seu estilo cognitivo individual ao descrever sua primeira visita à Sistina, a qual lhe proporcionou a primeira idéia do livro. A respeito do Pecado original, após menci- onar a figura feminina que tenta Adão e Eva, ele escreve (1991, p. 20-1): O que me chamou a atenção, porém, foi o pequeno toco árido com um ramo apon- tado para cima que está ao lado de Eva. Poucos dias antes, ainda me dedicava a um longo estágio que realizei no Departamento de Anatomia da Universidade. Nele, a prática me havia direcionado para a especialização cirúrgica. Com a mente im- pregnada dos estudos anatômicos, vi nesse toco um arco aórtico dissecado. A “curiosa importância” de Baxandall, aqui evidentemente não relacionada a um caso marginal da história da arte, passou ao largo dele e de seu fiel colega durante o único ano de estudos em que eles se dedicaram a descobrir as anatomias secretas de Michelangelo Buonarrotti. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 V Na filosofia da ciência, se aqui é permitida uma expressão do âmbito das chamadas “ciências duras”, chama-se isso de “impregnação teórica dos fatos”. No cotidiano, pessoas leigas sabem disso ao dizer que “cada pessoa vê o mundo com os olhos que tem”, mas elas não levam isso às suas últimas conseqüências, isto é, face a uma situação qualquer cada pessoa só pode conhecer a si mesma. Sem saber dis- so, elas não sabem a importância e mesmo a necessidade de transcender isso na busca por objetividade, como o livro em questão evidentemente mostra. A impregnação teó- rica é necessária para a elaboração de hipóteses, mas necessita ser confirmada por cri- térios intersubjetivos. Concordando com os autores (p. 27), a obra de Michelangelo não foi deixada para leigos. Por fim, resta indagar a respeito da edição do livro em questão, um livro ricamente ilustrado e de edição evidentemente cara como poucos livros de história da arte lança- dos no Brasil. Não cabe aqui supor más intenções por parte dos autores, pois seu texto expressa sinceramente a ingenuidade de duas pessoas fascinadas por suas descober- tas. Entretanto, apenas sinceridade não faz uma boa obra, qualquer que seja. Pode-se supor a falta de rigor de seus editores. O prefácio repleto de lugares comuns a respeito 441 scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 Eduardo Henrique Peiruque Kickhöfel do Renascimento, os quais lembram algum iluminista muito ingênuo e retardatário a maldizer mais uma vez a ainda tão maldita Idade Média, não resiste nem mesmo a mais caridosa das leituras, assim como os elogios presentes na contra-capa do livro, feitos por médicos sem cultura para comentar um livro que, mesmo contendo aspectos da medicina, é um livro de história. Pode-se pensar também a respeito de uma edição de feição comercial voltada a um público pouco exigente, o qual talvez se encante com suas interpretações inventivas e curiosas e com seu jogo de associações livres. Seja como for, mesmo com evidentes pretensões de influir nos estudos acadêmicos a respeito da obra de Michelangelo, o livro em questão não os compromete, pois historiadores não se deixam enganar por sua falta de rigor. Eduardo Henrique Peiruque Kickhöfel Membro do Projeto Temático “Estudos de filosofia e história da ciência” da FAPESP, doutorando do Departamento de Filosofia da Universidade de São Paulo. kickhofel@hotmail.com Eduardo Henrique Peiruque Kickhöfel Membro do Projeto Temático “Estudos de filosofia e história da ciência” da FAPESP, doutorando do Departamento de Filosofia da Universidade de São Paulo. kickhofel@hotmail.com V Entretanto, o livro é prejudicial à educação de pessoas leigas genuinamente interessadas em se instruir a respeito de história da arte e da cultura do Renascimento em geral, e sua falta de precisão das categorias da época e suas diferenças em relação às categorias atuais, especialmente as relacionadas à ciên- cia, contribui para a ignorância do senso comum relativa à ciência e a métodos rigoro- sos de investigação históricos, ignorância da qual o livro de Gilson Barreto e Marcelo G. de Oliveira também é vítima. Eduardo Henrique Peiruque Kickhöfel Membro do Projeto Temático “Estudos de filosofia e história da ciência” da FAPESP, doutorando do Departamento de Filosofia da Universidade de São Paulo. kickhofel@hotmail.com scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 referências bibliográficas Barocchi, P. (Ed.). Scritti d’arte del cinquecento. Milão/Nápoles, Riccardo Ricciardi Editore, 1971. v. 1. Barocchi, P. (Ed.). Scritti d’arte del cinquecento. Milão/Nápoles, Riccardo Ricciardi Editore, 1971. v. 1. Baxandall, M. O olhar renascente. Pintura e experiência social na Itália da Renascença. Trad. de M. C. P. R. Baxandall, M. O olhar renascente. Pintura e experiência social na Itália da Renascença. Trad. de M. C. P. R. Almeida. São Paulo, Paz e Terra, 1991 [1972]. Baxandall, M. O olhar renascente. Pintura e experiên Almeida. São Paulo, Paz e Terra, 1991 [1972]. Bernabeo, R. Artisti pre-vesaliani. In: Ciardi, R. P. & Tomasi, L. T. (Ed.). Immagini anatomiche e natura- listiche nei disegni degli Uffizi, secc. XVI e XVII. Florença, Gabineto Disegni e Stampe degli Uffizi LX/Leo S. Olschki Editore, 1984. p. 31-6. g g ff S. Olschki Editore, 1984. p. 31-6. 9 Brizio, A. M. The painter. In: Reti, L. (Ed.). The unknown Leonardo. Nova Iorque, Abradale Press/Harry N. Abrams Publishers, 1974. p. 20-55. Brizio, A. M. The painter. In: Reti, L. (Ed. N. Abrams Publishers, 1974. p. 20-55. N. Abrams Publishers, 1974. p. 20-55. Chastel, A. Treatise on painting. In: Reti, L. (Ed.). The unknown Leonardo. Nova Iorque, Abradale Press/ Harry N. Abrams Publishers, 1974. p. 216-39. Chastel, A. Treatise on painting. In: Reti, L Chastel, A. Treatise on painting. In: Reti, L. (Ed.). Harry N. Abrams Publishers, 1974. p. 216-39. , p g , ( ) Harry N. Abrams Publishers, 1974. p. 216-39. Harry N. Abrams Publishers, 1974. p. 216-39. 9 9 Ciardi, R. P. & Tomasi, L. T. (Ed.). Immagini anatomiche e naturalistiche nei disegni degli Uffizi, secc. XVI e XVII. Florença, Gabineto Disegni e Stampe degli Uffizi LX/Leo S. Olschki Editore, 1984. scientiæ zudia, São Paulo, v. 2, n. 3, p. 427-43, 2004 442 Uma falsa lição de anatomia... Condivi, A. Michelangelo: la vita raccolta dal suo discepolo Ascanio Condivi. Milão, P. d’Ancona, 1982. g p Galluzzi, P. Renaissance engineers from Brunelleschi to Leonardo. Florença, Giunti, 1996. Jaeger, W. Paideia. A formação do homem grego. Trad. de A. M. Parreira. São Paulo, Martins Fontes, 1979 [1936]. 9 Kristeller, P. O. Humanism. In: Schmitt, C. B.; Skinner, Q. & Kessler, E. (Ed.). The Cambridge history of Renaissance philosophy. Cambridge, Cambridge University Press, 1988. p. 113-37. Panofsky, E. Idea. Ein Beitrag zur Begriffsgeschichte der älteren Kunsttheorie. Leipzig, Teubner, 1924. (Studien der Bibliothek Warburg V). _______. referências bibliográficas Artist, scientist, genius: notes on the “Renaissance-Dämmerung”. In: _______. The Re _______. Artist, scientist, genius: notes on the “Renaissance-Dämmerung”. In: _______. The Renaissance. Six essays. Nova Iorque, Harper & Row Publishers, 1962. p. 121-82. Six essays. Nova Iorque, Harper & Row Publishers, 1962. p. 121-82. Perrig, A. Michelangelo’s drawings. Yale, Yale University Press, 1991. Reti, L. (Ed.). The unknown Leonardo. Nova Iorque, Abradale Press/Harry N. Abrams Publish Reti, L. (Ed.). The unknown Leonardo. Nova Iorque, Abradale Press/Harry N. Abrams Publishers, 1974. Schlosser, J. La letteratura artistica. Florença/Viena, La Nuova Italia/Kunstverlag Anton Schroll, 1994 [1924]. q y 974 Schlosser, J. La letteratura artistica. Florença/Viena, La Nuova Italia/Kunstverlag Anton Schroll, 1994 [1924]. 9 Schmitt, C. B.; Skinner, Q. & Kessler, E. (Ed.). The Cambridge history of Renaissance philosophy. Cambridge, Cambridge University Press, 1988. Singer, C. A study in early Renaissance anatomy, with a new text. The Anathomia of Hieronymo Manfredi (1490). In: _______. (Ed.). Studies in the history and method of science. Nova Iorque, Arno Press, 1975. v. 1. p. 80-164. C. de. Corpus dei disegni di Michelangelo. Novara, Istituto Geografico de Agostini, 1975. v. 1. Vasari, G. Le vite de’ piu’ eccellenti pittori, scultori e architettori scritte da Giorgio Vasari pittore a architetto aretino. Florença, Presso S. Audin, 1822 [1568]. v. 3. Wallace, W. A. Traditional natural philosophy. In: Schmitt, C. B.; Skinner, Q. & Kessler, E. (Ed.). The Cambridge history of Renaissance philosophy. Cambridge, Cambridge University Press, 1988. p. 201-35. White, M. Leonardo, o primeiro cientista. Rio de Janeiro, Record, 2002 [2000]. Wallace, W. A. Traditional natural philosophy. In: Schmitt, C. B.; Skinner, Q. & Kessler, E. (Ed.). The Cambridge history of Renaissance philosophy. Cambridge, Cambridge University Press, 1988. p. 201-35. White, M. Leonardo, o primeiro cientista. Rio de Janeiro, Record, 2002 [2000]. White, M. Leonardo, o primeiro cientista. Rio de Janeiro, Record, 2002 [2000]. 443 443
https://openalex.org/W4243766478	https://hrcak.srce.hr/file/302808	English	null	Investigating Parameter Interactions with the Factorial Design Method: Webster’s Optimal Cycle Length Model	Tehnički vjesnik/Tehnički vjesnik	2,018	cc-by	5,605	1 INTRODUCTION seemed to be of secondary importance [7]. Akgüngör performed another study with factorial design method to determine the parameter effects of Webster delay model. The study results showed that the cycle length and green time are the most effective two parameters for the delay model. Moreover, the cycle length - green time and traffic volume- cycle length have major effects for the multi- parameter interactions. Delay studies with factorial design method investigated by Akgüngör showed that cycle length is an important parameter for intersection performance [8]. Average vehicle delay at a signalized intersection can be minimized by a good estimate of the cycle length. Longer and shorter cycle lengths cause increased delay and queues. In fact, short cycle lengths do not provide adequate green time for all phases and result in cycle failures, while longer cycle lengths cause unused green time. This forces vehicles to wait longer than necessary, and results in more waiting time and queues. Therefore, estimation of optimal cycle length is the most important parameter for traffic signal system. Additionally, capacity, delay, start-up lost time are other important parameters for traffic signal system and they are related to optimal cycle length [9-11]. Çalışkenlli et al. [12] showed that start-up lost time is a significant parameter and it increases rapidly as cycle time increases. Sensitivity analysis of a model is useful to determine relative effects of model parameters on model results. The one-way sensitivity analysis, named as change of one- factor-at-a-time, is a widely used sensitivity analysis because of its simplicity. In this method, only one parameter is changed at one time by a given amount, and is examined for the impact that the change has on the model’s results. On the other hand, the major weakness of this method is its inability to identify multiple factor interactions among the model parameters. As an alternative approach, the factorial design method developed by Box et al. has been successfully employed in various sensitivity studies [1-5]. Unlike the standard "change one-factor-at-a- time" sensitivity approach, this method has the advantage of testing both the sensitivity of model results to changes in individual parameters, and to interactions within a group of parameters. p This method has successfully been applied to different models in traffic and transportation engineering. Akgüngör et al. investigated the parameter effects of HCM 2000 delay model by factorial design method. ISSN 1330-3651 (Print), ISSN 1848-6339 (Online) ISSN 1330-3651 (Print), ISSN 1848-6339 (Online) https://doi.org/10.17559/TV-20170908185847 Orginal scientific paper Investigating Parameter Interactions with the Factorial Design Method: Webster’s Optimal Cycle Length Model Ali Payıdar AKGÜNGÖR, Ersin KORKMAZ Abstract: Accurate estimation of cycle length is an important factor in the performance of a signalized intersection. Cycle length is determined by employing some parameters such as arrival flow, number of phase, lost time etc., but each parameter has different effects on a cycle length model. If the effects of parameters and their interactions in the cycle length model are known, the performance of the model can be effectively increased. In this study, the sensitivity of optimal cycle length model proposed by Webster and its parameters were analysed with the factorial design method. The reason for selecting this model is that the model has still been used in signal timing practice and has lead many studies of researchers over 50 years. The evaluation of sensitivity analysis shows that while arrival flow as single parameter has a major effect on the optimal cycle length model, the remaining single parameters of the model (i.e., the number of phase in a cycle length, saturation flow and lost time) have secondary importance. Additionally, two parameter interactions of arrival flow-saturation flow have major effect on the model results. For three parameter interactions, the number of phase-arrival flow-saturation flow interaction has a slightly larger effect than the other three-parameter interactions. As a result, the factorial design method is an effective tool to determine the importance of the model parameters for researchers, and it can be employed to other traffic engineering applications. Keywords: arrival flow; factorial design method; optimal cycle length; sensitivity analysis; Webster model 1 INTRODUCTION This study showed that the arrival flow, the saturation flow, and the green signal time are the main parameters that significantly affect the average control delay. Additionally, the multi-parameter interactions of the arrival flow-saturation flow and the arrival flow-green signal time have major effects on the model. This study also illustrated that the effect of parameters on the uniform delay showed that the green signal time and the cycle length appeared to significantly affect the uniform delay [6]. Akgüngör and Yıldız applied fractional factorial method to an accident prediction model. The evaluation of sensitivity analysis indicated that average daily traffic (ADT), lane width (W), width of paved shoulder (PA), median (H) and their interactions (i.e., ADT–W, ADT–PA and ADT–H) have significant effects on number of accidents. Based on the absolute value of parameter effects at the three- and two-standard deviation thresholds ADT was found to be of primary importance, while the remaining identified parameters Over the past 50 years, many cycle length models have been developed by researchers for optimal signal operations. One of the oldest models presented in the literature is optimal cycle length model derived by Webster in 1958 [13]. After Webster’s optimal cycle length expression, more representative models have been proposed by many researchers using various techniques. Miller [14] developed a formula for optimum cycle time to minimize the overall average delay. He used the lowest of the saturation flows for any of the representative movements. Lan [15] used cycle length and traffic flow parameters including duration of analysis period in his model and Lan and Gu [16] proposed a new formulation for optimal cycle length by employing non-linear regression analysis. Cheng et al. [17, 18] modified Webster’s optimal cycle length equation based on HCM 2000 and suggested a quick estimation method. Tehnički vjesnik 25, Suppl. 2(2018), 391-395 391 Ali Payıdar AKGÜNGÖR, Ersin KORKMAZ: Investigating Parameter Interactions with the Factorial Design Method: Webster’s Optimal Cycle Length Model prediction variable obtained from the ith experimental run and Nj shows the number of "+" signs in column j. The individual affects and interactions between the parameters in the experiment may be determined by the factorial design method. Thus, the experiment results are obtained in a shorter time with fewer experiments. However, this study is not an experimental design and is different from the studies in the literature. 1 INTRODUCTION Table 1 Design matrix for single parameters Design Matrix Prediction Variable Run ω ψ γ R 1 - - - R1 2 + - - R2 3 - + - R3 4 + + - R4 5 - - + R5 6 + - + R6 7 - + + R7 8 + + + R8 Table 2 Computation matrix for the three-parameter interactions Computation Matrix Run ω·ψ ω·γ ψ·γ ω·ψ·γ 1 + + + − 2 − − + + 3 − + − + 4 + − − − 5 + − − + 6 − + − − 7 − − + − 8 + + + + Table 1 Design matrix for single parameters Design Matrix Prediction Variable Run ω ψ γ R 1 - - - R1 2 + - - R2 3 - + - R3 4 + + - R4 5 - - + R5 6 + - + R6 7 - + + R7 8 + + + R8 Table 1 Design matrix for single parameters Design Matrix Prediction Variable Run ω ψ γ R 1 - - - R1 2 + - - R2 3 - + - R3 4 + + - R4 5 - - + R5 6 + - + R6 7 - + + R7 8 + + + R8 Table 2 Computation matrix for the three-parameter interactions Computation Matrix Run ω·ψ ω·γ ψ·γ ω·ψ·γ 1 + + + − 2 − − + + 3 − + − + 4 + − − − 5 + − − + 6 − + − − 7 − − + − 8 + + + + Table 2 Computation matrix for the three-parameter interactions Computation Matrix Run ω·ψ ω·γ ψ·γ ω·ψ·γ 1 + + + − 2 − − + + 3 − + − + 4 + − − − 5 + − − + 6 − + − − 7 − − + − 8 + + + + Table 2 Computation matrix for the three-parameter interactions Computation Matrix Run ω·ψ ω·γ ψ·γ ω·ψ·γ 1 + + + − 2 − − + + 3 − + − + 4 + − − − 5 + − − + 6 − + − − 7 − − + − 8 + + + + This paper is organized in the following way. 2 FACTORIAL DESIGN METHOD This method is used to examine the sensitivity of changes in singular parameters and interactions between groups of parameters as well. When the number of the parameters is equal to five or less a full factorial design can be carried out. In this method, a fixed number of possible values for each of the model parameters is tested considering specific perturbations of the magnitudes of the parameters (usually two levels: upper and lower) assigned before. Then, each parameter is identified and ranked according to some pre-established measure of model sensitivity by running the model through all possible combinations of the parameters. For instance, a factorial design at two levels for k parameters will result in 2k combinations of the model parameters. This is explained in the following 3-parameter (23 factorial) design with parameters ω, ψ and γ, and prediction variable R. Here, "+" and "−" signs represent the two possible values of each parameter (upper and lower levels, respectively). The effects of parameters and parameter interactions on the model results are predicted by utilizing Eq. (1) and the factorial design matrix given in Tab. 1. The signs of the parameter interactions are determined by the following rule: plus times minus produces a minus, and minus times minus or plus times plus produces a plus. According to this rule, the corresponding computation matrix for parameter interactions is given in Tab. 2. 1 INTRODUCTION In section 2, the factorial design method is explained. The application of Webster optimal cycle length model by the factorial design method is presented in Section 3. Section 4 is devoted to results and discussions. Finally, conclusions are given in the last section. After all, Ej values are computed by Eq. (1), the degree of importance of the parameters and their interactions can be evaluated. As recommended by Box et al. [1] the parameters with major effects are identified and ranked by plotting the effects on a normal probability scale. In this method, any outliers from the straight line on this probability paper could be regarded to influence the model results significantly. On the other hand, the other effects would result in variability in model results consistent with the result of random variation about a fixed mean. Here, it is assumed that higher order interactions are insignificant, in a manner similar to neglecting higher order terms in a Taylor series expansion. Henderson-Sellers [2-3] suggested another method to identify the parameters that have major effects on the model results. According to this method, thresholds are obtained two, three or four standard deviations from zero. Thus, any effects greater than the predicted thresholds are assumed to have significant effects on the model results. 1 INTRODUCTION In here, the factorial design method has been employed to determine the sensitivity of Webster optimal cycle length model widely used in traffic signalization. The reason for selecting this model is that it has still been used in signal timing practice and lead many studies of researchers over the past 50 years. In the light of this study, novel cycle length models are developed for future studies after determining effective parameters. Thus, this study shows that the factorial design method can be used for different purposes besides experiment design. Additionally, determining the effective parameters on cycle length models by the factorial design method is important for traffic researches and traffic signals practitioners. and Nj shows the number of "+" signs in column j. Technical Gazette 25, Suppl. 2(2018), 391-395 3 APPLICATION OF THE FACTORIAL DESIGN METHOD FOR WEBSTER’S OPTIMAL CYCLE LENGTH MODEL L C Y + = − (2) (2) 4 Saturation flow (veh/h) s 1500 1900 Table 4 The design matrix of the single parameters Model Parameters Run 1 2 3 4 1 − − − − 2 + − − − 3 − + − − 4 + + − − 5 − − + − 6 + − + − 7 − + + − 8 + + + − 9 − − − + 10 + − − + 11 − + − + 12 + + − + 13 − − + + 14 + − + + 15 − + + + 16 + + + + Four model parameters with parameter index numbers from 1 to 4 (1 : n, 2 : l, 3 : q, 4 : s) were selected for the sensitivity analysis with the two-level factorial design method. The upper and lower levels of model parameters given in Tab. 3 were chosen arbitrarily within their reasonable ranges. For the given number of parameters and perturbation levels, the design matrix for the main parameters is given in Tab. 4 and the computation matrix for the multiple parameter interactions is given in Tab. 5. Table 4 The design matrix of the single parameters Model Parameters Run 1 2 3 4 1 − − − − 2 + − − − 3 − + − − 4 + + − − 5 − − + − 6 + − + − 7 − + + − 8 + + + − 9 − − − + 10 + − − + 11 − + − + 12 + + − + 13 − − + + 14 + − + + 15 − + + + 16 + + + + Table 4 The design matrix of the single parameters where Co - optimal cycle length (s); L - total lost time (s); Y - sum of critical phase flow ratios The Webster cycle length model is used in the situations where the sum of flow ratios (the arrival flow to saturation flow ratio) is less than 1.0. However, this formulation has been proven to be ineffective over- saturated situations, and is inappropriate for other control type such as semi-actuated and full-actuated control. For the sensitivity analysis, the two-level factorial design method was applied to Webster’s optimal cycle length model given in Eq. (1). 3 APPLICATION OF THE FACTORIAL DESIGN METHOD FOR WEBSTER’S OPTIMAL CYCLE LENGTH MODEL Total lost time (L) and sum of critical phase flow ratios (Y) could not be selected as individual parameters for the sensitivity analysis because they are dependent parameters. Total lost time changes with the number of the phases. An increase in the number of phase in a cycle length results in an increase of total lost time in a given cycle length. Sum of critical phase flow ratios is expressed as a function of both traffic flow (q) and saturation flow (s). Hence, Webster’s optimal cycle length model defined in Eq. (2) can be rearranged as follows: Four model parameters with parameter index numbers from 1 to 4 (1 : n, 2 : l, 3 : q, 4 : s) were selected for the sensitivity analysis with the two-level factorial design method. The upper and lower levels of model parameters given in Tab. 3 were chosen arbitrarily within their reasonable ranges. For the given number of parameters and perturbation levels, the design matrix for the main parameters is given in Tab. 4 and the computation matrix for the multiple parameter interactions is given in Tab. 5. Four model parameters with parameter index numbers from 1 to 4 (1 : n, 2 : l, 3 : q, 4 : s) were selected for the sensitivity analysis with the two-level factorial design method. The upper and lower levels of model parameters given in Tab. 3 were chosen arbitrarily within their reasonable ranges. For the given number of parameters and perturbation levels, the design matrix for the main parameters is given in Tab. 4 and the computation matrix for the multiple parameter interactions is given in Tab. 5. Four model parameters with parameter index numbers from 1 to 4 (1 : n, 2 : l, 3 : q, 4 : s) were selected for the sensitivity analysis with the two-level factorial design method. The upper and lower levels of model parameters given in Tab. 3 were chosen arbitrarily within their reasonable ranges. For the given number of parameters and perturbation levels, the design matrix for the main parameters is given in Tab. 4 and the computation matrix for the multiple parameter interactions is given in Tab. 5. o 1 5 ( ) 5 1 . 3 APPLICATION OF THE FACTORIAL DESIGN METHOD FOR WEBSTER’S OPTIMAL CYCLE LENGTH MODEL A cycle length is a time required to complete the movements in the phases at a signalized intersection. The optimal cycle length provides sufficient capacity and minimum delay. Cycle lengths usually range between 40 and 150 seconds according to traffic flow and the number of the phase. On the other hand, short and long cycle lengths significantly cause the increase in delay. Traffic engineers and researchers have used Webster’s principles and expressions to determine the optimal cycle length and green split allocation at isolated intersections, regulated by traffic signals with a pre-timed control. Webster’s model has never lost its popularity for long decades since his principles are straightforward and easy. When developing his model, Webster accepted the following two principles for signal timing strategies. First, for a given cycle length, fixed time signals should have 1 n ij i j i j S R E N = = ∑ (1) (1) in which Ej indicates the effect of the jth factor, n represents the total number of experimental runs, Sij illustrates the sign in row i and column j, Ri describes the value of the Technical Gazette 25, Suppl. 2(2018), 391-395 392 Ali Payıdar AKGÜNGÖR, Ersin KORKMAZ: Investigating Parameter Interactions with the Factorial Design Method: Webster’s Optimal Cycle Length Model Table 3 Lower and upper levels of model parameters used in the sensitivity analysis Parameter Index No Parameter Name Symbol Lower Level Upper Level 1 Number of phases n 2 5 2 Lost time (s) l 2.5 4 3 Arrival flow (veh/h) q 700 1200 4 Saturation flow (veh/h) s 1500 1900 Table 3 Lower and upper levels of model parameters used in the sensitivity analysis their critical phase time for equal degrees of saturation. Second, effective green times of the phases are in the ratio of their respective flow ratio values. The optimal cycle length model proposed by Webster can be expressed as: o 1 5 5 1 . 3 APPLICATION OF THE FACTORIAL DESIGN METHOD FOR WEBSTER’S OPTIMAL CYCLE LENGTH MODEL n l C q s Σ ⋅ ⋅ + =   −     (3) (3) where Co - optimal cycle length (s); n - the number of phase; l - lost time in each phase (s); q - traffic flow (vph/h); s - saturation flow (vph/h). Table 5 The computation matrix of the multiple parameter interactions Multiple Parameter Interactions Run 12 13 14 23 24 34 123 124 134 234 1234 1 + + + + + + − − − − + 2 − − − + + + + + + − − 3 − + + − − + + + − + − 4 + − − − − + − − + + + 5 + − + − + − + − + + − 6 − + − − + − − + − + + 7 − − + + − − − + + − + 8 + + − + − − + − − − − 9 + + − + − − − + + + − 10 − − + + − − + − − + + 11 − + − − + − + − + − − 12 + − + − + − − + − − + 13 + − − − − + + + − − − 14 − + + − − + − − + − − 15 − − − + + + − − − + − 16 + + + + + + + + + + + 4 SENSIVITY RESULTS OF THE WEBSTER’S OPTIMAL CYCLE LENGTH MODEL AND DISCUSSION For the given factorial design, 16 runs in total were carried out and the results for the optimal cycle length model are listed in Tab. 6. Additionally, the parameter effects for single and multiple parameter interactions are presented in the same table. To identify parameters with major effects on the model results, the parameter effects were plotted on a standard normal probability scale as suggested by Box et al. [1] (see Fig. 1). As a single parameter effect, arrival flow on the model results has the highest effect when compared with other single parameters. As represented in the figure, the number of phase, saturation flow and the lost time in each phase were ranked as other single parameters, respectively. 3 APPLICATION OF THE FACTORIAL DESIGN METHOD FOR WEBSTER’S OPTIMAL CYCLE LENGTH MODEL Table 5 The computation matrix of the multiple parameter interactions Multiple Parameter Interactions Run 12 13 14 23 24 34 123 124 134 234 1234 1 + + + + + + − − − − + 2 − − − + + + + + + − − 3 − + + − − + + + − + − 4 + − − − − + − − + + + 5 + − + − + − + − + + − 6 − + − − + − − + − + + 7 − − + + − − − + + − + 8 + + − + − − + − − − − 9 + + − + − − − + + + − 10 − − + + − − + − − + + 11 − + − − + − + − + − − 12 + − + − + − − + − − + 13 + − − − − + + + − − − 14 − + + − − + − − + − − 15 − − − + + + − − − + − 16 + + + + + + + + + + + 4 SENSIVITY RESULTS OF THE WEBSTER’S OPTIMAL CYCLE LENGTH MODEL AND DISCUSSION 4 SENSIVITY RESULTS OF THE WEBSTER’S OPTIMAL CYCLE LENGTH MODEL AND DISCUSSION Table 6 Results of runs and parameter effects Run Optimal Cycle Length Parameter Index No Parameter Effects 1 23.438 1 40.850 2 44.531 2 21.996 3 31.875 3 46.948 4 65.625 4 −28.432 5 62.500 12 9.427 6 118.750 13 15.561 7 85.000 14 −9.424 8 175.000 23 8.379 9 19.792 24 −5.074 10 37.604 34 −21.997 11 26.917 123 3.591 12 55.417 124 −2.175 13 33.927 134 −7.291 14 64.464 234 −3.926 15 46.143 1234 −1.682 16 95.000 Table 6 Results of runs and parameter effects Table 7 Importance of identified parameters based on thresholds of \|4σ\|, \|3σ\| and \|2σ\| Outliers Primary Importance Secondary Importance \|4σ\| \|3σ\| \|2σ\| q  n  l  s  q-s  Table 7 Importance of identified parameters based on thresholds of \|4σ\|, \|3σ\| and \|2σ\| On the other hand, if arrival flow is less than the capacity some vehicles have to wait until the end of the cycle due to a fixed cycle length. Therefore, accurate estimation of arrival flow is important to determine the appropriate cycle length. The analysis results indicate that, as main parameters, the number of phase in a cycle length, saturation flow and lost time have relatively lower effects on optimal cycle length as compared to arrival flow. The number of phase comes after arrival flow as another effective parameter on cycle length. This result is also expected because increasing numbers of phase in a cycle length results in larger delay, lost time and longer cycle length. The Webster’s model proposes the use of a saturation flow which is the maximum rate of discharge and the lost time parameter. Sensitivity analysis results show that saturation flow is partly effective main parameter in the model. As known that saturation flow is defined as the number of vehicles per hour crossing signalized stop line if the signal remains green all the time. It is mainly depended on roadway and traffic conditions. While saturation flow is of primary importance for the capacity of intersection it is not so effective on cycle length, because it remains almost As explained before, using the iterative method suggested by Henderson-Sellers, the model parameters were classified into two categories as primary and secondary importance. More specifically, the importance of these parameters was ranked on the absolute value of their effects at the four-, three-, and two-standard thresholds deviation (i.e. 4 SENSIVITY RESULTS OF THE WEBSTER’S OPTIMAL CYCLE LENGTH MODEL AND DISCUSSION major effects on the model results, the parameter effects were plotted on a standard normal probability scale as suggested by Box et al. [1] (see Fig. 1). For the given factorial design, 16 runs in total were carried out and the results for the optimal cycle length model are listed in Tab. 6. Additionally, the parameter effects for single and multiple parameter interactions are presented in the same table. To identify parameters with As a single parameter effect, arrival flow on the model results has the highest effect when compared with other single parameters. As represented in the figure, the number of phase, saturation flow and the lost time in each phase were ranked as other single parameters, respectively. As a single parameter effect, arrival flow on the model results has the highest effect when compared with other single parameters. As represented in the figure, the number of phase, saturation flow and the lost time in each phase were ranked as other single parameters, respectively. Tehnički vjesnik 25, Suppl. 2(2018), 391-395 393 Ali Payıdar AKGÜNGÖR, Ersin KORKMAZ: Investigating Parameter Interactions with the Factorial Design Method: Webster’s Optimal Cycle Length Model Similarly as two parameter interactions arrival flow- saturation flow (q-s) were detected to be outliers which have major effects on the model results. For three parameter interactions, the effect of the number of the phase-arrival flow-saturation flow (n-q-s) is slightly higher than the other three parameter interactions as seen in Tab. 6. Figure 1 Parameter effects plotted on a normal probability scale Figure 1 Parameter effects plotted on a normal probability scale Table 6 Results of runs and parameter effects Run Optimal Cycle Length Parameter Index No Parameter Effects 1 23.438 1 40.850 2 44.531 2 21.996 3 31.875 3 46.948 4 65.625 4 −28.432 5 62.500 12 9.427 6 118.750 13 15.561 7 85.000 14 −9.424 8 175.000 23 8.379 9 19.792 24 −5.074 10 37.604 34 −21.997 11 26.917 123 3.591 12 55.417 124 −2.175 13 33.927 134 −7.291 14 64.464 234 −3.926 15 46.143 1234 −1.682 16 95.000 insufficient green time and cycle length. Thus, some vehicles have to wait the next cycle to be serviced. insufficient green time and cycle length. Thus, some vehicles have to wait the next cycle to be serviced. 4 SENSIVITY RESULTS OF THE WEBSTER’S OPTIMAL CYCLE LENGTH MODEL AND DISCUSSION 4σ, 3σ, and 2σ) as shown in Tab. 7. Referring to Tabs. 6 and 7, and Fig. 1, arrival flow is of primary importance, while the remaining parameters are of secondary importance. This is not surprising because arrival flow is known as an effective parameter in determining traffic signal control strategies. As known the arrival flow to a pre-timed signalized intersection may vary from cycle to cycle due to stochastic nature of traffic. In some cycles, if arrival flow is greater than the capacity of a signalized intersection, the cycle will fail because of Technical Gazette 25, Suppl. 2(2018), 391-395 394 Ali Payıdar AKGÜNGÖR, Ersin KORKMAZ: Investigating Parameter Interactions with the Factorial Design Method: Webster’s Optimal Cycle Length Model Factorial Design Method. Turkish Journal of Engineering and Environmental Sciences, 30(4), 259-267. constant after the initial lost time has been accommodated. Moreover, saturation flow depends on various factors such as lane with, number of lanes, grade, pedestrian activities etc. Lost time in the model appears to be the least effective main parameter in the examined cycle length model. The lost time is a portion of the cycle length which is not being fully employed. In other words, it is a time which is not effectively serving any traffic movements despite a green signal. Lost time almost remains fixed, regardless of cycle length, but shorter cycle lengths incorporates a large percentage of the lost time when compared to longer cycle lengths. According to sensitivity results, the interaction of arrival flow-saturation flow (q-s) seems to be relatively more effective on the selected cycle length model among the other two parameter interactions. [7] Akgüngör, A. P., & Yıldız, O. (2007). Sensitivity analysis of an accident prediction model by the fractional factorial method. Accident Analysis & Prevention, 39(1), 63-68. https://doi.org/10.1016/j.aap.2006.06.013 [8] Akgüngör, A. P. (2011). The investigation of the parameter effects in Webster delay model with factorial design method. 9th Transportation Congress / Istanbul, Turkey, May 16-18, 211-218 (in Turkish). [9] Hatami, H., & Aghayan, I. (2017). Traffic Efficiency Evaluation of Elliptical Roundabout Compared with Modern and Turbo Roundabouts Considering Traffic Signal Control. PROMET – Traffic & Transportation, 29(1), 1-11. https://doi.org/10.7307/ptt.v29i1.2053 [10] Hao, Y., Teng, J., Wang, Y., & Yang, X. (2016). Increasing Capacity of Intersections with Transit Priority. PROMET – Traffic & Transportation, 28(6), 627-637. https://doi.org/10.7307/ptt.v28i6.1999 5 CONCLUSION [11] Nassiri, H., Tabatabaie, S., & Sahebi, S. (2017). Delay-based Passenger Car Equivalent at Signalized Intersections in Iran. PROMET – Traffic & Transportation, 29(2), 135-142. https://doi.org/10.7307/ptt.v29i2.2040 In this study, the sensitivity analysis of optimal cycle length model proposed by Webster and its parameters was investigated by utilizing the factorial design method. The results of the sensitivity analysis indicated that arrival flow among the main parameters has a significant effect on the optimal cycle length model. The study results also revealed that the remaining main parameters in the model (number of phase, saturation flow and lost time) have relatively lower effects on optimal cycle length. Additionally, the analysis results illustrated that the two-parameter interaction of arrival flow-saturation flow has a major effect on the model results. For the three-parameter interaction, the effect of the number of the phase-arrival flow-saturation flow has a slightly larger effect than the other three-parameter interactions but it was not a significant level. In summary, this study shows that the factorial design method is an effective way of examining the relative importance of the selected optimal cycle length model, and it can be employed for similar application in traffic engineering. [12] Çalışkanelli, S. P., Coşkun Atasever, F., & Tanyel, S. (2017). Start-up Lost Time and its Effect on Signalized Intersections in Turkey. PROMET – Traffic & Transportation, 29(3), 321- 329. https://doi.org/10.7307/ptt.v29i3.2214 [13] Webster, F. V. (1958). Traffic signal settings, road research technical paper no. 39. Road Research Laboratory, London. [14] Miller, A. J. (1968). Australian Road Capacity Guide: Provisional Introduction and Signalized Intersections. Australian Road Research Board, 4. [15] Lan, C. (2004). New optimal cycle length formulation for pre-timed signals at isolated intersections. Journal of Transportation Engineering, 130(5), 637-647. https://doi.org/10.1061/(ASCE)0733-947X (2004) 130:5(637) [16] Lan, C. J., & Gu, X. (2005, September). Optimal signal controls and effects of flow uncertainty. In Intelligent Transportation Systems, 2005. Proceedings IEEE, 549-554. [17] Cheng, D., Messer, C. J., Tian, Z. Z., & Liu, J. (2003, January). Modification of Webster’s minimum delay cycle length equation based on HCM 2000. In the 81st Annual Meeting of the Transportation Research Board in Washington, DC. 6 REFERENCES [18] Cheng, D., Tian, Z. Z., & Messer, C. J. (2005). Development of an improved cycle length model over the highway capacity manual 2000 quick estimation method. Journal of transportation engineering, 131(12), 890-897. https://doi.org/10.1061/(ASCE)0733-947X(2005)131:12(890) [1] Box, G. E., Hunter, W. G., & Hunter, J. S. (1978). Statistics for experimenters: an introduction to design, data analysis, and model building (Vol. 1). New York: Wiley. [2] Henderson‐Sellers, A. (1992). Assessing the Sensitivity of A Land‐Surface Scheme to Parameters Used In Tropical‐ Deforestation Experiments. Quarterly Journal of the Royal Meteorological Society, 118(508), 1101-1116. https://doi.org/10.1002/qj.49711850805 Contact information: [3] Henderson-Sellers, A. (1993). A factorial assessment of the sensitivity of the BATS land-surface parameterization scheme. Journal of Climate, 6(2), 227-247. https://doi.org/10.1175/1520- 0442(1993)006<0227:AFAOTS>2.0.CO;2 Ali Payıdar AKGÜNGÖR Department of Civil Engineering Engineering Faculty, Kırıkkale University Kırıkkale/Turkey aakgungor@gmail.com [4] Henderson-Sellers, B., & Henderson-Sellers, A. (1996). Sensitivity evaluation of environmental models using fractional factorial experimentation. Ecological Modelling, 86(2), 291-295. https://doi.org/10.1016/0304-3800(95)00066-6 Ersin KORKMAZ (corresponding author) Department of Civil Engineering Engineering Faculty, Kırıkkale University Kırıkkale/Turkey ersin_korkmaz1@hotmail.com [5] Barros, A. P. (1996). An evaluation of model parameterizations of sediment pathways: a case study for the Tejo estuary. Continental Shelf Research, 16(13), 1725- 1749. https://doi.org/10.1016/0278-4343(96)00009-X [6] Akgüngör, A. P., Yildiz, O., & Demirel, A. (2006). A Sensitivity Analysis of the HCM 2000 Delay Model with the [6] Akgüngör, A. P., Yildiz, O., & Demirel, A. (2006). A Sensitivity Analysis of the HCM 2000 Delay Model with the Tehnički vjesnik 25, Suppl. 2(2018), 391-395 395
https://openalex.org/W2796138856	https://scindeks-clanci.ceon.rs/data/pdf/0352-5252/2017/0352-52521776741I.pdf	English	null	Trend of the incidence and mortality rate of lip cancer in Central Serbia from 1999-2014	Acta stomatologica Naissi/Acta Stomatologica Naissi	2,017	cc-by-sa	7,210	Abstract Uvod: Karcinom usne je najčešći oralni karcinom. Uvod: Karcinom usne je najčešći oralni karcinom. Background: Lip cancer is the most frequent oral cancer. Background: Lip cancer is the most frequent oral cancer. The aim of this study was to present and estimate trends in the incidence and mortality of lip cancer in Central Serbia during a 16-year period, from 1999 to 2014. j j Cilj ovog rada bio je da se prikaže i proceni trend incidencije i mortaliteta od karcinoma usne u centralnoj Srbiju u periodu od 16 godina, od 1999. do 2014. Cilj ovog rada bio je da se prikaže i proceni trend incidencije i mortaliteta od karcinoma usne u centralnoj Srbiju u periodu od 16 godina, od 1999. do 2014. g Materijal i metode: Epidemiološka studija zasnovana je na podacima koji su javno dostupni iz registra karcinoma “Incidencija i mortalitet od raka u centralnoj Srbiji” koji objavljuje Institut za javno zdravlje Srbije za period od 1999. do 2014. godine. Trend i godišnja procentualna promena (APC), sa odgovarajućim 95% intervalom poverenja, izračunati su pomoću joinpoint regresione analize. Materijal i metode: Epidemiološka studija zasnovana je na podacima koji su javno dostupni iz registra karcinoma “Incidencija i mortalitet od raka u centralnoj Srbiji” koji objavljuje Institut za javno zdravlje Srbije za period od 1999. do 2014. godine. Trend i godišnja procentualna promena (APC), sa odgovarajućim 95% intervalom poverenja, izračunati su pomoću joinpoint regresione analize. y p f Material and methods: This registry-based study was carried out based on the data extracted from publicly available Yearbooks of the Institute of Public Health of Serbia – Incidence and Mortality in Central Serbia, from 1999 to 2014. The trend and the annual percentage change (APC) of the incidence and mortality rate with the corresponding 95% confidence intervals were calculated by performing joinpoint regression analyses y p f Material and methods: This registry-based study was carried out based on the data extracted from publicly available Yearbooks of the Institute of Public Health of Serbia – Incidence and Mortality in Central Serbia, from 1999 to 2014. The trend and the annual percentage change (APC) of the incidence and mortality rate with the corresponding 95% confidence intervals were calculated by performing joinpoint regression analyses. Results: A total number of 1,922 cases (1,402 in men and 520 in women) of lip cancer were registered in Central Serbia from 1999 to 2014. Aleksandra M. Ignjatović1,2, Miodrag M. Stojanović1,2, Zoran G. Milošević1,2, Marija R. Anđelković-Apostolović1,2, Branislava B. Stojković3, Marija M. Topalović4, Suzana A. Otašević2,5 1 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET, PREDMET MEDICINSKA STATISTIKA I INFORMATIKA, NIŠ,SRBIJA 2 INSTITUT ZA JAVNO ZDRAVLJE NIŠ, NIŠ,SRBIJA 3 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET, KATEDRA ZA DEČIJU I PREVENTIVNU STOMATOLOGIJU, NIŠ,SRBIJA 4 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET NIŠ,SRBIJA 5 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET, KATEDRA ZA MIKROBIOLOGIJU I IMUNOLOGIJU, NIŠ,SRBIJA 1UNIVERSITY OF NIŠ, FACULTY OF MEDICINE, DEPARTMENT OF MEDICAL STATISTICS AND INFORMATICS NIŠ,SERBIJA 2 PUBLIC HEALTH INSTITUTE NIŠ NIŠ SERBIJA 4 UNIVERSITY OF NIŠ, FACULTY OF MEDICINE NIŠ,SERBIJA 5 UNIVERSITY OF NIŠ, FACULTY OF MEDICINE, DEPARTMENT OF MICROBIOLOGY AND IMMUNOLOGY NIŠ,SERBIJA 2017 Faculty of Medicine in Niš. Clinic of Dentistry in Niš. All rights reserved / © 2017. Medicinski fakultet Niš. Klinika za stomatologiju Niš. Sva prava zadržana. Acta Stomatologica Naissi Dec / Dec 2017, Vol. 33, br./num. 76 str./p. 1741-1753 Primljen/ Received on: 30.08.2017. Revidiran/ Revised on : 15.09.2017. Prihvaćen/ Accepted on :21.10.2017. TREND INCIDENCIJE I MORTALITETA OD KARCINOMA USNE U CENTRALNOJ SRBIJI U PERIODU OD 1999. DO 2014. TREND OF THE INCIDENCE AND MORTALITY RATE OF LIP CANCER IN CENTRAL SERBIA FROM 1999-2014 Aleksandra M. Ignjatović1,2, Miodrag M. Stojanović1,2, Zoran G. Milošević1,2, Marija R. Anđelković-Apostolović1,2, Branislava B. Stojković3, Marija M. Topalović4, Suzana A. Otašević2,5 1 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET, PREDMET MEDICINSKA STATISTIKA I INFORMATIKA, NIŠ,SRBIJA 2 INSTITUT ZA JAVNO ZDRAVLJE NIŠ, NIŠ,SRBIJA 3 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET, KATEDRA ZA DEČIJU I PREVENTIVNU STOMATOLOGIJU, NIŠ,SRBIJA 4 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET NIŠ,SRBIJA 5 UNIVERZITET U NIŠU, MEDICINSKI FAKULTET, KATEDRA ZA MIKROBIOLOGIJU I IMUNOLOGIJU, NIŠ,SRBIJA 1UNIVERSITY OF NIŠ, FACULTY OF MEDICINE, DEPARTMENT OF MEDICAL STATISTICS AND INFORMATICS NIŠ,SERBIJA 2 PUBLIC HEALTH INSTITUTE NIŠ, NIŠ,SERBIJA 3 UNIVERSITY OF NIŠ, FACULTY OF MEDICINE, DEPARTMENT OF PEDIACTRIC AND PREVENTIVE DENTISTRY NIŠ,SERBIJA 4 UNIVERSITY OF NIŠ, FACULTY OF MEDICINE NIŠ,SERBIJA 5 UNIVERSITY OF NIŠ FACULTY OF MEDICINE DEPARTMENT OF MICROBIOLOGY AND IMMUNOLOGY NIŠ SERBIJA ORIGINAL ARTICLE CLINICAL STUDY ORIGINALNI RAD KLINIČKA STUDIJA doi:10.5937/asn1776741I Dec / Dec 2017, Vol. 33, br./num. 76 str./p. 1741-1753 Acta Stomatologica Naissi Dec / Dec 2017, Vol. 33, br./num. 76 str ORIGINAL ARTICLE CLINICAL STUDY ORIGINALNI RAD KLINIČKA STUDIJA doi:10.5937/asn1776741I Primljen/ Received on: 30.08.2017. Revidiran/ Revised on : 15.09.2017. Prihvaćen/ Accepted on :21.10.2017. Abstract The crude rate (CR) and age-standardized rate (ASR-W) of the incidence in males decreased (p=0.001, p<0.001, respectively) during the study period with APC of 4.3%, 5.2%, respectively. In females, CR and ASR-W of incidence showed the only significant trend between 2003-2014, 2006-2014, respectively with APC 12.3, 24.3%, respectively. ASR-W of mortality rate in males showed the decreasing trend with APC 4.5%. Rezultati: U centralnoj Srbiji, u periodu 1999. - 2014. godine, registrovano je ukupno 1922 novoobolelih od karcinoma usne (1402 muškarca i 520 žena). Stopa incidencije i standardizovana stopa incidencije imaju statistički značajan trend pada u muškoj populaciji (p = 0,001, odnosno p < 0,001) tokom ispitivanog perioda, sa APC 4,3%, odnosno 5,2%. Kod žena, incidencija i standardizivana stopa incidencije pokazuju statistički značajan trend pada u periodu 2003.- 2011. godine, odnosno 2006. - 2011. godine, sa APC 12,3, odnosno 24,3%. Standardizovana stopa mortaliteta kod muškaraca pokazuje statistički značajan trend pada kod muškaraca sa APC, 4,5%. were calculated by performing joinpoint regression analyses. Results: A total number of 1,922 cases (1,402 in men and 520 in women) of lip cancer were registered in Central Serbia from 1999 to 2014. The crude rate (CR) and age-standardized rate (ASR-W) of the incidence in males decreased (p=0.001, p<0.001, respectively) during the study period with APC of 4.3%, 5.2%, respectively. In females, CR and ASR-W of incidence showed the only significant trend between 2003-2014, 2006-2014, respectively with APC 12.3, 24.3%, respectively. ASR-W of mortality rate in males showed the decreasing trend with APC 4.5%. Zaključak: U ispitivanom periodu trend incidencije kod muškaraca pada. Kod žena trend je divergentnog karaktera. Stopa mortaliteta karcinoma usne kod oba pola je niska i stabilna. Conclusions: Our results reveal that the incidence of lip cancer declines trough the study period. The mortality rate is low and stable. Klјučne reči: karcinom usne, incidencija, mortalitet, registar karcinoma, trend Key words: lip cancer, incidence rate, mortality rate, cancer registry, trends Corresponding author: Lecturer Aleksandra Ignjatović, MD,PhD, Faculty of Medicine, Department of Medical Statistics and Informatics, 81 Dr Zoran Đinđić Blvd, 18000 Niš, Serbia E-mail:drsalea@yahoo.com Corresponding author: Lecturer Aleksandra Ignjatović, MD,PhD, Faculty of Medicine, Department of Medical Statistics and Informatics, 81 Dr Zoran Đinđić Blvd, 18000 Niš, Serbia E-mail:drsalea@yahoo.com Uvod Razvoj kancera nastaje usled interakcije različitih faktora rizika, kako unutrašnjih, kao što je genetska predispozicija, tako i spoljašnjih, kao što su životne navike i faktori sredine. Karcinom usta, zajedno sa drugim oralnim i faringealnim karcinomima, šesti je najčešći karcinom 1,2. Karcinom usta se predominantno javlja kod muškaraca 3. Muška predominantnost je rezultat profesionalne i razlike u ponašanju između polova. Rad i aktivnosti na otvorenom su daleko češći kod muškaraca. Pored toga, najčešće je ovim karcinomom zahvaćana populacija starija od 60 godina. Preko 60% novoobolelih od karcinoma usta su stariji pacijenti 4. Cancer develops in the interaction of risk factors which include internal factors such as genetic susceptibility and hormonal factors and external factors such as life style and environmental factors. Lip cancer, along with other oral and pharyngeal cancer is the sixth most common cancer type worldwide 1, 2. Lip cancer is far more common in men 3. Male predominance is a result of occupational and behavioral differences among sexes. Outdoor working and outdoor activities are far more frequent in males. The population over 60 is the most affected with lip cancer. Over 60% of new cases of lip cancer occurred in older patients 4. Squamous cell carcinomas (SCCs) is the dominant histological type of lip cancer 5. SCCs are typical of the lower lip, while basal cell carcinoma (BCC) is more frequent in the upper lip 6, 7. Because of its location, lip cancer shares risk factors with skin cancers and cancers of the oral cavity and nasopharynx. Therefore, the main risk factors for lip cancer are: smoking, alcohol consumption and sun exposure, especially for external lip cancer 8. Squamous cell carcinomas (SCCs) is the dominant histological type of lip cancer 5. SCCs are typical of the lower lip, while basal cell carcinoma (BCC) is more frequent in the upper lip 6, 7. Because of its location, lip cancer shares risk factors with skin cancers and cancers of the oral cavity and nasopharynx. Therefore, the main risk factors for lip cancer are: smoking, alcohol consumption and sun exposure, especially for external lip cancer 8. Serbia was recognized as a country with a large proportion of potentially avoidable oral cancer9.The aim of this study was to present and estimate trends in the incidence and mortality of lip cancer in Central Serbia from 1999 to 2014. p j Skvamocelularni karcinom (SCCs) je dominantni histopatološki tip karcinoma usta 5. Corresponding author: Lecturer Aleksandra Ignjatović, MD,PhD, Faculty of Medicine, Department of Medical Statistics and Informatics, 81 Dr Zoran Đinđić Blvd, 18000 Niš, Serbia E-mail:drsalea@yahoo.com 1741 Acta Stomatologica Naissi Dec/Dec. 2017, Vol. 33, br./num. 76 Materials and Methods This registry-based study was carried out based on the data extracted from publicly available Yearbooks of the Institute of Public Health of Serbia – Incidence and Mortality in Central Serbia, from 1999 to 2014, in which the incidence and mortality rates were aggregated and stratified by sex, 5-year age groups, and diagnosis. The age-standardized incidence and mortality rate in both sexes were calculated based on those data. Lip cancer was coded according to the tenth Revision of International Classification of Disease (codes C00-C96) 10. Mortality to incidence ratio (MIR) was calculated by dividing the crude rate of mortality by crude rate of incidence for each year. The trend and the annual percentage change (APC) of the incidence and mortality rate with cor- responding 95% confidence intervals (95% CI) were calculated by performing join point Uvod SCCs je karakterističan za karcinom donje usne, dok se bazocelularni karcinomi (BCC) mnogo češće javljaju na gornjoj usni 6, 7. Zbog svoje lokalizacije, karcinom usta deli faktore rizika sa karcinomima kože, oralnim karcinomima i karcinomima nasofarinksa. Zbog toga su glavni faktori rizika za nastanak karcinoma usne: pušenje, konzumacija alkohola, izlaganje sunčevom zračenju, posebno za spoljašnji karcinom usne 8. Srbija je prepoznata kao zemlja sa velikim udelom oralnih karcinoma koji se potencijalno mogu izbegnuti 9. Zato je cilj ovog rada bio da prikaže i proceni trend incidencije i mortaliteta od karcinoma usne u centralnoj Srbiji u periodu od 1999. do 2014.godina. Rezultati A total number of 1,922 cases (1,402 in men and 520 in women) of lip cancer were registered in Central Serbia from 1999 to 2014. The men to women ratio was 2.8:1. Table 1 presents the distribution of new cases of lip cancer, the crude rate of incidence and mortality, the age-standardized incidence and mortality rate (ASR-W; to the world population) and mortality to incidence rate (MIR) to the specified rates according to the years of observation in males. The highest CR and ASR-W of incidence in males were in 2001 (5.0, 2.9 respectively), and the lowest values of those rates were in 2011 (1.5, 0.8 respectively). In males, the CR and ASR-W of mortality rate were highest in 2006 (0.9, 0.4, respectively), and lowest in 2012 (0.3, 0.2, respectively). U centralnoj Srbiji su u periodu 1999.- 2014. registrovana ukupno 1922 novoobolela od karcinoma usne (1402 muškarca i 520 žena). Odnos muškaraca prema ženama je 2,8:1. Tabela 1 prikazuje distribuciju novo-obolelih od ovog karcinoma, stopu incidencije i mortaliteta, standardizivane stope incidencije i mortaliteta (ASR-W, standardizovane na svetsku populaciju) i MIR za prikazane stope u ispitivanom periodu kod muškaraca u centralnoj Srbiji. Najveća stopa incidencije i standardizovana stopa incidencije kod muškaraca bila je 2001. godine (5,0, odnosno 2,9), a najniža stopa incidencije i standar- dizivana stopa incidenicije bila je 2011. godine (1,5, odnosno 0,8). Kod muškaraca, najviši mortalitet i standardizovana stopa mortaliteta bila je 2006. (0,9, odnosno 0,4), a najniža u 2012. godini (0,3, odnosno 0,2). Kod žena, najviša stopa incidencije i standardizovana stopa incidencije bila je u 2009. godini, a najniža vrednost ovih stopa bila je u 2010. i u 2011. godini (1,9, odnosno 1,0). Najviša vrednost mortaliteta i standardizovana stopa mortaliteta kod žena bile su u 1999. i 2012. godini (0,4, odnosno 0,2). Zajednička tačka regresione analize incidencije kod muškaraca pokazuje statistički značajan trend pada u periodu 1999.- 2014. sa APC 4,3% (95%CI - 6,5 - -2,0%, p=0,001). Kod žena, stopa incidencije pokazuje trend blagog porasta, koji nije statistički značajan u periodu 1999- 2003. i 2011-2014. godine, i trend statistički značajnog pada u periodu 2003.- 2011. sa APC 12,3% (95%CI -19.1 - -5,0%, p = 0,008) (Tabela 3, Slika 1). Stopa mortaliteta kod oba pola je pokazala trend pada vrednosti koji nije statistički značajan u ispitivanom periodu (Slika 2). Materijal i metode Ova epidemiološka studija je sprovedena na osnovu podataka dobijenih iz javno dostupnog registra kancera – “Incidencija i mortalitet od raka u centralnoj Srbiji” koji objavljuje Institut za javno zdravlje Srbije za period od 1999. do 2014. godine, u kojima su incidencija i mortalitet prikazani prema polu i starosnoj strukturi i prema dijagnozi. Standar- dizovane stope incidencije i mortaliteta su izračunate na osnovu ovih podataka za oba pola. Karcinom usne je kodiran prema Međunarodnoj klasifikaciji bolesti – deseta revizija 10. Odnos mortaliteta i incidencije (MIR) je dobijen deljenjem stope mortaliteta i incidencije i deljenjem standardizovane stope mortaliteta i standardizovane stope incidencije za svaku godinu. 1742 1742 Ignjatović et.al. RATE OF LIP CANCER Trend i godišnja procentualna promena incidencije i mortaliteta sa 95% intervalom poverenja (95%CI) izračunati su joinpoint regresionom analizom. Optimalan broj prelomnih tačaka dobijen je metodom Monte Karlo permutacije. Za analizu trenda korišćen je programski paket Joinpoint Regression Program verzija 4.1.0 (dostupan na http: //surveil- lance.cancer.gov/joinpoint). Promena trenda je smatrana statistički značajnom ukoliko je p- vrednost manja od 0,05. regression analyses. The optimal number of joinpoints was identified using the Monte Carlo permutation method. For trend analyses, the Joinpoint Regression Program version 4.1.0 was used (available at http: // surveillance.cancer.gov/joinpoint). The trend was considered to be significantly changing when the p-value was below 0.05 (p < 0.05). Rezultati In females, the highest CR and ASR- W of incidence in 2009 (1.9, 1.0, respectively), and the lowest values of those rates were in 2010 and 2011 (0.6, 0.2, respectively). In females, the CR and ASR- W of mortality rate were highest in 1999 and 2012 (0.4, 0.2, respectively). Joinpoint analysis of the crude rate of incidence of lip cancer in males showed a significantly decreasing trend between 1999 and 2014 with APC 4.3% (95%CI -6.5 - - 2.0%, p = 0.001). In females, the crude rate of the incidence of lip cancer showed a non- significantly increasing trend between 1999 and 2003 and between 2011 and 2014, and a significantly decreasing trend between 2003- 2011 with APC 12.3% (95%CI -19.1- -5.0%, p=0.008) (Table 3, Figure 1). Crude rates of mortality of lip cancer in both sexes showed a non-significant decreasing trend between 1999 and 2014 (Figure 2). 1743 Acta Stomatologica Naissi Dec/Dec. 2017, Vol. 33, br./num. 76 Tabela 1. Broj novoregistovanih pacijenata, stopa incidencije i mortaliteta i standardizovana stopa incidecnije i mortaliteta karcinoma usta kod muškaraca u centralnoj Srbiji od 1999. do 2014. Tabela 1. Broj novoregistovanih pacijenata, stopa incidencije i mortaliteta i standardizovana stopa incidecnije i mortaliteta karcinoma usta kod muškaraca u centralnoj Srbiji od 1999. do 2014. j g p j p j a incidecnije i mortaliteta karcinoma usta kod muškaraca u centralnoj Srbiji od 1999. 2014. able 1. CR – crude rate, ASR – W – the age - standardized rate Rezultati The number of newly registered cases, the crude rate and the age-standardized cidence and mortality rate of lip cancer in males from Central Serbia from 1999-2014 Year Incidence Mortality MIR of CR MIR of ASR- W Count CR ASR- W Count CR ASR- W 1999 105 3.9 2.1 18 0.7 0.4 0.18 0.19 2000 107 4.0 2.3 16 0.6 0.3 0.15 0.13 2001 132 5.0 2.9 13 0.5 0.3 0.10 0.10 2002 108 4.1 2.3 17 0.6 0.3 0.15 0.13 2003 93 3.5 2.0 10 0.4 0.2 0.11 0.10 2004 115 4.3 2.5 12 0.5 0.2 0.12 0.08 2005 92 3.5 1.8 17 0.6 0.3 0.17 0.17 2006 91 3.5 1.8 23 0.9 0.4 0.26 0.22 2007 74 2.8 1.6 13 0.5 0.2 0.18 0.13 2008 75 2.9 1.5 22 0.8 0.4 0.28 0.27 2009 90 3.5 1.8 12 0.5 0.2 0.14 0.11 2010 72 2.8 1.5 10 0.4 0.2 0.14 0.13 2011 40 1.5 0.8 14 0.5 0.2 0.33 0.25 2012 77 3.0 1.6 8 0.3 0.1 0.10 0.06 2013 58 2.3 1.1 10 0.4 0.2 0.17 0.18 2014 73 2.9 1.4 15 0.6 0.2 0.21 0.14 Table 1. The number of newly registered cases, the crude rate and the age-standardized incidence and mortality rate of lip cancer in males from Central Serbia from 1999-2014 Table 1. The number of newly registered cases, the crude rate and the age-standardized ncidence and mortality rate of lip cancer in males from Central Serbia from 1999-2014 1744 Ignjatović i sar. STOPA KARCINOMA USNE Tabela 2. Broj novoregistovanih pacijenata, stopa incidencije i mortaliteta i standardizovana stopa incidecnije i mortaliteta karcinoma usta kod žena u centralnoj Srbiji od 1999. do 2014. Table 2. CR – crude rate, ASR-W – the age-standardized rate Rezultati Joinpoint analysis of the trend in the crude rate of incidence and mortality rate of cancer in Central Serbia from 1999-2014 Crude rate - Incidence Segment Period APC 95% p-value Male - 0 Joinpoints 1 1999-2014 -4.3^ -6.5 - -2.0 0.001 Female - 2 Joinpoints 1 1999-2003 20.3 -0.5 -45.6 0.055 Female - 2 Joinpoints 2 2003-2011 -12.3^ -19.1 - -5.0 0.008 Female - 2 Joinpoints 3 2011-2014 19.4 -11.7 -61.4 0.052 Crude rate -mortality Male - 0 Joinpoints 1 1999-2014 -2.1 -5.1-1.0 0.163 Female - 0 Joinpoints 1 1999-2014 -2.0 -8.1-4.5 0.511 ASR-W Incidence rate Male - 0 Joinpoints 1 1999-2014 -5.2^ -7.4 - -2.9 <0.001 Female - 2 Joinpoints 1 1999-2006 5.2 -5.2 – 16.7 0.296 Female - 2 Joinpoints 2 2006-2011 -24.3^ -40.9 - -3.1 0.031 Female - 2 Joinpoints 3 2011-2014 26.9 -14.1 – 87.4 0.197 ASR-W Mortality rate Male – 0 Joinpoints 1 1999-2014 -4.5^ -7.8 – -2.8 0.015 Female - Not calculated ^ indicates that the Annual Percent Change (APC) is significantly different from zero at the alp 0.05 level, 95%CI – 95% confidence interval, trend for ASR-W mortality rate for females wa calculated 3. Analiza trenda incidencije i mortaliteta karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Tabela 3. Analiza trenda incidencije i mortaliteta karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. 3. Joinpoint analysis of the trend in the crude rate of incidence and mortality rate of lip cancer in Central Serbia from 1999-2014 ^ indicates that the Annual Percent Change (APC) is significantly different from zero at the alpha = 0.05 level, 95%CI – 95% confidence interval, trend for ASR-W mortality rate for females was not calculated The age-standardized incidence rate in males showed a significantly decreasing trend between 1999 and 2014 with APC 5.2% (95%CI -7.4 - -2.9%, p < 0.001). In females, the age-standardized incidence rate of lip cancer showed the only significant trend between 2006 and 2011 with APC - 24.3% (-95%CI -40.9 - -3.1, p=0.031)(Figure 3). The age-standardized mortality rate of lip cancer in males showed a significantly decreasing trend during the period from 1999-2014 with APC -4.5% (95%CI -7.8 - - 2.8, p = 0.015)( Figure 4). Standardizovana stopa incidencije kod muškaraca je pokazala statistički značajan trend pada vrednosti u periodu 1999.-2014. sa APC 5,2% (95%CI -7,4 - - 2,9%, p < 0,001). Rezultati The number of new cases, the crude rate and the age-standardized incidence and mortality rate of lip cancer in females from Central Serbia from 1999-2014 Year Incidence Mortality MIR of CR MIR of ASR-W Count CR ASR-W Count CR ASR-W 1999 25 0.9 0.5 10 0.4 0.2 0.44 0.40 2000 27 1.0 0.5 4 0.1 0.1 0.10 0.20 2001 43 1.5 0.7 9 0.3 0.1 0.20 0.14 2002 48 1.7 0.8 7 0.2 0.1 0.12 0.13 2003 53 1.9 1.0 3 0.1 0.0 0.05 0.00 2004 33 1.2 0.5 9 0.3 0.1 0.25 0.20 2005 43 1.5 0.8 8 0.3 0.1 0.20 0.13 2006 36 1.3 0.7 4 0.1 0.0 0.08 0.00 2007 39 1.4 0.6 5 0.2 0.0 0.14 0.00 2008 33 1.2 0.5 3 0.1 0.0 0.08 0.00 2009 28 1.0 0.4 5 0.2 0.0 0.20 0.00 2010 17 0.6 0.2 6 0.2 0.1 0.33 0.50 2011 17 0.6 0.2 7 0.3 0.1 0.50 0.50 2012 21 0.8 0.3 10 0.4 0.1 0.50 0.33 2013 28 1.0 0.3 3 0.1 0.0 0.10 0.00 2014 29 1.1 0.4 3 0.1 0.0 0.09 0.00 1745 Acta Stomatologica Naissi Dec/Dec. 2017, Vol. 33, br./num. 76 Tabela 3. Analiza trenda incidencije i mortaliteta karcinoma usta u centralnoj Srbiji u peri od 1999. do 2014. Table 3. Discussion The main finding of the study is a decreasing trend of the incidence of lip cancer in men and low and stable mortality in both sexes. In the study period, the incidence rate in females showed divergent trends, and after a significant decrease in the incidence between 2006 and 2011, there was the trend of a non-significant increase until the end of the study period. The results of significantly decreased rates are also coherent with the data published for lip cancer in Belgrade from 1999-2010 11. Lip cancer profile in Belgrade almost matches Central Serbia profile. In global terms, a declining incidence trend of lip cancer in Central Serbia parallels observations from Europe (3) and the United States 12. The male:female ratio (2.8) is similar to the global M/F ratio (2.5)4. Generally, a declining trend of M/F ratio was noticed partly due to decreasing of incidence in males, and partly due to rising of incidence in females, especially the incidence of the upper lip cancer 3, 13. 9 Glavni nalaz ove studije je trend pada incidencije karcinoma usne kod muškaraca i niska, bez promena, vrednost mortaliteta kod oba pola. U ispitivanom periodu kod žena postoji divergenti trend, nakon statistički značajnog pada incidencije u periodu 2006.- 2011, postoji trend blagog porasta koji nije statistički značajan sve do kraja ispitivanog perioda. Rezultati pada trenda incidencije karcinoma usne su u skladu sa podacima koji su objavljeni za grad Beograd za period 1999.-2011 11. Profil karcinoma usne u Beogradu se gotovo kompletno poklapa sa profilom za centralnu Srbiju koji je dobijen u ovom istraživanju. Globalno gledano, smanjenje incidenicije u centralnoj Srbiji se poklapa sa podacima publikovanim za Evropu (3) i Sjedinjenje Američke Države 12. Odnos muškarac/žena (2,8) približan je vrednostima koje su dobijene na globalnom nivou (2,5) 4. Generalno, smanjenje ovog odnosa je registrovano delom zbog smanjenja incidencije kod muškaraca, a delom i zbog povećanja incidencije kod žena, posebno zbog porasta incidencije karcinoma gornje usne 3, 13. 3 A recently published study 9 established that 60% of cancers of the oral cavity, the oesophagus, the larynx, the corpus uteri and the bladder are considered to be avoidable which indicates great opportunities for prevention. Therefore, the focus of public health activities should be on measures of prevention and reduction of lip cancer risk factors. The main modifiable risk factors for lip cancers are smoking 3, 14 and alcohol consumption 15, 16 and sun exposure. Discussion The first step in risk factors reduction at the national level was done with the creation of a legislation framework for tobacco control 17. j g j Nedavno publikovana studija 3 je utvrdila da je 60% oralnih karcinoma, karcinoma jednjaka, larinksa, grlića materice i mokraćne bešike moguće izbeći, odnosno prevenirati. Stoga, fokus javnog zdravlja treba da bude na merama prevencije i smanjenja učestalosti faktora rizika. Glavni faktori rizika za karcinom usne su: pušenje 3,14, konzumiranje alkohola 15, 16 i izlaganje sunčevom zračenju. Prvi korak u smanjenju faktora rizika za nastanak ovog karcinoma na nacionalnom niovu je postignut donošenjem zakona o zaštiti stanovništva od izloženosti duvanskom dimu 17. j g j Nedavno publikovana studija 3 je utvrdila da je 60% oralnih karcinoma, karcinoma jednjaka, larinksa, grlića materice i mokraćne bešike moguće izbeći, odnosno prevenirati. Stoga, fokus javnog zdravlja treba da bude na merama prevencije i smanjenja učestalosti faktora rizika. Glavni faktori rizika za karcinom usne su: pušenje 3,14, konzumiranje alkohola 15, 16 i izlaganje 3,14, konzumiranje alkohola 15, 16 i izlaganje sunčevom zračenju. Prvi korak u smanjenju faktora rizika za nastanak ovog karcinoma na nacionalnom niovu je postignut donošenjem zakona o zaštiti stanovništva od izloženosti duvanskom dimu 17. g Yet, Serbia still has a reputation of a country with a very high prevalence of smoking. The smoking cessation is also related to socio-economic factors. It was established that the most addicted people were the poorest women 18. Smoking and alcohol consumption are well established risk factors 1, 19. Smoking and drinking alcohol have an individual and synergistic effect. The combination of smoking and heavy drinking creates 38 times higher risk for lip cancer compared to the risk of abstainers from both habits 20. The high lip cancer incidence matched the great alcohol consumption in Central and Eastern Europe4. The above-mentioned findings suggest that high prevalence of risk factors for lip cancer was registered in Serbia only in a few studies, which implies that improved and expanded surveillance of risk factors at the national level is needed. Ipak, Srbija i dalje ima reputaciju zemlje sa vrlo visokom prevalencijom pušenja, smatra se delom jer je pušenje povezano sa brojnim socio-ekonomskim faktorima. Studija sprovedena u našoj zemlji je pokazala da su najveći zavisnici od pušenja bile najsiro- mašnije žene 18. Pušenje i konzumiranje alkohola su dobro utvrđeni faktori rizika 1, 19. Pušenje i konzumiranje alkohola imaju i individualno i sinergističko dejstvo. Rezultati Kod žena je standardizovana stopa incidencije pokazala statistički značajan trend pada u periodu 2006.-2011. godine, sa APC - 24,3% (95%CI -40.9 - - 3.1, p = 0,031) (Slika 3). Standardizovana stopa mortaliteta od karcinoma usne kod muškaraca je pokazala statistički značajan trend pada u ispitivanom periodu sa APC 4,5% (95%CI - 7.8 - -2,8, p = 0,015) (Slika 4). 1746 Ignjatović et.al. RATE OF LIP CANCER Slika 1. Trend stope incidencije karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 1 The trend of crude rate of incidence of lip cancer in Central Serbia from 1999-2014 Slika 1. Trend stope incidencije karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 1. The trend of crude rate of incidence of lip cancer in Central Serbia from 1999-2014 ure 1. The trend of crude rate of incidence of lip cancer in Central Serbia from 1999-201 Slika 2. Trend mortaliteta karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 2. The trend of crude rate of mortality of lip cancer in Central Serbia from 1999-2014 Slika 2. Trend mortaliteta karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 2. The trend of crude rate of mortality of lip cancer in Central Serbia from 1999-2014 1747 Acta Stomatologica Naissi Dec/Dec. 2017, Vol. 33, br./num. 76 Slika 3. Trend stanadardizovane stope incidencije karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 3. The trend of ASR-W of incidence of lip cancer in Central Serbia from 1999-2014 Slika 3. Trend stanadardizovane stope incidencije karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. gure 3. The trend of ASR-W of incidence of lip cancer in Central Serbia from 1999-2014 Figure 3. The trend of ASR-W of incidence of lip cancer in Central Serbia from 1999-2014 Slika 4. Trend stanadardizovane stope mortaliteta karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 4. The trend of ASR-W of mortality of lip cancer in Central Serbia from 1999-2014 Slika 4. Trend stanadardizovane stope mortaliteta karcinoma usta u centralnoj Srbiji u periodu od 1999. do 2014. Figure 4. The trend of ASR-W of mortality of lip cancer in Central Serbia from 1999-2014 1748 Ignjatović i sar. STOPA KARCINOMA USNE Discussion Nedavno objavljeni podaci pokazuju da je ukupna HPV prevalencija i 16/18 HPV prevalencija u Srbiji kod žena zdravog izgleda i normalnog citološkog nalaza na grliću materice vrlo visoka 24. Time se zaokružuje slika o Srbiji kao zemlji sa visokom prevalencijom faktora rizika za karcinom usne. Almost the whole of the study period is characterized by low mortality/incidence ratio, which implies that lip cancer was timely diagnosed in Central Serbia. The early detection and treatment of lip cancer are related to better prognosis. Therefore, the development of screening program is needed, especially in younger population that is affected 25. Gotovo ceo ispitivani period karakteriše nizak MIR, što ukazuje na pravo- vremenu dijagnozu ovog karcinoma u centralnoj Srbiji. Svakako da su rana detekcija i lečenje povezani sa boljim ishodom, zato je razvoj skrining programa neophodan, posebno u mlađoj populaciji 25. j p p j Globalna standardizivana stopa inci- dencije za karcinom usne u 2012. godini je procenjena na 0,3/100000, sa vrlo izraženim regionalnim razlikama. Prema GLOBOCAN- ovim procenama, incidencija ovog karcinoma kod muškaraca u Srbiji je u 2012. godini rangirana na 12. mestu, a na 21. kod žena. Razmatrajući stopu mortaliteta, prema GLOBOCAN-ovim procenama, Srbija se nalazi na 15. mestu po smrtnosti u muškoj populaciji i na 14. po smrtnosti u ženskoj populaciji. Centralna i istočna Evropa se smatraju regionom sa visokom stopom prevalencije karcinoma usne 19, 26. Gotovo petina novih slučajeva (19,2%) se javlja na prostoru centralne i istočne Evrope 4. Dakle, Srbija je okružena zemljama sa vrlo visokom učestalošću ovog karcinoma. Najveće stope karcinoma usne su posebno karakteristične za sledeće zemlje istočne Evrope, posebno u muškoj populaciji: u Bugarskoj, Češkoj, Mađarskoj, Poljskoj, Rumuniji i Slovačkoj 27. Izrazito visoke stope smrtnosti su registrovane u Mađarskoj devedesetih godina prošlog veka. The global ASR-W of lip cancer in 2012 was estimated at 0.3/100,000 with pronounced regional differences. Regarding the GLOBOCAN estimates in 2012, Serbia was ranked at the 12th place by the incidence rate of lip cancer in males and at the 21st place by the incidence of lip cancer in females. Considering the mortality rate, this comparison with GLOBOCAN 2012 estimates showed that Serbia was placed at the 15th place in males and at the 14th place in females. Central and Eastern Europe is considered as a region with a relatively high incidence rate of lip cancer 19, 26. Discussion Kombinacije pušenja i teškog alkoholizma stvara 38 puta veću šansu za razvoj karcinoma usne u odnosu na one koji ne upražnjavaju nijednu od ovih navika 20. Osim toga, visoka stopa incidencije ovog karcinoma se poklapa teritorijalno sa velikom konzumacijom alkohola u centralnoj i istočnoj Evropi 4. 1749 Acta Stomatologica Naissi Dec/Dec. 2017, Vol. 33, br./num. 76 The decreasing trend of lip cancer incidence in males might be the first result of controlling tobacco use. Additionally, the downward trend of lip cancer incidence in males may also be caused by less sunlight exposure due to a reduction in the population living in rural areas 21. A similar effect could be achieved with a reduction of outdoor occupancy. Sunlight and radiation were established as strongly suggestive risk factors for lip cancer 19.The effect of sunlight exposure on the carcinoma mechanism depends on the time of exposure and different activities. Long-term and constant exposure to sunlight during work is a predictor of lip cancer, but exposure during holidays and leisure seems to have a protective effect 22. In the group of possible risk factors of lip cancer, special attention was drawn to the HPV infection 23. Recent data showed that the prevalence of the overall and 16/18 HPV infections in Serbian women was high in women with a healthy appearance and a cytologically normal cervix 24. The above-mentioned facts complete the image of Serbia as a region with the high prevalence of lip cancer risk factors. Gore pomenuti nalazi ukazuju da je visoka prevalencija faktora rizika za nastanak ovog karcinoma registrovana u Srbiji u samo nekoliko studija i da je potreban poboljšani i prošireni nadzor ovih faktora rizika na nacionalnom nivou. Trend pada incidencije karcinoma usne kod muškaraca može biti prvi rezultat kontrole pušenja u našoj zemlji. Osim toga, može biti i posledica manje izloženosti sunčevom zračenju usled smanjenja populacije u ruralnim oblastima 21. Sličan efekat se može dobiti i smanjenjem aktivnosti na otvorenom. Sunčeva svetlost i zračenje se ubrajaju u snažne sugestivne faktore rizika 19. Efekat izlaganja sunčevom zračenju na sam mehanizam nastanka karcinoma usne zavisi od vremena izlaganja i tipa aktivnosti. Dugoročno izlaganje suncu na poslu je prediktor za nastanak karcinoma usne, a izlaganje tokom odmora i slobodnog vremena izgleda da ima zaštitini efekat 22. U grupu potencijalnih faktora rizika posebnu pažnju privlači HPV infekcija 23. Discussion Almost one fifth of new cases (19,2%) occurred in Central and Eastern Europe 4. Serbia is surrounded with countries with the high prevalence of this cancer. The highest rates of lip cancer mortality are typical for Eastern Europe 27, particularly for males from Bulgaria, the Czech Republic, Hungary, Poland, Romania and Slovakia. An extremely high rate of mortality was observed in Hungary in the mid-1990’s. 1750 Ignjatović et.al. RATE OF LIP CANCER Limitations of the study Several possible limitations of this study should be considered. Strategies for cancer control are based on the epidemiology data. Therefore, the quality of cancer registers is crucial. The potential classification problem arises from the position of lip cancer 3. This cancer is a form of oral cancer located at the junction between the oral cavity and the skin which could be misdiagnosed as skin cancer28. The quality of data related to the cause of death in Serbia was assessed as moderate by the World Health Organization29. The percentage of unknown and ill-defined cancer deaths for the most recent year indicated that cause-of-death data in Serbia were of moderate quality 30. Furthermore, GLOBOCAN 2012 categorized data from Serbia as B2 which means high quality regional data (coverage between 10% and 50%) for the incidence rates and medium quality for complete vital registration for the mortality rates. Potrebno je napomenuti nekoliko mogu- ćih nedostataka ove studije. Strategije u kontroli kancera su vezane za epidemiološke podatke. Zato je kvalitet podataka u registrima karcinoma vrlo značajan. Potencijalni problemi u klasifikaciji karcinoma usne javljaju se zbog same lokalizacije promene 3. Ovaj karcinom je tip oralnog karcinoma koji je lokalizovan na spoju usne šupljine i kože, što može dovesti do toga da bude dijagnostikovan kao karcinom kože 28. Ipak, Svetska zdravstvena oragnizacija je kvalitet podataka koji se odnose na uzork smrti u Srbiji ocenila kao umeren 29. Takođe, procenat nepoznatih i nedefinisanih uzroka smrti kod obolelih od karcinoma pokazuje da su podaci o uzrocima smrti u Srbiji umerenog kvaliteta 30. Osim toga, GLOBOCAN 2012 je kategorisao podatke iz Srbije kao B2, što podrazumeva visoko kvalitetne regionalne podatke za stope incidencije i srednjeg kvaliteta podatke vezane za smrtnost. LITERATURA /REFERENCES LITERATURA /REFERENCES 11. Videnović G, Ilić D, Miljuš D, Krasić D, Vlahović Z, Živković S, et al. Lip, oral cavity and pharyngeal cancers in the population of the city of Belgrade in the period 1999-2010. Vojnosanitetski pregled. 2016;73(1):53-8. 1. Maruccia M, Onesti M, Parisi P, Cigna E, Troccola A, Scuderi N. Lip cancer: a 10-year retrospective epidemiological study. Anticancer research. 2012;32(4):1543-6. y g p Vojnosanitetski pregled. 2016;73(1):53-8 12 Shib ki CH Shib ki SC Sil S ( ) 2. Warnakulasuriya S. Causes of oral cancer-an appraisal of controversies. British dental journal. 2009;207(10):471. 12. Shiboski CH, Shiboski SC, Silverman S. Trends in oral cancer rates in the United States, 1973– 1996. Community dentistry and oral epidemiology. 2000;28(4):249-56. 3. Moore S, Johnson N, Pierce A, Wilson D. The epidemiology of lip cancer: a review of global incidence and aetiology. Oral diseases. 1999 Jul;5(3):185-95. PubMed PMID: 10483063. p gy 13. Czerninski R, Zini A, Sgan Cohen H. Lip cancer: incidence, trends, histology and survival: 1970–2006. British Journal of Dermatology. 2010;162(5):1103-9. 4. Shield KD, Ferlay J, Jemal A, Sankaranarayanan R, Chaturvedi AK, Bray F, et al. The global incidence of lip, oral cavity, and pharyngeal cancers by subsite in 2012. CA: a cancer journal for clinicians. 2017;67(1):51-64. 14. Çankaya H, Garça MF, Bozan N, Isik D, Kiroglu AF. Epidemiological features of the lip cancers and it's relation with smoking. Eastern Journal of Medicine. 2013;18(2):64. ( ) 5. Casal D, Carmo L, Melancia T, Zagalo C, Cid O, Rosa-Santos J. Lip cancer: a 5-year review in a tertiary referral centre. Journal of Plastic, Reconstructive & Aesthetic Surgery. 2010;63(12):2040-5. 15. Maruccia M, Onesti MG, Parisi P, Cigna E, Troccola A, Scuderi N. Lip cancer: a 10-year retrospective epidemiological study. Anticancer research. 2012 Apr;32(4):1543-6. p 16. Radoï L, Paget-Bailly S, Cyr D, Papadopoulos A, Guida F, Schmaus A, et al. Tobacco smoking, alcohol drinking and risk of oral cavity cancer by subsite: results of a French population-based case–control study, the ICARE study. European Journal of Cancer Prevention. 2013;22(3):268- 76. ; ( ) 6. Zitsch RP, Lee BW, Smith RB. Cervical lymph node metastases and squamous cell carcinoma of the lip. Head & neck. 1999;21(5):447-53. p 7. Khuder SA. Etiologic clues to lip cancer from epidemiologic studies on farmers. Scandinavian journal of work, environment & health. 1999:125-30. 17. O'Rourke M, Djukic J. Welcome to Serbia: feel free to smoke. Tobacco control. 2008 Dec;17(6):428-30. PubMed PMID: 18827037. 8. Conclusion Rezultati naše studije ukazuju da je profil karcinoma usne u centralnoj Srbiji stabilan i ima trend pada u ispitivanom periodu. Nakon 2011. godine, pokazuje se blagi trend porasta incidencije u ženskoj popu- laciji. Smrtnost od karcinoma usne u istom periodu prati trend incidencije. Na osnovu ovoga može se zaključiti da epidemiološka slika ovog karcinoma na ovoj teritoriji nije dramatična. Ali, GLOBOCAN predikcije ukazuju na porast incidencije ovog karcinoma usled porasta i starenja populacije. Pored toga, zemlje u razvoju su bile pogođene tranzicijom od zaraznih ka nezaraznim bolestima 31, što ukazuje na mogući trend rasta u narednom periodu. Stoga, visoka prevalencija faktora rizika, naročito pušenje i konzumiranje alkohola, ukazuju da postoji prostor za javno- zdravstvene aktivnosti na nacionalnom nivou. Ove aktivnosti treba da budu usmerene na ekonomski efikasne mere u redukciji pušenja i smanjenju konzumiranja alkohola, kao i na uporno širenje svesti o rizicima preteranog izlaganja suncu. Pooled evidence implies that the profile of lip cancer in Central Serbia is stable and declines in the study period. After 2011, there are indications of a slight increase in female incidence. The mortality for the same period follows the pattern of incidence. Based on that, it can be concluded that the epidemiological picture of lip cancer is not dramatic. However, predictions for this cancer indicate its increase based on population growth and aging. Additionally, developing countries were affected by transition from infectious diseases to noncom- municable diseases 31, which implicates further possible increasing trend. Therefore, the high prevalence of risk factors, especially smoking habits, alcohol consumption, indicates that there is a place for public health activities at the national level. Those activities should be focused on cost-effective policies to reduce alcohol and tobacco use and persistent dissemination of information about the risks of sun over-exposure. 1751 Acta Stomatologica Naissi Dec/Dec. 2017, Vol. 33, br./num. 76 Zahvalnica The authors thank the staff of the Institute of Public Health of Serbia whose efforts to collect accurate and complete data have made this report possible. This study was funded by the Serbian Ministry of Education and Science (grant No 41018). The authors declare that there is no conflict of interest. Autori se zahvaljuju osoblju Instituta za javno zdravlje Srbije, bez čijeg napora u tačnom i preciznom prikupljanju podataka ova studija ne bi bila moguća. Ova studija je finansijski podržana od strane Ministarstva prosvete, nauke i tehnološkog razvoja Republike Srbije, broj III 41018. Autori izja- vljuju da ne postoji sukob interesa. LITERATURA /REFERENCES Lopez EP-M, Minarro-Del Moral R, Martinez- Garcia C, Zanetti R, Rosso S, Serrano S, et al. Lifestyles, environmental and phenotypic factors associated with lip cancer: a case–control study in southern Spain. British journal of cancer. 2003;88(11):1702. 18. Djikanovic B, Marinkovic J, Jankovic J, Vujanac V, Simic S. Gender differences in smoking experience and cessation: do wealth and education matter equally for women and men in Serbia? Journal of Public Health. 2011;33(1):31- 8. 9. Soerjomataram I, de Vries E, Pukkala E, Coebergh JW. Excess of cancers in Europe: a study of eleven major cancers amenable to lifestyle change. International journal of cancer. 2007 Mar 15;120(6):1336-43. 19. Warnakulasuriya S. Global epidemiology of oral and oropharyngeal cancer. Oral oncology. 2009 Apr-May;45(4-5):309-16. 10. World Health Organization. International classification of diseases and related health problems, 10th revision. Volume 1st edition. http://www.who.int/ classifications/icd/en/. p y ( ) 20. Blot WJ, McLaughlin JK, Winn DM, Austin DF, Greenberg RS, Preston-Martin S, et al. Smoking and drinking in relation to oral and pharyngeal cancer. Cancer research. 1988;48(11):3282-7. 1752 Ignjatović i sar. STOPA KARCINOMA USNE 26. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. International journal of cancer. 2015;136(5). 21. Blomberg M, Nielsen A, Munk C, Kjaer SK. Trends in head and neck cancer incidence in Denmark, 1978–2007: focus on human papillomavirus associated sites. International journal of cancer. 2011;129(3):733-41. j ( ) 27. La Vecchia C, Lucchini F, Negri E, Levi F. Trends in oral cancer mortality in Europe. Oral oncology. 2004;40(4):433-9. j ( ) 22. Perea-Milla Lopez E, Minarro-Del Moral RM, Martinez-Garcia C, Zanetti R, Rosso S, Serrano S, et al. Lifestyles, environmental and phenotypic factors associated with lip cancer: a case-control study in southern Spain. British journal of cancer. 2003 Jun 02;88(11):1702-7. gy ( ) 28. Macfarlane GJ, Boyle P, Evstifeeva T, Scully C. Epidemiological aspects of lip cancer in Scotland. Community dentistry and oral epidemiology. 1993;21(5):279-82. j ; ( ) 23. Galyon SW, Frodel JL. Lip and perioral defects. Otolaryngologic clinics of North America. 2001;34(3):647-66. p gy ( ) 29. Mathers CD, Fat DM, Inoue M, Rao C, Lopez AD. Counting the dead and what they died from: an assessment of the global status of cause of death data. Bulletin of the World Health Organization. 2005 Mar;83(3):171-7. ( ) 24. LITERATURA /REFERENCES Malisic E, Brotto K, Krivokuca A, Cavic M, Jankovic R. Overall human papilloma virus and types 16/18 prevalence in women with normal cervical cytology in Serbia: is it time for human papillomavirus testing and/or vaccination. Journal of BUON : official journal of the Balkan Union of Oncology. 2014;19(4):973-9. g 30. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JW, Comber H, et al. Cancer incidence and mortality patterns in Europe: estimates for 40 countries in 2012. European journal of cancer. 2013 Apr;49(6):1374-403. gy ( ) 25. Llewellyn CD, Johnson NW, Warnakulasuriya S. Factors associated with delay in presentation among younger patients with oral cancer. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology. 2004;97(6):707- 13. 31. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. The lancet oncology. 2012;13(8):790-801. 1753
https://openalex.org/W2261430354	https://www.nature.com/articles/srep11959.pdf	English	null	Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre-diabetes and type 2 diabetes	Scientific reports	2,015	cc-by	7,503	Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre- diabetes and type 2 diabetes received: 13 March 2015 accepted: 11 June 2015 Published: 07 July 2015 Chiranjit Ghosh1, Prabuddha Mukhopadhyay2, Shibendu Ghosh3 & Manik Pradhan1 New strategies for an accurate and early detection of insulin resistance are important to delay or prevent the acute onset of type 2 diabetes (T2D). Currently, insulin sensitivity index (ISI0,120) is considered to be a viable invasive method of whole-body insulin resistance for use in clinical settings in comparison with other invasive sensitivity indexes like homeostasis model assessment (HOMA), and quantitative insulin sensitivity check index (QUICKI). To investigate how these sensitivity indexes link the 13C/12C-carbon isotopes of exhaled breath CO2 to pre-diabetes (PD) and type 2 diabetes in response to glucose ingestion, we studied excretion dynamics of 13C/12C-isotopic fractionations of breath CO2. Here, we show that 13C/12C-isotope ratios of breath CO2 were well correlated with blood glucose, insulin, glycosylated-hemoglobin as well as with HOMA-IR and 1/QUICKI. Conversely, the strongest correlation was observed between 1/ISI0,120 and breath CO2 isotopes. Consequently, we determined several optimal diagnostic cut-off points of 1/ISI0,120 and 13CO2/12CO2-isotope ratios to distinctively track the evolution of PD prior to the onset of T2D. Our findings suggest that isotopic breath CO2 is a novel method for accurate estimation of ISI0,120 and thus may open new perspectives into the isotope-specific non-invasive evaluation of insulin resistance for large-scale real-time diabetes screening purposes. Type 2 diabetes mellitus (T2D), the most common deleterious metabolic disease at present all over the world, is usually preceded by the combined effects of pancreatic β -cell dysfunction and insulin resist- ance1,2. Several lines of evidence suggest that insulin resistance is the key risk factor in the pathogenesis of T2D. Hence an accurate and early detection of insulin resistance is important to delay or prevent the acute onset of T2D. However, there is still a real challenge on when or how to evaluate individuals who are at high-risk for developing insulin resistance or during the preclinical phase of T2D. g p g g p p To quantify the insulin resistance (i.e. inverse of insulin sensitivity), the hyperinsulinemic-euglycemic clamp (HEC) study has been considered over the decades to be the “gold standard” method. But due to several complications like continuous infusion of insulin, frequent blood glucose infusion and overall withdrawals of blood samples at different time intervals, it is difficult to apply in large-scale screening purposes3,4. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre- diabetes and type 2 diabetes Therefore, the different surrogate techniques such as the quantitative insulin sensitivity check index (QUICKI) and homeostasis model assessment (HOMA)5–7, derived primarily from the measure- ments of fasting blood glucose and fasting plasma insulin levels, are well accepted alternative methods for assessing insulin resistance3. More recently, another invasive surrogate index called insulin sensitivity index, ISI0,120, that exploits both the fasting (0 min) and post-dose (120 min) plasma insulin and blood 1Department of Chemical, Biological and Macromolecular Sciences, S. N. Bose National Centre for Basic Sciences, Salt Lake, JD Block, Sector III, Kolkata-700098, India. 2Department of Medicine, Vivekananda Institute of Medical Sciences, 99 Sarat Bose Road, Kolkata-700027, India. 3Department of Medicine, Raipur Institute of Medical Sciences, Raipur-492006, Chhattisgarh, India. Correspondence and requests for materials should be addressed to M.P. (email: manik.pradhan@bose.res.in) Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 1 www.nature.com/scientificreports/ glucose concentrations, has been proposed to be a viable method of whole-body insulin sensitivity for use in clinical settings8. The ISI0,120 is calculated by the following formula: 1/ISI0,120 = log mean insu- lin/MCR, where mean of 0 and 120 min blood glucose values are utilized to calculate the metabolic clearance rate (MCR)9. Results obtained by utilizing ISI0,120 were well correlated with the HEC study in comparison to the other surrogate sensitivity indexes. This is because of the fact that during oral glucose tolerance test (OGTT), the 0 and 120 min samples are more reproducible than other intermediate time points10. Moreover, HOMA-IR and QUICKI both estimate the hepatic glucose resistance, whereas the peripheral insulin resistance can be evaluated from the ISI0,120. Besides, the calculations for estimating HOMA-IR and other sensitivity indexes consider the insulin secretory capability, whereas ISI0,120 evalu- ates the insulin sensitivity in body. Nevertheless, the samples required for ISI0,120 measurement are less and it estimates better insulin sensitivity compared to the other sensitivity index formulas8. Therefore, the ability to non-invasively evaluate insulin sensitivity index (ISI0,120) for diagnosis of pre-diabetes and type 2 diabetes has a substantial clinical significance. Recently, it has been proposed that 13C-glucose breath test (13C-GBT) may be a non-invasive approach for quantifying insulin resistance by contrast with the direct invasive HEC study5. The 13C-GBT exploits the carbohydrate metabolism of orally admin- istered 13C-labelled glucose substrate. The substrate is metabolized and produces 13C-labelled carbon dioxide (13CO2). This 13CO2 is then transported to the lungs through the blood stream, and finally it is excreted in exhaled breath. Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre- diabetes and type 2 diabetes During OGTT, the glucose utilization for cellular fuel oxidation is strongly dependent on the insulin sensitivity. Postprandially, the alteration of insulin sensitivity in PD and T2D significantly decreases the cellular glucose uptake. In individuals with pre-diabetes and type 2 diabetes, when a dose containing isotopically labelled glucose [U-13C6] is ingested, the cellular response to this exogenous glucose is remarkably blunted. Individuals with insulin resistance or T2D will exhibit less carbon-13 isotopes of exhaled breath CO2 because of impaired glucose uptake by the cells6,7. However, the potential link between the exhaled breath 13CO2/12CO2 isotope ratios and ISI0,120 for the pre-diabetes and type 2 diabetes is not currently known. The 13CO2 / 12CO2 isotope ratios in the exhaled breath are usually expressed as delta-over-baseline (DOB) values described in the following way: δ =       −       × ( ) R R C‰ 1 1000 1 13 sample standard ( ) 1 δ ( ) = (δ ) −(δ ) ( ) = = C t ‰ C‰ C‰ 2 DOB 13 13 t t min 13 t 0 min ( ) 2 where Rsample and Rstandard are defined as 13C/12C isotope ratio of the sample and international standard Pee Dee Belemnite (PDB) value (0.0112372) respectively. The isotope tracer, δ DOB 13C(t)‰, quantified from the exhaled breath samples, can be followed to accurately evaluate the insulin resistance prior to the onset of T2D. Therefore, there is a pressing need to evaluate the clinical efficacy of the carbon iso- topic fractionations of breath CO2 during the glucose metabolism for large-scale screening of individuals with insulin resistance and type 2 diabetes. Moreover, unravelling the potential link between the stable isotopes of carbon in breath CO2 and ISI0,120 may specifically track the pathogenesis of the preclinical phase of T2D and hence may introduce a new strategy for non-invasive evaluation of insulin resistance.i p y gy To find the association between the 13C/12C-isotope ratios of breath CO2 and the ISI0,120, we have analysed the exhaled breath carbon dioxide isotopes for the accurate and fast non-invasive assessment of insulin resistance in PD and T2D by means of a laser-based high-precision cavity-enhanced integrated cavity output spectroscopy (ICOS) system. In this article, we have demonstrated the associations between exhaled breath carbon-13 isotopes of CO2 and invasive parameters like blood glucose, insulin and HbA1c levels. Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre- diabetes and type 2 diabetes Furthermore, we also determined several diagnostic parameters of the breath isotope analysis including sensitivity, specificity, optimal diagnostic cut-off points along with positive and negative pre- dictive values to accurately evaluate the insulin resistance as well as the precise metabolic transition from normal to PD and then on to T2D. Results and Discussion The individuals with PD and T2D accordingly produce less amount of 13CO2 in exhaled breath samples compared with NDC. Our results support that the breath tests are more direct measure of intracellular glucose metabolism and impairments of exogenous (oral) glucose. Thus the monitoring of stable carbon isotopes may assist in non-invasive assessment of NDC, PD and T2D individuals by providing an alternative approach for large-scale screening purposes without the need for invasive repeated blood samplings. g g p p p p g We next explored whether there were any correlations of δ DOB 13C‰ values in exhaled breath at 120 min with the variables related to the insulin resistance, such as absolute changes in blood glucose levels, and plasma insulin levels at the particular time during the OGTT, in addition to glycated hae- moglobin (HbA1c) measurements. Figure 2 depicts the inverse correlations of δ DOB 13C‰ values with all these measured parameters. The correlation of δ DOB 13C (t = 120 min)‰ with HbA1c (%) was strong with a correlation coefficient of r = − 0.71 (p < 0.001) compared with the blood glucose (r = − 0.64, p < 0.001) and plasma insulin levels (r = − 0.5, p < 0.001). p ( p ) We further investigated the association of the δ DOB 13C‰ values at 120 min in breath samples with the three surrogate methods of HOMA-IR, QUICKI and ISI0,120 to measure the insulin resistance in all individuals with different metabolic states. Significant inverse correlations (p < 0.01) between breath δ DOB 13C‰ values and all the measured indexes were observed, as shown in Fig. 3. However, the best correlation was observed between δ DOB 13C‰ values and 1/ISI0,120 (r = − 0.81, p < 0.001), suggesting a significant association between the results of δ DOB 13C‰ and ISI0,120 index. i , Recent studies demonstrated that ISI0,120 is superior to HOMA-IR for the assessment of insulin resist- ance because of essentially two intrinsic drawbacks of HOMA model12,13. Firstly, HOMA index is based on the assumption that beta cell function is normal and therefore it can’t be applied for patients with type 2 diabetes where beta cell is destroyed in many cases. Secondly, HOMA-IR considers that the rela- tionship between insulin and glucose is linear, but practically it is parabolic1. The QUICKI has also the similar disadvantages as like HOMA-IR12. Results and Discussion We initially investigated the time-dependent excretion kinetics of stable carbon-13 isotopes, expressed as δ DOB 13C(t)‰ values, in exhaled breath samples by ICOS method to investigate the distribution of 13CO2 isotopic abundance in breath samples associated with isotopically labelled glucose metabolism for NDC (non-diabetic control), PD and T2D and the results have been illustrated in Fig. 1a. We observed that the mean δ DOB 13C(t)‰ values for the group with T2D were significantly lower (p < 0.01) compared to the groups with PD and NDC between 90 min and 210 min after the glucose load. These findings suggest that carbon isotopic fractionations of breath CO2 is capable of detecting marked differences in δ DOB 13C(t)‰ values in exhaled breath samples within 90 min of a 5 h-OGTT among the groups with NDC, PD and T2D. However, there were no statistically significant differences (p > 0.05) in mean δ DOB 13C(t)‰ values among all the groups from 240 min in response to glucose ingestion. We also simultaneously studied how the concentration of blood glucose changes with time in response to oral glucose ingestion.i g g p g g Figure 1b indicates that blood glucose levels were significantly (p < 0.001) higher in T2D and groups compared with the NDC group, as expected. For an individual with T2D or PD, the impa Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 2 www.nature.com/scientificreports/ / p / Figure 1. Kinetics study of breath carbon-13 isotope excretions and blood glucose levels after administration of 13C-glucose. a, Breath δ DOB 13C (‰) enrichments. b, blood glucose concentrations for normal (NDC), pre-diabetes (PD) and Type 2 diabetes (T2D) individuals at every 30 min interval during 5 h-OGTT. indicates statistically significant difference (p < 0.05). Data are expressed in terms of mean ± SEM. Figure 1. Kinetics study of breath carbon-13 isotope excretions and blood glucose levels after administration of 13C-glucose. a, Breath δ DOB 13C (‰) enrichments. b, blood glucose concentrations for normal (NDC), pre-diabetes (PD) and Type 2 diabetes (T2D) individuals at every 30 min interval during 5 h-OGTT. indicates statistically significant difference (p < 0.05). Data are expressed in terms of mean ± SEM. glucose uptake plays an important role for blunted glucose oxidation in cells because of diminished pancreatic insulin secretion or impaired insulin action on the target tissues11. Results and Discussion Furthermore, during the OGTT, the majority of the insulin mediated glucose disposal takes place in peripheral tissues14, suggesting that the estimation of insulin resistance in peripheral tissue is more important to correlate with 13CO2 excretions from isotopically labelled glucose disposal for NDC, PD and T2D. It is noteworthy that HOMA-IR and QUICKI evaluate the hepatic insulin resistance essentially in fasting state, whereas ISI0,120 estimates the peripheral insulin resistance which is primarily responsible for exogenous glucose uptake. Therefore ISI0,120 may be the best predictor of insulin resistance and hence the correlation of δ DOB 13C ‰ with ISI0,120 index is more practical to evaluate the efficacy of breath test to measure the insulin resistance.i fi y Taken together, these findings suggest that monitoring stable 13C/12C isotopes of breath CO2 in response to glucose ingestion may be an easy and non-invasive approach to evaluate the insulin resist- ance in individuals. In view of the present results, we posit that the measurements of ISI0,120 index as a surrogate marker of insulin resistance and thus may distinctively track the evolution of pre-diabetes prior Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 3 www.nature.com/scientificreports/ Figure 2. Linear regression plots to show the correlations of δDOB 13C (‰) in breath with different invasive parameters. a, breath δ DOB 13C(‰) with glycosylated haemoglobin (HbA1c %). b, plasma insulin levels (∆ Insulin Level) from the baseline during 2 h-OGTT. c, shows correlation of blood glucose concentrations (∆ Blood Glucose). The data are statistically significant different (p < 0.001). Figure 2. Linear regression plots to show the correlations of δDOB 13C (‰) in breath with different invasive parameters. a, breath δ DOB 13C(‰) with glycosylated haemoglobin (HbA1c %). b, plasma insulin levels (∆ Insulin Level) from the baseline during 2 h-OGTT. c, shows correlation of blood glucose concentrations (∆ Blood Glucose). The data are statistically significant different (p < 0.001). to the onset of T2D, as depicted in Fig. 4. The distributions of δ DOB 13C‰ and 1/ISI0,120 values illustrate that individuals can be classified into three distinct zones with NDC in green zone, PD (moderately insulin resistance) in yellow zone and T2D (sufficiently higher insulin resistance) in red zone. Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 Results and Discussion We hav determined the cut-off values based on the Indian populations utilizing the mentioned labelled glucos Therefore, it would be interesting to estimate the cut-off values within the subject-variability and also elucidate the probable dietary effects on breath isotope analysis in future studies. Figure 3. Correlations of δDOB 13C(‰) in breath with currently available insulin resistance assessment methods. a, HOMA-IR. b, 1/QUICKI. c, 1/ISI0,120. The data are statistically significant different (p < 0.01). Figure 4. Distribution of δDOB 13C‰ and 1/ISI0,120 values against HbA1c (%) in normal (NDC), pre- diabetes (PD) and type 2 diabetes (T2D). Plot represents the clear transitions of δ DOB 13C and 1/ISI0,120 from NDC (green) to PD (yellow) and T2D (red). Data are expressed as mean ± SD. Figure 3. Correlations of δDOB 13C(‰) in breath with currently available insulin resistance assessment methods. a, HOMA-IR. b, 1/QUICKI. c, 1/ISI0,120. The data are statistically significant different (p < 0.01). Figure 3. Correlations of δDOB 13C(‰) in breath with currently available insulin resistance assessment methods. a, HOMA-IR. b, 1/QUICKI. c, 1/ISI0,120. The data are statistically significant different (p < 0.01). Figure 3. Correlations of δDOB 13C(‰) in breath with currently available insulin resistance assessment methods. a, HOMA-IR. b, 1/QUICKI. c, 1/ISI0,120. The data are statistically significant different (p < 0.01). Figure 4. Distribution of δDOB 13C‰ and 1/ISI0,120 values against HbA1c (%) in normal (NDC), pre- diabetes (PD) and type 2 diabetes (T2D). Plot represents the clear transitions of δ DOB 13C and 1/ISI0,120 from NDC (green) to PD (yellow) and T2D (red). Data are expressed as mean ± SD. Figure 4. Distribution of δDOB 13C‰ and 1/ISI0,120 values against HbA1c (%) in normal (NDC), pre- diabetes (PD) and type 2 diabetes (T2D). Plot represents the clear transitions of δ DOB 13C and 1/ISI0,120 from NDC ( ) PD ( ll ) d T2D ( d) D d ± SD Figure 4. Distribution of δDOB 13C‰ and 1/ISI0,120 values against HbA1c (%) in normal (NDC), pre- diabetes (PD) and type 2 diabetes (T2D). Plot represents the clear transitions of δ DOB 13C and 1/ISI0,120 from NDC (green) to PD (yellow) and T2D (red). Data are expressed as mean ± SD. different populations of various countries and the isotopic compositions of labelled glucose. We have determined the cut-off values based on the Indian populations utilizing the mentioned labelled glucose. Results and Discussion yfi y g Finally, to investigate the precise metabolic transition from NDC to PD and then on to T2D, we deter- mined the optimal diagnostic cut-off points of 1/ISI0,120 and δ DOB 13C‰ values in exhaled breath using receiver operating characteristics curve (ROC) analysis (Fig. 5a and b). ROC curves were generated by plotting the true positive rate (sensitivity) against the false positive rate (1-specficity) using the values of 1/ISI0,120 and δ DOB 13C‰. The highest values of sensitivity and specificity were used to calculate the optimal diagnostic cut-off points. A diagnostic cut-off point of 1/ISI0,120 = 0.0237 between individuals with PD and T2D, exhibited the sensitivity and specificity of 96.1% and 95.1%, respectively, whereas 1/ISI0,120 = 0.0149 accurately diagnosed individuals with NDC and PD. To apply breath analysis for the diagnosis of insulin resistant PD and T2D, we have calculated the optical diagnostic cut-off values of carbon-13 isotopes of exhaled breath CO2. We observed that individuals with δ DOB 13C‰ < 24.4 and δ DOB 13C‰ > 29.4 were considered as T2D and NDC respectively, whereas subjects with 29.4 > δ DOB 13C‰ > 24.4 were suggested to be PD. These cut-off points corresponded to the similar levels of diagnostic sensitivity and specificity as shown in Table 1. We have calculated these cut-off values with 95% confidence intervals. Thus the analyses of stable isotopes of the major metabolite of human breath CO2 establish a broad clinical fea- sibility as a sufficiently robust non-invasive detection method for an accurate diagnosis of PD and T2D with different metabolic states of insulin resistance. We also finally explored the positive and negative predictive values (PPV and NPV) for the diagnostic assessment. These two parameters essentially indi- cate the probabilities of getting diseases once the actual test results of the patients are known15. The pres- ent method demonstrated diagnostic PPV of 98% between NDC vs PD, and 96% between PD vs T2D. It also exhibited diagnostic NPV of 94% between NDC vs PD, and 95% between PD vs T2D, indicating an excellent diagnostic accuracy for the accurate evaluation of insulin resistance in different metabolic states. However, it is important to note that the cut-off values may depend on the food habits in the Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 4 www.nature.com/scientificreports/ w.nature.com/scientificreports/ different populations of various countries and the isotopic compositions of labelled glucose. Results and Discussion Therefore, it would be interesting to estimate the cut-off values within the subject-variability and also to elucidate the probable dietary effects on breath isotope analysis in future studies. different populations of various countries and the isotopic compositions of labelled glucose. We have determined the cut-off values based on the Indian populations utilizing the mentioned labelled glucose. Therefore, it would be interesting to estimate the cut-off values within the subject-variability and also to elucidate the probable dietary effects on breath isotope analysis in future studies. Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 5 www.nature.com/scientificreports/ Figure 5. Receiver operating characteristic (ROC) curves to determine optimal diagnostic cut-off points. a, δ DOB 13C‰ and b, 1/ISI0,120 for clinical diagnosis of normal (NDC), pre-diabetes (PD) and type 2 diabetes (T2D). Figure 5. Receiver operating characteristic (ROC) curves to determine optimal diagnostic cut-off points. a, δ DOB 13C‰ and b, 1/ISI0,120 for clinical diagnosis of normal (NDC), pre-diabetes (PD) and type 2 diabetes (T2D). Groups Cut-off points of δDOB 13C‰ Sensitivity Specificity PPV NPV AUC Accuracy NDC vs PD 29.4 94.7% 97.2% 97.3% 94.6% 0.97 95.9% PD vs T2D 24.4 97.9% 92.1% 93.8% 97.2% 0.98 95.3% Groups Cut-off points of 1/ ΙSI0,120 Sensitivity Specificity PPV NPV AUC Accuracy NDC vs PD 0.0149 92.1% 100% 100% 92.3% 0.963 95.9% PD vs T2D 0.0237 95.7% 95.7% 95.7% 94.7% 0.954 95.3% Table 1. Determination of important diagnostic parameters related to cut-off values of 1/ISI0,120 and δDOB 13C (‰) by ICOS method for screening individuals with non-diabetic control (NDC), pre-diabetes (PD) and type 2 diabetes (T2D). AUC: area under the curve; PPV: positive predictive value; NPV: negative predictive value. Table 1. Determination of important diagnostic parameters related to cut-off values of 1/ISI0,120 and δDOB 13C (‰) by ICOS method for screening individuals with non-diabetic control (NDC), pre-diabetes (PD) and type 2 diabetes (T2D). AUC: area under the curve; PPV: positive predictive value; NPV: negative predictive value. Materials and Methods Subjects. A total of 116 human subjects (31 non-diabetic controls, 37 pre-diabetes and 48 type 2 diabetes) were recruited for the study. Individuals with hypertension, previous history of diabetes, taking any medication that may affect lipid or glucose metabolism, were excluded from the study. According to the guideline of the American Diabetes Association (ADA)16, the subjects were classified into three groups: non-diabetic controls with glycosylated hemoglobin (HbA1c %) < 5.7 and 2-hr blood glu- cose < 140 mg/dL during the oral glucose tolerance test (2h-OGTT), pre-diabetes with 5.7 ≤ HbA1c (%) < 6.5 and 140 mg/dL ≤ 2-hr OGTT glucose < 200 mg/dL and type 2 diabetes with HbA1c (%) ≥ 6.5 and 2-hr OGTT ≥ 200 mg/dL. Individuals were selected from the treatment-naive population, where they were totally undiagnosed about the status of their diabetes. The clinical parameters are described in Table 2. The whole study protocol was approved by the Institutional Ethics Committee of Vivekananda Institute of Medical Sciences (Registration No. ECR/62/Inst/WB/2013) and the methods were carried out in accordance with the approved guidelines. Informed consent was collected from each partici- pant according to the protocol approved by Institutional Ethics Committee of Vivekananda Institute of Medical Sciences. The study also received administrative approval by the S.N. Bose Centre, Kolkata (Ref. No: SNB/PER-2-6001/13-14/1769). Breath and blood samples collection and measurements. Subjects were asked for overnight fasting (~12 hours) before the study. A baseline breath sample was taken in a breath sample collec- tion bag (QUINTRON, USA, SL No.QT00892), whereas blood sample was drawn from each participant in EDTA-vial. Thereafter, participants were instructed to ingest a test meal containing 75 mg U-13C6 labelled D-glucose (CIL-CLM-1396-CTM, Cambridge Isotope Laboratories, Inc., USA) along with the 75 g normal glucose dissolved in 150 mL water. From this starting point, the post-dose breath and blood samples were collected in every 30 minutes interval up to 5 h. Breath bags were designed in such a way that oral-breath first passed into a dead space and endogenously produced end-tidal breath entered into the reservoir bag through a one-way valve. During analysis, the breath samples were taken from the reservoir bags by Quintron air-tight syringe fitted onto the bag. Blood samples were utilized to measure the plasma glucose spectro-photometrically (2300 STAT Plus Glucose Analyzer) and insulin concen- trations by using monoclonal antibody coated immunoassay DIA source INS-EASIA Kits (DIAsource ImmunoAssays, Belgium). Conclusion In conclusion, our study confirms the clinical feasibility of the exhaled breath carbon dioxide isotopes analysis for estimating insulin resistance and thereafter the diagnosis of non-diabetic control, pre-diabetes and type 2 diabetes. When ISI0,120 has been suggested as the most correlated alternative insulin resistance (1/insulin sensitivity) parameter with respect to euglycemic hyperinsulinemic clamp study, our observa- tions demonstrated a strong correlation of ISI0,120 index with breath carbon dioxide isotopes, suggesting a new perspective into the non-invasive evaluation of insulin resistance rather than the traditional inva- sive measurements. Additionally, we first estimated the cut-off values of 1/ISI0,120 for the diagnosis of pre-diabetes and type 2 diabetes, thus making it a potentially robust approach for accurate evaluation of insulin resistance. Overall, our calculated optimal cut-off values of both 1/ISI0,120 and carbon-13 isotopes of breath CO2 also suggest that they may serve as useful methods for early detection and follow-up of individuals who are at high-risk for developing insulin resistance prior to the onset of type 2 diabetes that threaten modern society. Although, many important gaps may remain in understanding the poten- tial link between ISI0,120 index and breath CO2 isotopes in the present study, our results, however, have significant implications in the isotope-specific molecular diagnosis of insulin resistant pre-diabetes and type 2 diabetes with broad clinical applications. Besides, this non-invasive approach for estimation of Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 6 www.nature.com/scientificreports/ Parameters Non-diabetic control (NDC) (n = 31) Pre-diabetes (PD) (n = 38) Type 2 Diabetes (T2D) (n = 47) p values Sex (M/F) 18/13 26/12 31/16 Age (Years) 35.48 ± 6.8 34.55 ± 6.6 38.63 ± 9.8 0.632 Weight (kg) 65.01 ± 6.5 65.93 ± 7.1 65.92 ± 5.6 0.789 BMI (kg/m 2) 24.43 ± 2.7 24.21 ± 2.5 23.93 ± 2 0.807 Fasting Plasma Glucose (mg/dL) 90.1 ± 12.5 107.2 ± 9.1 196 ± 17.6 < 0.001* Fasting Plasma Insulin (μIU/mL) 7.57 ± 3.5 10.51 ± 2.6 22.94 ± 4.1 < 0.001* 2-hr Post-dose Plasma Glucose (mg/dL) 144.3 ± 11.2 182.1 ± 16.1 372.1 ± 26.5 < 0.001* HbA1c (%) 5.26 ± 0.2 6.07 ± 0.2 7.64 ± 0.5 < 0.001* HOMA-IR 1.68 ± 0.76 2.79 ± 0.8 7.92 ± 1.5 < 0.001* 1/QUICKI 3.49 ± 0.08 3.04 ± 0.11 2.79 ± 0.19 < 0.001* 1/ISI 0,120 0.014 ± 0.002 0.023 ± 0.002 0.046 ± 0.012 < 0.001* Table 2. Clinical characteristic of the study subjects. Conclusion Represents statistically significant differences among non-diabetic control (NDC), pre-diabetes (PD) and type 2 diabetes (T2D). M and F stand for male and female, respectively. p < 0.05 (statistically significant). Results are mean ± SD. Table 2. Clinical characteristic of the study subjects. Represents statistically significant differences among non-diabetic control (NDC), pre-diabetes (PD) and type 2 diabetes (T2D). M and F stand for ma and female, respectively. p < 0.05 (statistically significant). Results are mean ± SD. insulin resistance may assist in detecting the pre-diabetes stage of asymptomatic type 2 diabetic subjects in preclinical phase. This point-of-care diagnostic method may also help to overcome the current com- pliance of invasive techniques for screening diabetes mellitus in future days. Therefore, our proposed breath isotopes analysis may be a new method to prevent or treat the deleterious effect of the most common metabolic syndrome in the world. Finally, as this breath analysis approach is safe, simple and non-invasive, it could be an attractive option for large-scale screening purposes in a wide variety of individuals including children, pregnant women and seniors. insulin resistance may assist in detecting the pre-diabetes stage of asymptomatic type 2 diabetic subjects in preclinical phase. This point-of-care diagnostic method may also help to overcome the current com- pliance of invasive techniques for screening diabetes mellitus in future days. Therefore, our proposed breath isotopes analysis may be a new method to prevent or treat the deleterious effect of the most common metabolic syndrome in the world. Finally, as this breath analysis approach is safe, simple and non-invasive, it could be an attractive option for large-scale screening purposes in a wide variety of individuals including children, pregnant women and seniors. Materials and Methods × ( ) / = ( / ) + ( / ) / ( ) , L L L L 1 ISI log mean insulin IU MCR; where log mean insulin IU [ log fasting insulin IU log 2h post dose insulin IU 2] and MCR metabolic clearance rate m mean glucose m [75000 fasting glucose mg L 2h post dose glucose mg L 0 19 weight kg ] 120 and mean glucose [fasting glucose mmol L glucose at 2h mmol L ] 2 5 0 120 = ( ( (µ / ) + ( − (µ / ))/ ( ) , L L [ log fasting insulin IU log 2h post dose insulin IU 2] and MCR metabolic clearance rate = ( ( (µ / ) + ( − (µ / ))/ ( ) = , = + (( ( / )− ( / )) × . × ( ) / L L [ log fasting insulin IU log 2h post dose insulin IU 2] and MCR metabolic clearance rate m mean glucose m [75000 fasting glucose mg L 2h post dose glucose mg L 0 19 weight kg ] 120 and mean glucose ( ) 5 = ( / ) + ( / ) / [fasting glucose mmol L glucose at 2h mmol L ] 2 Statistical Method. Data were usually expressed as mean ± SE. To compare the non-parametric dis- tribution, we applied Mann-Whitney and Kruskal-Wallis tests, whereas the normal distributed data were analyzed by one way analysis of variance (ANOVA). To determine the optimal cut-off points, we utilized receiver operating characteristic curve (ROC)20 analysis. The optimal cut-off value was considered as the cut-off point where we obtained maximum diagnostic sensitivity (true positive rate) and specificity (true negative rate). Data were analyzed by Origin Pro 8.0 (Origin Lab Corporation, USA) and Analyse-it Method Evaluation software (Analyse-it Software Ltd, UK, version 2.30). A two-sided p values < 0.05 was considered as statistically significant. Here, p values have been calculated to check the differences among the three individual groups. Materials and Methods A continuous wave (cw) diode laser (distributed feedback) operating at near infrared ~2.05 μ m was utilized to scan over 20 GHz to record the absorption features of 12C16O16O and 13C16O16O isotopes at the wave numbers of 4874.448 cm−1 and 4874.086 cm−1 respectively, in the (2,00,1)←(0,00,0) vibrational combination band of the CO2 molecule. The laser light moves back and forth inside the optical cavity and provides an effective optical path length of ~3 km. The measure- ment cell of ICOS instrument was maintained at a temperature of 46 °C by a resistive heater and feed- back control system. The typical pressure of the measurement cell was ~30 Torr which was achieved by a diaphragm pump. The real-time ro-vibrational spectral features were fitted with Voigt profiles and the absolute concentrations of 12C16O16O and 13C16O16O isotopes in exhaled breath samples were determined using Beer’s law. The ICOS system was calibrated by the three standard certified gases of 5% CO2 in air with different δ 13C‰ values (δ 13C‰ = − 22.8‰, − 13.22‰ and − 7.3‰, Cambridge Isotope Laboratory, CIL, USA). The accuracy and precision of the δ DOB 13C‰ using the ICOS system were in the range of 98–99% and 0.15‰, respectively. Insulin resistance index. HOMA-IR, QUICKI and ISI0,120 were calculated from the blood glucose and insulin levels in the fasting (0 min) and post-dose (120 min) conditions. Materials and Methods Breath samples were analyzed by a high-resolution isotopic CO2 integrated cavity output spectrometer as described below. HbA1c (%) was measured by utilizing an HbA1c analyzer (D-10, USA) in HPLC method. 7 Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 www.nature.com/scientificreports/ Integrated cavity output spectrometer (ICOS) for breath analysis. We have employed a high-precision carbon dioxide isotope analyzer (CCIA 36-EP, Los Gatos research, USA) exploiting integrated cavity output spectroscopy (ICOS) to analyze the isotopic compositions of CO2 in exhaled breath samples. The technical details and its working principle have been described elsewhere17–19. In brief, the ICOS system consists of two high reflectivity mirrors (R ~ 99.98%) placed at the two ends of a high-finesse optical cavity (59 cm long). A continuous wave (cw) diode laser (distributed feedback) operating at near infrared ~2.05 μ m was utilized to scan over 20 GHz to record the absorption features of 12C16O16O and 13C16O16O isotopes at the wave numbers of 4874.448 cm−1 and 4874.086 cm−1 respectively, in the (2,00,1)←(0,00,0) vibrational combination band of the CO2 molecule. The laser light moves back and forth inside the optical cavity and provides an effective optical path length of ~3 km. The measure- ment cell of ICOS instrument was maintained at a temperature of 46 °C by a resistive heater and feed- back control system. The typical pressure of the measurement cell was ~30 Torr which was achieved by a diaphragm pump. The real-time ro-vibrational spectral features were fitted with Voigt profiles and the absolute concentrations of 12C16O16O and 13C16O16O isotopes in exhaled breath samples were determined using Beer’s law. The ICOS system was calibrated by the three standard certified gases of 5% CO2 in air with different δ 13C‰ values (δ 13C‰ = − 22.8‰, − 13.22‰ and − 7.3‰, Cambridge Isotope Laboratory, CIL, USA). The accuracy and precision of the δ DOB 13C‰ using the ICOS system were in the range of 98–99% and 0.15‰, respectively. Integrated cavity output spectrometer (ICOS) for breath analysis. We have employed a high-precision carbon dioxide isotope analyzer (CCIA 36-EP, Los Gatos research, USA) exploiting integrated cavity output spectroscopy (ICOS) to analyze the isotopic compositions of CO2 in exhaled breath samples. The technical details and its working principle have been described elsewhere17–19. In brief, the ICOS system consists of two high reflectivity mirrors (R ~ 99.98%) placed at the two ends of a high-finesse optical cavity (59 cm long). References 1. Kahn, S. E., Hull, R. L. & Utzschneider, K. M. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 444, 860–846 (2006). 2. Lowell, B. B. & Shulman, G. I. Mitochondrial dysfunction and type 2 diabetes. Science 307, 384–387 (2005). 3. Katz, A. et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab 85, 2402–2410 (2000). 4. Muniyappa, R., Lee, S., Chen, H. & Quon, M. J. Current approaches for assessing insulin sensitivity and resistance in vivo: advantages, limitations, and appropriate usage. Am J Physiol Endocrinol Metab 294, E15–E26 (2008). 5. Lewanczuk, R. Z., Paty, B. W. & Toth, E. L. Comparison of the [13C] glucose breath test to the hyperinsulinemic-euglycemic clamp when determining insulin resistance. Diabetes Care 27, 441–447 (2004). 6. Ibarra-Pastrana, E., Plata, M. D. C. C., Alvarez, G. & Valencia, M. E. Estimation of insulin resistance in Maxican adults by the [13C] glucose breath test corrected for total endogenous CO2 production. Int. J. Endocrinol. 2012, 1–7 (2012). 7. Mizrahi, M., Lalazar, G., Adar, T., Raz, I. & Ilan, Y. Assessment of insulin resistance by a 13C glucose breath test: a new tool for early diagnosis and follow-up of high-risk patients. Nutrition Journal 9, 1–9 (2010). References 1. Kahn, S. E., Hull, R. L. & Utzschneider, K. M. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 444 860–846 (2006). 2 L ll B B & Sh l G I Mit h d i l d f ti d t 2 di b t S i 307 384 387 (2005) 2. Lowell, B. B. & Shulman, G. I. Mitochondrial dysfunction and type 2 diabetes. Science 307, 384–387 (2005). 3. Katz, A. et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans J Clin Endocrinol Metab 85, 2402–2410 (2000). g g 7. Mizrahi, M., Lalazar, G., Adar, T., Raz, I. & Ilan, Y. Assessment of insulin resistance by a 13C glucose breath test: a new tool fo early diagnosis and follow-up of high-risk patients. Nutrition Journal 9, 1–9 (2010). ( ) 2. Lowell, B. B. & Shulman, G. I. Mitochondrial dysfunction and type 2 diabetes. Science 307, 384–387 (2005). p g ( ) 6. Ibarra-Pastrana, E., Plata, M. D. C. C., Alvarez, G. & Valencia, M. E. Estimation of insulin resistance in Maxican adults by the [13C] glucose breath test corrected for total endogenous CO2 production. Int. J. Endocrinol. 2012, 1–7 (2012). Materials and Methods Insulin resistance indexes obtained from insulin and glucose concentrations during 2 h-OGTT study and these were measured by utilizing the following equations9: − = ( / ) × (µ / ) ( ) L HOMA IR Fasting glucose mg dL Fasting insulin IU 405 3 / = ( ( / ) × (µ / )) ( ) L 1 QUICKI log Fasting glucose mg dL Fasting insulin IU 4 − = ( / ) × (µ / ) ( ) L HOMA IR Fasting glucose mg dL Fasting insulin IU 405 3 / = ( ( / ) × (µ / )) ( ) L 1 QUICKI log Fasting glucose mg dL Fasting insulin IU 4 ( ) 3 ( ) 4 / = ( (µ / ))/ ( (µ / )) = ( ( (µ / ) + ( − (µ / ))/ ( ) , L L L L 1 ISI log mean insulin IU MCR; where log mean insulin IU [ log fasting insulin IU log 2h post dose insulin IU 2] and MCR metabolic clearance rate 0 120 / = ( (µ / ))/ ( (µ / )) = ( ( (µ / ) + ( − (µ / ))/ ( ) = , = + (( ( / ) ( / )) , L L L L 1 ISI log mean insulin IU MCR; where log mean insulin IU [ log fasting insulin IU log 2h post dose insulin IU 2] and MCR metabolic clearance rate m mean glucose m [75000 f ti l L 2h t d l L 0 120 / = ( (µ / ))/ ( (µ / )) = ( ( (µ / ) + ( − (µ / ))/ ( ) = , = + (( ( / )− ( / )) × . 4. Muniyappa, R., Lee, S., Chen, H. & Quon, M. J. Current approaches for assessing insulin sensitivity and resistance in vivo advantages, limitations, and appropriate usage. Am J Physiol Endocrinol Metab 294, E15–E26 (2008). g pp p g y 5. Lewanczuk, R. Z., Paty, B. W. & Toth, E. L. Comparison of the [13C] glucose breath test to the hyperinsulinemic-euglycemic clamp when determining insulin resistance. Diabetes Care 27, 441–447 (2004). 6 Ib P t E Pl t M D C C Al G & V l i M E E ti ti f i li i t i M i d lt b th 3. Katz, A. et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab 85, 2402–2410 (2000). 4. Muniyappa, R., Lee, S., Chen, H. & Quon, M. J. Current approaches for assessing insulin sensitivity and resistance in vivo: advantages, limitations, and appropriate usage. Am J Physiol Endocrinol Metab 294, E15–E26 (2008). 5. Lewanczuk, R. Z., Paty, B. W. & Toth, E. L. Comparison of the [13C] glucose breath test to the hyperinsulinemic-euglycemic clamp when determining insulin resistance. Diabetes Care 27, 441–447 (2004). 6. Ibarra-Pastrana, E., Plata, M. D. C. C., Alvarez, G. & Valencia, M. E. Estimation of insulin resistance in Maxican adults by the [13C] glucose breath test corrected for total endogenous CO2 production. Int. J. Endocrinol. 2012, 1–7 (2012). 7. Mizrahi, M., Lalazar, G., Adar, T., Raz, I. & Ilan, Y. Assessment of insulin resistance by a 13C glucose breath test: a new tool for References Use of laser spectroscopy to measure the 13C/12C and 18O/16O composition of carbonate minerals. Anal. Chem. 83, 2220–2226 (2011). 8. Ghosh, C. et al. Oxygen-18 isotope of breath CO2 linking to erythrocytes carbonic anhydrase activity: a biomarker for pre diabetes and type 2 diabetes. Sci. Rep. 5, 8137 (2015). doi: 10.1038/srep08137. 18. Ghosh, C. et al. Oxygen-18 isotope of breath CO2 linking to erythrocytes carbonic anhydrase activity: a biomarker for pre diabetes and type 2 diabetes. Sci. Rep. 5, 8137 (2015). doi: 10.1038/srep08137. y p 9. Banik, G. D. et al. Diagnosis of small intestinal bacterial overgrowth in irritable bowel syndrome patients using high-precision stable 13CO2/12CO2 isotope ratios in exhaled breath. J. Anal. At. Spectrom. 29, 1918–1924 (2014). 19. Banik, G. D. et al. Diagnosis of small intestinal bacterial overgrowth in irritable bowel syndrome patients usin stable 13CO2/12CO2 isotope ratios in exhaled breath. J. Anal. At. Spectrom. 29, 1918–1924 (2014). 20. Swets, J. A. Measuring the accuracy of diagnostic systems. Science 240, 1285–1293 (1988). stable 13CO2/12CO2 isotope ratios in exhaled breath. J. Anal. At. Spectrom. 29, 1918–1924 (2014 20. Swets, J. A. Measuring the accuracy of diagnostic systems. Science 240, 1285–1293 (1988). 2 2 p p 20. Swets, J. A. Measuring the accuracy of diagnostic systems. Science 240, 1285–1293 (1988). 2 2 p p 20. Swets, J. A. Measuring the accuracy of diagnostic systems. Science 240, 1285–1293 (1988). 20. Swets, J. A. Measuring the accuracy of diagnostic systems. Science 240, 1285–1293 (1988). References Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 8 www.nature.com/scientificreports/ 8. Gutt, M. et al. Validation of the insulin sensitivity index (ISI0,120): comparison with other measures. Diabetes Res Clin Prac 177–184 (2000). 9. Qureshi, K., Clements, R. H., Saeed, F. & Abrams, G. A. Comparative evaluation of whole body and hepatic insulin resistance using indices from oral glucose tolerance test in morbidly obese subjects with nonalcoholic fatty liver disease. Journal of Obesity 2010, 1–7 (2010). using indices from oral glucose tolerance test in morbidly obese subjects with nonalcoholic fatty liver disease. Journal of Obesity 2010, 1–7 (2010). 10 Dix D Cohen P Barzegar S & Striefler M The oral glucose tolerance test: A comparison of the time points on the basis of 2010, 1 7 (2010). 10. Dix D., Cohen P., Barzegar S. & Striefler, M. The oral glucose tolerance test: A comparison of the time points on the basis of limit values normal dispersion and reproducibility Acta Diabetol Lat 18 21 26 (1981) ( ) 10. Dix D., Cohen P., Barzegar S. & Striefler, M. The oral glucose tolerance test: A comparison of the time points on the ba limit values, normal dispersion, and reproducibility. Acta Diabetol. Lat. 18, 21–26 (1981). 11. Schwartz, M. W. & Kahn, S. E. Diabetes: insulin resistance and obesity. Nature 402, 860–861 (1999). 11. Schwartz, M. W. & Kahn, S. E. Diabetes: insulin resistance and obesity. Nature 402, 860–861 (1999). ll T C & h d b f O A d ll b C ( ) 12. Wallace, T. M., Levy, J. C. & Matthews, D. R. Use and abuse of HOMA modelling. Diabetes Care 27, 1487–1495 (2004). 13. Matsuda, M. & Defronzo, R. A. Insulin sensitivity index obtained from oral glucose tolerance testing: comparison wit euglycemic insulin camp. Diabetes Care 22, 1462–1470 (1999).t 4. Shapiro, E. T. et al. Insulin secretion and clearance: comparison after oral and intravenous glucose. Diabetes 36, 1365–1371 (1987). 5. Grzybowski, M. & Younger, J. G. Statistical methodology: III. receiver operating characteristic (ROC) curves. Academic Emergency Medicine 4, 818–826 (1997). 16. American Diabetes Association. Standards of medical care in diabetes-2014. Diabetes care 37, Suppl 1:S14–S80 (2014). k l f l h 13 12 d 18 16 16. American Diabetes Association. Standards of medical care in 7. Barker, S. L., Dipple, G. M., Dong, F. & Baer, D. S. Author Contributions M.P. provided the funding; M.P. and C.G. designed the whole study and provided the conception; M.P., P.M. and S.G. supervised the overall study; C.G. collected and analysed the samples; All authors drafted the manuscript with critical revision. Acknowledgementsh g This work was supported by the Department of Biotechnology (DBT), Government of India for the award of “Rapid Grant for Young Investigators” (Ref. No. BT/PR6683/GBD/27/477/2012). The author (C. Ghosh) acknowledges S.N. Bose Centre for PhD fellowship. We are also extremely grateful to all volunteers for participating in this study. Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 Additional Informationi Competing financial interests: The authors declare no competing financial interests. Competing financial interests: The authors declare no competing financial interests. How to cite this article: Ghosh, C. et al. Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre-diabetes and type 2 diabetes. Sci. Rep. 5, 11959; doi: 10.1038/srep11959 (2015). How to cite this article: Ghosh, C. et al. Insulin sensitivity index (ISI0, 120) potentially linked to carbon isotopes of breath CO2 for pre-diabetes and type 2 diabetes. Sci. Rep. 5, 11959; doi: 10.1038/srep11959 (2015). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Com- mons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports \| 5:11959 \| DOI: 10.1038/srep11959 9
https://openalex.org/W2773717538	https://mau.diva-portal.org/smash/get/diva2:1419491/FULLTEXT01.pdf	English	null	Performance of Smoothing Methods for Reconstructing NDVI Time-Series and Estimating Vegetation Phenology from MODIS Data	Remote sensing	2,017	cc-by	10,920	Zhanzhang Cai 1,* ID , Per Jönsson 2, Hongxiao Jin 1 ID and Lars Eklundh 1 Zhanzhang Cai 1,* ID , Per Jönsson 2, Hongxiao Jin 1 ID and Lars Eklundh 1 1 Department of Physical Geography and Ecosystem Science, Lund University, Sölvegatan 12, SE-223 62 Lund, Sweden; hongxiao.jin@nateko.lu.se (H.J.); lars.eklundh@natkeo.lu.se (L.E.) 2 Department of Materials Science and Applied Mathematics, Malmö University, SE-205 06 Malmö, Sweden; per.jonsson@mah.se * Correspondence: zhanzhang.cai@nateko.lu.se; Tel.: +46-46-222-31-52 1 Department of Physical Geography and Ecosystem Science, Lund University, Sölvegatan 12, SE-223 62 Lund, Sweden; hongxiao.jin@nateko.lu.se (H.J.); lars.eklundh@natkeo.lu.se (L.E.) 2 Department of Materials Science and Applied Mathematics, Malmö University, SE-205 06 Malmö, Sweden; per.jonsson@mah.se Sweden; hongxiao.jin@nateko.lu.se (H.J.); lars.eklundh@natkeo.lu.se (L.E.) 2 Department of Materials Science and Applied Mathematics, Malmö University, SE-205 06 Malm per.jonsson@mah.se * Correspondence: zhanzhang.cai@nateko.lu.se; Tel.: +46-46-222-31-52 Received: 8 September 2017; Accepted: 4 December 2017; Published: 7 December 2017 * Correspondence: zhanzhang.cai@nateko.lu.se; Tel.: +46-46-222-31-52 * Correspondence: zhanzhang.cai@nateko.lu.se; Tel.: +46-46-222-31-52 Received: 8 September 2017; Accepted: 4 December 2017; Published: 7 December 2017 Abstract: Many time-series smoothing methods can be used for reducing noise and extracting plant phenological parameters from remotely-sensed data, but there is still no conclusive evidence in favor of one method over others. Here we use moderate-resolution imaging spectroradiometer (MODIS) derived normalized difference vegetation index (NDVI) to investigate ﬁve smoothing methods: Savitzky-Golay ﬁtting (SG), locally weighted regression scatterplot smoothing (LO), spline smoothing (SP), asymmetric Gaussian function ﬁtting (AG), and double logistic function ﬁtting (DL). We use ground tower measured NDVI (10 sites) and gross primary productivity (GPP, 4 sites) to evaluate the smoothed satellite-derived NDVI time-series, and elevation data to evaluate phenology parameters derived from smoothed NDVI. The results indicate that all smoothing methods can reduce noise and improve signal quality, but that no single method always performs better than others. Overall, the local ﬁltering methods (SG and LO) can generate very accurate results if smoothing parameters are optimally calibrated. If local calibration cannot be performed, cross validation is a way to automatically determine the smoothing parameter. However, this method may in some cases generate poor ﬁts, and when calibration is not possible the function ﬁtting methods (AG and DL) provide the most robust description of the seasonal dynamics. Keywords: normalized difference vegetation index (NDVI); smoothing methods; gross primary production (GPP); phenology; TIMESAT; MODIS remote sensing Keywords: normalized difference vegetation index (NDVI); smoothing methods; gross primary production (GPP); phenology; TIMESAT; MODIS 1. Introduction Accurate time-series of vegetation indices from satellite remote sensing are vital for long term vegetation monitoring, especially for studies of vegetation phenology [1] and carbon balance [2,3]. However, the signals received by the satellite sensors are affected by noise due to geometric misregistration, anisotropic reflectance effects, electronic errors, artifacts due to data resampling, atmosphere, and clouds [4]. In order to reconstruct the vegetation seasonal growth course from noisy satellite signals a multitude of time-series processing methods have been used. Some early methods are e.g., the best index slope extraction [5], the repeated moving-window median filter [6–8], harmonic functions [9] and median filters [10]. More recently and widely used methods are e.g., Savitzky-Golay filtering [11], least-squares fits to asymmetric Gaussian functions [12] and double logistic functions [13–16], and variations of spline smoothing [17,18] including the Whittaker smoother [19], and wavelet transforms [20–22]. The processing methods can be divided into two broad categories: (1) methods that enable precise capturing of short-term variations during the growing season, generally by employing different types of statistical filters [5,6,10]; and (2), methods that fit mathematical functions (sinusoidal, logistic, Gaussian etc. [12–16,20–22]) to part of Remote Sens. 2017, 9, 1271; doi:10.3390/rs9121271 www.mdpi.com/journal/remotesensing 2 of 18 Remote Sens. 2017, 9, 1271 a season, the full season, or even a sequence of seasons. We can in a loose sense label these as “local” and “global”, respectively, denoting their flexibility to match to details in the time-series or to generalize across the full growing season. Spline methods [17–19] have the ability to take on both local and global character depending on a smoothing parameter. Due to the difficulty of existing physiological process-based phenology models, e.g., various budburst models using chilling or forcing temperature requirements, in accurately simulating phenology from climate data [23], time-series of smoothed vegetation indices measured by satellite are an important data resource for estimating phenology [24,25]. Many studies have employed remote sensing time-series to extract and monitor the timing of plant phenological events [13,15,26–29]. Since these studies have used different smoothing methods to estimate phenology, it is hard to draw any firm conclusions about which methods are the most suitable. White, et al. [30] compared 10 different phenology retrieval methods applied to an Advanced Very High Resolution Radiometer (AVHRR) NDVI dataset in North America and concluded that although smoothing methods applied to satellite NDVI could generally capture vegetation phenology, no method was universally better than the others. 1. Introduction Atkinson, et al. [31] and Chen, et al. [32] concluded that the performance of smoothing methods in estimating vegetation phenology varies spatially and temporally, and that, because of bias and random errors due to cloud influence, no single method exhibited a consistently superior performance. Cong, et al. [33] argued that the reason for not finding a single best method could be different definitions of phenology parameters. Kandasamy and Fernandes [34] highlighted that the performance of different smoothing methods depends on both land surface condition and the clear sky identification approach adopted. Many studies have investigated the performance of smoothing methods in reconstructing consecutive time-series of vegetation indices, leaf area index and fractional absorbed photosynthetically active radiation (fAPAR) [31,35–39]. Besides vegetation growth time-series from satellite remote sensing, ground-based optical measurements, e.g., Phenocam [40] and multispectral sensors [41,42], have been successfully applied in vegetation phenology studies for different biomes [29] and for linking to satellite remote sensing measurements [43,44]. Also scene-based methods, including geostatistical methods, have been used to investigate the effects of time-series smoothing [19]. However, despite the multitude of studies there is a lack of agreement regarding the choice of time-series smoothing methods. To investigate the usefulness of time-series smoothing methods it is appropriate to test the ability of the methods to reconstruct the seasonal variation of the temporal signal. This can best be done by directly relating it to ground measurements. Today, growing networks of ground measured reflectance measurements exist, suitable for process studies and validation of remotely sensed vegetation index data [45,46]. Furthermore, some indirect data may provide useful time-series records for establishing the ability of different time-series methods to portray the vegetation dynamics. One of these is gross primary productivity (GPP) estimated from carbon ﬂux measurements. This is based on the assumption that GPP is related to the satellite derived NDVI. The relationship between NDVI and fAPAR has been investigated by Asrar, et al. [47] and Sellers, et al. [48], and it has been used as major input in models to estimate GPP [2,3,49]. This relationship implies that GPP is a useful reference for evaluating the abilities of different methods for smoothing satellite NDVI. Also spatial data may be useful for investigating the different methods. It is well known that topography inﬂuences vegetation phenology due to adiabatic temperature difference with height. 2.1. MODIS NDVI Data and Ground Reference Data The MODIS/Terra MOD09Q1 land surface reflectance dataset of 8-day interval and 250 m spatial resolution [54] was used as an example in this study. We used the binary MODIS quality flags to set weights for each observation. NDVI time-series from 2000 to 2014 were calculated using near-infrared (NIR) and red reflectance from MOD09Q1 for the 10 field sites and the test area with elevation reference data. For each field site, the reflectance data of single pixels were used rather than the mean value of several neighboring pixels, so that the performance of smoothing methods on single pixel time-series data could be fully evaluated. For ground reference, NDVI data from sensors mounted on masts at 10 sites (Figure S1 in Supplementary Material), located in Scandinavia, Greenland and Africa, were used [42,55]. These sites cover different environments: arctic, sub-arctic, deciduous forest, evergreen forest, grassland, savanna, and clear-cut forest (Table 1) and the sites were established in homogeneous areas in terms of vegetation coverage and topography. The sensor pairs were calibrated following Jin and Eklundh [56]. Ground-based NDVI was computed from daily noontime reflectance in red and NIR bands, and it agreed very well with MODIS data during the growing season [42]. For further information about the field measurements, see Eklundh, et al. [42]. To agree with the time step of MOD09Q1 NDVI, daily NDVI were converted to 8-day NDVI values by median filtering. Due to the high latitude of sites 1–7, snow cover in winter disturbed both spectral NDVI sensors on the ground and MODIS. For this reason, we only compared them during the non-winter period: April–October at the Hyytiälä and Norunda sites and June–September at the Abisko and Zackenberg sites [57–60]. Table 1. Ground-based spectral observation sites and ﬂux tower sites (more information available in Table S1). Site Latitude Longitude Elevation a.s.l. 1. Introduction At high elevation of temperate and boreal regions, slow thermal accumulation leads to delayed canopy development [50,51] and generally shorter growing seasons than those at low elevation. Based on this concept, a number of studies have used elevation gradients to spatially investigate phenology from satellite data e.g., [15,52,53]. Given the difficulty in drawing firm conclusions about different smoothing methods from previous studies, this study aims to address the following questions: (1) is it possible to define one smoothing method that is always best suited, irrespective of situation; (2) if not, what capability do the different methods have for smoothing data in different situations. To achieve this aim, we evaluate the performance 3 of 18 Remote Sens. 2017, 9, 1271 of smoothing algorithms for representing seasonal vegetation growth with 8-day MODIS NDVI data by employing a variety of reference data sets. We systematically analyze five different smoothing methods: Savitzky-Golay filtering (SG), locally weighted regression scatterplot smoothing (LO), spline smoothing (SP), least-squares fitting to asymmetric Gaussian functions (AG), and least-squares fitting to double logistic functions (DL), against three different reference data sets: ground measured NDVI from ten locations, ground measured GPP from four sites, and elevation data from an uphill area with a vegetation gradient. Altogether these analyses will support our aim and allow better understanding of how to choose suitable time-series smoothing methods for remote sensing phenology studies. * Ground-based spectral observation sites with ﬂux measurements. * Ground-based spectral observation sites with ﬂux measurements. 2.1. MODIS NDVI Data and Ground Reference Data (m) Spectral Sensor Footprint Area (m2) Biome Year of Spectral (CO2 Flux) Data Abisko Delta 68.36 18.80 340 2000 Deciduous forest 2009–2011 Abisko Stordalen 68.36 19.05 360 2000 Sub-arctic mire 2009–2011 Fäjemyr * 56.27 13.55 140 3000 Permanent wetlands 2009–2011 (2005–2009) Zackenberg * 74.48 −20.56 1300 10,000 Arctic mire 2008–2010 (2008–2009) Hyytiälä * 61.85 24.29 170 7000 Coniferous forest 2009–2012 (2005–2012) Norunda Clear-cut 60.09 17.47 70 300 Coniferous forest (Clear-cut in 2009) 2011–2014 Norunda Forest * 60.09 17.48 70 1500 Coniferous forest 2010–2012 (2005–2007) Sudan Barah 13.82 30.36 500 10,000 Grassland 2008–2010 Sudan Dilling 11.86 29.72 720 10,000 Grassland 2008–2009 Sudan Demokeya 13.27 30.49 530 10,000 Cropland 2008–2009 * G d b d t l b ti it ith ﬂ t le 1. Ground-based spectral observation sites and ﬂux tower sites (more information available in Table S1). Table 1. Ground-based spectral observation sites and ﬂux tower sites (more information a Remote Sens. 2017, 9, 1271 4 of 18 We used ﬂux-tower measured GPP from sites Fäjemyr, Zackenberg, Hyytiälä and Norunda Forest (Table 1) to evaluate the performance of the time-series smoothing methods for estimating vegetation temporal dynamics from the satellite images. Weekly (8-day) GPP data for these sites were acquired from Fluxnet (http://www.eosdis.ornl.gov/FLUXNET). Besides using ground reference data from point locations, we used spatial data from the Ammar region, located in Västerbotten, Sweden, and having a signiﬁcant elevation gradient (approximately 1980 m). Elevation data (50 m spatial resolution) were obtained from the Swedish mapping, cadastral and land registration authority, Lantmäteriet (Figure 1). It was resampled to 250 m and matched to MOD09Q1 pixels. The reason for selecting the Ammar region is that the Ammar region represents a gradient of changing elevation from low to high, and vegetation varying from dense to sparse with elevation. In addition, the region has stable land cover (mainly forest and natural vegetation classes) with few logging activities or pest disturbances reported during the studied period. Figure 1. Elevation of the Ammar region. (Source: National Land Survey of Sweden, Dnr: I2014/00579). Figure 1. Elevation of the Ammar region. (Source: National Land Survey of Sweden, Dnr: I2014/00579). 2.2. Time-Series Smoothing Methods Since it is beyond our capacity to investigate all possible time-series smoothing methods we have selected some common methods from the two broad categories of local vs global methods. From the local category we have chosen adaptive Savitzky-Golay ﬁltering (SG) [11,61] and adaptive LOESS ﬁltering (LO) [62], and from the global category we have chosen least-squares ﬁts to asymmetric Gaussian functions (AG) [12] and double logistic functions (DL) [13,14]. In addition, we test spline smoothing (SP) [63,64] that has capacity to act both as a local and a global smoother. SG, AG and DL are available in the current TIMESAT version 3.3 [16,65]. LO and SP were integrated into a development version of TIMESAT for this study. The adaption to the upper envelope of the original NDVI time-series was applied to all smoothing methods for reducing negative bias of satellite NDVI due to clouds or poor atmospheric conditions [11,16]. Upper-envelope adaptation makes the smoothed function approach the upper NDVI envelope by iteratively generating new curves with data weights that have lower values for the data points below the previous smoothed curve [11,12]. This upper-envelope adaptation is in TIMESAT determined by two parameters: the number of iterations u for which weights are being modiﬁed, and the adaptation strength a, which ﬁne-tunes the strength of the upper-envelope weighting in each iteration (Figure 2). 5 of 18 Remote Sens. 2017, 9, 1271 Figure 2. Examples of the effects of modifying parameters for upper-envelope adaptation in TIMESAT using double logistic (DL) ﬁts to data. u: the number of iterations, and a: is the adaptation strength. Figure 2. Examples of the effects of modifying parameters for upper-envelope adaptation in TIMESAT using double logistic (DL) ﬁts to data. u: the number of iterations, and a: is the adaptation strength. 2.2.1. Polynomial Local Filtering (SG and LO) The principles of SG and LO are similar in that both methods replace each data value by a combination of adjacent values in a window using a least-squares second-order polynomial ﬁt (Figures 3 and 4). The size of the ﬁltering window determines the degree of smoothing. If rapid variations are encountered in the window, both methods are in TIMESAT modiﬁed to automatically decrease the window size. This enables quite sudden variations in the NDVI trajectory to be modelled. In this study, the deﬁnition of the ﬁltering window size r in both smoothing methods is the same as in TIMESAT, i.e., r = 2q + 1, where q is the number of time steps from right/left to the middle point m. Figure 3. Examples of Savitzky-Golay (SG) smoothing with different window sizes, r = 2q + 1. Figure 4. Examples of LOESS (LO) smoothing with different window sizes, r = 2q + 1. Figure 3. Examples of Savitzky-Golay (SG) smoothing with different window sizes, r = 2q + 1. Figure 3. Examples of Savitzky-Golay (SG) smoothing with different window sizes, r = 2q + 1. Figure 4. Examples of LOESS (LO) smoothing with different window sizes, r = 2q + 1. Remote Sens. 2017, 9, 1271 6 of 18 The difference between SG and LO is in the weights of each data point in the ﬁltering window. In SG, all points of the ﬁtting window have weight 1.0, whereas in LO weight wi of point i decreases outwards from the middle point m of the ﬁltering window wi = 1 − \|m −i\| r 3!3 (1) (1) If the same window size is applied to both methods, the different weighting systems cause LO to have lower total data weight than SG. In other words, LO has to have a larger window size than SG to result in the same data smoothing effect as SG. A beneﬁt of LO is that it prevents some of the sharp peaks often encountered in SG ﬁltering. In this study, we tested varying q in SG from 2 to 7, and equivalently varying q in LO from 4 to 12 (Table 2). Table 2. Number of parameters of each smoothing method. Altogether, 987 simulations were made for each test site. Table 2. Number of parameters of each smoothing method. Altogether, 987 simulations were made for each test site. 2.2.1. Polynomial Local Filtering (SG and LO) Smoothing Method Upper-Envelope Iterations (u) Upper-Envelope Strength (a) Smoothing Parameter (q/log(p)) Stiffness Strength (l) Total Number of Settings SG 1–3 1–10 2–7 - 126 LO 1–3 1–10 4–12 - 189 SP 1–3 1–10 (−1)–1 0.1–0.9 630 DL 1–3 1–10 - - 21 AG 1–3 1–10 - - 21 Note: u: number of iterations of upper-envelope, a: adaption strength of upper-envelope iterations (only valid for u > 1), q: half window size, p: smoothing parameter, l: adaption strength of spline stiffness. Note: u: number of iterations of upper-envelope, a: adaption strength of upper-envelope iterations (only valid for u > 1), q: half window size, p: smoothing parameter, l: adaption strength of spline stiffness. te: u: number of iterations of upper-envelope, a: adaption strength of upper-envelope iterations (only valid for > 1), q: half window size, p: smoothing parameter, l: adaption strength of spline stiffness. 2.2.2. Spline Smoothing (SP) We adapted the Fortran code by Woltring [64] to achieve spline smoothing based on the description in Craven and Wahba [63]. The Whittaker smoother [66,67] can be considered as a special case of spline smoother [68]. Spline smoothing is employed to ﬁnd a spline function Sp(t) that minimizes the value of a criterion function Cp for a selected smoothing parameter p ≥0, Cp = n ∑ i=1 wi yi −Sp(ti) 2 + vip Z S ′′ p(t) 2dt (2) (2) where ti, i = 1, . . . , n is a regularly spaced time grid and y(ti) ≡yi the corresponding NDVI values. Each point in the time-series is associated with a weight w(ti) ≡wi. v is a vector containing an adaptive stiffness weight for each i. The smoothing parameter p controls the shape of the spline, varying from an exactly interpolating spline (p = 0) to a straight line ( p →∞) (Figure 5). where ti, i = 1, . . . , n is a regularly spaced time grid and y(ti) ≡yi the corresponding NDVI values. Each point in the time-series is associated with a weight w(ti) ≡wi. v is a vector containing an adaptive stiffness weight for each i. The smoothing parameter p controls the shape of the spline, varying from an exactly interpolating spline (p = 0) to a straight line ( p →∞) (Figure 5). Instead of manually selecting suitable p, an alternative automatic approach is to use generalized Cross-Validation (CV), which was described by Craven and Wahba [63] and implemented in the Fortran code based on the improvement by Woltring [64]. In this study, we tested SP with CV (SPCV) and without CV, respectively. We also implemented CV in our code for the SG and LO methods to automatically choose the q parameter and to be able to compare the effect of CV on these methods with the effect on SP. The range of p in SP was limited to produce local ﬁts, which means that we forced the performance of SP to be fairly close to SG or LO. In this case p was set to the range 0.1–10 with exponential intervals, i.e., 0.1, 0.25, 0.63, 1.58, 3.98 and 10 (Table 2). 7 of 18 Remote Sens. 2017, 9, 1271 Figure 5. Examples of spline function smoothers (SP) with different smoothing parameters (p) and cross validation (SPCV). Figure 5. 2.2.2. Spline Smoothing (SP) A large value of p gives a stiff spline, and a low value of p gives a spline that is ﬂexible. However, a large value of p may result in low absolute values of the derivative in the spring and the autumn, resulting in poor ﬁts during these seasons. The idea with the adaptive method is to give each point i a stiffness by multiplying the global stiffness parameter p with an adaptation factor vi. The adaptation factor vi should be such that it is close to 1 for points in the summer/winter and smaller, i.e., 0.1, at points in the spring/autumn, where the season changes rapidly (Figure 6), see also Lamm [69]. In this study, l was tested at values 0.1, 0.3, 0.5, 0.7 and 0.9 (Table 2). While iterating to adapt to the upper-envelope in SPCV, the p values calculated from the ﬁrst iteration were locked and applied in the remaining iterations. Figure 6. Examples of spline functions with cross validation (SPCV) and different stiffness strength l. Figure 6. Examples of spline functions with cross validation (SPCV) and different stiffness strength l. 2.2.2. Spline Smoothing (SP) Examples of spline function smoothers (SP) with different smoothing parameters (p) and cross validation (SPCV). Differently from the upper-envelope iterations in the other smoothing methods, the upper- envelope iterations in SP and SPCV also updates each vi based on the formula: vi = 1 −l ∥Sq p(ti)∥ ∥Sp∥ (3) (3) where ∥Sp∥is the amplitude of all Sp, and ∥Sq p(ti)∥is the amplitude of Sp in the window from ti −q to ti + q, in which q was set to 5, i.e., a quarter of one year’s data; l is the stiffness adaptation strength, which is in the range from 0 to close to 1, where 0 will give no local adaptation and close to 1 will give very high local adaptation. The adaptation factor vi was integrated into SP to locally adjust the curve shape. In the original code spline stiffness is governed by p. A large value of p gives a stiff spline, and a low value of p gives a spline that is ﬂexible. However, a large value of p may result in low absolute values of the derivative in the spring and the autumn, resulting in poor ﬁts during these seasons. The idea with the adaptive method is to give each point i a stiffness by multiplying the global stiffness parameter p with an adaptation factor vi. The adaptation factor vi should be such that it is close to 1 for points in the summer/winter and smaller, i.e., 0.1, at points in the spring/autumn, where the season changes rapidly (Figure 6), see also Lamm [69]. In this study, l was tested at values 0.1, 0.3, 0.5, 0.7 and 0.9 (Table 2). While iterating to adapt to the upper-envelope in SPCV, the p values calculated from the ﬁrst iteration were locked and applied in the remaining iterations. where ∥Sp∥is the amplitude of all Sp, and ∥Sq p(ti)∥is the amplitude of Sp in the window from ti −q to ti + q, in which q was set to 5, i.e., a quarter of one year’s data; l is the stiffness adaptation strength, which is in the range from 0 to close to 1, where 0 will give no local adaptation and close to 1 will give very high local adaptation. The adaptation factor vi was integrated into SP to locally adjust the curve shape. In the original code spline stiffness is governed by p. 2.2.3. Least-Squares Fits to Model Functions (AG and DL) Two model function-ﬁtting methods, AG and DL, employ least-squares ﬁtting to full or half growing seasons [12,14]. No smoothing parameter is required when applying either AG or DL (Figure 7). 8 of 18 Remote Sens. 2017, 9, 1271 Figure 7. Examples of least-squares ﬁts to asymmetric Gaussian (AG) and DL model functions. Figure 7. Examples of least-squares ﬁts to asymmetric Gaussian (AG) and DL model functions. 2.3. Performance Assessment of the Time-Series Smoothing Methods 2.3. Performance Assessment of the Time-Series Smoothing Methods 2.3.1. Ground Measured NDVI Time-Series vs. Smoothed Satellite NDVI 2.3.1. Ground Measured NDVI Time-Series vs. Smoothed Satellite NDVI The root mean squared error (RMSE) between smoothed satellite NDVI time-series and ground-based NDVI time-series was used to evaluate the accuracy of each smoothing method. In most cases, the method generating the minimum RMSE can be considered the most accurate, and it thereby indicates the optimal capability of the method. However, for unknown areas where ground reference data are lacking, which is commonly the case in operational remote sensing applications, the optimal parameter settings for each method is unknown. Therefore, we also computed the median RMSE and the range of RMSE from all possible settings for a single method. Together, the minimum, median and range provide a general description of the performance of each method. The minimum indicates the best performance of each method given that calibration data are available. When no calibration data are available, the median and range are better indicators, providing information on the average performance and the associate range of each method. A large range indicates a ﬂexible method, meaning that both low and high accuracy could be attained. 3.1. Ground Measured NDVI Time-Series vs. Smoothed Satellite NDVI Time-Series RMSE values for the 987 runs with different settings for the smoothing of MODIS NDVI are displayed as averages (Table 3) and as boxplots for each of the sites and methods (Figure S2). Overall, for the ten test sites, average RMSE of the raw MODIS NDVI and ground-measured NDVI was 0.14 (Table 3). After applying smoothing, RMSE was on average reduced to 0.08, and the range of RMSE was 0.06. Most sites (89% of the 987 simulations) showed reduced RMSE compared to raw data, except at Sudan Barah. At the sites with high NDVI values and long growing season, such as Abisko Delta, Abisko Stordalen, Fäjemyr, Hyytiäla and Norunda forest, the reductions in RMSE were larger than at the other sites that have lower NDVI or shorter growing seasons. The difference can be partly attributed to cloud effects. Clouds contribute to the majority of noise in the satellite received vegetation signal during the growing season. Hence, sites with high NDVI values are more strongly affected by cloud contamination than sites with low NDVI. In addition, the dry Sudanian sites have short growing season, and there is generally lower cloud contamination than in the humid areas. As a result, smoothing methods have a weaker effect on the RMSE than at the humid sites with high NDVI values and long growing seasons. Table 3. The ten sites’ average of minimum, median and range of root mean squared error (RMSE) between smoothed moderate-resolution imaging spectroradiometer (MODIS) derived normalized difference vegetation index (NDVI) from all settings of each smoothing algorithm and ground measured NDVI. Table 3. The ten sites’ average of minimum, median and range of root mean squared error (RMSE) between smoothed moderate-resolution imaging spectroradiometer (MODIS) derived normalized difference vegetation index (NDVI) from all settings of each smoothing algorithm and ground measured NDVI. Statistic SG LO SP AG DL Average Minimum 0.063 0.062 0.064 0.066 0.069 0.065 Median 0.076 0.075 0.085 0.076 0.077 0.078 CV 0.077 0.077 0.076 - - - Range 0.069 0.068 0.071 0.038 0.035 0.056 Note: the RMSE between raw MODIS NDVI and ground measured NDVI is 0.142. When studying each smoothing method, we found that the minimum and median RMSEs of all methods differed relatively little, but that overall LO generated slightly lower RMSE than the other four methods (Table 3). 2.3.2. Ground Measured GPP Time-Series vs. Smoothed Satellite NDVI Time-Series 2.3.2. Ground Measured GPP Time-Series vs. Smoothed Satellite NDVI Time-Series We employed Spearman’s rank correlation coefﬁcient (ρ) to measure the synchronized variation (not necessarily linear) between ground measured GPP and smoothed NDVI time-series. When GPP and NDVI are unsynchronized, the Spearman correlation is 0, and when they co-vary monotonically it is ±1.0. We used non-winter period GPP data for comparison only during the vegetative season. 2.3.3. Elevation vs. Phenology Parameters from Satellite NDVI Data 2.3.3. Elevation vs. Phenology Parameters from Satellite NDVI Data Besides evaluating the performance of the smoothing methods for reconstructing time-series, we evaluated their abilities for extracting phenological metrics across a spatial gradient. The assumption of the experiment was that in this area the variation of vegetation phenology, such as start of the season (SOS), end of the season (EOS), length of the season (LOS) and small seasonal integral (SI) along the gradient, are generally related to average air temperature [23,51], and that this is partly controlled by the local elevation. Other factors, such as solar radiation, hydrology, soil type, etc., also affect the relationship [23], but we assume that for comparing the different smoothing methods the inﬂuence of these is negligible. We extracted SOS, EOS, LOS and SI by TIMESAT based on smoothed MODIS NDVI time-series. To avoid inter-annual variations we averaged the phenology variables over the 10 years between 2001 and 2010, which reduces inﬂuences of trends and extreme events, and emphasizes phenological variation along the elevation gradient. To avoid effects of non-linearity we used Spearman’s rank correlation coefﬁcient for detecting the relationships. For extracting phenological metrics from time-series, suitable thresholds to deﬁne start and end of season are required. Since NDVI has a strong response to snow [70] and MOD09Q1 does not have snow ﬂags, using high amplitude thresholds is necessary to avoid capturing snow dynamics rather 9 of 18 Remote Sens. 2017, 9, 1271 than dynamics of vegetation. Therefore, the thresholds for start and end of season were set to a high value of 72% of the amplitude, which was based on Fennoscandian studies by [71,72]. For the same reason we did not use derivatives for determining SOS and EOS, which would have been strongly inﬂuenced by snow seasonality during early and late parts of the seasons. 3.1. Ground Measured NDVI Time-Series vs. Smoothed Satellite NDVI Time-Series Although window size q in SG/LO and smoothing parameter p in SP have different deﬁnitions, these methods produced similar smoothed time series (Figure S2). DL generated the smallest RMSE range of 0.035. The function ﬁtting methods AG and DL also generated very similar results to one another. SP was closer to the local ﬁltering methods. Cross validation (SGCV, LOCV and SPCV) generated RMSE values lower than median RMSE in most cases (87%), which means that CV is in many cases able to assist in selecting satisfying settings for a single pixel. In four out of the 30 cases, CV resulted in very poor results. When comparing the methods across the ten test sites, SG and LO produced the eight lowest RMSE values out of the ten test sites, except at the Hyytiälä site and the Demokeya site where AG and SP produced the lowest RMSE respectively. The lowest median RMSEs varied between methods at the different sites, without any clear pattern. The smallest RMSE ranges were produced by the function-ﬁtting methods (AG and DL) for nine sites. At one site, Sudan Barah, which had the lowest NDVI and the shortest growing season of all, this was not true. This may be due to the large NDVI uncertainties from measurement when NDVI is small [56]. The time-series patterns generated from each method are shown in Figure S3. SG, LO, SP were able to describe the variation 10 of 18 10 of 18 Remote Sens. 2017, 9, 1271 of NDVI during the growing seasons (e.g., year 2011 in Figure 8). However, AG and DL were more robust in reconstructing the general seasonal patterns when MODIS NDVI was noisy (e.g., year 2010 in Figure 8). For details of time-series patterns generated from each method for all 10 sites, see Figure S3. Figure 8. Smoothed time-series of each method at all test sites based on the settings that provide the minimum RMSE values at Fäjemyr site. Figure 8. Smoothed time-series of each method at all test sites based on the settings that provide the minimum RMSE values at Fäjemyr site. In conclusion, this experiment showed that applying smoothing methods to MODIS NDVI reduced differences with ground-measured NDVI. 3.1. Ground Measured NDVI Time-Series vs. Smoothed Satellite NDVI Time-Series The local ﬁltering methods (SG and LO) had better performance in representing ground NDVI from MODIS NDVI with optimized setting parameters, but the function ﬁtting methods (AG and DL) were more robust in the sense that they generated less varying results depending on how the settings were chosen. CV for automatically choosing smoothing parameters generated better-than-average ﬁts for the majority of SG, LO and SP runs, however it failed in a few cases. 3.3. Elevation vs. Phenology Parameters from Satellite NDVI Data 3.3. Elevation vs. Phenology Parameters from Satellite NDVI Data The correspondences between elevation and phenological parameters are shown in Figure 9 and Table 5. Positive ρ values for SOS means that SOS occurs later with increasing elevation, whereas negative ρ values for EOS, LOS and SI means that EOS occurs earlier, LOS is shorter, and SI is smaller with increasing elevation. DL produced the highest absolute maximum and median ρ values (Table 5 and Table S4). The smallest ranges of ρ were generated with AG and DL. The ranges of ρ produced by AG and DL were much smaller than those produced by SG, LO and SP. CV was able to generate similar or higher absolute ρ for LO and SP in most cases, and for SP results from CV were very good. However, for SG it generated much lower absolute ρ. Phenological maps generated with optimized DL settings (Figure 10) show coherent spatial patterns and correspondence with elevation in Figure 2. In general, the experiment showed that the function ﬁtting methods, especially DL, performed better than the other methods for extracting generalized vegetation phenology. Figure 9. Boxplot of the Spearman’s rank correlation coefﬁcient between elevation in the Ammar region and ten years’ average vegetation phenology parameters, estimated from smoothed MODIS NDVI time-series by each method with different settings for the period 2001–2010. The black squares in SG, LO and SP are generated from auto-selecting parameters by CV. Note: inverted y-scale on EOS, LOS and SI: lowest negative ρ at the top of the graphs, and the outliers are not shown. Figure 9. Boxplot of the Spearman’s rank correlation coefﬁcient between elevation in the Ammar region and ten years’ average vegetation phenology parameters, estimated from smoothed MODIS NDVI time-series by each method with different settings for the period 2001–2010. The black squares in SG, LO and SP are generated from auto-selecting parameters by CV. Note: inverted y-scale on EOS, LOS and SI: lowest negative ρ at the top of the graphs, and the outliers are not shown. Table 5. The median of ρ between elevation and vegetation phenology parameters extracted by each smoothing algorithm and all settings. Numbers in parenthesis are generated from auto-selecting parameters by CV. Table 5. The median of ρ between elevation and vegetation phenology parameters extracted by each smoothing algorithm and all settings. Numbers in parenthesis are generated from auto-selecting parameters by CV. 3.2. Ground Measured GPP Time-Series vs. Smoothed Satellite NDVI Time-Series At the four ﬂux measurement sites, ρ of non-smoothed 8-day MODIS NDVI and GPP was 0.338 on average (Table 4). After applying smoothing, ρ on average increased up to 0.513, and the range between the best and the worst case was 0.228 (Table 4). In 98% of the runs, the tested methods successfully increased ρ values at the four test sites (Table S3). The method that produced the highest maximum ρ values varied between the sites. Table 4. The four sites’ average maximum, median and range of ρ between smoothed MODIS NDVI from all settings of each smoothing algorithm and ground measured GPP. Table 4. The four sites’ average maximum, median and range of ρ between smoothed MODIS NDVI from all settings of each smoothing algorithm and ground measured GPP. Statistic SG LO SP AG DL Average Maximum 0.635 0.635 0.658 0.673 0.603 0.641 Median 0.488 0.498 0.505 0.555 0.503 0.510 CV 0.453 0.528 0.534 - - - Range 0.313 0.245 0.243 0.215 0.165 0.236 Note: the ρ between raw MODIS NDVI and ground measured NDVI is 0.338. When comparing the smoothing methods, we note that the highest maximum and median ρ were generated by AG of 0.673 and 0.555, respectively (average for four sites). However, they varied among sites (Table S3). CV automatically selected fairly suitable smoothing parameter for SP but sometimes failed for SG and LO. In contrast to the previous experiment, the local ﬁltering methods did not generate the highest ρ even with the optimized settings. The smallest range of ρ was generated by DL, which was similar to the ﬁrst experiment. In conclusion, this experiment showed that applying smoothing methods to MODIS NDVI improved covariation with ground-measured GPP. No single method always generated the best results Remote Sens. 2017, 9, 1271 11 of 18 in representing NDVI time-series highly correlated with GPP time-series, but AG and DL were more robust in the sense of being less inﬂuenced by individual settings. in representing NDVI time-series highly correlated with GPP time-series, but AG and DL were more robust in the sense of being less inﬂuenced by individual settings. 3.3. Elevation vs. Phenology Parameters from Satellite NDVI Data Phenology Parameter SG LO SP AG DL SOS 0.682 (0.335) 0.715 (0.786) 0.693 (0.875) 0.887 0.891 EOS −0.585 (0.002) −0.637 (−0.667) −0.668 (−0.753) −0.767 −0.785 LOS −0.732 (−0.130) −0.765 (−0.788) −0.776 (−0.861) −0.873 −0.874 SI −0.518 (−0.082) −0.566 (−0.587) −0.558 (−0.569) −0.630 −0.655 12 of 18 Remote Sens. 2017, 9, 1271 Figure 10. Maps of start of the season (SOS), end of the season (EOS), length of the season (LOS) and small seasonal integral (SI) for the Ammar region, generated using the optimized settings of DL. Figure 10. Maps of start of the season (SOS), end of the season (EOS), length of the season (LOS) and small seasonal integral (SI) for the Ammar region, generated using the optimized settings of DL. 4. Discussion In this study, we demonstrated the degree to which ﬁve smoothing methods (SG, LO, SP, AG and DL) improved MODIS NDVI time-series data for representing ground-measured NDVI, ground measured GPP, and for extracting phenological parameters correlated with topographical elevation. To cover a range of possible settings chosen by users we employed a total of 987 different settings for the ﬁve smoothing methods in each experiment. The three independent experiments together provide insights in the choice of methods across different sites. 4.1. Which Smoothing Method is the Most Accurate The local ﬁltering methods SG and LO have very similar deﬁnitions, although LO has a more advanced weighting system than SG. Despite this, the results from LO were not clearly better than those from SG. The explanation may be that LO assigns very high weight to the middle data point in the smoothing window and gives low weights to the other points. This leads to a more pronounced ‘local’ characteristic for LO, which may be negative in the presence of strong noise. On the other hand, this ‘local’ characteristic may be an advantage for detecting detailed changes during a growing season. In many cases the local methods were superior for reconstructing ground NDVI. However, the differences in RMSE were small and may be due to random chance. Spline smoothing has the advantage of having only one parameter, p, to change the degree of smoothness of a curve. However, the performance of SP was not better than the other methods in any of the three experiments. This could partly be explained by a limited number of tested p values. p can be set to any positive natural number, but in this paper p was only tested for six selected values, which means that the full capability of SP was not utilized. SPCV generated more robust results than SP, but it had poorer performance in generating results in optimized parameters. When applying SPCV to images it optimizes p for every pixel. However, it can be questioned whether is it reasonable to allow 13 of 18 Remote Sens. 2017, 9, 1271 different degrees of smoothing for different pixels, irrespective of their class membership. The third experiment showed that SPCV performed better than SG, LO and SP but worse than AG and DL. Contrary to the other methods, AG and DL smooth data in a static window (half/full seasons), which is the reason for their robustness. In addition, AG and DL tend to eliminate noise at peaks and valleys in a time-series. Although this elimination limits the ability of AG and DL in detecting detailed variations during periods near the season peak, it helps to automatically and robustly extract temporal phenology variables, e.g., SOS, EOS, and LOS (third experiment). The results of our tests point to the superiority of these function ﬁtting methods for extracting phenological parameters over large areas where calibration of the smoothing parameters is not possible. 4.2. How to Choose Smoothing Methods Since there is no single method that is always optimal, selecting a suitable smoothing method depends on the data quality, signal dynamics, and the intended level of generalization of the smoothing. We found that in general, the function ﬁtting methods are robust for extracting phenological parameters or reconstructing strongly generalized seasonal curves. Function ﬁtting methods are also the prior choice for smoothing data with very low quality (e.g., frequent noise and gaps). However, these methods have a limitation when it comes to capturing curve details and risk of underﬁtting, thereby compelling us to sometimes use local ﬁltering methods. If the purpose is to analyze intra-seasonal variations (e.g., in disturbance analyses), local ﬁltering methods are recommended, because they are able to capture seasonal details. Also, if there are multiple peaks during the season (e.g., due to multiple harvesting in agricultural lands), local ﬁltering methods may be more useful because their smoothing process is freer. Function ﬁtting methods may work also with multiple cropping but a difﬁculty is to provide initial positions and shapes for the functions of all seasons. 4.3. How to Choose Smoothing Parameters The efﬁciency of each smoothing method depended on the choice of parameters. For example, applying upper-envelope iterations generally improved the results. This emphasizes the need to further compensate time-series of MODIS 8-day composite NDVI data for remaining negative bias due to cloud and aerosols in the atmosphere. At the sites Barah and Demokeya, where upper-envelope iterations did not improve results, vegetation density was very low, which means that the data were probably more affected by soil background characteristics, which would lead to decreased noise bias. When comparing against GPP, upper-envelope iterations also generally improved results, except when using DL and SPCV at the Norunda forest site. In the third experiment, upper-envelope iteration of SPCV, AG and DL did not improve ρ in correlating end of season timing and elevation, which could help explain why upper-envelope iterations also did not improve DL and SPCV ﬁts at the Norunda forest site. The smoothing parameters, q in SG and LO, and p in SP, were much more inﬂuential than the envelope parameters u and a in the ﬁrst experiment. As a consequence, more attention needs to be paid to adjusting q and p. In addition, the ability to vary the smoothing parameters q and p means that SG, LO, and SP have larger ﬂexibility than AG, and DL, but this ﬂexibility means that the methods are less robust to user selection of parameter values. A higher ﬂexibility means that a method is more capable of matching satellite data with ground reference data, but accurate matching also requires ground observations for calibration. In most situations, this calibration is impossible due to a lack of ground observations, and, as seen from the large ranges observed in the second and third experiments, a robust method with fewer parameters may be preferred. Automatically selecting smoothing parameters by CV can be useful for SG, LO and SP if no ground reference data are available for calibration, as demonstrated by the above-average results achieved with this technique. However, applying CV can sometimes lead to very poor results, as seen in a few cases (e.g., Abisko and Hyytiälä in Figure S2). CV may also result in different parameter settings being used for different pixels of the same land cover class, and it can be discussed if this is Remote Sens. 2017, 9, 1271 14 of 18 reasonable. 4.4. Limitations of this Study and Future Possible Directions 4.4. Limitations of this Study and Future Possible Directions Besides NDVI, there are many optional indexes used to describe vegetation time-series dynamic such as the Enhanced Vegetation Index [73] and the plant phenology index [60]. Different indexes could present different time-series patterns for the same pixel, which could lead to different conclusions for different smoothing methods. Moreover, the difference between data from different satellite sensors, such as MODIS and NDVI3g, could also lead to different estimates of phenological parameters [74]. In addition, although the tested sites represent several different biomes, it cannot be ruled out that other vegetation types with different seasonal proﬁles might yield differences in relationships with ground reference data. Furthermore, although the sites used in the ﬁrst and second experiments were generally spatially homogeneous, uncertainties in footprint matching may have affected the relationships. In the last experiment, it is recognized that several factors other than elevation affect the general phenology in the area. It is unknown if and to what degree these different factors affect differences in performance between the different smoothing methods. The results of this study are valid for the tested data and sites, and relationships may change with other data or under different conditions. However, the consistency of the results indicates a general validity of the conclusions also outside the tested domain. 4.3. How to Choose Smoothing Parameters It can also be noted that CV had different effects on SP as compared to SG and LO. In SP, the smoothing parameter is continuous, whereas it is discrete in SG and LO. The discrete parameters in SG and LO may lead to unsatisfactory results if the data quality is poor and the smoothing parameter signiﬁcantly inﬂuences the ﬁtting, which could be the reason that CV is less reliable for SG and LO than for SP. Although most smoothing methods improved the ability of the satellite time-series to represent the ground reference data, it was not possible to ﬁnd a single parameter setting that was always able to generate suitable results for all environments, which supports the ﬁnding by Hird and McDermid [35] and Atkinson, et al. [31]. Supplementary Materials: The following material is available online at www.mdpi.com/2072-4292/9/12/1271/s1, Figure S1: Geographic distributions of ground-based spectral observation sites and Ammar region, Figure S2: Boxplots of RMSEs between ground-based measured NDVI and smoothed satellite NDVI, Figure S3: Plots of smoothed curves of each method at all testing sites based on the settings that provide the minimum RMSE values, Table S1: Additional information of ground-based spectral observation sites and ﬂux tower sites, Table S2: The minimum, median and range of RMSE between smoothed MODIS NDVI from all settings of each smoothing algorithm and ground measured NDVI from all spectral sensor tower sites, Table S3: Maximum, median and 5. Conclusions In this paper, we utilized three ground reference data sets for evaluating the performance of ﬁve smoothing methods in processing MODIS NDVI time-series. We conclude that the smoothing methods reduced the error between MODIS NDVI and ground-measured NDVI in 89% of the simulations, and that they improved the covariation between MODIS NDVI and ground-measured GPP in 98% of the simulations. It is possible to achieve very good similarity with ground NDVI data with the local ﬁltering methods (Savitzky-Golay and LOESS smoothing) when parameters are ﬁnely tuned. However, on average the differences between the methods were small and it is not possible to conclusively recommend a single method in favor of the others. Spline smoothing did not generate better results than any of the other methods, but using cross validation generally improved the spline ﬁts and gave fairly good results. Cross validation was also useful for selecting parameters for Savitzky-Golay and LOESS smoothing; however, in some cases the method failed. The function ﬁtting methods (double logistic and asymmetric Gaussian functions) showed the best performance for estimating average phenology across a topographic gradient. The function ﬁtting methods were also found to generally reduce the risk of achieving very poor results (shown by the smaller range of output results), thus they are safer than the other methods to use when it is not possible to carry out any calibration against ground measurements. Supplementary Materials: The following material is available online at www.mdpi.com/2072-4292/9/12/1271/s1, Figure S1: Geographic distributions of ground-based spectral observation sites and Ammar region, Figure S2: Boxplots of RMSEs between ground-based measured NDVI and smoothed satellite NDVI, Figure S3: Plots of smoothed curves of each method at all testing sites based on the settings that provide the minimum RMSE values, Table S1: Additional information of ground-based spectral observation sites and ﬂux tower sites, Table S2: The minimum, median and range of RMSE between smoothed MODIS NDVI from all settings of each smoothing algorithm and ground measured NDVI from all spectral sensor tower sites, Table S3: Maximum, median and 15 of 18 Remote Sens. References 1. Schwartz, M.D. Phenology: An Integrative Environmental Science; Springer: Dordrecht, The Netherlands, 2003; Volume 346. 2. Prince, S. A model of regional primary production for use with coarse resolution satellite data. Int. J. Remote Sens. 1991, 12, 1313–1330. [CrossRef] 3. Running, S.W.; Nemani, R.R.; Heinsch, F.A.; Zhao, M.; Reeves, M.; Hashimoto, H. A continuous satellite-derived measure of global terrestrial primary production. Bioscience 2004, 54, 547–560. [CrossRef] 4. Goward, S.N.; Markham, B.; Dye, D.G.; Dulaney, W.; Yang, J. Normalized difference vegetation index measurements from the advanced very high resolution radiometer. Remote Sens. Environ. 1991, 35, 257–277. [CrossRef] 5. Viovy, N.; Arino, O.; Belward, A. The Best Index Slope Extraction (BISE): A method for reducing noise in NDVI time-series. Int. J. Remote Sens. 1992, 13, 1585–1590. [CrossRef] 6. van Dijk, A.; Callis, S.L.; Sakamoto, C.M.; Decker, W.L. Smoothing vegetation index proﬁles: An alternative method for reducing radiometric disturbance in NOAA/AVHRR data. Photogramm. Eng. Remote Sens. 1987, 53, 1059–1067. 7. Menenti, M.; Azzali, S.; Verhoef, W.; van Swol, R. Mapping agroecological zones and time lag in vegetation growth by means of Fourier analysis of time series of NDVI images. Adv. Space Res. 1993, 13, 233–237. [CrossRef] 8. Olsson, L.; Eklundh, L. Fourier series for analysis of temporal sequences of satellite sensor imagery. Int. J. Remote Sens. 1994, 15, 3735–3741. [CrossRef] 9. Zhu, Z.; Woodcock, C.E.; Olofsson, P. Continuous monitoring of forest disturbance using all available Landsat imagery. Remote Sens. Environ. 2012, 122, 75–91. [CrossRef] 10. Reed, B.C.; Brown, J.F.; VanderZee, D.; Loveland, T.R.; Merchant, J.W.; Ohlen, D.O. Measuring phenological variability from satellite imagery. J. Veg. Sci. 1994, 5, 703–714. [CrossRef] 11. Chen, J.; Jönsson, P.; Tamura, M.; Gu, Z.; Matsushita, B.; Eklundh, L. A simple method for reconstructing a high-quality NDVI time-series data set based on the Savitzky–Golay ﬁlter. Remote Sens. Environ. 2004, 91, 332–344. [CrossRef] 12. Jönsson, P.; Eklundh, L. Seasonality extraction by function ﬁtting to time-series of satellite sensor data. IEEE Trans. Geosci. Remote Sens. 2002, 40, 1824–1832. [CrossRef] 13. Zhang, X.; Friedl, M.A.; Schaaf, C.B.; Strahler, A.H.; Hodges, J.C.; Gao, F.; Reed, B.C.; Huete, A. Monitoring vegetation phenology using MODIS. Remote Sens. Environ. 2003, 84, 471–475. [CrossRef] 14. Beck, P.S.; Atzberger, C.; Høgda, K.A.; Johansen, B.; Skidmore, A.K. Improved monitoring of vegetation dynamics at very high latitudes: A new method using MODIS NDVI. Remote Sens. Environ. 2006, 100, 321–334. [CrossRef] 15. Fisher, J.; Mustard, J.; Vadeboncoeur, M. 5. Conclusions 2017, 9, 1271 range of ρ between smoothed MODIS NDVI from all settings of each smoothing algorithm and ground measured GPP, Figure S4: Boxplots of Spearman’s rank correlation coefﬁcient of tower GPP and smoothed MODIS NDVI for each method, showing the distribution with different settings at test sites, Table S4: The maximum, median and range of ρ between elevations and land surface phenology parameters extracted by each smoothing algorithm of all settings on MODIS NDVI. Acknowledgments: The research was funded by a grant from the Swedish National Space Board to Lars Eklundh and Per Jönsson, Dnr. 119/13. We acknowledge access to ﬂux data from ICOS Abisko station, Hyytiälä station, Norunda station Sweden, ICOS Finland, and Fluxnet. We thank Dr. Jonas Ardö for providing spectral data for the Sudan sites. The MOD09Q1 data product was retrieved from the online Data Pool, courtesy of the NASA EOSDIS Land Processes Distributed Active Archive Center (LP DAAC), USGS/Earth Resources Observation and Science (EROS) Center, Sioux Falls, South Dakota, https://lpdaac.usgs.gov/data_access/data_pool. Author Contributions: Z.C., P.J. and L.E. conceived and designed the experiments; Z.C., H.J. and L.E. performed the experiments; Z.C. and L.E. analyzed the data; P.J. and L.E. contributed analysis tools and ﬁeld equipment; Z.C. and H.J. wrote the manuscript draft with further contributions by L.E. and P.J. Conﬂicts of Interest: The authors declare no conﬂict of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results. References Green leaf phenology at Landsat resolution: Scaling from the ﬁeld to the satellite. Remote Sens. Environ. 2006, 100, 265–279. [CrossRef] Remote Sens. 2017, 9, 1271 16 of 18 16 of 18 16. Jönsson, P.; Eklundh, L. TIMESAT—A program for analyzing time-series of satellite sensor data. Comput. Geosci. 2004, 30, 833–845. [CrossRef] 17. Hermance, J.F.; Jacob, R.W.; Bradley, B.A.; Mustard, J.F. Extracting phenological signals from multiyear AVHRR NDVI time series: Framework for applying high-order annual splines with roughness damping. IEEE Trans. Geosci. Remote Sens. 2007, 45, 3264–3276. [CrossRef] 18. Bradley, B.A.; Jacob, R.W.; Hermance, J.F.; Mustard, J.F. A curve ﬁtting procedure to derive inter-annual phenologies from time series of noisy satellite NDVI data. Remote Sens. Environ. 2007, 106, 137–145. [CrossRef] 19. Atzberger, C.; Eilers, P.H.C. Evaluating the effectiveness of smoothing algorithms in the absence of ground reference measurements. Int. J. Remote Sens. 2011, 32, 3689–3709. [CrossRef] 20. Sakamoto, T.; Yokozawa, M.; Toritani, H.; Shibayama, M.; Ishitsuka, N.; Ohno, H. A crop phenology detection method using time-series MODIS data. Remote Sens. Environ. 2005, 96, 366–374. [CrossRef] 21. Lu, X.; Liu, R.; Liu, J.; Liang, S. Removal of noise by wavelet method to generate high quality temporal data of terrestrial MODIS products. Photogramm. Eng. Remote Sens. 2007, 73, 1129–1139. [CrossRef] 2. Campos, A.N.; di Bella, C.M. Multi-temporal analysis of remotely sensed information using wave J. Geogr. Inf. Syst. 2012, 4, 383–391. 3. Olsson, C.; Bolmgren, K.; Lindström, J.; Jönsson, A.M. Performance of tree phenology models alon bioclimatic gradient in Sweden. Ecol. Model. 2013, 266, 103–117. [CrossRef] 24. Richardson, A.D.; Anderson, R.S.; Arain, M.A.; Barr, A.G.; Bohrer, G.; Chen, G.; Chen, J.M.; Ciais, P.; Davis, K.J.; Desai, A.R.; et al. Terrestrial biosphere models need better representation of vegetation phenology: Results from the North American Carbon Program Site Synthesis. Glob. Chang. Biol. 2012, 18, 566–584. [CrossRef] 25. Olsson, C.; Jönsson, A.M. Process-based models not always better than empirical models for simulating budburst of Norway spruce and birch in Europe. Glob. Chang. Biol. 2014, 20, 3492–3507. [CrossRef] [PubMed] 25. Olsson, C.; Jönsson, A.M. Process-based models not always better than empirical models for simulating budburst of Norway spruce and birch in Europe. Glob. Chang. Biol. 2014, 20, 3492–3507. [CrossRef] [PubMed] 26. Schwartz, M.D.; Reed, B.C.; White, M.A. Assessing satellite-derived start-of-season measures in the J y p g budburst of Norway spruce and birch in Europe. Glob. Chang. Biol. 2014, 20, 3492–3507. [CrossRef] [PubMed] 26. References Schwartz, M.D.; Reed, B.C.; White, M.A. Assessing satellite-derived start-of-season measures in the conterminous USA. Int. J. Climatol. 2002, 22, 1793–1805. [CrossRef] 26. Schwartz, M.D.; Reed, B.C.; White, M.A. Assessing satellite-derived start-of-season measures in the conterminous USA. Int. J. Climatol. 2002, 22, 1793–1805. [CrossRef] 27. Soudani, K.; le Maire, G.; Dufrêne, E.; François, C.; Delpierre, N.; Ulrich, E.; Cecchini, S. Evaluation of the onset of green-up in temperate deciduous broadleaf forests derived from Moderate Resolution Imaging Spectroradiometer (MODIS) data. Remote Sens. Environ. 2008, 112, 2643–2655. [CrossRef] 28. Zhang, X.; Friedl, M.A.; Schaaf, C.B. Global vegetation phenology from Moderate Resolution Imaging Spectroradiometer (MODIS): Evaluation of global patterns and comparison with in situ measurements. J. Geophys. Res. Biogeosci. 2006, 111. [CrossRef] 29. Soudani, K.; Hmimina, G.; Delpierre, N.; Pontailler, J.-Y.; Aubinet, M.; Bonal, D.; Caquet, B.; de Grandcourt, A.; Burban, B.; Flechard, C. Ground-based Network of NDVI measurements for tracking temporal dynamics of canopy structure and vegetation phenology in different biomes. Remote Sens. Environ. 2012, 123, 234–245. [CrossRef] 30. White, M.A.; Beurs, D.; Kirsten, M.; Didan, K.; Inouye, D.W.; Richardson, A.D.; Jensen, O.P.; O'Keefe, J.; Zhang, G.; Nemani, R.R. Intercomparison, interpretation, and assessment of spring phenology in North America estimated from remote sensing for 1982–2006. Glob. Chang. Biol. 2009, 15, 2335–2359. [CrossRef] 31. Atkinson, P.M.; Jeganathan, C.; Dash, J.; Atzberger, C. Inter-comparison of four models for smoothing satellite sensor time-series data to estimate vegetation phenology. Remote Sens. Environ. 2012, 123, 400–417. [CrossRef] 32. Chen, W.; Foy, N.; Olthof, I.; Latifovic, R.; Zhang, Y.; Li, J.; Fraser, R.; Chen, Z.; McLennan, D.; Poitevin, J.; et al. Evaluating and reducing errors in seasonal proﬁles of AVHRR vegetation indices over a Canadian northern national park using a cloudiness index. Int. J. Remote Sens. 2013, 34, 4320–4343. [CrossRef] 33. Cong, N.; Wang, T.; Nan, H.; Ma, Y.; Wang, X.; Myneni, R.B.; Piao, S. Changes in satellite-derived spring vegetation green-up date and its linkage to climate in China from 1982 to 2010: A multimethod analysis. Glob. Chang. Biol. 2013, 19, 881–891. [CrossRef] [PubMed] 34. Kandasamy, S.; Fernandes, R. An approach for evaluating the impact of gaps and measurement errors on satellite land surface phenology algorithms: Application to 20year NOAA AVHRR data over Canada. Remote Sens. Environ. 2015, 164, 114–129. [CrossRef] 35. Hird, J.N.; McDermid, G.J. Noise reduction of NDVI time series: An empirical comparison of selected techniques. Remote Sens. Environ. 2009, 113, 248–258. References [CrossRef] 17 of 18 Remote Sens. 2017, 9, 1271 17 of 18 36. Julien, Y.; Sobrino, J.A. Comparison of cloud-reconstruction methods for time series of composite NDVI data. Remote Sens. Environ. 2010, 114, 618–625. [CrossRef] 37. Kandasamy, S.; Baret, F.; Verger, A.; Neveux, P.; Weiss, M. A comparison of methods for smoothing and gap ﬁlling time series of remote sensing observations – application to MODIS LAI products. Biogeosciences 2013, 10, 4055–4071. [CrossRef] 38. Moreno, Á.; García-Haro, F.; Martínez, B.; Gilabert, M. Noise reduction and gap ﬁlling of fAPAR time Series using an adapted local regression ﬁlter. Remote Sens. 2014, 6, 8238–8260. [CrossRef] 39. Geng, L.; Ma, M.; Wang, X.; Yu, W.; Jia, S.; Wang, H. Comparison of eight techniques for reconstructing multi-satellite sensor time-series NDVI data sets in the Heihe river basin, China. Remote Sens. 2014, 6, 2024–2049. [CrossRef] 40. Richardson, A.D.; Jenkins, J.P.; Braswell, B.H.; Hollinger, D.Y.; Ollinger, S.V.; Smith, M.L. Use of digital webcam images to track spring green-up in a deciduous broadleaf forest. Oecologia 2007, 152, 323–334. [CrossRef] [PubMed] 41. Balzarolo, M.; Anderson, K.; Nichol, C.; Rossini, M.; Vescovo, L.; Arriga, N.; Wohlfahrt, G.; Calvet, J.-C.; Carrara, A.; Cerasoli, S.; et al. Ground-Based Optical Measurements at European Flux Sites: A Review of Methods, Instruments and Current Controversies. Sensors 2011, 11, 7954–7981. [CrossRef] [PubMed] 2. Eklundh, L.; Jin, H.; Schubert, P.; Guzinski, R.; Heliasz, M. An optical sensor network for vegeta phenology monitoring and satellite data calibration. Sensors 2011, 11, 7678–7709. [CrossRef] [PubMed] 43. Hufkens, K.; Friedl, M.; Sonnentag, O.; Braswell, B.H.; Milliman, T.; Richardson, A.D. Linking near-surface and satellite remote sensing measurements of deciduous broadleaf forest phenology. Remote Sens. Environ. 2012, 117, 307–321. [CrossRef] 44. Hmimina, G.; Dufrêne, E.; Pontailler, J.Y.; Delpierre, N.; Aubinet, M.; Caquet, B.; de Grandcourt, A.; Burban, B.; Flechard, C.; Granier, A.; et al. Evaluation of the potential of MODIS satellite data to predict vegetation phenology in different biomes: An investigation using ground-based NDVI measurements. Remote Sens. Environ. 2013, 132, 145–158. [CrossRef] 45. Gamon, J.A.; Rahman, A.; Dungan, J.; Schildhauer, M.; Huemmrich, K. Spectral Network (SpecNet)—What is it and why do we need it? Remote Sens. Environ. 2006, 103, 227–235. [CrossRef] 46. Porcar-Castell, A.; Arthur, A.M.; Rossini, M.; Eklundh, L.; Pacheco-Labrador, J.; Anderson, K.; Balzarolo, M.; Martín, M.; Jin, H.; Tomelleri, E.; et al. EUROSPEC: At the interface between remote-sensing and ecosystem CO2 ﬂux measurements in Europe. Biogeosciences 2015, 12, 6103–6124. [CrossRef] 47. References Asrar, G.; Fuchs, M.; Kanemasu, E.; Hatﬁeld, J. Estimating absorbed photosynthetic radiation and leaf area index from spectral reﬂectance in wheat. Agron. J. 1984, 76, 300–306. [CrossRef] 48. Sellers, P.J.; Berry, J.A.; Collatz, G.J.; Field, C.B.; Hall, F.G. Canopy reﬂectance, photosynthesis, and transpiration. III. A reanalysis using improved leaf models and a new canopy integration scheme. Remote Sens. Environ. 1992, 42, 187–216. [CrossRef] 49. Myneni, R.; Williams, D. On the relationship between FAPAR and NDVI. Remote Sens. Environ. 1994, 49, 200–211. [CrossRef] 50. Jin, H.; Jönsson, A.M.; Bolmgren, K.; Langvall, O.; Eklundh, L. Disentangling remotely-sensed plant phenology and snow seasonality at northern Europe using MODIS and the plant phenology index. Remote Sens. Environ. 2017, 198, 203–212. [CrossRef] 51. Singh, R.K.; Svystun, T.; AlDahmash, B.; Jönsson, A.M.; Bhalerao, R.P. Photoperiod-and temperature-mediated control of phenology in trees–A molecular perspective. New Phytol. 2017, 213, 511–524. [CrossRef] [PubMed] 52. Pellikka, P. Application of vertical skyward wide-angle photography and airborne video data for phenological studies of beech forests in the German Alps. Int. J. Remote Sens. 2001, 22, 2675–2700. [CrossRef] 52. Pellikka, P. Application of vertical skyward wide-angle photography and airborne video data for phenological studies of beech forests in the German Alps. Int. J. Remote Sens. 2001, 22, 2675–2700. [CrossRef] 53. Klisch, A.; Atzberger, C. Evaluating Phenological Metrics derived from the MODIS Time Series over the 53. Klisch, A.; Atzberger, C. Evaluating Phenological Metrics derived from the MODIS Time Series over the European Continent Ableitung und Evaluierung phänologischer Kenngrößen aus MODIS-Zeitreihen für den Europäischen Kontinent. Photogramm. Fernerkund. Geoinf. 2014, 2014, 409–421. [CrossRef] 54. Vermote, E. MYD09Q1 MODIS/Aqua Surface Reﬂectance 8-Day L3 Global 250 m SIN Grid V006; NASA EOSDIS Land Processes DAAC: Sioux Falls, SD, USA, 2015; Available online: https://doi.org/10.5067/modis/ myd09q1.006 (accessed on 5 December 2017). 18 of 18 18 of 18 Remote Sens. 2017, 9, 1271 55. Tagesson, T.; Mastepanov, M.; Tamstorf, M.P.; Eklundh, L.; Schubert, P.; Ekberg, A.; Sigsgaard, C.; Christensen, T.R.; Ström, L. High-resolution satellite data reveal an increase in peak growing season gross primary production in a high-Arctic wet tundra ecosystem 1992–2008. Int. J. Appl. Earth Obs. Geoinf. 2012, 18, 407–416. [CrossRef] 56. Jin, H.; Eklundh, L. In situ calibration of light sensors for long-term monitoring of vegetation. IEEE Trans. Geosci. Remote Sens. 2015, 53, 3405–3416. [CrossRef] 57. Suni, T.; Rinne, J.; Reissell, A.; Altimir, N.; Keronen, P.; Rannik, U.; Maso, M.; Kulmala, M.; Vesala, T. References Long-term measurements of surface ﬂuxes above a Scots pine forest in Hyytiala, southern Finland, 1996–2001. Boreal Environ. Res. 2003, 8, 287–302. 58. Kohler, J.; Brandt, O.; Johansson, M.; Callaghan, T. A long-term Arctic snow depth record from Abisko, northern Sweden, 1913–2004. Polar Res. 2006, 25, 91–113. [CrossRef] 59. Hinkler, J.; Hansen, B.U.; Tamstorf, M.P.; Sigsgaard, C.; Petersen, D. Snow and Snow-Cover in Central Northeast Greenland. Adv. Ecol. Res. 2008, 40, 175–195. 60. Jin, H.; Eklundh, L. A physically based vegetation index for improved monitoring of plant phenology. Remote Sens. Environ. 2014, 152, 512–525. [CrossRef] 61. Savitzky, A.; Golay, M.J. Smoothing and differentiation of data by simpliﬁed least squares procedures. Anal. Chem. 1964, 36, 1627–1639. [CrossRef] 62. Cleveland, W.S.; Devlin, S.J. Locally weighted regression: An approach to regression analysis by local ﬁtting. J. Am. Stat. Assoc. 1988, 83, 596. [CrossRef] 63. Craven, P.; Wahba, G. Smoothing noisy data with spline functions. Numer. Math. 1979, 31, 377–403. [CrossRef] 64. Woltring, H.J. A Fortran package for generalized, cross-validatory spline smoothing and differentiation. Adv. Eng. Softw. 1978 1986, 8, 104–113. [CrossRef] 65. Eklundh, L.; Jönsson, P. Timesat 3.3 with Seasonal Trend Decomposition and Parallel Processing—Software Manual; Lund University: Lund, Sweden, 2017; p. 88. 66. Eilers, P.H. A perfect smoother. Anal. Chem. 2003, 75, 3631–3636. [CrossRef] [PubMed] 67. Whittaker, E.T. On a new method of graduation. Proc. Edinb. Math. Soc. 1923, 41, 63–75. 68. Paul, H.C.E.; Marx, B.D. Flexible smoothing with B-splines and penalties. Stat. Sci. 1996, 11, 89–102 69. Lamm, P.K. Variable-smoothing regularization methods for inverse problems. In Theory and Practice of Control and Systems; Perdon, A.C.a.A.M., Ed.; World Scientiﬁc: Singapore, 1999. Variable-smoothing regularization methods for inverse 70. Jönsson, A.M.; Eklundh, L.; Hellström, M.; Bärring, L.; Jönsson, P. Annual changes in MODIS vegetation indices of Swedish coniferous forests in relation to snow dynamics and tree phenology. Remote Sens. Environ. 2010, 114, 2719–2730. [CrossRef] 71. Høgda, K.; Tømmervik, H.; Karlsen, S. Trends in the Start of the Growing Season in Fennoscandia 1982–2011. Remote Sens. 2013, 5, 4304–4318. [CrossRef] 72. Karlsen, S.R.; Høgda, K.A.; Wielgolaski, F.E.; Tolvanen, A.; Tømmervik, H.; Poikolainen, J.; Kubin, E. Growing-season trends in Fennoscandia 1982–2006, determined from satellite and phenology data. Clim. Res. 2009, 39, 275–286. [CrossRef] 73. Huete, A.; Liu, H.; Batchily, K.; van Leeuwen, W. A comparison of vegetation indices over a global set of TM images for EOS-MODIS. Remote Sens. Environ. 1997, 59, 440–451. [CrossRef] 74. References Atzberger, C.; Klisch, A.; Mattiuzzi, M.; Vuolo, F. Phenological metrics derived over the European Continent from NDVI3g data and MODIS time series. Remote Sens. 2013, 6, 257–284. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
W4327966574.txt	https://www.mdpi.com/1422-0067/24/6/5817/pdf?version=1679130206	en		Circulating miRNA-451a and miRNA-328-3p as Potential Markers of Coronary Artery Aneurysmal Disease	International journal of molecular sciences	2,023	cc-by	7,842	International Journal of Molecular Sciences Article Circulating miRNA-451a and miRNA-328-3p as Potential Markers of Coronary Artery Aneurysmal Disease Sylwia Iwańczyk 1, * , Tomasz Lehmann 2 , Artur Cieślewicz 3 , Katarzyna Malesza 3 , Patrycja Woźniak 1 , Agnieszka Hertel 2 , Grzegorz Krupka 1 , Paweł P. Jagodziński 2 , Marek Grygier 1 , Maciej Lesiak 1 and Aleksander Araszkiewicz 1 1 2 3 * Citation: Iwańczyk, S.; Lehmann, T.; Cieślewicz, A.; Malesza, K.; Woźniak, P.; Hertel, A.; Krupka, G.; Jagodziński, P.P.; Grygier, M.; Lesiak, 1st Department of Cardiology, Poznan University of Medical Sciences, 61-848 Poznań, Poland Department of Biochemistry and Molecular Biology, Poznan University of Medical Sciences, 60-781 Poznań, Poland Clinical Pharmacology, Poznan University of Medical Sciences, 61-848 Poznań, Poland Correspondence: sylwia.iwanczyk@skpp.edu.pl; Tel.: +48-662-712-627 Abstract: MicroRNAs (miRNAs) are currently investigated as crucial regulatory factors which may serve as a potential therapeutic target. Reports on the role of miRNA in patients with coronary artery aneurysmal disease (CAAD) are limited. The present analysis aims to confirm the differences in the expression of previously preselected miRNAs in larger study groups and evaluate their usefulness as potential markers of CAAD. The study cohort included 35 consecutive patients with CAAD (Group 1), and two groups of 35 patients matched Group 1 regarding sex and age from the overall cohort of 250 patients (Group 2 and Group 3). Group 2 included patients with angiographically documented coronary artery disease (CAD), while Group 3 enrolled patients with normal coronary arteries (NCA) assessed during coronary angiography. We applied the RT-qPCR method using the custom plates for the RT-qPCR array. We confirmed that the level of five preselected circulating miRNAs was different in patients with CAAD compared to Group 2 and Group 3. We found that miR-451a and miR-328 significantly improved the CAAD prediction. In conclusion, miR-451a is a significant marker of CAAD compared to patients with CAD. In turn, miR-328-3p is a significant marker of CAAD compared to patients with NCA. M.; et al. Circulating miRNA-451a and miRNA-328-3p as Potential Keywords: microRNAs; coronary aneurysm; coronary artery disease; atherosclerosis Markers of Coronary Artery Aneurysmal Disease. Int. J. Mol. Sci. 2023, 24, 5817. https://doi.org/ 10.3390/ijms24065817 Academic Editors: Kyriacos N. Felekkis and Christos Papaneophytou Received: 1 March 2023 Revised: 12 March 2023 Accepted: 13 March 2023 Published: 18 March 2023 Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 1. Introduction Circulating non-coding RNAs have been proposed as a new class of biomarkers and are investigated for their potential in many areas of cardiology, including coronary artery disease (CAD) [1]. As critical regulatory factors, microRNAs (miRNAs) may serve as a potential therapeutic target [2,3]. Reports on the role of miRNA in patients with coronary artery aneurysmal disease (CAAD) are limited and relate mainly to patients with Kawasaki disease. CAAD is a rare condition affecting epicardial arteries. It occurs in 1.2% to 4.9% of patients undergoing coronary angiography [4]. CAAD includes two phenotypes: coronary artery aneurysm (CAA) and coronary artery ectasia (CAE) [5]. CAA is defined as a focal dilatation of the artery with a diameter of 1.5 times the adjacent normal coronary artery, whereas CAE describes more diffuse dilatation, often involving the predominant part of the artery and more than one vessel [6]. Moreover, CAA is classified into fusiform type if the longitudinal dimension is greater than the transverse or saccular type when the transverse dimension exceeds the longitudinal [7]. In symptomatic patients, the most common clinical presentation of CAAD is stable angina pectoris. It is significantly less frequent that myocardial infarction (MI) and aneurysm complications occur, i.e., thrombus formation following distal embolization, 4.0/). Int. J. Mol. Sci. 2023, 24, 5817. https://doi.org/10.3390/ijms24065817 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2023, 24, 5817 2 of 15 development of a fistula, and aneurysm rupture [6–8]. Moreover, CAAD is associated with poor long-term outcomes, regardless of concomitant atherosclerotic coronary disease [9,10]. The most common cause of abnormal dilatation of the coronary arteries is the atherosclerotic process followed by positive remodeling of the vessel. Approximately 20 to 30% of CAADs are considered inflammatory or congenital [11,12]. However, pathogenesis is multifactorial and can be influenced by the disturbance of physiological processes at the molecular level. Understanding these pathomechanisms is crucial for diagnosing and preventing CAAD complications. According to a growing body of literature, miRNAs significantly regulate human physiological processes, including gene expression in the cardiovascular system [13]. Based on our previous study [14], we selected the potential miRNA markers to validate them in a larger independent patient cohort. The present analysis aims to confirm the differences in the expression of five preselected miRNA markers (miR-125b-5p, miR-2103p, miR-328-3p, miR-425-3p, and miR-483-5p), and three reference miRNAs (miR-16-5p, miR-30d-5p, miR-320d) in larger study groups and evaluate their usefulness as potential markers of CAAD occurrence. 2. Results 2.1. Clinical Characteristics of the Study Population We presented baseline clinical characteristics in Table 1. Patients were predominantly male at the mean age of 65.9 ± 9.4, 67.1 ± 8.3, and 64.6 ± 12.3 in Group 1, Group 2, and Group 3, respectively. The prevalence of major cardiovascular risk factors did not differ significantly between the study groups. Despite the comparable incidence of renal failure, serum creatinine levels were significantly higher in Groups 1 and 2 compared to Group 3. In addition, Group 3 patients had no history of myocardial infarction (MI) or percutaneous coronary intervention (PCI). Therefore, a significantly lower group percentage was treated with aspirin and P2Y12 inhibitor. Moreover, beta-blocker and statin use were significantly less frequent in Group 3 than in Group 2. There were no differences in pharmacotherapy between Groups 1 and 2. Table 1. Baseline clinical characteristics. Variable Sex, male Age, years BMI, kg/m2 Hypertension Hyperlipidemia Diabetes mellitus HF, NYHA ≥ 2 LVEF, % Stroke Smoking CKD, eGFR ≤ 60 mL/min Previous MI Previous PCI Previous CABG Group 1 (n = 33) Group 2 (n = 35) Group 3 (n = 35) 28 (84.8) 65.9 ± 9.4 29.7 (26.5–32.9) 29 (87.9) 20 (60.6) 28 (80.0) 67.1 ± 8.3 28.9 (25.3–31.9) 30 (85.7) 29 (82.9) 10 (30.3) p-Value Group 1 vs. Group 2 Group 1 vs. Group 3 Group 2 vs. Group 3 27 (77.1) 64.6 ± 12.3 30.1 (27.2–33.7) 31 (88.6) 27 (77.1) 0.41 0.88 0.72 0.81 0.09 0.69 0.85 0.72 0.90 0.26 0.18 0.55 0.63 0.72 0.55 10 (28.6) 11 (31.4) 0.97 0.77 0.79 6 (18.2) 57.5 (55–60) 4 (12.1) 11 (33.3) 9 (25.7) 60 (50–60) 3 (8.6) 11 (31.4) 4 (11.4) 60 (55–60) 1 (2.9) 8 (22.8) 0.12 0.87 0.90 0.80 0.40 0.87 0.30 0.30 0.12 0.87 0.60 0.42 2 (6.1) 7 (20) 9 (25.7) 0.45 0.88 0.20 10 (33.3) 10 (33.3) 1 (3.0) 13 (37.1) 20 (57.1) 3 (8.6) 0 0 0 0.61 0.03 0.67 0.01 <0.01 0.46 <0.01 <0.01 0.24 Int. J. Mol. Sci. 2023, 24, 5817 3 of 15 Table 1. Cont. Variable Group 1 (n = 33) Group 2 (n = 35) Group 3 (n = 35) p-Value Group 1 vs. Group 2 Group 1 vs. Group 3 Group 2 vs. Group 3 0.60 0.24 0.84 0.95 0.78 0.96 0.90 0.01 <0.01 0.15 0.12 0.82 0.33 0.74 0.05 <0.01 0.04 0.02 1.00 0.29 1.00 6.5 (5.5–7.7) 9 (8.4–9.8) 201 (176.5–249.5) 0.58 0.45 0.15 0.85 0.05 0.45 0.91 0.91 0.91 Drugs administration Aspirin P2Y12 inhibitor Beta-blocker Statin ACEI/ARB OHA Insulin 25 (75.8) 14 (42.4) 23 (69.7) 26 (78.8) 25 (75.8) 7 (21.2) 2 (6.1) 26 (74.3) 20 (57.1) 25 (71.4) 28 (80.0) 26 (74.3) 6 (17.1) 1 (2.9) 22 (62.9) 1 (2.9) 22 (62.9) 30 (85.7) 31 (88.6) 12 (34.3) 3 (8.6) Laboratory test 109 /L WBC, HGB, mmol/L PLT, 109 /L LDL-C, mmol/L Creatinine, mmol/L Glucose, mmol/L 7.2 (5.9–9.1) 9.2 (8.3–9.9) 223.5 (188.5–257) 7.5 (6.1–9.6) 9 (8.1–9.5) 204 (180–239) 2.1 (1.6–2.6) 1.9 (1.6–3.2) 2.8 (2.0–3.3) 0.78 0.33 0.44 85.7 (77.6–99.5) 93 (80–99.5) 80.9 (69.4–86.0) 0.45 0.04 <0.01 5.6 (5.1–6.3) 5.8 (5.2–6.6) 5.8 (5.3–6.4) 0.86 0.86 0.86 We presented continuous variables as a mean ± s.d. or median (interquartile range), and categorical variables as a number (percentage). Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; CABG, coronary artery bypass graft; CCB, calcium channel blocker; CKD, chronic kidney disease; HGB, hemoglobin; HF, heart failure; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricle ejection fraction; MI, myocardial infarction; NYHA, New York Heart Association; OHA, oral hypoglycemic agents; PCI, percutaneous coronary intervention; PLT, platelet count; s.d., standard deviation; WBC, white blood cells; significant difference = p < 0.05. 2.2. Expression Profile of miRNAs in the Plasma of Group 1 Versus Group 2 and Group 3 We confirmed that the concentration of five preselected plasma miRNAs was different in patients with CAAD (Group 1) compared to patients with CAD (Group 2) and patients without significant coronary artery disease (Group 3) (Table 2). We identified a specific signature of plasma miRNAs up- or down-regulated in individual groups. Table 2. Deregulated miRNAs in the circulation of all studied groups. Up-Down Regulation Symbol hsa-miR-451a hsa-miR-23a-3p hsa-miR-125b-5p hsa-miR-210-3p hsa-miR-328-3p hsa-miR-425-3p hsa-miR-483-5p hsa-miR-16-5p hsa-miR-30d-5p hsa-miR-320d Group 2 Compared to Group 1 Group 3 Compared to Group 1 Group 3 Compared to Group 2 Fold Regulation p-Value Fold Regulation p-Value Fold Regulation p-Value −2.18 1.54 −1.51 −2.07 −1.08 −1.96 −1.28 −1.34 −1.18 −1.38 0.03 0.01 0.28 0.22 0.53 0.17 0.45 0.07 0.96 0.27 −2.64 2.11 −1.78 −4.71 −2.25 −3.30 −1.66 −1.23 −1.49 −1.46 0.01 <0.01 0.07 <0.01 0.03 0.01 0.30 0.15 0.15 0.04 −1.51 1.10 −1.51 −1.50 −1.88 −1.63 −1.47 −1.15 −1.59 −1.38 0.39 0.59 0.14 0.03 0.03 0.14 0.02 0.49 0.18 0.10 The calculation does not include data for the spike-in controls (UniSP2-6 and cel-miR-39-3p) or any blank spots. MiRNA markers selected for further analysis are marked in red. Int. J. Mol. Sci. 2023, 24, 5817 4 of 15 2.3. Comparison of Group 1 with Group 2 The univariate logistic regression analysis revealed three independent variables potentially increasing the risk of CAAD (Table 3). Table 3. The univariate and multivariate logistic regression analysis of clinical variables and miRNA expression profile in Group 1 compared to Group 2. Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW Variable Previous PCI Dyslipidemia Univariate Analysis Multivariate Analysis OR [95% CI] p-Value OR [95% CI] p-Value 0.3 [0.1; 0.9] 0.2 [0.1; 1.1] 0.03 0.06 0.2 [0.1; 0.7] 0.2 [0.1; 0.8] 0.01 0.02 We included0.7all[0.5; of0.9] them hsa-miR-451a in the0.03 multivariate logistic 0.7 [0.5; 0.99] regression 0.049 analysis. Mo percutaneous coronary inte (PCI) and dyslipidemia. Despite the low model fit (R2Pseudo = 0.11), it was sta We included all of them in the multivariate logistic regression analysis. Model 1 significant (p = 0.004). Both previous PCI and dyslipidemia significantly reduced included two independent clinical variables, previous percutaneous coronary intervention of CAAD. In Model Despite 2, miR-451a the model fit (R2Pseudo = 0.16) and w (PCI) and dyslipidemia. the lowimproved model fit (R2Pseudo = 0.11), it was statistically ciated with significant reduction ofdyslipidemia the risk of significantly CAAD (Table 3). the risk of significant (p = a 0.004). Both previous PCI and reduced CAAD.Comparative In Model 2, miR-451a improved the model fit (R2Pseudo = 0.16) and was associated and th analysis of Model 1 and Model 2 based on R2Pseudo with a significant reduction of the risk of CAAD (Table 3). hood-Ratio Test (LRT) (p = 0.038) revealed a significantly better fit of Model 2, in Comparative analysis of Model 1 and Model 2 based on R2Pseudo and the LikelihoodmiR-451a, than Model 1 (AIC = 83.6, AIC = 85.9, for Model 2 andmiR-451a, Model 1, resp Ratio Test (LRT) (p =of 0.038) revealed a significantly better fit of Model 2, including (Figure 1). than of Model 1 (AIC = 83.6, AIC = 85.9, for Model 2 and Model 1, respectively) (Figure 1). Abbreviations: OR,independent odds ratio; PCI, percutaneous intervention. cluded two clinicalcoronary variables, previous Figure 1. Comparison of logistic regression models between Group 1 and Group Figure 1. Comparison of logistic regression models between Group2.1 and Group 2. 2.4. Comparison between Group 1 and Group 3 2.4.The Comparison Group 1analysis and Group 3 six independent clinical variables univariate between logistic regression revealed and five miRNAs potentially increasing the risk of CAAD (Table 4). included all of The univariate logistic regression analysis revealed sixWe independent clinical v them in the multivariate logistic regression analysis. and five miRNAs potentially increasing the risk of CAAD (Table 4). We includ them in the multivariate logistic regression analysis. Table 4. The univariate logistic regression analysis of clinical variables and miRNA expres file in Group 1 compared to Group 3. Univariate Analysis Int. J. Mol. Sci. 2023, 24, 5817 5 of 15 Table 4. The univariate logistic regression analysis of clinical variables and miRNA expression profile in Group 1 compared to Group 3. Variable Univariate Analysis OR [95% CI] p-Value Previous MI 7.5 [1.5; 37.6] 0.01 P2Y12 inhibitor Aspirin WBC, 109 /L RBC, 109 /L Creatinine, mmol/L hsa-miR-328-3p hsa-miR-451a hsa-miR-23a-3p hsa-miR-210-3p hsa-miR-425-3p 36.3 [4.4; 229.0] 5.5 [1.4; 21.8] 1.2 [0.9; 1.6] 3.2 [1.1; 10.0] 1.02 [1.0; 1.1] 0.7 [0.5; 0.9] 0.6 [0.4; 0.9] 2.6 [1.5; 4.7] 0.7 [0.5; 0.8] 0.8 [0.6; 0.9] <0.01 0.01 0.06 0.04 0.03 <0.01 <0.01 <0.01 <0.01 0.01 Abbreviations: MI, myocardial infarction; RBC, red blood cells; WBC, white blood cells; OR, odds ratio. Model 1 included all independent clinical variables, such as previous myocardial infarction (MI), P2Y12 inhibitor and aspirin taking, white blood cell (WBC) count, red blood cell (RBC) count, and serum creatinine level. It was significantly associated with the risk of CAAD compared to patients without significant coronary artery lesions (p < 0.01). In addition, the goodness-of-fit was quite high (R2Pseudo = 0.46). Among the independent variables included in the model, only taking the P2Y12 inhibitor significantly increased the risk of CAAD (OR [95% CI] 80.3 [5.8; 1108.4, p < 0.01). In Model 2, miRNAs were added separately to Model 1. Only miR-328-3p improved the model fit (R2Pseudo = 0.62) and was associated with a significant reduction in the risk of CAAD (OR [95% CI] 0.21 [0.07; 0.67], p = 0.01) (Table 5). Table 5. The multivariate logistic regression analysis of clinical variables and miRNA expression profile in Group 1 compared to Group 3 (Model 2). Variable Previous MI P2Y12 inhibitor Aspirin WBC, 109 /L RBC, 109 /L Creatinine, mmol/L hsa-miR-328-3p Multivariate Analysis OR [95% CI] p-Value 7.0 [0.9; 51.7] 80.3 [5.8; 1108.4] 3.7 [0.4; 29.8] 1.3 [0.9; 1.9] 4.5 [0.8; 24.6] 1.0 [0.97; 1.1] 0.2 [0.1; 0.7] 0.06 <0.01 0.22 0.08 0.11 0.54 <0.01 Abbreviations: MI, myocardial infarction; RBC, red blood cells; WBC, white blood cells; OR, odds ratio. Comparative analysis of Model 1 and Model 2 based on R2Pseudo and LRT (p < 0.01) revealed a significantly better fit of Model 2, including miR-328-3p, than of Model 1 (AIC = 49.0, AIC = 61.4, for Model 2 and Model 1 respectively) (Figure 2). 2.5. Functional Enrichment Analysis Recent research reports that exogenous, circulating miRNAs can enter recipient cells and play a regulatory role in cells or organisms [15]. Assuming that miRNA is significantly deregulated in CAAD and CAD samples, and may be transported into epithelial, smooth muscle, or myocyte cells, we undertook in silico functional enrichment analysis. To identify biochemical signaling pathways potentially affected by miRNA modulated in CAAD/CAD patients, we analyzed the predicted target genes of the five significantly affected miRNAs (hsa-miR-451a, hsa-miR-23a-3p, hsa-miR-210-3p, hsa-miR-328-3p, hsa-miR-425-3p, hsamiR-483-5p) (Table 6). The target genes of the five identified miRNAs were extracted from Variable Int. J. Mol. Sci. 2023, 24, 5817 OR [95% CI] p-V Previous MI 7.0 [0.9; 51.7] 0. P2Y12 inhibitor 80.3 [5.8; 1108.4] <0 Aspirin 3.7 [0.4; 29.8] 0. 6 of 15 1.3 [0.9; 1.9] 0. WBC, 109/L 4.5 [0.8; 24.6] 0. RBC, 109/L Creatinine, mmol/L 1.0 [0.97; 1.1] 0. hsa-miR-328-3p [0.1; 0.7] <0 the miRWalk database (Table S1), and the intersection of identified0.2target genes from GSEA tool programs was chosen [15,16]. infarction; We correlated predicted target genes withwhite the KEGG Abbreviations: MI, myocardial RBC, red blood cells; WBC, blood cells; O (Kyoto ratio. Encyclopedia of Genes and Genomes) to identify potential biochemical pathways as the functional output of five deregulated miRNAs. The enrichment analysis represents pathways significantly enriched with target genes for miRNAs in CAAD and CAD patients Comparative analysis of Model 1 and Model 2 based on R2Pseudo and LRT ( in relation to the control group. Significantly (FDR < 0.05) enriched KEGG pathways are revealedinaTable significantly better of Model including miR-328-3p, than of Mode presented 6. Annotated notes fit about predicted2,target genes for the five dysregulated miRNAs in CAA and CAD groups 2 areand presented (Tables = 49.0, AIC = 61.4, for Model Modelin1supplementary respectively)tables (Figure 2).S1–S5). Figure 2. 2. Evaluation of logistic regression models tomodels compareto Group 1 and Group Figure Evaluation of logistic regression compare Group3.1 and Group 3. Table 6. Pathways regulated by miRNAs hsa-miR-451a, hsa-miR-23a-3p, hsa-miR-210-3p, hsa-miR-3282.5. Functional Enrichment Analysis 3p, hsa-miR-425-3p. The target genes of the five identified miRNAs were extracted from the miRWalk Recent thattarget exogenous, miRNAs can enter database, and theresearch intersectionreports of identified genes fromcirculating GSEA tool programs was chosen. We recipi correlated identified target genesrole with the Encyclopedia[15]. of Genes and Genomes) to point and play a regulatory in KEGG cells (Kyoto or organisms Assuming that miRNA is to biochemical pathways as potentially functional outputs of five deregulated miRNAs. cantly deregulated in CAAD and CAD samples, and may be transported into ep smooth muscle, we undertook silico functional enrichment a Pathwaysor myocyte cells, Genes p-Value,in Adjusted p-Value (BH) To identify biochemical signaling pathways potentially affected by miRNA modu IL6R NDUFV3 CAAD/CAD patients, we analyzed the predicted target genes of the five signific Non-alcoholic fatty liver PIK3R1 <0.01 0.01hsa-miR-328fected disease miRNAs (hsa-miR-451a,SDHD hsa-miR-23a-3p, hsa-miR-210-3p, (NAFLD) UQCRFS RAC1 AMPK signaling pathway TSC1, PIK3R1 PPARGC1A FOXO3 SCD <0.01 0.01 Int. J. Mol. Sci. 2023, 24, 5817 7 of 15 Table 6. Cont. Pathways Genes p-Value, Adjusted p-Value (BH) Autophagy TSC1 ATG7 PIK3R1 HMGB1 VMP1 <0.01 <0.01 Regulation of actin cytoskeleton VAV3 EZR PIK3R1 MYH10 RAC1 0.02 0.06 Thermogenesis TSC1 NDUFV3 SDHD PPARGC1A UQCRFS1 0.03 0.06 Huntington disease NDUFV SDHD PPARGC1A ACTR1A UQCRFS1 0.05 0.08 PI3K-Akt signaling pathway TSC1 IL6R PIK3R1 FOXO3 RAC1 0.12 0.18 Pathways in cancer MITF IL6R TPM3 RUNX1T1 PIK3R1 RAC1 0.20 0.25 Herpes simplex virus 1 infection TSC1 ZNF275 ZNF268 ZNF701 PIK3R1 0.29 0.32 Metabolic pathways NDUFV3 HMGCS1 GLS SDHD STT3A MGAT5 FUT4 SCD UQCRFS1 STT3B 0.62 0.62 The KEGG enrichment analysis indicated that the five miRNAs targeted genes are involved in ten pathways. The AMPK signaling pathway is involved in vast processes in the heart [17]: e.g., modulation of vascular inflammation and vascular smooth muscle cells phenotype [18]. Several miR-23-3p potential targets, in the AMPK pathway, were selected: TSC1, PIK3R1, PPARGC1A, FOXO3, and one miR-328-3p target SCD. Endothelial cells, vascular smooth muscle cells, and macrophages play important roles in the development of CAD. In the etiology of atherosclerosis, these cells undergo Int. J. Mol. Sci. 2023, 24, 5817 8 of 15 two types of programmed cell death apoptosis and autophagy [19,20]. Numerous studies have revealed recently that autophagy-related genes (ATGs) are involved in endothelial and smooth-muscle inappropriate autophagy in the processes of atherogenesis [21]. In our in silico analysis, several genes were selected as potential targets for miRNAs. The following miRNAs were selected as regulators of autophagy-related genes: miR-328-3p HMGB1, miR-23-3p TSC1, PIK3R1, miR-210-3p ATG7, and VMP1. An example is HMGB1, the intranuclear non-histone DNA binding protein, which contributes to the stabilization of nucleosomes and the development of atherosclerosis [22]. 3. Discussion Coronary artery aneurysmal disease (CAAD) is a progressive, gradual coronary artery abnormality that mainly results in recurrent angina or, less commonly, severe complications such as distal embolization and myocardial infarction. The critical factor that initiates and deteriorates the course of CAAD is atherosclerosis. Asymptomatic course and progression make an early diagnosis of CAAD difficult. miRNAs play a role in the mechanism of atherosclerosis, and atherosclerotic coronary artery disease is characterized by specific signatures of miRNA in tissue and plasma assays [23–26]. However, scarce data concerning miRNAs involved in CAAD pathology prompted us to study five circulating miRNAs: MiR-125b-5p, miR-210-3p, miR-328-3p, miR-425-3p, and miR-483-5p, which we preselected and described in our former study [14] as potentially associated with CAAD. This study evaluated the predictive role of circulating miRNAs for CAAD. Through a sampling process, we found miR-451a and miR-328 to be the most effective for CAAD prediction in a general Greater Poland population. Adding miR-451a and miR-328-3p to traditional risk factors significantly improved the detection of CAAD. We found that a higher circulating level of miR-328-3p was significantly associated with a higher risk of CAAD than the control group. On the other hand, the higher plasma level of miR-451a was predictive of the higher risk of CAAD compared to CAD. As described in the methodology, miRNAs appeared as stable molecules and were recognized as an advantage of good potential markers for cardiovascular issues [27]. Selecting miRNA markers from dozens of circulating miRNA requires a multistep procedure. We overcame this difficulty and applied RT-qPCR to select miRNA markers for CAAD. Since the presence of miRNAs in plasma was first mentioned in the literature, circulating miRNAs have been reported as a potential biomarker of heart failure and CAD [23,28]. In contrast to the numerous studies in CAD cohorts, the predictive value of miRNAs in CAAD is yet to be discovered. CAAD is usually detected during coronary angiography. Although data suggest a few CAD biomarkers, the need for reliable CAAD biomarkers, discriminating it from CAD, remains a limitation in the clinical practice. Our study suggested miRNA as a minimally invasive biomarker to improve pre-procedure diagnosis of CAAD. We applied the RT-qPCR method using a PCR array system. The custom plates for the RT-qPCR array were designed as in our former study [14]. We confirmed that the concentration of five preselected plasma miRNAs was different in patients with CAAD (Group 1) compared to patients with ischemic heart disease (Group 2) and patients without significant coronary artery disease (Group 3). By applying logistic regression, we compared two models without and with miRNA. The model without miRNA to compare CAAD (Group 1) and CAD (Group 2) included two independent variables: previous percutaneous coronary intervention (PCI) and dyslipidemia. By adding miR-451a to the model, the prediction increased. miR-451a appeared to be a significant marker of CAAD compared to the group of patients with CAD. The model without miRNA to compare CAAD (Group 1) and the control (Group 3) included six independent variables, previous MI, P2Y12 inhibitor or aspirin taking, WBC count, RBC count, and serum creatinine level. By adding miR-328-3p to the model, the prediction increased. miR-328-3p appeared to be a significant marker of CAAD compared to patients without significant CAD. Int. J. Mol. Sci. 2023, 24, 5817 9 of 15 The available papers concerning miRNA as a marker of CAAD are scarce. RT-qPCR evaluated let-7i-3p as a biomarker to identify CAAD patients [29]. Serum exosomal let-7i-3p was proposed as a diagnostic marker for CAAD in patients with Kawasaki disease [29]. miRNAs miR-451a and miR-328, which we elaborated on in the present study, have already been proposed as potential markers of heart failure (HF) and CAD. miR-451a expression is elevated in the plasma of patients with acute myocardial infarction (AMI) compared with unstable angina pectoris and healthy control groups [30]. Data suggest that miR-451a regulates ischemic heart injury [31]. Overexpression of miR451a in mice significantly attenuated Ang II-induced cardiac fibrosis and inflammation. miR-451a directly targeted transcription factor TBX1 expression, thus decreasing TBX1dependent TGFB1 gene expression. In our study, we observed a higher plasma level of miR-451a in CAAD patients compared to CAD patients. Therefore, we conclude that miR451a could be involved in CAAD pathomechanism since CAAD is a severe consequence of coronary vessel inflammation and atherosclerosis. Recently, miR-451a has been also found to inhibit the expression of both MMP-2 and MMP-9 in cardiomyocytes by macrophage migration inhibitory factor (MIF) [32]. In our study, miR-328 was increased in the CAAD group compared to the control group. An increased miR-328 plasma level was reported in patients with AMI, strongly associated with increased risk of mortality or heart failure [33]. Elevated expression of miR-328 is also a marker of cardiac fibrosis during MI. This observation has recently been demonstrated in vitro and explained by the paracrine mechanism. miR-328 is secreted by cardiomyocytes promoting fibrosis by regulating adjacent fibroblasts [34]. Transgenic mice with miR-328 overexpression in the heart induced cardiac hypertrophy [35]. miR-328 reduces the transcript of sarco/endoplasmic reticulum Ca2+ -ATPase (SERCA2a, ATP2A2), a transporter of Ca2+ from the cytosol into the sarcoplasmic reticulum [35–37]. miR-23a, miR-210, and miR-425 were significantly regulated in the CAAD group versus the control group, but did not significantly influence a logistic regression model improving prediction. Our study demonstrated decreased miR-23a in the CAAD group versus the control. Previous studies suggested increased plasma miR-23a in CAD patients [38,39]. It has been suggested to be used to predict the presence and severity of coronary lesions in patients with CAD. In our study, miR-23a has not been significantly modulated in the CAD group. So far, several studies evaluating the prognostic value of single circulating miRNAs in cardiovascular disease have been reported [40]. miR-210 derived from peripheral blood predict mortality and represent a valuable biomarker for risk estimation in CAD [41,42]. Circulating miR-210 improved the prediction model, including AUC (area under the receiver-operating characteristic curve). In our study, miR-210 was increased in CAAD vs. control but did not improve the model, including clinical parameters. miRbase miR-425-3p is a minor transcript in mammalian cells [43]. Several studies demonstrated that predominant miR-425-p functions as a negative regulator of cardiac fibrosis by suppressing TGFB1 (encoding TGFβ1) expression [44]. Our finding shows a new look at the miRNA signatures for plasma samples from CAAD patients. It provides a composed clinical and molecular predictive model, benefiting the diagnosis of CAAD patients. Developing a method using miRNAs as CAAD markers could improve the early diagnosis of CAAD and identify patients at increased risk of progression and complications. We proposed a new model of CAAD prediction using miR-451a and miR-328 in addition to independent clinical variables. Our approach with logistic regression models revealed a significant increase in the predictability of discrimination between the CAAD group and the CAD group, as well as the CAAD group against the control group, by adding the status of miR-451a and miR-328, respectively.. This preclinical assessment of potential miRNA markers could be developed in the future in reliable tests for early diagnosis of CAAD. Int. J. Mol. Sci. 2023, 24, 5817 10 of 15 Our achievement has limits in the number of patients compared to the study. According to several studies, the condition, time of the day, and several other parameters could influence miRNA levels and should be controlled and standardized in future studies. In future endeavours is the question of how miR-451a and miR-328 contribute to CAAD development. The important question also concerns target cells for miRNA and the source of abnormal levels of selected markers in plasma. 4. Materials and Methods 4.1. Study Design and Patient Selection The overall cohort enrolled two hundred fifty patients undergoing coronary angiography for angina symptoms between April 2020 and December 2021. The study cohort included 35 consecutive patients with CAAD (Group 1), and two groups of 35 patients Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 10 of 14 matched Group 1 in terms of sex and age from the overall cohort (Group 2 and Group 3) (Figure 3). Group 2 included patients with angiographically documented CAD, while Group 3 enrolled patients with normal coronary arteries (NCA) assessed during coronary disease; (v) severe inflammatory diseases; systemic connective tissue angiography. Due to hemolysis inand the malignant two CAAD patient (vi) samples, we finally analyzed disease; (vii) biological treatment; and (viii) no informed consent. 33 patients. Figure3.3.Flowchart Flowchartof ofthe thestudy studypopulation. population. CAAD, CAAD, coronary coronary artery arteryaneurysmal aneurysmal disease; disease; CAD, CAD, Figure coronary artery disease; NCA, normal coronary arteries. coronary artery disease; NCA, normal coronary arteries. 4.2. Blood Sampleswere Collection and Evaluation of miRNA Profiles according to the European All patients qualified for coronary angiography, Society of previous Cardiology (ESC) guidelines [45]. According to the available data,samples CAAD [14]. is a Our study precisely described obtaining and preparing blood diffuse or focal dilatation of the coronary artery with a diameter of 1.5 times the adjacent We treated the new collection of blood samples according to the same procedure denormal also enrolled patients(10 with concomitant coronary stenosis in scribedsegment. here. We We collected blood samples mL) from patients on theartery first day after the Group 1. The angiographic criteria for CAD included: coronary artery stenosis > 90% or coronary angiography and processed them within 30 min. Samples were centrifuged at intermediate stenosis with documented ischemia orwas hemodynamically 1300× g for 15 min at(50–90%) room temperature. The supernatant transferred to significant, RNase-free defined as either fractional flow reserve (FFR) ≤ 0.80 or an instantaneous wave-free ratio tubes and then stored at −80 °C. (iFR) ≤ 0.89. Patients included in Group 3 had normal and echocardiography The miRNAs were isolated from individual 200 ECG μL frozen plasma using theapart miRfrom noSerum/Plasma significant lesions in the coronary In addition, showedGermany) no evidence Neasy Advanced Kit (Cat.arteries. No 217204, Qiagen, they Dusseldorf, acof ischemia during non-invasive exercise testing. cording to the manufacturer’s instructions. To correct sample-to-sample variation, a synthetic set of spike-ins, UniSp2, UniSp4, and UniSp5 was applied (RNA Spike-In Kit for reverse transcription (RT) Cat. No 339390, Qiagen, Dusseldorf, Germany). Three of the RNA spike-in templates (UniSp2, UniSp4, and UniSp5) were provided premixed in one vial, each at a different concentration in 100-fold increments (UniSp2 (2 fmol/μL), UniSp4 (0.02 fmol/μL), and UniSp5 (0.0002 fmol/μL)). Before starting the RNA isolation proce- Int. J. Mol. Sci. 2023, 24, 5817 11 of 15 Patients meeting the following criteria were excluded from the study: (i) acute coronary syndrome; (ii) elevated levels of myocardial injury markers (troponin I or creatine kinase); (iii) severe hepatic and renal dysfunction present or history; (iv) hematological disease; (v) severe inflammatory and malignant diseases; (vi) systemic connective tissue disease; (vii) biological treatment; and (viii) no informed consent. 4.2. Blood Samples Collection and Evaluation of miRNA Profiles Our previous study precisely described obtaining and preparing blood samples [14]. We treated the new collection of blood samples according to the same procedure described here. We collected blood samples (10 mL) from patients on the first day after the coronary angiography and processed them within 30 min. Samples were centrifuged at 1300× g for 15 min at room temperature. The supernatant was transferred to RNase-free tubes and then stored at −80 ◦ C. The miRNAs were isolated from individual 200 µL frozen plasma using the miRNeasy Serum/Plasma Advanced Kit (Cat. No 217204, Qiagen, Dusseldorf, Germany) according to the manufacturer’s instructions. To correct sample-to-sample variation, a synthetic set of spike-ins, UniSp2, UniSp4, and UniSp5 was applied (RNA Spike-In Kit for reverse transcription (RT) Cat. No 339390, Qiagen, Dusseldorf, Germany). Three of the RNA spike-in templates (UniSp2, UniSp4, and UniSp5) were provided premixed in one vial, each at a different concentration in 100-fold increments (UniSp2 (2 fmol/µL), UniSp4 (0.02 fmol/µL), and UniSp5 (0.0002 fmol/µL)). Before starting the RNA isolation procedure, one microliter of this RNA spike-in mix per RNA prepared with 60 µL Lysis Buffer RPL was aliquoted. Approximate RNA quantity and quality were estimated using a NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). Four microliters of each total RNA were reversely transcribed with the miRCURY LNA RT Kit (Cat. No 339340, Qiagen, Dusseldorf, Germany). The UniSp6 RNA spike-in from the miRCURY LNA RT Kit and cel-miR-39-3p RNA from miRCURY LNA RT Kit were applied as controls. Twenty microliters of cDNA product of the miRCURY LNA RT Kit were applied for each quantitative PCR (qPCR). qPCR was performed using miRCURY LNA SYBR Green PCR Kit (Cat. No 339345, 339346, 339347, Qiagen, Dusseldorf, Germany) with miRCURY LNA Custom PCR Panels (Qiagen) (Table 7). Table 7. miRCURY LNA Custom PCR Panels (Qiagen) primers on a 96-well plate. Columns 1 and 2 were repeated six times. Six samples were analyzed on a single 96-well plate. A B C D E F G H 1 2 UniSp2 UniSp4 UniSp5 UniSp6 cel-miR-39-3p UniSp3 hsa-miR-451a hsa-miR-23a-3p hsa-miR-125b-5p hsa-miR-210-3p hsa-miR-328-3p hsa-miR-425-3p hsa-miR-483-5p hsa-miR-16-5p hsa-miR-30d-5p hsa-miR-320d Table 8 summarizes the genes analyzed in the custom plate, including preselected miRNA markers and spike-in controls applied in the analysis: the control, hemolysis markers, and reference markers. We provided a detailed description of the spike-in controls in the appendix (Appendix A). Int. J. Mol. Sci. 2023, 24, 5817 12 of 15 Table 8. Genes analyzed on the Custom plate. Position Mature ID Meaning of the Gene 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 UniSp2 UniSp4 UniSp5 UniSp6 cel-miR-39-3p UniSp3 hsa-miR-451a hsa-miR-23a-3p hsa-miR-125b-5p hsa-miR-210-3p hsa-miR-328-3p hsa-miR-425-3p hsa-miR-483-5p hsa-miR-16-5p hsa-miR-30d-5p hsa-miR-320d Control Control Control Control Control Control Hemolysis marker Hemolysis marker Marker Marker Marker Marker Marker Reference marker Reference marker Reference marker 4.3. Statistical Analysis We presented all continuous variables as means with standard deviation for normal distribution or medians (upper and lower quartile) for non-normal distribution. We tested the normality of the distribution of variables using the Kolmogorov–Smirnov test and presented categorical variables as counts and percentages or frequencies. We assessed the significance of differences between the mean values of the continuous data consistent with the normal distribution using one-way ANOVA and Tukey’s Test. To compare the continuous data inconsistent with the normal distribution, we used Mann–Whitney and Kruskal–Wallis tests with Benjamini–Hochberg post hoc analysis and compared categorical variables using the χ2 test. We performed an analysis of array data with GENEGLOBE online software (https://geneglobe.qiagen.com/pl/analyze, accessed on 12 January 2023). We used the global mean as a reference for the PCR-array analysis and calculated the p values based on a Student’s t-test of replicate 2(−∆CT) values for each miRNA in the control and treatment groups. Subsequently, we based the p-value calculation on a parametric, two-sample equal variance, unpaired, two-tailed analysis. In the current analysis, the selection criteria to elaborate potential miRNA markers were the fold-regulation ≥ 2 and p ≤ 0.05. We conducted the univariate logistic regression analysis to evaluate confounding factors for each CAAD comparison (Group 1 vs. Group 2 and Group 1 vs. Group 3). The confounding factor was selected based on a p value < 0.1. The multivariate logistic regression analysis was conducted for each comparison with confounding factors that met the criteria in univariate analysis (Model 1). The following step was to create the Model 2 by adding the analyzed miRNA to Model 1, which had an expression that was significantly different between the compared groups (Table 2). We assessed the model fit for Model 1 and Model 2 using the Akaike information criterion (AIC) and pseudo-R-squared values. The model with the lowest AIC value was considered the best. The higher pseudo-Rsquared value indicated a better model fit. All statistical analyses were two-sided and were conducted with PQStat Software (PQStat v.1.8.0.476, Poznań, Poland). A p value < 0.05 was considered a statistical significance. 5. Conclusions The clinical use of the studied miRNAs as markers requires validation on a bigger cohort or advanced sequencing techniques. However, preliminary analysis indicates that miR-451a is a significant marker of CAAD compared to patients without concomitant aneurysms. In turn, miR-328-3p is a significant marker of CAAD compared to patients without significant changes in the coronary arteries. Int. J. Mol. Sci. 2023, 24, 5817 13 of 15 Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijms24065817/s1. Author Contributions: Conceptualization, S.I., T.L. and A.A.; methodology, S.I. and T.L.; software, A.C. and G.K.; validation, A.H. and T.L.; formal analysis, T.L. and S.I.; investigation, S.I. and T.L.; resources, M.G. and M.L.; data curation, A.C., K.M. and G.K.; writing—original draft preparation, S.I. and T.L.; writing—review and editing, P.W., A.C., P.P.J., M.G., M.L. and A.A.; visualization, S.I.; supervision, P.P.J., M.G., M.L. and A.A.; project administration, S.I.; funding acquisition, A.A. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by Polish Cardiac Society, grant number 09/2018. Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Ethics Committee of Poznan University of Medical Sciences (protocol code, 985/18; date of approval, 11 October 2018). Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available due to confidential genetic data. Conflicts of Interest: The authors declare no conflict of interest. Appendix A The following spike-in controls were applied: (1) the RNA isolation control spike-in templates UniSp2, UniSp4, and UniSp5 are provided as different concentrations in 100-fold increments, and their primers are present on plates miRCURY LNA miRNA QC PCR Panel, miRNeasy Serum/Plasma Kit (Cat. No 217184) and our custom plates; (2) the cDNA synthesis controls the UniSp6 RNA spike-in template in combination with cel-miR-39-3p, a synthetic version of a C. elegans miRNA; their primers are present on plates miRCURY LNA miRNA QC PCR Panel, miRNeasy Serum/Plasma Kit (Cat. No 217184) and our custom plates; (3) the PCR efficiency control UniSp3, with their primers being present on plates miRCURY LNA miRNA QC PCR Panel, miRNeasy Serum/Plasma Kit (Cat. No 217184) and our custom plates; (4) the hemolysis markers, the red-blood-cell-specific miR451, and the stable miR-23a were used to monitor hemolysis. A difference of crossing-point Ct values for miR-23a and miR-451 (∆Cq) < 5 in human serum or plasma represents nonhemolyzed samples. If the ∆Cq ≥ 7, there is an increased risk that the samples are affected by hemolysis; their primers are present on plates miRCURY LNA miRNA QC PCR Panel, miRNeasy Serum/Plasma Kit (Cat. No 217184) and our custom plates; (5) three reference markers were chosen to normalize variation of RNA extraction efficiency by GeneGlobe using protocols geNorm and Normfinder (Table 2). References 1. 2. 3. 4. 5. 6. 7. 8. Fichtlscherer, S.; Zeiher, A.M.; Dimmeler, S. Circulating MicroRNAs: Biomarkers or Mediators of Cardiovascular Diseases? Arterioscler. Thromb. Vasc. Biol. 2011, 31, 2383–2390. [CrossRef] [PubMed] Huntzinger, E.; Izaurralde, E. Gene Silencing by MicroRNAs: Contributions of Translational Repression and MRNA Decay. Nat. Rev. Genet. 2011, 12, 99–110. [CrossRef] Rocchi, A.; Chiti, E.; Maiese, A.; Turillazzi, E.; Spinetti, I. MicroRNAs: An Update of Applications in Forensic Science. Diagnostics 2020, 11, 32. [CrossRef] Swaye, P.S.; Fisher, L.D.; Litwin, P.; Vignola, P.A.; Judkins, M.P.; Kemp, H.G.; Mudd, J.G.; Gosselin, A.J. Aneurysmal Coronary Artery Disease. Circulation 1983, 67, 134–138. [CrossRef] Baman, T.S.; Cole, J.H.; Devireddy, C.M.; Sperling, L.S. Risk Factors and Outcomes in Patients with Coronary Artery Aneurysms. Am. J. Cardiol. 2004, 93, 1549–1551. [CrossRef] [PubMed] Warisawa, T.; Naganuma, T.; Tomizawa, N.; Fujino, Y.; Ishiguro, H.; Tahara, S.; Kurita, N.; Nojo, T.; Nakamura, S.; Nakamura, S. High Prevalence of Coronary Artery Events and Non-Coronary Events in Patients with Coronary Artery Aneurysm in the Observational Group. Int. J. Cardiol. Heart Vasc. 2016, 10, 29–31. [CrossRef] [PubMed] Markis, J.E.; Joffe, C.D.; Cohn, P.F.; Feen, D.J.; Herman, M.V.; Gorlin, R. Clinical Significance of Coronary Arterial Ectasia. Am. J. Cardiol. 1976, 37, 217–222. [CrossRef] Kühl, M.; Varma, C. A Case of Acute Coronary Thrombosis in Diffuse Coronary Artery Ectasia. J. Invasive Cardiol. 2008, 20, E23–E25. Int. J. Mol. Sci. 2023, 24, 5817 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 14 of 15 Doi, T.; Kataoka, Y.; Noguchi, T.; Shibata, T.; Nakashima, T.; Kawakami, S.; Nakao, K.; Fujino, M.; Nagai, T.; Kanaya, T.; et al. Coronary Artery Ectasia Predicts Future Cardiac Events in Patients With Acute Myocardial Infarction. Arterioscler. Thromb. Vasc. Biol. 2017, 37, 2350–2355. [CrossRef] Janus, M.; Iwańczyk, S.; Stanisławska, K.; Mularek-Kubzdela, T.; Lesiak, M. Endovascular Closure of Coronary Artery to Pulmonary Artery Fistula with Giant Aneurysm. Kardiol. Pol. 2022, 80, 621–622. [CrossRef] Iwańczyk, S.; Borger, M.; Kamiński, M.; Chmara, E.; Cieślewicz, A.; Tykarski, A.; Radziemski, A.; Krasiński, Z.; Lesiak, M.; Araszkiewicz, A. Inflammatory response in patients with coronary artery ectasia and coronary artery disease. Kardiol. Pol. 2019, 77, 713–715. [CrossRef] [PubMed] Iwańczyk, S.; Lehmann, T.; Cieślewicz, A.; Radziemski, A.; Malesza, K.; Wrotyński, M.; Jagodziński, P.P.; Grygier, M.; Lesiak, M.; Araszkiewicz, A. Involvement of Angiogenesis in the Pathogenesis of Coronary Aneurysms. Biomedicines 2021, 9, 1269. [CrossRef] [PubMed] Wojciechowska, A.; Braniewska, A.; Kozar-Kamińska, K. MicroRNA in Cardiovascular Biology and Disease. Adv. Clin. Exp. Med. 2017, 26, 865–874. [CrossRef] Iwańczyk, S.; Lehmann, T.; Cieślewicz, A.; Radziemski, A.; Malesza, K.; Wrotyński, M.; Jagodziński, P.P.; Grygier, M.; Lesiak, M.; Araszkiewicz, A. Circulating MicroRNAs in Patients with Aneurysmal Dilatation of Coronary Arteries. Exp. Ther. Med. 2022, 23, 404. [CrossRef] Wang, L.; Liu, J.; Xu, B.; Liu, Y.-L.; Liu, Z. Reduced Exosome MiR-425 and MiR-744 in the Plasma Represents the Progression of Fibrosis and Heart Failure. Kaohsiung J. Med. Sci. 2018, 34, 626–633. [CrossRef] [PubMed] Subramanian, A.; Tamayo, P.; Mootha, V.K.; Mukherjee, S.; Ebert, B.L.; Gillette, M.A.; Paulovich, A.; Pomeroy, S.L.; Golub, T.R.; Lander, E.S.; et al. Gene Set Enrichment Analysis: A Knowledge-Based Approach for Interpreting Genome-Wide Expression Profiles. Proc. Natl. Acad. Sci. USA 2005, 102, 15545–15550. [CrossRef] Zaha, V.G.; Young, L.H. AMP-Activated Protein Kinase Regulation and Biological Actions in the Heart. Circ. Res. 2012, 111, 800–814. [CrossRef] [PubMed] Lei, L.; Zhou, Y.; Wang, T.; Zheng, Z.; Chen, L.; Pan, Y. Activation of AMP-Activated Protein Kinase Ablated the Formation of Aortic Dissection by Suppressing Vascular Inflammation and Phenotypic Switching of Vascular Smooth Muscle Cells. Int. Immunopharmacol. 2022, 112, 109177. [CrossRef] Zhao, F.; Satyanarayana, G.; Zhang, Z.; Zhao, J.; Ma, X.-L.; Wang, Y. Endothelial Autophagy in Coronary Microvascular Dysfunction and Cardiovascular Disease. Cells 2022, 11, 2081. [CrossRef] [PubMed] Ni, D.; Mo, Z.; Yi, G. Recent Insights into Atherosclerotic Plaque Cell Autophagy. Exp. Biol. Med. 2021, 246, 2553–2558. [CrossRef] [PubMed] Henderson, J.M.; Weber, C.; Santovito, D. Beyond Self-Recycling: Cell-Specific Role of Autophagy in Atherosclerosis. Cells 2021, 10, 625. [CrossRef] Naglova, H.; Bucova, M. HMGB1 and Its Physiological and Pathological Roles. Bratisl. Lekárske Listy 2012, 113, 163–171. [CrossRef] [PubMed] Vavassori, C.; Cipriani, E.; Colombo, G.I. Circulating MicroRNAs as Novel Biomarkers in Risk Assessment and Prognosis of Coronary Artery Disease. Eur. Cardiol. Rev. 2022, 17, e06. [CrossRef] Samra, M.; Srivastava, K. Non-Coding RNA and Their Potential Role in Cardiovascular Diseases. Gene 2023, 851, 147011. [CrossRef] [PubMed] Kabłak-Ziembicka, A.; Badacz, R.; Przewłocki, T. Clinical Application of Serum MicroRNAs in Atherosclerotic Coronary Artery Disease. J. Clin. Med. 2022, 11, 6849. [CrossRef] Bergami, M.; Fabin, N.; Cenko, E.; Bugiardini, R.; Manfrini, O. MicroRNAs as Potential Biomarkers in Coronary Artery Disease. Curr. Top. Med. Chem. 2023, 23. [CrossRef] Shen, N.-N.; Wang, J.-L.; Fu, Y.-P. The MicroRNA Expression Profiling in Heart Failure: A Systematic Review and Meta-Analysis. Front. Cardiovasc. Med. 2022, 9, 856358. [CrossRef] Sigutova, R.; Evin, L.; Stejskal, D.; Ploticova, V.; Svagera, Z. Specific MicroRNAs and Heart Failure: Time for the next Step toward Application? Biomed. Pap. Med. Fac. Univ. Palacky Olomouc Czechoslov. 2022, 166, 359–368. [CrossRef] Wang, Y.-F.; Lian, X.-L.; Zhong, J.-Y.; Su, S.-X.; Xu, Y.-F.; Xie, X.-F.; Wang, Z.-P.; Li, W.; Zhang, L.; Che, D.; et al. Serum Exosomal MicroRNA Let-7i-3p as Candidate Diagnostic Biomarker for Kawasaki Disease Patients with Coronary Artery Aneurysm. IUBMB Life 2019, 71, 891–900. [CrossRef] Xu, L.; Tian, L.; Yan, Z.; Wang, J.; Xue, T.; Sun, Q. Diagnostic and Prognostic Value of MiR-486-5p, MiR-451a, MiR-21-5p and Monocyte to High-Density Lipoprotein Cholesterol Ratio in Patients with Acute Myocardial Infarction. Heart Vessel. 2023, 38, 318–331. [CrossRef] [PubMed] Deng, H.-Y.; He, Z.-Y.; Dong, Z.-C.; Zhang, Y.-L.; Han, X.; Li, H.-H. MicroRNA-451a Attenuates Angiotensin II-Induced Cardiac Fibrosis and Inflammation by Directly Targeting T-Box1. J. Physiol. Biochem. 2022, 78, 257–269. [CrossRef] Scrimgeour, N.R.; Wrobel, A.; Pinho, M.J.; Høydal, M.A. MicroRNA-451a Prevents Activation of Matrix Metalloproteinases 2 and 9 in Human Cardiomyocytes during Pathological Stress Stimulation. Am. J. Physiol. Cell Physiol. 2020, 318, C94–C102. [CrossRef] [PubMed] He, F.; Lv, P.; Zhao, X.; Wang, X.; Ma, X.; Meng, W.; Meng, X.; Dong, S. Predictive Value of Circulating MiR-328 and MiR-134 for Acute Myocardial Infarction. Mol. Cell. Biochem. 2014, 394, 137–144. [CrossRef] [PubMed] Int. J. Mol. Sci. 2023, 24, 5817 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 15 of 15 Zhao, D.; Li, C.; Yan, H.; Li, T.; Qian, M.; Zheng, N.; Jiang, H.; Liu, L.; Xu, B.; Wu, Q.; et al. Cardiomyocyte Derived MiR-328 Promotes Cardiac Fibrosis by Paracrinely Regulating Adjacent Fibroblasts. Cell. Physiol. Biochem. 2018, 46, 1555–1565. [CrossRef] Li, C.; Li, X.; Gao, X.; Zhang, R.; Zhang, Y.; Liang, H.; Xu, C.; Du, W.; Zhang, Y.; Liu, X.; et al. MicroRNA-328 as a Regulator of Cardiac Hypertrophy. Int. J. Cardiol. 2014, 173, 268–276. [CrossRef] [PubMed] Lu, Y.; Zhang, Y.; Wang, N.; Pan, Z.; Gao, X.; Zhang, F.; Zhang, Y.; Shan, H.; Luo, X.; Bai, Y.; et al. MicroRNA-328 Contributes to Adverse Electrical Remodeling in Atrial Fibrillation. Circulation 2010, 122, 2378–2387. [CrossRef] [PubMed] Du, W.; Liang, H.; Gao, X.; Li, X.; Zhang, Y.; Pan, Z.; Li, C.; Wang, Y.; Liu, Y.; Yuan, W.; et al. MicroRNA-328, a Potential Anti-Fibrotic Target in Cardiac Interstitial Fibrosis. Cell. Physiol. Biochem. 2016, 39, 827–836. [CrossRef] Ghafouri-Fard, S.; Gholipour, M.; Taheri, M. Role of MicroRNAs in the Pathogenesis of Coronary Artery Disease. Front. Cardiovasc. Med. 2021, 8, 632392. [CrossRef] Wang, S.; He, W.; Wang, C. MiR-23a Regulates the Vasculogenesis of Coronary Artery Disease by Targeting Epidermal Growth Factor Receptor. Cardiovasc. Ther. 2016, 34, 199–208. [CrossRef] Guan, Y.; Song, X.; Sun, W.; Wang, Y.; Liu, B. Effect of Hypoxia-Induced MicroRNA-210 Expression on Cardiovascular Disease and the Underlying Mechanism. Oxidative Med. Cell. Longev. 2019, 2019, 4727283. [CrossRef] Karakas, M.; Schulte, C.; Appelbaum, S.; Ojeda, F.; Lackner, K.J.; Münzel, T.; Schnabel, R.B.; Blankenberg, S.; Zeller, T. Circulating MicroRNAs Strongly Predict Cardiovascular Death in Patients with Coronary Artery Disease-Results from the Large AtheroGene Study. Eur. Heart J. 2017, 38, 516–523. [CrossRef] [PubMed] Li, T.; Cao, H.; Zhuang, J.; Wan, J.; Guan, M.; Yu, B.; Li, X.; Zhang, W. Identification of MiR-130a, MiR-27b and MiR-210 as Serum Biomarkers for Atherosclerosis Obliterans. Clin. Chim. Acta 2011, 412, 66–70. [CrossRef] [PubMed] Landgraf, P.; Rusu, M.; Sheridan, R.; Sewer, A.; Iovino, N.; Aravin, A.; Pfeffer, S.; Rice, A.; Kamphorst, A.O.; Landthaler, M.; et al. A Mammalian MicroRNA Expression Atlas Based on Small RNA Library Sequencing. Cell 2007, 129, 1401–1414. [CrossRef] Wei, F.; Ren, W.; Zhang, X.; Wu, P.; Fan, J. MiR-425-5p Is Negatively Associated with Atrial Fibrosis and Promotes Atrial Remodeling by Targeting CREB1 in Atrial Fibrillation. J. Cardiol. 2022, 79, 202–210. [CrossRef] [PubMed] Neumann, F.J.; Sousa-Uva, M.; Ahlsson, A.; Alfonso, F.; Banning, A.P.; Benedetto, U.; Byrne, R.A.; Collet, J.P.; Falk, V.; Head, S.J.; et al. 2018 ESC/EACTS Guidelines on myocardial revascularization. Eur. Heart J. 2019, 40, 87–165. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
https://openalex.org/W4283759783	https://hal.science/hal-03874607/document	English	null	Tracking transitional probabilities and segmenting auditory sequences are dissociable processes in adults and neonates	Developmental science	2,022	cc-by	14,109	Tracking transitional probabilities and segmenting auditory sequences are dissociable processes in adults and neonates Lucas Benjamin, Ana Fló, Marie Palu, Shruti Naik, Lucia Melloni, Ghislaine Dehaene-Lambertz Lucas Benjamin, Ana Fló, Marie Palu, Shruti Naik, Lucia Melloni, Ghislaine Dehaene-Lambertz To cite this version: Lucas Benjamin, Ana Fló, Marie Palu, Shruti Naik, Lucia Melloni, et al.. Tracking transitional probabilities and segmenting auditory sequences are dissociable processes in adults and neonates. Developmental Science, 2022, 2022, pp.e13300. 10.1111/desc.13300. hal-03874607 Tracking transitional probabilities and segmenting auditory sequences are dissociable processes in adults and neonates KEY WO RDS KEY WO RDS EEG, language learning, neonates, prosody, sequence learning, statistical learning Tracking transitional probabilities and segmenting auditory sequences are dissociable processes in adults and neonates Lucas Benjamin1 Ana Fló1 Marie Palu1 Shruti Naik1 Lucia Melloni2,3 Ghislaine Dehaene-Lambertz1 1Cognitive Neuroimaging Unit, CNRS ERL 9003, INSERM U992, CEA, Université Paris-Saclay, NeuroSpin Center, Gif-sur-Yvette, Île-de-France, France R E S E A R C H A R T I C L E R E S E A R C H A R T I C L E HAL Id: hal-03874607 https://hal.science/hal-03874607v1 Submitted on 28 Nov 2022 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0 International License Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0 International License Received: 9 November 2021 Revised: 31 May 2022 Accepted: 11 June 2022 Received: 9 November 2021 Revised: 31 May 2022 Accepted: 11 June 2022 DOI: 10.1111/desc.13300 DOI: 10.1111/desc.13300 Correspondence KEY WO RDS EEG, language learning, neonates, prosody, sequence learning, statistical learning Developmental Science. 2022;e13300. https://doi.org/10.1111/desc.13300 Abstract Since speech is a continuous stream with no systematic boundaries between words, how do pre-verbal infants manage to discover words? A proposed solution is that they might use the transitional probability between adjacent syllables, which drops at word boundaries. Here, we tested the limits of this mechanism by increasing the size of the word-unit to four syllables, and its automaticity by testing asleep neonates. Using markers of statistical learning in neonates’ EEG, compared to adult behavioral performances in the same task, we confirmed that statistical learning is automatic enough to be efficient even in sleeping neonates. We also revealed that: (1) Success- fully tracking transition probabilities (TP) in a sequence is not sufficient to segment it. (2) Prosodic cues, as subtle as subliminal pauses, enable to recover words segmenting capacities. (3) Adults’ and neonates’ capacities to segment streams seem remarkably similar despite the difference of maturation and expertise. Finally, we observed that learning increased the overall similarity of neural responses across infants during expo- sure to the stream, providing a novel neural marker to monitor learning. Thus, from birth, infants are equipped with adult-like tools, allowing them to extract small coher- ent word-like units from auditory streams, based on the combination of statistical analyses and auditory parsing cues. 2Department of Neuroscience, Max Planck Institute for Empirical Aesthetics, Frankfurt am Main, Hessen, Germany Correspondence Lucas Benjamin, Cognitive Neuroimaging Unit, CNRS ERL 9003, INSERM U992, CEA, Université Paris-Saclay, NeuroSpin center, 91191 Gif/Yvette, France. Email: lucas.benjamin@cea.fr 2 of 16 1 INTRODUCTION 2 of 16 BENJAMIN ET AL. One of the limitations of statistical learning, already reported in the literature, is its interaction with alternative segmentation cues (Black & Bergmann, 2017), especially its embedding in prosodic units. A word cannot straddle a prosodic boundary. Therefore, even if two syllables are always presented in succession, they are not attributed to the same word if a prosodic boundary separates them. This property is observed in adults (Shukla et al., 2007) and in 5–8-month-old babies (Johnson & Tyler, 2010; Shukla et al., 2011). This result is not surprising given the importance of prosody to structure the speech signal. A hierarchy of prosodic units (Nespor & Vogel, 2006) roughly parallel to the syntactic tree is used to improve speech comprehension in adults and to favor language acquisition in infants. For example, even at 2 months of age, infantsmemorize better the phonetic contentofa sentence witha well- formed prosodic contour relative to a word list (Mandel et al., 1994). This advantage can be explained because statistical computations are limited to a few elements within the prosodic unit, relieving memory. Prosodic units also provide perceptual anchors, which help infants note the reproducible location of certain words at their edges, such as arti- cles or proper name. Finally, the higher frequency of function words relative to content words has also been proposed as anchors favor- ing word discovery (Hochmann et al., 2010). To succeed in the complex task of constructing a lexicon from natural speech, infants have a tool- box of procedures at their disposal, whose relative contributions are currently underspecified. One of the main challenges encountered by infants to learn their native language and construct their lexicon is that words are rarely produced in isolation. Instead, words are embedded in sentences with no system- atic silence or clear acoustic boundaries between them. Subtle acous- tical markers such as the lengthening of the last syllable, pitch change, slowing-down of the syllabic rate and less coarticulation between syl- lables can signal words ending. But adults rely mainly on lexical knowl- edge and sentential context to retrieve words in their native language (Mattys et al., 2005) and in an unknown language, they have great dif- ficulty in correctly segmenting sentences into words (Wakefield et al., 1974). 2 of 16 1 INTRODUCTION However, when the experimental task is simplified by using an artificial stream of concatenated words, these acoustical cues can be used to discover the possible words as shown by their above-chance accuracy in forced-choice tasks (Bagou & Frauenfelder, 2018). Simi- larly, neonates can detect these subtle variations in a binary situation in which they have to discriminate pseudo-words constituted of syllables either coming from inside a word (e.g., /mati/ from mathematicien) or from two successive words (e.g., /mati/ from pyjama tissé) (Christophe et al., 1994). However, the relative weights of these markers vary across languages (Ordin et al., 2017) and within a language (i.e., they depend on the position of the word in the sentence and interact with other prosodic features such as lexical stress). Therefore, the robust- ness of these word-boundary cues is commonly estimated as insuffi- cient for infants to segment natural speech in successive word units. Here we investigated another putative limitation of the statistical learning mechanism: the size of the words that can be learned. In fact, most, if not all, studies in infants have used tri-syllabic words. Is it due to particular experimental choices? Or is there a hard limit to segmenta- tion based on statistical computation, especially in immature infants? If the latter, can subtle prosodic cues rescue segmentation and word learning, allowing memory processes to deploy (Fló et al., 2022)? To investigate these questions, we created a first artificial stream con- sisting of four quadri-syllabic words, pseudo-randomly concatenated without any prosodic cue, and a second one strictly identical to the first one but with a 25 ms pause between each word, every four syllables. In previous studies of artificial streams with this short pause, adults reported not noticing it and were at chance when they had to choose which of the two streams had pauses (Peña et al., 2002). Neverthe- less, pauses significantly improved their performances (Buiatti et al., 2009; Peña et al., 2002). The pause was probably perceived as a vowel lengthening, a universal ending cue for words and musical segments (Tyler & Cutler, 2009). In adults, final syllable lengthening improved tri- syllabic word segmentation (Saffran et al., 1996). The authors proposed a putative hierarchy in using these cues, i.e., infants first rely on transi- tionalprobabilities,thennoticethatsyllablelengtheningcoincideswith a word ending to finally learn this new cue. Research Highlights ∙Successfully tracking transitional probabilities in a sequence is not always sufficient to segment it. ∙Word segmentation solely based on transitional probability is limited to bi- or tri- syllabic elements. ∙Prosodic cues, as subtle as subliminal pauses, enable to recover chunking capacities in sleeping neonates and awake adults for quadriplets. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2022 The Authors. Developmental Science published by John Wiley & Sons Ltd. Developmental Science. 2022;e13300. wileyonlinelibrary.com/journal/desc 1 of 16 https://doi.org/10.1111/desc.13300 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2022 The Authors. Developmental Science published by John Wiley & Sons Ltd. D l t l S i 2022 13300 il li lib /j l/d 1 f 16 2 of 16 1 INTRODUCTION 3 of 16 FIGURE 1 (a) Experimental design: Participants were first exposed to a structured stream comprising four quadri-syllabic words (called ABCD) then presented with 3 types of isolated test words. (b) Experimental procedure: Neonates were tested asleep using high-density EEG (128 channels) while they were presented with random stream, structured stream, and isolated words. Short Str: short structured streams were presented to the neonates to maintain learning. Adults were tested on a web platform. After familiarization with the structured stream, adults were asked to rank the familiarity on a scale (1–6). To avoid a bias to quadri-syllabic words, they were also presented with foils corresponding to three other types of bi-syllabic test words (see Section 5 for more details) BENJAMIN ET AL. 3 of 16 FIGURE 1 (a) Experimental design: Participants were first exposed to a structured stream comprising four quadri-syllabic words (called FIGURE 1 (a) Experimental design: Participants were first exposed to a structured stream comprising four quadri-syllabic words (called ABCD) then presented with 3 types of isolated test words. (b) Experimental procedure: Neonates were tested asleep using high-density EEG (128 channels) while they were presented with random stream, structured stream, and isolated words. Short Str: short structured streams were presented to the neonates to maintain learning. Adults were tested on a web platform. After familiarization with the structured stream, adults were asked to rank the familiarity on a scale (1–6). To avoid a bias to quadri-syllabic words, they were also presented with foils corresponding to three other types of bi-syllabic test words (see Section 5 for more details) response, which is observed as a power increase at the frequency of the syllable presentation. If the syllables are perceived grouped in a quadri-syllabic word, the power should increase at ¼ of the syllable frequency (⅓if tri-syllabic words are used). Such a power increase at the word frequency has indeed been reported for tri-syllabic words in adults (Batterink & Choi, 2021; Benjamin et al., 2021; Buiatti et al., 2009), in 6–8-month-old infants and in neonates (Choi et al., 2020; Fló et al., 2022; Kabdebon et al., 2015). We also presented neonates with “random” streams constituted of pseudo-randomly concatenated sylla- bles, with and without a pause every four syllables, to control whether the pause itself was sufficient to induce a 4-syllable-rhythm. 2 of 16 1 INTRODUCTION Yet, this hypothesis remains untested because the relative contribution of transitional probabilities and this subtle prosodic cue was not assessed in this study. A second mechanism, based on the analysis of the transitions between syllables, has thus been proposed. Within a word, syllables have a fixed order, whereas any syllable can follow the last syllable of a word. Thus, a word boundary corresponds to a drop in the tran- sition probabilities (TP) between consecutive syllables. To prove that the concept could apply for word learning in infancy, Saffran, Aslin,and Newport (1996) used a mini-language of four tri-syllabic words and tested 8-month-old infants who listened for 3 mins to a continuous stream in which these words were concatenated with a flat intonation. The authors reported that infants were subsequently able to distin- guish two different lists of isolated tri-syllables pseudo-words: one corresponding to the Words (i.e., ABC:TP were equal to 1 between each syllable) and the other to PartWords formed by the two last syl- lables of a word and the first syllable of the next word for example (i.e., BCA’:TP were equal to 1 and 0.33). This result has been replicated mul- tiple times (Black & Bergmann, 2017) and extended to non-linguistic stimuli (Saffran et al., 1999; Schön et al., 2008) and to the visual domain (Fiser & Aslin, 2002). Sensitivity to statistics in sequences is also observed in animals (Hauser et al., 2001; James et al., 2020; Toro & Trobalón, 2005) indicating that the capacity of extracting transitional probabilities between subsequent elements is a robust general mech- anism. Additionally, it has been reported in asleep neonates (Fló et al., 2019; Fló, Benjamin, et al., 2022; Teinonen et al., 2009); and to some extent, in inattentive adults (Batterink & Choi, 2021; Benjamin et al., 2021; Fernandes et al., 2010; Toro et al., 2005). Yet the limits of this mechanism and the influence of development and expertise on the performances are still poorly known. We used high-density EEG (128 channels) to evaluate segmenta- tion processes in neonates. EEG allows not only to observe different responses to test-words after learning, but also to track learning while neonates are listening to the artificial stream. As the syllables have exactly the same length, their perception creates a regular evoked BENJAMIN ET AL. 2 of 16 1 INTRODUCTION In adults, inserting such a pause in a random stream did not produce any increase of power at the pause frequency (Buiatti et al., 2009). Therefore, in the case of successful segmentation, we expected a significant power increase at the word frequency in the structured streams relative to the random streams. No change, or perhaps a decrease, was expected at the syllabic rate, in line with previous reports in adults in which per- ceiving the word induced a decrease of the entrainment at the syllabic rate (Batterink & Choi, 2021; Benjamin et al., 2021). indicates an ill-formed word. In ShuffleWords, the two middle sylla- bles of a Word were inverted, violating local position. Thus, while all the transitional probabilities were zero, all syllables were always heard in close proximity during the learning stream. This temporal proximity might induce memory errors and a wrong recognition of ShuffleWords as possible words. Indeed, in longer words of six-syllables, neonates are not able to detect a shuffle of the middle syllables, whereas they detect a shuffle of the edges syllables (Ferry et al., 2016). Thus, our experimental design provides several markers of transi- tional probability computation and word segmentation that might be differently associated, opening the possibility to disentangle several steps or hypotheses of this classical learning task. (H1: TP computa- tion) If infants computed TP and memorized the TP matrix, they should reject Words from Shuffle- Words (1+1+1 vs. 0+0+0) but marginally Words from PartWords (1+1+1 vs. 1+0.33+1). (H2: segmentation) Stream segmentation should create an increase of neural entrainment at the word frequency. (H3: complete memorization of the word) should create a difference between words on one side and PartWords and ShuffleWords on the other side. (H4: memory errors) If Words are seg- mented and swap errors occur, ShuffleWords should not differ from Words due to the temporal proximity of the syllables belonging to the same Word. (H5: first syllable memorization only). This hypothesis could explain why words are preferred over PartWords in many statistical learning studies. As the typical trisyllabic paradigm compares Words (ABC) to PartWords (BCA’), the difference observed could result from After the learning phase, three types of test-words were presented in isolation: Words, PartWords, and ShuffleWords (Figure 1). Success- ful word segmentation is commonly revealed by a significant difference between the measured response to Words and PartWords. 2.1 Adults As described in other studies on neural entrainment (Fló, et al., 2022; Hochmann et al., 2010; Kabdebon et al., 2015), there was a signifi- cant increase in power and Phase Locking Value (PLV) at the frequency of the syllables presentation compared to neighboring frequencies in both groups (continuous and with pauses) and stream types (random and structured) (all ps < 0.05 FDR corrected). No interaction was observed between groups and streams indicating a similar signal-to- noise ratio and comparable experimental conditions in the two groups (all ps > 0.05 FDR corrected). The power analysis for each condition and group is presented in the supplementary material. Two groups of adults were tested online on a web platform (n = 43). After having listened to the continuous stream, or to the stream with 25 ms subliminal pauses depending on the group, the participants had to judge the familiarity of three types of words (Words, Part-Words and Shuffle-Words). Exposure (3.3 mn) and test were performed two times and results of the two tests were aggregated (see Section 5 and Figure S5 for results in each test). We analyzed the responses by items in a linear mixed-effects model in each group with FDR correction. For the stream without subliminal pause at the end of the word, Words and Part-Words were similarly rated (p = 0.26) and estimated more famil- iar than the ShuffleWords (W vs. SW p < 0.001, PW vs. SW p < 0.01). When subliminal pauses were added at the end of the words in the stream, all types of words were ranked differently (all p < 0.001) with the following order: Words were judged more familiar than PartWords themselves more familiar than ShuffleWords (Figure 1a,b). To better visualize the difference in segmentation performances between the two groups, we calculated the difference in mean familiarity ranking given by each participant to Words and PartWords, and performed an unpaired unidirectional, t-test t(41) = 2.3, p = 0.013. The segmentation effect, seen as a positive value in Figure 2c, was larger when subliminal pauses were present (Figure 2c). Stream segmentation should be revealed by a significant increase of power and/or PLV at the frequency of the words (i.e., ¼ of the syllabic frequency) relative to the random stream. by a subliminal acoustic cue. Yet even when there was no pause, they rejected ShuffleWords because of null transitional probabilities. the encoding of the first syllable only (A vs. B). In a recent study with tri-syllabic words, we indeed observed an ERP difference between words and PartWords for the first syllable, whereas no difference was recorded when the last syllable was incorrect (Fló, et al., 2022). 2.2 Finally, comparing the two groups of neonates, one listening to the stream without pauses (continuous group) and the other to the stream with pauses (with pauses group), should clarify the relative contribution of auditory parsing cues and transitional probabilities to word segmentation at that age. This comparison should disentangle whether pauses rescue segmentation, subsequently allowing the com- putation of transitional probabilities on smaller segments, or whether the computation of TP is done independently of the segmentation process. EEG was recorded in two groups of healthy full-term neonates (n = 52 after rejection procedure, see Section 5) while they were listening to streams without pauses for the first group and with 25 ms pauses every four syllables for the second group. For each group, neonates were exposed first to ∼7 min of a “random” stream in which syllables were randomly concatenated with a flat TP of 0.33 with and without a subliminal pause every four syllables depending on the group, fol- lowed by ∼13.5 min of the word-structured stream. After the exposure learning phase, a test phase followed in which they were exposed to isolated quadri-syllables sequences (Words, PartWords, and Shuffle- Words). To avoid interference with learning in the testing phase and to reinforce the learning of the structured materials, 40s-short segments of the structured stream were presented every 12 words during this phase. Finally, another ∼7 min of the random stream was presented again after this testing phase. The division of the random stream into two periods was done to avoid a time confound in the comparison between random and structured streams. We used a longer exposure time than usual statistical learning experiments in order to perform pattern similarity analyses as done by Henin et al (2021). Neonates are two-decades far from a mature state in terms of linguistic abilities but also in terms of memory capacities. To our knowl- edge, no adult equivalent of the paradigm proposed here was available. Thus, we collected adult behavioral data as a mature model of the mechanisms we explored in neonates. We adapted the paradigm to collect behavioral responses via a web-based procedure and short- ened the habituation to avoid over-learning already reported in similar experiments in adults (Peña et al., 2002) (see Figure 1). Despite different procedures and learning indicators, the results were sur- prisingly congruent with those obtained in infants, especially showing comparable limitations. 4 of 16 BENJAMIN ET AL. 2 of 16 1 INTRODUCTION In Words, all transitional probabilities between syllables equal 1, while in Part- Words (straddling two words), a drop in transitional probabilities 2.2.3 Syllable pattern similarity analysis In a recent paper, Henin et al. (2021) proposed that pattern similar- ity between syllables can vary with learning in a similar task. More specifically, using electrocorticography in epileptic adult patients who listened to a structured stream composed of the concatenation of four trisyllabic words, they computed different patterns of similar- ity between the 12 syllables. They took advantage of the high spatial resolution of electrocorticography and observed different clusters of electrodes sensitive either to TP transitions (low vs. high TP), the ordi- nal position (1st vs. 2nd vs. 3rd syllable), or the word identity (word 1 vs. word 2 vs. word 3 vs. word 4) in different brain areas. We computed a similar analysis on the responses to the syllables during the structured stream and showed that the similarity pattern for syllables belonging to the same words was significantly increased in the pause group com- pared to the continuous group (p = 0.012). However, we failed to find an increase in pattern similarity for low TP (DA’) and ordinal position (AA’,BB’,CC’,andDD’)betweenthetwogroups.Toinvestigateifthesig- nificant increase in similarity for syllables belonging to the same word was only due to an increase in high TP pairs or all pairs belonging to the same word, we separated the word condition in two sub-conditions: Consecutive (AB, BC, CD) and non-Adjacent (AC, BD & AD). Interest- ingly,Consecutiveandnon-adjacentpairsshowedasignificantincrease BENJAMIN ET AL. 5 of 16 BENJAMIN ET AL. 5 of 16 FIGURE 2 Adult behavioral results. (a,b) Results of the familiarity ranking tests in adults subjects for each item. Both test sessions were aggregated. Results for each session are presented in Figure S5. (c) Interaction at the subject level between both groups on the main effect of segmenting (Word - PartWord). Dots represent familiarity ranking difference between Words and PartWords in each subject. p-Value is estimated using one-way unpaired t-test FIGURE 2 Adult behavioral results. (a,b) Results of the familiarity ranking tests in adults subjects for each item. Both test sessions were aggregated. Results for each session are presented in Figure S5. (c) Interaction at the subject level between both groups on the main effect of segmenting (Word - PartWord). Dots represent familiarity ranking difference between Words and PartWords in each subject. p-Value is estimated using one-way unpaired t-test continuous group (Figure 4a Right). No difference was observed during the random streams (Figure 4b Right). frequencies of interest over sliding time windows of 2 min with a 1.5 s step, similarly to Fló, et al. (2022). We observed no change at the word frequencyalong time for either group (Figure S3). The poor signal/noise ratio at these low frequencies might explain the poor sensitivity of this analysis. FIGURE 2 Adult behavioral results. (a,b) Results of the familiarity ranking tests in adults subjects for each item. Both test sessions were aggregated. Results for each session are presented in Figure S5. (c) Interaction at the subject level between both groups on the main effect of segmenting (Word - PartWord). Dots represent familiarity ranking difference between Words and PartWords in each subject. p-Value is estimated using one-way unpaired t-test 2.1 Adults In the first group (continu- ous stream), we failed to find this result, contrary to the second group (stream with pause), in whom a significant increase of both power and PLV was observed in several electrodes. Finally, the interaction between groups and streams was significant for both power and PLV on several electrodes (Figure 3, all ps < 0.05 cluster corrected, Figure S2 for the results for each stream in each group). It has been described that the power at the syllabic rate decreased when adults segment the stream (Buiatti et al., 2009). However, we did not find any modulation of the power or PLV at the syllabic fre- quency in the structured stream compared to the random one. We also performed a time-course analysis of the neural entrainment at the Thus, adults were able to distinguish Words from PartWords, indi- cating that they had correctly segmented the stream only if helped 5 of 16 s FIGURE 3 Neural entrainment analyses at the word frequency in the structured stream minus the random stream (Power and phase locking value [PLV]) in the two groups of neonates (continuous and with pauses). Top rows: the presentation of stimuli with a fixed duration evoked a reproducible time-locked neural response that can be recovered as a neural oscillation at the frequency of stimulation. If infants segment the structure streams based on the quadri-syllabic words, an increase at the word frequency should be observed relative to the random stream. It is what is seen in the second group of neonates (with pauses) who listened to the streams with sub-liminal pauses. The acoustical effect of pauses was controlled by also adding a pause every four syllables in the random stream in this group. The last column shows the interaction between groups and frequencies. Dots locate the electrodes showing a significant result at p < 0.05 uncorrected, and larger dots after cluster correction (cluster p < 0.05). Power during structured and random streams are presented separately in each group in Figure S2 ing in the occipital ROI (0.05 < p < 0.1). Informed by the adult results, we compared Words and PartWords against ShuffleWords—the only condition rejected by the adults. This contrast revealed a significant difference in the frontal ROI (time: [1253–1644]ms, p = 0.025), and a trend was observed in the occipital ROI (0.05 < p < 0.1) (Figure 5 first panel). In the group with pauses, as shown in the second row of Figure 5, the response to Words seems to stand out from the other two conditions. ERPs to Words indeed significantly differed from those to PartWords in the two ROIs (all p < 0.05 Frontal: [1224–1780]ms, Occipital: [1196–1768]ms), and from those to Shuffle words in the occipital ROI ([1196–1768]ms) whereas two clusters were showing a trend (0.05 < p < 0.1) in the frontal ROI. in pattern similarity (both p < 0.05 FDR corrected) with pauses com- pared to the continuous group. The differences in pattern similarity between the two groups for each condition are reported in Figure 4c. 6 of 16 BENJAMIN ET AL. FIGURE 3 Neural entrainment analyses at the word frequency in the structured stream minus the random stream (Power and phase locking value [PLV]) in the two groups of neonates (continuous and with pauses). Top rows: the presentation of stimuli with a fixed duration evoked a reproducible time-locked neural response that can be recovered as a neural oscillation at the frequency of stimulation. If infants segment the structure streams based on the quadri-syllabic words, an increase at the word frequency should be observed relative to the random stream. It is what is seen in the second group of neonates (with pauses) who listened to the streams with sub-liminal pauses. The acoustical effect of pauses was controlled by also adding a pause every four syllables in the random stream in this group. The last column shows the interaction between groups and frequencies. Dots locate the electrodes showing a significant result at p < 0.05 uncorrected, and larger dots after cluster correction (cluster p < 0.05). Power during structured and random streams are presented separately in each group in Figure S2 BENJAMIN ET AL. 6 of 16 6 of 16 BENJAMIN ET A FIGURE 3 Neural entrainment analyses at the word frequency in the structured stream minus the random stream (Power and phase locking value [PLV]) in the two groups of neonates (continuous and with pauses). Top rows: the presentation of stimuli with a fixed duration evoked a reproducible time-locked neural response that can be recovered as a neural oscillation at the frequency of stimulation. If infants segment the structure streams based on the quadri-syllabic words, an increase at the word frequency should be observed relative to the random stream. It is what is seen in the second group of neonates (with pauses) who listened to the streams with sub-liminal pauses. The acoustical effect of pauses wa controlled by also adding a pause every four syllables in the random stream in this group. The last column shows the interaction between groups and frequencies. Dots locate the electrodes showing a significant result at p < 0.05 uncorrected, and larger dots after cluster correction (cluster p < 0.05). Power during structured and random streams are presented separately in each group in Figure S2 6 of 16 BENJ FIGURE 3 Neural entrainment analyses at the word frequency in the structured stream minus the random stream (Power and phase value [PLV]) in the two groups of neonates (continuous and with pauses). Top rows: the presentation of stimuli with a fixed duration evok reproducible time-locked neural response that can be recovered as a neural oscillation at the frequency of stimulation. If infants segmen structure streams based on the quadri-syllabic words, an increase at the word frequency should be observed relative to the random stre what is seen in the second group of neonates (with pauses) who listened to the streams with sub-liminal pauses. The acoustical effect of p controlled by also adding a pause every four syllables in the random stream in this group. The last column shows the interaction between and frequencies. Dots locate the electrodes showing a significant result at p < 0.05 uncorrected, and larger dots after cluster correction p < 0.05). Power during structured and random streams are presented separately in each group in Figure S2 2.2.2 Between-subjects correlation analysis Because the exact same stream was used in each participant, we were able to analyze whether learning increased the global neural synchro- nization between neonates beyond neural entrainment. To do so, we tested whether the correlation between participants increased over time more when they listened to the structured stream. We then com- pared at each time the topography of each subject with the average of the other subjects’ topographies at that time. We observed a pro- gressive increase of the mean correlation between subjects in neural activity only in the second group with pauses (Figure 4a Left). Indeed, the increase was higher for the second group (with pauses) than the first one (continuous) (cluster corrected p < 0.01, time [88-820]s; Figure 4a). During the random streams, the correlations were flat rel- ative to baseline in both groups (Figure 4b Left). To confirm this effect, we computed a linear regression of the variation of subject correlation with the group with time at the subject level during each stream and compared the slopes in the two groups. Only when neonates listened to the structured stream with pauses (second group), the slope was signif- icantly positive and significantly greater than the same measure in the 2.2.4 ERP analysis For this analysis, we used the auditory localizer to determine regions of interest (ROI) for each group (see Section 5). A click preceding each word allows to isolate the two poles of the auditory response in each group and the time series of the electrodes of these data-driven ROIs were averaged together before we performed permutation cluster- based analyses on time as implemented in Fieldtrip (Oostenveld et al., 2011). In the continuous group, the response was flatter for Shuffle- Words compared to the other two conditions. When the conditions were contrasted two by two, we only observed that Word versus Shuf- fleWord tended to differ during two time-periods in the frontal ROI ([1308–1652]ms, p = 0.03; [1780–2000]ms, p = 0.05) but only trend- We also tested the interaction between groups and the main effect Words – Part Words and found significant interactions (p < 0.05) on both the frontal and occipital ROIs (Figure 5 last panel). Using a permu- tation cluster-based method (field trip cluster analysis) between 250 and 2000 ms without previous ROI extractions, we obtained similar results (Figure S4). BENJAMIN ET AL. 7 of 16 FIGURE 4 Correlation analysis. (a) Comparison of the correlation between neonates in the two groups during the structured stream. Left: Evolution of the correlation across neonates with time. Right: comparison of the slopes of the linear regression with time in each group (orange: ontinuous, blue: with pauses). (b) Similar analysis for the first (plain lines) and second random (dotted lines) streams in both groups. RND = Random. (c) Pattern Similarity analysis: We computed the increase of pattern similarity between the EEG response to each syllable in the wo groups during the structured stream. The similarity significantly increased for syllables belonging to the same word (for adjacent pairs: AB, BC, CD, and non-adjacent pairs AC, BD, and AD) BENJAMIN ET AL. 7 of 16 FIGURE 4 Correlation analysis. (a) Comparison of the correlation between neonates in the two groups during the structured stream. Left: Evolution of the correlation across neonates with time. Right: comparison of the slopes of the linear regression with time in each group (orange: continuous, blue: with pauses). (b) Similar analysis for the first (plain lines) and second random (dotted lines) streams in both groups. RND = Random. 3 DISCUSSION ous group contrary to what we obtained in a similar paradigm with tri-syllabicwords.Becausethe“noise”levelduetothebackgroundneu- ral activity is exponentially growing with low frequencies in EEG data, notably in neonates, the lack of neural entrainment for quadruplets might have been due to a lack of sensitivity of the method at 1 Hz, com- pared to 1.33 Hz in the case of triplets. However, the significant word entrainment in the group listening to the stream with pauses and the interaction between both groups confirm that neural entrainment is a sensitive method even at these low frequencies. In natural speech, many signal-derived cues might assist segmenta- tion (Wakefield et al., 1974), but none is robustly consistent to be systematically used by infants. Therefore, the computation of the transitions probabilities between syllables has been proposed as a possible solution (Saffran, Aslin, et al., 1996). We presented here a stream comprising quadri-syllabic words to investigate the efficiency of this strategy for longer words. Whereas tri-syllabic words are eas- ily extracted from a flat speech stream using TP between syllables in adults (Saffran, Newport, et al., 1996), infants (Saffran, Aslin, et al., 1996), and even sleeping neonates (Fló, et al., 2022), this single cue seemed insufficient here for quadri-syllabic words even in awake vigilant adults, revealing a clear limitation of the statistical learning mechanism in a segmentation task. Whereas the power increase at the syllabic frequency was large during all streams, we did not record a significant neural entrainment at the word frequency in the continu- In natural speech, many signal-derived cues might assist segmenta- tion (Wakefield et al., 1974), but none is robustly consistent to be systematically used by infants. Therefore, the computation of the transitions probabilities between syllables has been proposed as a possible solution (Saffran, Aslin, et al., 1996). We presented here a stream comprising quadri-syllabic words to investigate the efficiency of this strategy for longer words. Whereas tri-syllabic words are eas- ily extracted from a flat speech stream using TP between syllables in adults (Saffran, Newport, et al., 1996), infants (Saffran, Aslin, et al., 1996), and even sleeping neonates (Fló, et al., 2022), this single cue seemed insufficient here for quadri-syllabic words even in awake vigilant adults, revealing a clear limitation of the statistical learning mechanism in a segmentation task. 2.2.4 ERP analysis (c) Pattern Similarity analysis: We computed the increase of pattern similarity between the EEG response to each syllable in the two groups during the structured stream. The similarity significantly increased for syllables belonging to the same word (for adjacent pairs: AB, BC CD, and non-adjacent pairs AC, BD, and AD) FIGURE 5 ERP analysis. Grand average ERPs to the test-words in both groups and to the word minus part word difference. The ROIs correspond to the two poles of the response to the auditory localizer preceding the test word in each group. Dark gray areas identify significant temporal clusters. Light shaded areas surrounding the thick lines represent the standard error across neonates. W = Words (ABCD), PW = Part Words (CDA’B’) and SW = ShuffleWords (ACBD). Gray lines at the bottom of the plots indicate the time windows on which statistics were performed FIGURE 5 ERP analysis. Grand average ERPs to the test-words in both groups and to the word minus part word difference. The ROIs correspond to the two poles of the response to the auditory localizer preceding the test word in each group. Dark gray areas identify significant temporal clusters. Light shaded areas surrounding the thick lines represent the standard error across neonates. W = Words (ABCD), PW = Part Words (CDA’B’) and SW = ShuffleWords (ACBD). Gray lines at the bottom of the plots indicate the time windows on which statistics were performed of a pause but to a genuine learning process, as it was not observed for the random stream that also included pauses every four sylla- bles. Second, neural synchrony increased between participants only for the structured stream with pauses, further suggesting that neonates were following a similar learning process constraining their brain state in this condition. Again, this phenomenon was only observed for the structured stream and notfor the random stream withpauses. Itunder- scores that it was not a general difference between the two groups of neonates but was related to the learning process engaged when they listened tothe structured stream. Third, ERPstoWordsand PartWords were significantly different after the stream with pauses, a classi- cal indicator of word segmentation in this type of paradigm. Finally, adults also ranked Words higher than PartWords when they listened to streams with pauses relative to streams without pauses, confirming an undeniable advantage for the former over the latter. All these indica- tors of successful segmentation were not only lacking when the pause cue was not present, but for all of them, the differences between the two groups were significant in both infants and adults. 3 DISCUSSION Whereas the power increase at the syllabic frequency was large during all streams, we did not record a significant neural entrainment at the word frequency in the continu- The word segmentation failure is also not explained by the higher number of syllables to be memorized (16 syllables here vs. 12 for tri- syllabic words) and the larger word size since the same material, just with the addition of sub-liminal pauses, rescued the word extraction process. Furthermore, we recorded several other indicators of learning in the second group who listened to the stream with pauses: First, the increase in power and phase locking-value observed at the word fre- quency in the structured stream was not related to the mere presence BENJAMIN ET AL. 8 of 16 3.1 Word segmentation based on statistical learning is limited by the word size Part-Words created by the pauses faded away (Figure S5) suggest- ing that the weight attributed to each parameter might not be strictly hierarchized but dependent on the strength of the evidence provided. It was proposed that the computation of the TP might be used to seg- ment a speech stream, either through boundary markers—a TP drop creates a prediction error, and the surprise allows to memorize the syl- lable following the drop (i.e., the first syllable of the following word)—or because adjacent events acquired a similar representation. However, neither the local drop of TP nor the temporal proximity within a chunk were sufficient to structure the stream, not even after 13 mins of expo- sure when the unit size was four syllables (1s long). On the contrary, sleeping neonates perceived a tri-syllabic rhythm in the same circum- stances and only after 2 mins of exposure, and memorized the set of possible first syllables (Fló, et al., 2022). It remains possible that longer exposure to the continuous stream might eventually allow word segmentation. However, compared to the segmentation of tri-syllabic words tested under similar conditions, both neonates and adults had considerable difficulty performing the task with quadri-syllabic words. We also observed in infants that similarity between syllables within the word was increased relative to the continuous stream without pauses (Figure 4c). Not only similarity between adjacent syllables within a word was stronger in the stream with pauses than without pauses, but similarity also increased between more distant syllables belonging to the same word. We cannot disentangle whether this increase in similarity between syllables in a word induced the seg- mentation as in a clustering strategy that is opposed to a bracketing strategy in which splitting points are looked for (Swingley, 2005), or the reverse, that is, because syllables were perceived in the same chunk, their similarity increased. It is also interesting to note that perceiving the stream at a more complex level of representations increased neural synchrony between the neonates. Whereas the syllabic rate itself, which affects many chan- nels (see Figure 2 in appendix), already creates a strong and similar entrainment across participants, it is not this low-level cue that was predominant in the neural synchrony between neonates but the per- ception of a higher level of organization of the stream. 3.2 Rescuing segmentation with sub-liminal pauses Adding a subliminal pause at the end of the word radically affects the performances at both ages. Although not consciously perceived, pauses act as other word boundary markers (e.g., lengthening of the last syllable, pitch drop) that neonates can perceive (Christophe et al., 1994). Our result demonstrates that such a word boundary marker is not only perceived but is effectively used to segment a stream from birth on, that is, before infants have perceived many isolated words. The use of word-ending cues, at least when it is a pause as here, does not need that infants first learn words as it was postulated (Saffran, Newport, et al., 1996) but is part of the auditory/linguistic percep- tive system. This observation is in agreement with the proposal of a hierarchical framework in weighting the multiple segmentation cues (Wakefield et al., 1974) and the subordination of statistical learning to many other cues, such as coarticulation (Fernandes et al., 2007), prosodic contour (Shukla et al., 2007; Shukla et al., 2011), and top- down contextual parsing (Wang et al., 2020). However, in adults as exposure lengthens and the absolute frequency of all possible tran- sitions increased, the familiarity advantage for Words relative to Adding a subliminal pause at the end of the word radically affects the performances at both ages. Although not consciously perceived, pauses act as other word boundary markers (e.g., lengthening of the last syllable, pitch drop) that neonates can perceive (Christophe et al., 1994). Our result demonstrates that such a word boundary marker is not only perceived but is effectively used to segment a stream from birth on, that is, before infants have perceived many isolated words. 3.1 Word segmentation based on statistical learning is limited by the word size This synchrony measure probably captures a wider cross-subject convergence beyond neural entrainment at the two frequencies of interest in specific chan- nels. It reveals that neonates’ brain states are not purely entrained by the physical features of the stimulation, which remain similar along the stream but also constrained by learning mechanisms that led to more synchronous responses across neonates. The opposite hypothesis is that the neonates might have learned as quickly as in the case of trisyllabic words but that as time passed, this learning faded away because even low probability transitions became familiar. This overlearning effect has been reported in adults (Peña et al., 2002). The analysis of the neural entrainment along time of expo- sure was not sensitive to figure out the learning timeline even in the group with pauses, probably because of the very low signal to noise ratio in low frequencies. However, the group difference in the corre- lation between neonates’ recordings increased from around the first minute of exposure showing that the two groups were diverging early on between a learning condition (stream with pauses) and a no-learning condition (continuous stream). Finally, the performances between the test phase during which iso- lated quadri-syllabic sequences were presented were also massively affected by the stream condition, suggesting that once segmentation was done, memory encoding was improved. Words and PartWords were indeed only discriminated after the stream with pauses. However, in a similar experimental paradigm but after a stream of concatenated tri-syllabic words, Words were recognized since the first syllable (Fló et al., 2022), whereas here, the difference was developing from around 500 ms to become significant only after the end of the word. The lack of first syllable effect was confirmed by the absence of difference between PartWords and ShuffleWords, although the latter started with a correct first syllable. It is also consistent with the lack of sim- ilarity increase between both groups within the set of first syllables (Figure 4c right), which contrasts with the result reported in adults by Henin et al. (2021). Thus, contrary to the tri-syllabic stream, the ordinal position of the syllables was not encoded, and the difference between correct and incorrect chunks took longer to develop. Yet, even if participants were not able to segment the words in the continuous structured stream, both adults and neonates rejected Shuf- fleWords, which contained the exact same syllables as the Words, but in the wrong order. This result reveals that the participants computed TP and were not misled by the temporal proximity of the syllables, but this computation was not sufficient to trigger stream segmenta- tion. Interestingly, attentive adults appeared not better than sleeping neonates in the task: They also failed to segment the stream without the help of acoustical cues. Thus, tracking TP does not always result in word segmentation. 9 of 16 BENJAMIN ET AL. 3.4 Similarity between adults and neonate cognitive abilities The similar drop in performance in neonates and awake linguisti- cally productive adults suggests a structural limitation in the number of items that can be stored in the STM. This limit of four in STM has been proposed as explaining several higher order linguistic observations, such as the size of phrasal verbs and idioms predominantly used in spo- ken languages such as English, the mean length of continuous discourse without pauses (Green, 2017), and the drop in mutual information scores after four words in many languages (Pothos & Juola, 2007). It also seems compatible with the observed word length inferior to four syllables in many languages (Zipf, 1935, Sigurd et al, 2004), suggesting that this chunk size limitation we observe here might be fine-tuned to real language word size. This limitation also reveals that TP computa- tion might not be robust enough to be the proposed general-purpose mechanism for word segmentation in all speech streams without being complemented by other indices. Despite very different measuring methods and attentional state in this set of experiments, the results in neonates and adults pointed to similar successes and failures in terms of TP computation and stream segmentation. This is somehow surprising given the fact that many of the structures that support sequence learning (Henin et al., 2021)—hippocampus, dorsal linguistic pathway, the superior temporal region—change rapidly in the first year of life; but the classic assump- tion that immature means poorly functional is increasingly challenged by brain imaging methods that provide markers of learning in young children. FMRI remains difficult in infants, but some results support the hypothesis of early efficiency despite immaturity. In a recent, paper Ellis et al. tested 3–24 month-old infants on a statistical learning task in the visual domain with fMRI and reported activation in the hippocam- pus associated with segmentation. Dehaene-Lambertz et al. (2002) reported activations in temporal and frontal areas in 3-month-olds listening to speech showing that regions usually reported in adults dur- ing statistical learning tasks (Henin et al., 2021) are, to some extent, already functional in infants. If neural entrainment during the stream reflects the chunking and word encoding, the ERPs to the isolated chunks in the test phase tested the participants’ recognition and familiarity with the different condi- tions. to the global familiarity of the word rather than noticing a particular error. In adults, Henin et al. (2021) confirmed using similarity analy- ses on ECOG recordings, that the ordinal position of the syllables was encoded. Adults are nevertheless better than neonates, encoding not only which syllables were first but also which were second and which were last. Here, we tried similar analyses in the neonates’ data despite the sparser resolution of EEG. We observed an increase of similarity of the ERPs to the adjacent and non-adjacent syllables belonging to the same word in the stream with pause compared to the continuous stream. However, we found no evidence of an increase of similarity between the words first syllables. Thus, the particular status of the first syllables observed in neonates in the case of tri-syllablic words (Fló, et al., 2022) had no support in this study when quadri-syllabic words were used. This result might just reflect a lack of power of our analy- sis, or it might be explained by the difference in perception of the drop of TP in a tri-syllabic word stream. The TP drop, which can induce a surprise following a prediction error, might favor encoding these syl- lables at a particular position (i.e., the first position of the next word). These results might thus suggest that depending on the segmenting cue, different memory processes are engaged in neonates and that TP computation might favor a more precise encoding of the chunking ele- ments, starting with the first syllable and progressing from one syllable to the next. Such an intriguing hypothesis should be further tested in experiments specifically designed to contrast these two cues and the word-size at this age and also in adults. 3 to 4 syllables, raised interesting questions regarding both word seg- mentation during the stream and subsequent memory encoding of the word unit. Although this experiment does not directly test this question, we propose that recovering words in a stream is based on short-term memory (STM). Indeed, if TP between syllables can be locally computed within the auditory cortex, the integration of the successive syllables within a word requires a longer temporal window of integration. The sharp difference between tri- and quadri-syllabic words seems remi- niscent of the 4 ± 1 unit limit of the auditory STM (Cowan, 2001) and suggests that the TP drop leading to word segmentation might only be noticeable when all the elements of a word plus the next syllable are present at once in the STM. Several studies suggest that adults use STM, and more specifically working memory, in such statistical learn- ing tasks. For instance, their performance improves when speech is slowed down, an observation at odds with a decay-time in a purely sensory buffer that should be detrimental as the time between sylla- bles increases, but in favor of maintenance of the successive syllabic items (Palmer & Mattys, 2016). Performances also drop when partic- ipants perform a concurrent two-back task (Palmer & Mattys, 2016) suggesting competition for general resources. These observations are coherent with the activations reported by Henin et al. (2021) along the dorsal linguistic pathway, and notably in the inferior frontal region. In the neonates, no explicit rehearsal was possible because they were asleep and, in any case, unable at that age to repeat syllables, but even in adults, STM effects may remain implicit (Hassin et al., 2009). If statis- tical learning is improved when adult participants are actively doing the task, the task remains feasible when they are distracted and unaware of the task (Fernandes et al., 2007; Palmer & Mattys, 2016). 3.3 Why a sharp distinction between tri and quadri-syllabic words? The differences, in terms of neural entrainment during familiarization and ERPs during test in infants as the drop of performances in adults, between our two word-segmentation studies (Fló, etal., 2022), in which we used a similar paradigm except that the word size increased from BENJAMIN ET AL. 10 of 16 4 CONCLUSION Human neonates display sequence learning abilities even during sleep, based on TP computations and segmenting helped by acous- tic/prosodic cues. The similarities with adults’ successes and fail- ures were remarkable, revealing early powerful capacities to process speech. A speech stream is not a uniform landscape for infants, but different cues might help them to chunk it into smaller units, opening the possibility to discover the linguistic regularities and the productive properties of speech. Human neonates display sequence learning abilities even during sleep, based on TP computations and segmenting helped by acous- tic/prosodic cues. The similarities with adults’ successes and fail- ures were remarkable, revealing early powerful capacities to process speech. A speech stream is not a uniform landscape for infants, but different cues might help them to chunk it into smaller units, opening the possibility to discover the linguistic regularities and the productive properties of speech. 5.1.1 Participants A total of 43 adults were recruited via social media and mailing (21 males, age distribution = [18-25]: 9, [25–40]:16, [40–60]: 17, 60+: 1]) with no reported auditory issue or language related troubles. They were randomly assigned to one of the streams with the instruction to carefully listen for ∼3 min to a nonsense language composed of non- sense words that they have to learn because they will have to answer questions on the words afterward. The learning/test procedure was repeated twice. The study was coded in javascript using jspsych toolbox (de Leeuw, 2015) and played audio mp3 pre-loaded and pre-created in MATLAB (see below) to avoid latencies during the presentation. Subjects vol- untarily participated on their computer. They were asked to wear headphones, sit in a quiet environment, and stay focused during the whole task. The Ethical research committee of Paris Saclay University approved the protocol under the reference CER-Paris-Saclay-2019-063. 5 MATERIALS AND METHODS Together, our results show that behavioral subjective ranking and EEG analyses provide powerful tools to investigate statistical learning and segmenting tasks. There was a neat congruency between the behav- ioral results in adults and the neural markers observed in neonates. Moreover, EEG data enables the investigation of such questions in pre- verbal and non-verbal subjects with different levels of attention (e.g., neonates, sleeping subjects, comatose patients). Power and PLV during the stream as well as ERP during isolated test words, were already pro- posed as reliable neural markers in this task (Kabdebon et al., 2015; Fló, Benjamin, et al., 2022). However, to our knowledge, between-subject correlation as a function of time had not been shown to capture learn- ing in infants successfully. Our results confirm that despite the noise in infant EEG data, a significant part of the variance cannot be only explained by low-level bottom-up activation to external stimuli but instead by a more sustained learning effect. Although this first attempt might have been still noisy, we might hope that this method could more accurately quantify the average amount of learning of a group over time or even characterize learning at the subject level. One drawback of this method is that, to compare across subjects, all subjects have to be exposed to the exact same stimuli, which presents a risk of con- found in the experimental design. Here we minimized this risk by taking two precautions. We first carefully designed and balanced the auditory material on acoustic aspects (see SI). Secondly, we ran two groups with a minimal change (a subliminal pause every four syllables) so that, if any bias persists, it would be the same in both groups and thus cannot explain differences between groups. sleeping adults even for tri-syllabic words (Farthouat et al., 2018; Bat- terink&Zhang, 2022),andtheirlearningremainedlimitedtobi-syllabic words, that is, to classical associative learning. Infants might perform better than adults during sleep due to the different organization of sleep-wake cycles. At this age, sleep comprises only two clear stages, quiet (∼40% of a sleep cycle at birth) and active sleep (50%–60% of a sleep-cycle at birth) with many micro-arousal periods within and between sleep stages (Scher, 2008). The short periods of wakefulness are immediately followed by active sleep, which is the equivalent of REM sleep in adults. In adults, learning has been shown to exist dur- ing REM (Andrillon & Kouider, 2016) and also that a task started during wake can continue during REM (Andrillon et al., 2016), opening the possibilities that neonates might learn and consolidate more efficiently than later, thanks to the closer wake-REM sleep alternations. such as the placement of the net on the infant’s head which is more variable at this age due to birth-related head deformation, and can introduce between groups differences; (2) eventually long-tail effects of the previous trials on the topography that can affect the baseline. BENJAMIN ET AL. 11 of 16 3.4 Similarity between adults and neonate cognitive abilities In the tri-syllabic experiment (Fló et al., 2022), neonates during the test-phase were no more sensitive to TP (i.e., no distinctive ERP response for triplets containing a TP = 0) and were reacting to an incor- rect first syllable. Here, they were rejecting ShuffleWords, thus were still sensitive to TP, but did not react particularly fast to the incorrect first syllable (Word vs. Part-Word), suggesting a more general response A major distinction between adults and neonates seems to be the capacity of computing such a task during sleep. Indeed, with both three and quadri-syllabic experiments, we showed that sleeping neonates were able to process and segment the streams, under the correct circumstances. However, recent studies report a learning failure in 5.1.3 Procedure After listening to the structured stream, participants were asked to rank the familiarity of the individual words (from “Completely unfa- miliar” to “Completely familiar” on a six-step scale). Learning and test phases were repeated twice. Data of the two tests sessions were aggre- gated in the main analysis (see separated analysis of each session in Figure S5). Six conditions (three bi-syllabic as foils and three quadri- syllabic conditions) were used to avoid any bias based on the length of the test words, with four trials in each of the six conditions. To avoid phonological similarity effects that could bias toward one or the other condition, participants were assigned to one of two groups where con- ditions were reversed. Four different pairs of structured streams per group were also generated, and participants were randomly assigned to one pair to avoid any given particularity of a stream driving the results. Three conditions were studied: Words, PartWords, and Shuf- fleWords. Words corresponded to the words that were embedded in the structured streams (ABCD), while PartWords corresponded to the two last syllables of a word and the two first of another word (CDA’B’). Thus, although PartWords were heard during the structured stream, they violated chunking based on TP. ShuffleWords corresponded to words in which the second and third syllables were inverted, creating a null TP between all syllables (none of the transitions were heard dur- ing the structured stream). However, the first and last syllables were correct. 5.2.1 Participants Two groups of healthy full-term neonates were tested between days 1 and 3. There was no problem during pregnancy and delivery, birth- weight > 2500 g, term > 38 wGA, APGAR ≥6 and 8 at 1′ and 5′, normal audition tested with otoacoustic emission. Parents provided informed consent, and the Ethical Committee (CPP Tours Region Centre Ouest 1) approved the study (EudraCT/ID RCB: 2017-A00513-50). In the first group (continuous), 34 neonates were tested. Among them, seven were excluded because they did not complete the experimental protocol or technical issues leaving 27 infants (14 males). In the second group, 34 infants were tested (with pauses). Nine were excluded because they did not complete the experimental protocol or technical issues, leaving 25 infants (13 males). 5.2.2 Stimuli We used the same 16 isolated syllables generated with MBROLA as in the adult experiment to construct four different streams (structured and random, with and without pause). The random stream consisted of 1600 pseudo-randomly concatenated syllables (6.7 mn). Each sylla- ble could be followed by three others from the pool leading to a flat TP during the stream. This pseudo-random stream offers a more con- trolled stimulus than the random streams used previously because the TPs were fixed to 1/3 (instead of 1/15), a similar value than the TP between words in the structured stream. The structured stream was comprised of 3200 syllables (13.3 mn). All streams were ramped up and down during the first and last 5 s to prevent the beginning and the end of the streams from being used as anchors. We created only one syllabic order for each stream to obtain learning markers better comparable between infants. For the second group, a pause was added every four syllables in both the structured (duration: 13.7 mn) and the random streams (duration: 6.8 mn). Thus, the sequences were identi- cal for all infants in both groups except for the 25-ms subliminal pauses every four syllables in the second group. Because all subjects had the same auditory materials, we carefully controlled for low-level acoustic- phonetic properties. We equilibrated the characteristics of consonants and vowels in the different words and at the different syllabic positions within words to avoid learning based on low-level acoustic cues (See Figure S1 for more details). As in adults, three types of test words were created: Word (ABCD), PartWord (CBA’B’), and ShuffleWords (ACBD) (Table 1 and Figure S1). 5.2 Infant EEG experiment In a previous experiment with similar streams with and without 25 ms pauses, Peña et al. (2002) showed that participants were at chance when they had to choose which of the two streams had pauses. To confirm that the pauses were not consciously perceived, eight adults listened to 20 streams (40 syllables – 10 s) presented randomly (10 without pauses and 10 with a 25 ms-pause every four syllables) and were unable to indicate which stream had pauses or not (mean = 49% (range [40, 59]%); p = 0.89). 12 of 16 BENJAMIN ET AL. mean familiarity ranking for each condition in each subject and sub- tracted the PartWord ranking from the word ranking within each subject. We then performed a one-way unpaired t-test between the two groups. pause, a 25-ms pause was inserted every 4-syllables (total duration 3.4 mn). All streams were ramped up and down during the first and last 5 s to avoid the start and end of the streams serving as perceptual anchors. We used the same syllables and words for the infant experiment. To avoid phonological similarity effects that could bias toward one or the other condition, Words and PartWords were reversed for half of the subjects. 5.1.2 Stimuli All speech stimuli were generated with the MBROLA text-to-speech software (Dutoit et al., 1996) using French diphones. The duration of all syllables was equalized to 250 ms with flat intonation and no coar- ticulation between syllables. Each experiment was composed of 800 syllables (3.3 mn) of an artificial monotonous stream of concatenated syllables that correspond to the four possible words randomly con- catenated with the only restriction that the same word could not be presented twice in a row. The same vocabulary (sixteen syllables) was used in the two streams, with and without pause. In the stream with We also implemented what we believe to be an improvement for ERP analysis. Before the presentation of isolated words, we presented a short audio click as an auditory localizer. In this way, we were able to extract ROIs for analysis with a data-driven approach instead of liter- ature driven. We performed a cluster based-permutation analysis on all data against zero during the click presentation to extract the audi- tory ERP ROI. Moreover, this localizer cluster was representative of the auditory response in this particular group of subjects taking into account non-relevant variations due to (1) Experimental conditions 5.2.5 Neural entrainment The recordings from the structured and random streams were seg- mented into consecutive non-overlapping epochs of 15 words (corre- sponding to 15 s in the continuous group and 15.375 s in the pause group). All subjects having 10 good epochs or more in each condition were included in this analysis (25 neonates in the continuous group, 21 in the pause group). We averaged the activity over artifact-free epochs for each neonate and electrode and computed the Fourier Transform using the fast Fourier transform algorithm (FFT) as implemented in MATLAB. We then computed the power of the FFT. The PLV between trials was computed on the FFT of single trials. Those values were nor- malized with neighboring frequency bins [−8:1,1:8]. The frequencies of interest were selected as the inverse of the duration of a word (f = 1 Hz for the continuous group f = 0.975 for the second group with pauses) and one-quarter of a word (i.e., roughly a syllabic rate, f = 4 Hz for the first group, f = 3.9 Hz for the second). To assess the significance of the power/PLV at the two frequencies of interest, we computed a con- trast between the power/PLV during the structured stream compared to the random streams for each electrode. As we expect learning dur- ing the structured stream to elicit a word rate oscillation, we computed a one-way (structured > random) paired t-test on each electrode. We corrected for multiple comparisons using a cluster corrected approach (alpha = 0.05). To look for a potential difference between groups, we computed an interaction between the previously described contrasts of both groups (difference of the structure minus random contrast in each group). Specifically, we ran a one-way unpaired t-test on each electrode and the clustering approach for the interaction. Both groups followed the same procedure (Figure 1). A first con- trol stream of a pseudorandom concatenation of 1600 syllables was followed by a structured stream composed of 3200 syllables grouped in words of four syllables. Infants then heard eight repetitions of short structuredstreams(160syllables)followedby12testwordspresented in isolation (four in each condition: Word, PartWords and Shuffle- Words, ISI 2-2.5s) for a total of 96 test-words (32 in each condition). The short streams were added to maintain learning because 2/3 of the test-words violated the learned structure. Each test word was pre- ceded by a short click 200 ms before its onset. TABLE 1 Words used in the experiment s were presented in isolation, four words, four ShuffleWords and four PartWords out of the eight possible. Note: During each small test block, 12 test words were presented in isolation, four words, four ShuffleWords and four PartWords out of the eight possible. Following a reviewer’s requirement, we tested adults on the same material with the same exposure duration (∼13mn). Their behavioral results are presented in Figure S5. lished for each subject and electrode as two interquartile ranges away from the 3rd quartile. This gave a logical matrix of the size of the recording, indicating bad data. Electrodes were definitely rejected if they were marked as bad more than 50% of the recording time, and time-samples were marked as bad if more than 35% of the electrodes were marked bad at this time-sample. For the ERP analysis, we then performed spatial interpolation of missing channels, and the data were mathematically referenced to the average of the 128 channels. 5.2.5 Neural entrainment The click was added as a task unrelated auditory localizer and to reset the baseline with a neutral event to avoid long-range drifts following the words. Finally, a second control stream was presented. Thus, the two random-streams were sandwiching the structured stream to control for habituation to the auditory stimulation, change in sleep stage, and any confounding time effect. 5.2.3 Procedure EEG was recorded with 128 electrodes (EGI geodesic sensor net), care- fully placed on the neonates’ heads by trained researchers to increase the consistency of the net placement. Three nets with different radii were used to fit infants’ heads. For the continuous group, infants were tested while asleep in the experimenter or parent’s arms. Due to COVID restrictions, the second group of babies was tested asleep in the crib. This slightly increased the noise level in the second group and might have marginally decreased the sensitivity of our analysis for this group. 5.1.4 Data processing Each answer was converted to a numerical value from 1 (completely unfamiliar) to 6 (completely familiar). The responses to the bisyllabic trials were not considered. All data, from both test sessions, were aggregated together in each group to compute a linear mixed-effects model on items (y ∼condition + (1\|subject)) to take the subject effect into account. The p-values were then FDR corrected. To compare sub- jects’ segmenting performances for both streams, we computed the BENJAMIN ET AL. 13 of 16 TABLE 1 Words used in the experiment Words (W) Part words (PW) Shuffle words (SW) RaFiBouNeu BouNeuNonLo/BouNeuVouDon RaBouFiNeu GuMaReuZo ReuZoVouDon/ReZoNonLo GuReuMaZo NonLoSanBi SanBiGuMa/SanBiRaFi NonSanLoBi VouDonMuLan MuLanRaFi/MuLanGuMa VouMuDonLan Note: During each small test block, 12 test words were presented in isolation, four words, four ShuffleWords and four PartWords out of the eight possible. 13 of 16 BENJAMIN ET AL. 5.2.6 Correlation analysis In both experiments, all subjects heard the exact same auditory mate- rial avoiding differences in stimulation between participants. We could thus compute the instantaneous correlation between each participant and the others. For each subject at each time during the streams, we computed the correlation at the topographical level between the topography of subject i at time t (a vector of 128 voltage values at time t corresponding to the 128 electrodes) and the topogra- phy of the grand average excluding subject i at time t (a vector of 128 values corresponding to the average across the other subjects at time t for each of the 128 electrodes). It gave, for each sub- ject, a vector of correlation between its own topography and the mean topography of all other subjects throughout time. Bad data were replaced by zeros and not taken into account for the average topographies across subjects. Time points with only bad data gave NaN correlation results. We hypothesized that learning should lead to an increase with time in the correlation between neonates as they learn the same material. To test it, we used two different methods. In the first one, we smoothed the correlation signal using a 400s- sliding-average-window in each neonate and stream, then computed a cluster-based analysis to reveal a significant cluster of time during which one stream showed a greater correlation than the other one. In the second one, we computed the slope of the linear regression with time in each subject and then considered the slope as a vari- able for the structured and random conditions in t-test comparing both groups. To extract ROI corresponding to the functional auditory localizer of each group, we measured the auditory event-related potential associated with the click presentation at the beginning of each trial by running a cluster-based analysis against zero to extract auditory ERP (5000 randomizations, two-tailed t-test, alpha < 0.01, cluster- alpha < 0.01, between −200 and 0 ms). This procedure identified a positive frontal and a negative occipital cluster in each group, on which we restricted the ERP analyses. Therefore, the voltage was averaged across electrodes in each of the two clusters in each neonate and condition. A cluster-based analysis was performed on the obtained time- series (10000 randomizations two-tailed t-test alpha < 0.05, cluster alpha < 0.05) between 250 ms (end of the first syllable) and 2000 ms to compare all pairs of conditions. ACKNOWLEDGMENTS We thank Simon Henin for help and comments about pattern similarity analysis. We also thank all the families who participated in our study as well as the maternity hospitals of Port-Royal and Orsay. This research has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 695710). 5.2.8 ERP analysis shorter experiments at the expense of more data manipulations. Both approaches were quite similar, confirming the validity of both that can be better adapted depending on the amount of available data. Data were segmented in 2850 ms long epochs ([−750 +2100]ms relative to word onset), averaged in the three conditions (Words, PartWords, and ShuffleWords), and baseline-corrected with the mean voltage value inthe interval [−750 to0] ineachneonate. Neonateswith less than 20 remaining trials in total were excluded from analysis (none in the continuous group, 1 in the pause group). 5.2.6 Correlation analysis Because of the adults’ behavioral results, we added the contrast "heard" (average of Word and Part- Word) vs. "non heard" (ShuffleWord) in the continuous group. Finally, we computed the interaction between groups and conditions (Word- PartWord) during the time window in which the previous analysis revealed a significant effect. 14 of 16 14 of 16 ETHIC APPROVAL Ethical research committee of Paris Saclay University under the ref- erence CER-Paris-Saclay-2019-063 & Ethical Committee CPP Tours Region Centre Ouest 1 (EudraCT/ID RCB: 2017- A00513-50) COMPETING INTERESTS To compute pattern similarity between syllables, we epoched each syllable from the structured stream from −100 ms to 350 ms. We removed the 100 first syllables to give enough time for participants to learn the task. The remaining epochs were averaged by syllables for each subject and a correlation matrix between each pair of syllables was computed with all the electrodes between 0 and 350 ms. We then separated the pairs of syllables into five conditions: First syllable (AA’), Ordinal position (BB’ or CC’ or DD’), Word and TP (AB or BC or CD), Word only (AC or AD or BD) and Low TP (DA’). We then averaged the similarity per condition and subtracted the correlation between all the other pairs. We then compared if pattern similarity between groups of syllables was increased differently across groups (One-way t-test with pauses > continuous). All other authors declare they have no competing interests. All other authors declare they have no competing interests. DATA AND MATERIALS AVAILABILITY All data and analysis are available upon request. BENJAMIN ET AL. 5.2.4 Data processing EEG recordings were band-pass filtered between 0.2 and 15 Hz for all analyses. Artifact rejection was performed on the non-epoched record- ing session using APICE pipeline (Fló, et al., 2022) based on the EEGLAB toolbox (Delorme & Makeig, 2004). Artifacts were identified on contin- uous data, based on voltage amplitude, variance, first derivative, and running average. The variance algorithm was applied in sliding time windowsof500mswith100mssteps. Adaptivethresholdswereestab- Additionally, we also replicated the neural entrainment effects with a slightly different method as proposed in Fló, et al. (2022). With this approach, the signal is decomposed on 1s long epochs and reconstructed in longer meta-epochs composed of several non- necessarily consecutive segment. It allows to save more data for REFERENCES (2009). Investigating the neu- ral correlates of continuous speech computation with frequency-tagged neuroelectric responses. Neuroimage, 44(2), 509–519. Choi, D., Batterink, L. J., Black, A. K., Paller, K. A., & Werker, J. F. (2020). Preverbal infants discover statistical word patterns at similar rates as adults: Evidence from neural entrainment. Psychological Science, 31(9), 1161–1173. Hochmann, J. R., Endress, A. D., & Mehler, J. (2010). Word frequency as a cue for identifying function words in infancy. Cognition, 115(3), 444–457. James, L. S., Sun, H., Wada, K., & Sakata, J. T. (2020). Statistical learning for vocal sequence acquisition in a songbird. Scientific Reports, 10(1), 1–18. Christophe, A., Dupoux, E., Bertoncini, J., & Mehler, J. (1994). Do infants perceive word boundaries? An empirical study of the bootstrapping of lexical acquisition. Journal of the Acoustical Society of America, 95(3), 1570–1580. Johnson, E. K., & Tyler, M. D. (2010). Testing the limits of statistical learning for word segmentation. Developmental Science, 13(2), 339–345. Kabdebon, C., Pena, M., Buiatti, M., & Dehaene-Lambertz, G. (2015). Electrophysiological evidence of statistical learning of long-distance de- pendencies in 8-month-old preterm and full-term infants. Brain and Language, 148, 25–36. Cowan, N. (2001). The magical number 4 in short-term memory: A recon- sideration of mental storage capacity. Behavioral and Brain Sciences, 24(1), 87–114. Mandel, D. R., Jusczyk, P. W., & Kemler Nelson, D. G. (1994). Does senten- tial prosody help infants organize and remember speech information?, Cognition, 53(2), 155–180. Dehaene-Lambertz, G., Dehaene, S., & Hertz-Pannier, L. (2002). Functional neuroimaging of speech perception in infants. Science, 298(5600), 2013– 2015. Mattys, S. L., White, L., & Melhorn, J. F. (2005). Integration of multiple speech segmentation cues: A hierarchical framework. Journal of Experimental Psychology General, 134(4), 477–500. de Leeuw, J. R. (2015). jsPsych: A JavaScript library for creating behavioral experiments in a Web browser. Behav Res, 47, 1–12. https://doi.org/10. 3758/s13428-014-0458-y Nespor, M., & Vogel, I. (2006). Prosodic phonology. Delorme, A., & Makeig, S. (2004). EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods, 134(1), 9–21. Oostenveld, R., Fries, P., Maris, E., & Schoffelen, J. M. (2011). FieldTrip: Open source software for advanced analysis of MEG, EEG, and Invasive Electrophysiological Data. Computational Intelligence and Neuroscience, 2011(2011), Article ID 156869, https://doi.org/10.1155/2011/156869 Dutoit, T., Pagel, V., Pierret, N., Bataille, F., & van der Vrecken, O. (1996). REFERENCES Fiser, J., & Aslin, R. N. (2002). Statistical learning of new visual feature com- binations by infants. Proceedings of the National Academy of Sciences of the United States of America, 99(24), 15822–15826. Andrillon, T., & Kouider, S. (2016). Implicit memory for words heard during sleep. Neuroscience of Consciousness, 2016(1), niw014. Fló, A., Benjamin, L., Palu, M., & Dehaene-Lambertz, G. (2022). Sleeping neonates track transitional probabilities in speech but only retain the first syllable of words. Scientific Reports, 12(1), 4391. https://doi.org/10. 1038/s41598-022-08411-w Andrillon, T., Poulsen, A. T., Hansen, L. K., Léger, D. L., & Kouider, S. (2016). Neural markers of responsiveness to the environment in human sleep. Journal of Neuroscience, 36(24), 6583–6596. Bagou, O., & Frauenfelder, U. H. (2018). Lexical segmentation in artificial word learning: The effects of converging sublexical cues. Language and Speech, 61(1), 3–30. Fló, A., Brusini, P., Macagno, F., Nespor, M., Mehler, J., & Ferry, A. L. (2019). Newborns are sensitive to multiple cues for word segmentation in continuous speech. Developmental Science, 22(4), e12802. Batterink, L. J., & Choi, D. (2021). Optimizing steady-state responses to index statistical learning: Response to Benjamin and colleagues. Cortex, 142, 379–388. Fló, A., Gennari, G., Benjamin, L., & Dehaene-Lambertz, G. (2022). Auto- mated Pipeline for Infants Continuous EEG (APICE): A flexible pipeline for developmental cognitive studies. Developmental Cognitive Neuro- science, 54, 101077. https://doi.org/10.1016/j.dcn.2022.101077 Batterink, L. J., & Zhang, S. (2022). Simple statistical regularities pre- sented during sleep are detected but not retained. Neuropsychologia, 164, 108106. https://doi.org/10.1016/j.neuropsychologia.2021.108106 Green, C. (2017). Usage-based linguistics and the magic number four. Cognitive Linguistics, 28(2), 209–237. Benjamin, L., Dehaene-Lambertz, G., & Fló, A. (2021). Remarks on the analysis of steady-state responses: Spurious artifacts introduced by overlapping epochs. Cortex, 142(2), 370–378. Hassin, R. R., Bargh, J. A., Engell, A. D., & McCulloch, K. C. (2009). Implicit working memory. Consciousness and Cognition, 18(3), 665–678. Hauser, M. D., Newport, E. L., & Aslin, R. N. (2001). Segmentation of the speech stream in a non-human primate: Statistical learning in cotton-top tamarins. Cognition, 78(3), 53–64. Black, A., & Bergmann, C. (2017). Quantifying infants’ statistical word seg- mentation: A meta-analysis. Proceedings of the 39th Annual Conference of the Cognitive Science Society, (3), 124–129. Henin, S., Turk-Browne, N. B., Friedman, D., Liu, A., Dugan, P., Flinker, A., Doyle, W., Devinsky, O., & Melloni, L. (2021). Learning hierarchi- cal sequence representations across human cortex and hippocampus. Science Advances, 7(8), 1–13. Buiatti, M., Peña, M., & Dehaene-Lambertz, G. BENJAMIN ET AL. 15 of 16 ORCID LucasBenjamin https://orcid.org/0000-0002-9578-6039 AnaFló https://orcid.org/0000-0002-3260-0559 16 of 16 BENJAMIN ET AL. Scher, M. S. (2008). Ontogeny of EEG-sleep from neonatal through infancy periods. Sleep Medicine, 9, 615–636. Wakefield, J. A., Doughtie, E. B., & Lee Yom, B. H. (1974). The identifi- cation of structural components of an unknown language. Journal of Psycholinguistic Research, 3(3), 261–269. Schön, D., Boyer, M., Moreno, S., Besson, M., Peretz, I., & Kolinsky, R. (2008). Songs as an aid for language acquisition. Cognition, 106(2), 975–983. Wang, F. H., Zevin, J. D., Trueswell, J. C., & Mintz, T. H. (2020). Top-down grouping affects adjacent dependency learning. Psychonomic Bulletin & Review, 27(5), 1052–1058. Shukla, M., Nespor, M., & Mehler, J. (2007). An interaction between prosody and statistics in the segmentation of fluent speech. Cognitive Psychology, 54(1), 1–32. Zipf, G. K. (reprinted 1965). (1935). The psycho-biology of language. MIT Press. Shukla, M., White, K. S., & Aslin, R. N. (2011). Prosody guides the rapid mapping of auditory word forms onto visual objects in 6-mo-old infants. Proceedings of the National Academy of Sciences of the United States of America, 108(15), 6038–6043. REFERENCES The MBROLA project: Towards a set of high quality speech synthesizers free of use for non commercial purposes. Proceeding of Fourth Interna- tional Conference on Spoken Language Processing. ICSLP ’96, 3, 1393–1396. https://doi.org/10.1109/ICSLP.1996.607874 Ordin, M., Polyanskaya, L., Laka, I., & Nespor, M. (2017). Cross-linguistic dif- ferences in the use of durational cues for the segmentation of a novel language. Memory and Cognition, 45(5), 863–876. Palmer, S. D., & Mattys, S. L. (2016). Speech segmentation by statisti- cal learning is supported by domain-general processes within working memory. Quarterly Journal of Experimental Psychology (Hove), 69(12), 2390–2401. Ellis, C. T., Skalaban, L. J., Yates, T. S., Bejjanki, V. R., Córdova, N. I., & Turk-Browne, N. B.. Evidence of hippocampal learning in human infants. Current Biology, 31(15), 3358–3364.e4. Farthouat, J., Atas, A., Wens, V., De Tiege, X., & Peigneux, P. (2018). Lack of frequency-tagged magnetic responses suggests statistical regularities remain undetected during NREM sleep. Science Reports, 8(1), 11719. Peña, M., Bonatti, L. L., Nespor, M., & Mehler, J. (2002). Signal-driven computations in speech processing. Science, 298(5593), 604–607. Pothos, E. M., & Juola, P. (2007). Characterizing linguistic structure with mutual information. British Journal of Psychology, 98(2), 291–304. Fernandes, T., Kolinsky, R., & Ventura, P. (2010). The impact of attention load on the use of statistical information and coarticulation as speech segmentation cues. Atten Percept Psychophys, 72(6), 1522–1532. Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996). Statistical learning by 8- month-old infants. Science, 274(5294), 1926–1928. Fernandes, T., Ventura, P., & Kolinsky, R. (2007). Statistical information and coarticulation as cues to word boundaries: A matter of signal quality. Perception and Psychophysics, 69(6), 856–864. Saffran, J. R., Johnson, E. K., Aslin, R. N., & Newport, E. L. (1999). Statistical learning of tone sequences by human infants and adults. Cognition, 70(1), 27–52. Ferry, A. L., Fló, A., Brusini, P., Cattarossi, L., Macagno, F., Nespor, M., & Mehler, J. (2016). On the edge of language acquisition: Inherent constraints on encoding multisyllabic sequences in the neonate brain. Developmental Science, 19(3), 488–503. Saffran, J. R., Newport, E. L., & Aslin, R. N. (1996). Word segmentation: The role of distributional cues. Journal of Memory and Language, 35(4), 606– 621. SUPPORTING INFORMATION Sigurd, B., Eeg-Olofsson, M., & van de Weijer, J. (2004). Word length, sen- tence length and frequency – Zipf revisited, Studia Linguistica, 58(1), 37–52. Additional supporting information can be found online in the Support- ing Information section at the end of this article. Swingley, D. (2005). Statistical clustering and the contents of the infant vocabulary. Cognitive Psychology, 50(1), 86–132. Teinonen, T., Fellman, V., Näätänen, R., Alku, P., & Huotilainen, M. (2009). Statistical language learning in neonates revealed by event-related brain potentials. BMC Neuroscience, 10, 21. How to cite this article: Benjamin, L., Fló, A., Palu, M., Naik, S., Melloni, L., & Dehane-Lambertz, G. (2022). Tracking transitional probabilities and segmenting auditory sequences are dissociable processes in adults and neonates. Developmental Science, e13300. https://doi.org/10.1111/desc.13300 Toro, J. M., Sinnett, S., & Soto-Faraco, S. (2005). Speech segmentation by statistical learning depends on attention. Cognition, 97(2), 25–34. Toro, J. M., & Trobalón, J. B. (2005). Statistical computations over a speech stream in a rodent. Perception and Psychophysics, 67(5), 867–875. Tyler, M. D., & Cutler, A. (2009). Cross-language differences in cue use for speech segmentation. The Journal of the Acoustical Society of America, 126(1), 367–376.
https://openalex.org/W3013356715	http://pure-oai.bham.ac.uk/ws/files/95825387/er.5347.pdf	English	null	High performance cooling of a HVDC converter using a fatty acid ester‐based phase change dispersion in a heat sink with double‐layer oblique‐crossed ribs	International journal of energy research	2,020	cc-by	15,540	General rights U l li General rights Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes permitted by law. •Users may freely distribute the URL that is used to identify this publication. Users may freely distribute the URL that is used to identify this publication. •Users may download and/or print one copy of the publication from the University of Birmingham research portal fo study or non-commercial research. study or non-commercial research. •User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) •Users may not further distribute the material nor use it for the purposes of commercial gain. •User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) •Users may not further distribute the material nor use it for the purposes of commercial gain. here a licence is displayed above, please note the terms and conditions of the licence govern your use of this document played above, please note the terms and conditions of the licence govern your use of this document. When citing, please reference the published version. High performance cooling of a HVDC converter using a fatty acid ester‐based phase change dispersion in a heat sink with double‐layer oblique‐crossed ribs Li, Qi; Fischer, Ludger; Qiao, Geng; Mura, Ernesto; Li, Chuan; Ding, Yulong DOI: 10.1002/er.v44.7 License: Creative Commons: Attribution (CC BY) Document Version Publisher's PDF, also known as Version of record Citation for published version (Harvard): Li, Q, Fischer, L, Qiao, G, Mura, E, Li, C & Ding, Y 2020, 'High performance cooling of a HVDC converter using a fatty acid ester‐based phase change dispersion in a heat sink with double‐layer oblique‐crossed ribs', International Journal of Energy Research, vol. 44, no. 7, pp. 5819-5840. https://doi.org/10.1002/er.v44.7 Citation for published version (Harvard): Li, Q, Fischer, L, Qiao, G, Mura, E, Li, C & Ding, Y 2020, 'High performance cooling of a HVDC converter using a fatty acid ester‐based phase change dispersion in a heat sink with double‐layer oblique‐crossed ribs', International Journal of Energy Research, vol. 44, no. 7, pp. 5819-5840. https://doi.org/10.1002/er.v44.7 Link to publication on Research at Birmingham portal University of Birmingham R E S E A R C H A R T I C L E R E S E A R C H A R T I C L E Summary Summary The paper concerns with a high performance cooling method for a HVDC con- verter using fatty acid ester-based phase change dispersion (PCD) in a heat sink with double-layer oblique-crossed ribs. Thermo-physical properties of PCDs were first characterized under both solid and liquid states, and the cooling per- formance of the heat exchanger was then experimentally examined, by heating two copper blocks clamped closely to the aluminium heating surfaces. A three- dimensional Euler-Euler multiphase approach was further performed to evalu- ate the thermal performance under different operating conditions including heating power, flowrate and PCD concentration. The results showed that the viscosity of PCD can be a 100 times that of water, but the increased pumping power was only ~17.01% on average. The use of the PCD achieved a lower tem- perature of heat sink and fluid than that of water under the same set of condi- tions due to the latent heat of the PCM, thus enabling a safer and cooler environment for temperature-sensitive HVDC components such as insulated gate bipolar transistors (IGBT). An optimal set of working conditions was pro- posed and a flowrate of 8 L/min under a heating power of 1.1 kW and a PCM concentration of 25% was recommended for industrial cooling operations. 2Lucerne University of Applied Sciences and Arts, Horw, Switzerland 3Global Energy Interconnection Research Institute Europe GmbH, Berlin, Germany Qi Li1 \| Ludger Fischer2 \| Geng Qiao3 \| Ernesto Mura3 \| Chuan Li1 \| Yulong Ding1 1Birmingham Centre for Energy Storage (BCES) & School of Chemical Engineering, University of Birmingham, Birmingham, UK 2Lucerne University of Applied Sciences and Arts, Horw, Switzerland 3Global Energy Interconnection Research Institute Europe GmbH, Berlin, Germany 1Birmingham Centre for Energy Storage (BCES) & School of Chemical Engineering, University of Birmingham, Birmingham, UK Highlights g g • A novel cooling method for HVDC converter using a fatty acid based PCD presented. • A 3D Euler-Euler modelling performed and compared with experiments. • Effects of operating conditions and PCM loading on cooling behaviour discussed. • Three calculating methods of heat transfer coefficients compared. • Optimal PCD cooling condition given based on overall performance evaluation. • Optimal PCD cooling condition given based on overall performance evaluation. K E Y W O R D S Take down policy Take down policy While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been uploaded in error or has been deemed to be commercially or otherwise sensitive. If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access to the work immediately and investigate. Download date: 24. Oct. 2024 Received: 24 December 2019 Revised: 23 February 2020 Accepted: 28 February 2020 Received: 24 December 2019 Revised: 23 February 2020 Accepted: 28 February 2020 DOI: 10.1002/er.5347 DOI: 10.1002/er.5347 Correspondence Chuan Li and Yulong Ding, Birmingham Centre for Energy Storage (BCES) & School of Chemical Engineering, University of Birmingham, UK B15 2TT. Email: c.li.4@bham.ac.uk (C. L.) and y. ding@bham.ac.uk (Y. D.) Funding information tate Grid Corporation of China and Global Energy Interconnection Research Institute Europe GmbH, Grant/Award Number: SGRIWLZXQT[2017]882 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. g p p y © 2020 The Authors. International Journal of Energy Research published by John Wiley & Sons Ltd Int J Energy Res. 2020;44:5819–5840. High performance cooling of a HVDC converter using a fatty acid ester-based phase change dispersion in a heat sink with double-layer oblique-crossed ribs Qi Li1 \| Ludger Fischer2 \| Geng Qiao3 \| Ernesto Mura3 \| Chuan Li1 \| Yulong Ding1 1 \| INTRODUCTION 5820 This could lead to system thermal runaway because silicon as workhorse material could not bear the crucial operational conditions.7 To avoid the thermal runaway, new semiconductor materials such as SiC or GaN have been developed.7,8 However, applica- tions of such materials are limited by available power module packages,9 package agents,10 peripheral compo- nents11 and economic considerations. Moreover, persis- tent heat still generates in the presence of new material because multiple processing by power electronics con- verters before the end use. These reasons push the rapid development of novel thermal management technologies. Additionally the smaller the heat sinks leads to a higher volumetric power density,12 which requires a more advanced and efficient cooling method. The two most commonly used technologies in cooling electronic devices are air cooling and water cooling. The air based cooling systems, classified as natural or forced air flux by the way air flows, are normally suffice for lower power electronic devices with a power dissipation rate below ~1500 W.13 The introduction of a liquid but water, which is often a dielectric, a high cooling effi- ciency or lower/stabilized temperature can be achieved. Among various coolants, phase change material (PCM) offers more isothermal operation through storing heat at a high energy density within a small temperature change.14-18 Further thermal performance enhancements can be real- ized by dispersing and transporting PCM in a thermally active fluid that is immiscible with the PCM, the so-called phase change dispersion (PCD). An incredible number of PCM slurries have been developed by combining various proportions of PCMs, modifiers, nucleation agents, etc. This allows customizing thermal-physical properties with an adaptability that is hardly possible for other class of sol- vents. 1 \| INTRODUCTION 5820 LI ET AL. LI ET AL. 5820 To ensure long-term lifespan in a high voltage envi- ronment, the PCD component is best to be electrically insulating for cooling HVDC converters. As a type of organic dielectric material, paraffin wax have been widely used as PCM due to the widely studied interaction between paraffin and classical emulsifier.23-25 However, paraffin often yields different crystalline phases during phase change, which may result in contamination on the wall of heat exchanger and require strong solvents for cleaning at a high temperature. The low thermal conduc- tivity, unpleasant odour and flammability make them even harder for industrial applications. The long-chain fatty acid ester from renewable sources offers an alterna- tive PCM to paraffin.26 Such a material normally works in the ambient-to-moderate temperature range27 (20-100C), which is suitable for HVDC electronic devices cooling28 (around 60C). PCD emulsions are often fabri- cated in a high energy environment provided by a high- shear mixer or an ultrasound generator. The multiphase suspension shows Newtonian behaviour at low emulsion concentrations, but becomes highly non-Newtonian at an emulsion concentration, with a clear demarcation hard to be determined empirically. For example, a 30 w/w % tetradecane emulsion29 was found to be Newtonian, whereas a paraffin/water emulsion30 showed pseudo- plastic behaviour with a mass fraction between 15% and 50%. The viscosity of emulsions can be 2-550 times that of water,31,32 yet the pressure drop has been found not as big as the viscosity increase and hence a small increase in the pumping work.33,34 As a result, a detailed material characterization of the PCD under both solid and liquid conditions of the PCM is required to enable an accurate prediction of thermal performance of the heat sink using PCDs. Rapid growing demand for electric power and continuous effort of miniaturization of high-voltage direct current (HVDC) devices drive heat flux to the megawatt range.1,2 In a high power transmission system, devices such as HVDC converters that transform alternating current (AC) to direct current (DC) and vice versa via controlla- ble electric switches, high power dissipation can be up to 9.72 kW, which is nearly 30% of the total HVDC power loss.3,4 Under such a high dissipation, sensitive power semiconductor devices such as insulated gate bipolar transistor (IGBT) in the HVDC converter could experi- ence a heat flux over 2 MW/m25 and a temperature increase to 90C,6 depending on the source voltage and switching frequency. K E Y W O R D S Euler-Euler multiphase modelling, fatty acid ester, HVDC converter cooling, multi-stream heat sink, oblique-crossed ribs, phase change dispersion wileyonlinelibrary.com/journal/er 5819 wileyonlinelibrary.com/journal/er 5 Int J Energy Res. 2020;44:5819–5840. 5821 showed a decreased junction temperature up to 5 K for a 40 W heating power with air cooling, but the fanning fric- tion factor was 19 times higher than a straight channel at high Re numbers. In industrial practical applications, the series and parallel arrangement of heat sinks can amplify the increased pressure drop. To avoid unnecessary energy consumption but keep the advantage of thermal perfor- mance enhancement, a multi-stream plate heat exchanger can be used due to compact structure, high efficiency, low cost and ease of handling multiple streams.40-42 Radwan et al.43 developed a monolithic double-layer microchannel heat sink for concentrated photovoltaic cooling under par- allel and counter flows, and found a single-phase-liquid parallel flow could effectively cool the PV at a higher flowrate. et al.51 studied ice slurry (<25 v/v %) in horizontal, vertical and 90 elbow pipes using the Euler-Euler model, and found a relative errors of ±20% with respect to the mea- surements. Ma et al.20 investigated the hydraulic and heat transfer characteristics of PCD in a circular smooth pipe using the Euler-Euler model, and showed a good agree- ment with previous experimental results. However, the use of the Eulerian-Eulerian model to study heat sinks is still limited due to large computation resource demand. Another noteworthy fact of heat sink is that there is still lack of unified standard of temperatures used for heat transfer coefficient calculation. The wall temperature can either be derived from wetted wall or outer wall of the heat sink, and the fluid temperature is either the bulk mean temperature or average value between inlet and outlet. The temperature difference used to drive the convective heat transfer coefficient can be calculated by taking the differ- ence between the wall and fluid temperature,36,52 or from log mean temperature difference.53,54 A large deviation between the two can occur at small temperature differ- ences under a high heat flux. Thus, an accurate and more convenient calculation method is needed. Metals such as aluminium alloys and copper are widely used to manufacture heat sinks, due to their light- weight and high thermal conductivity.44-46 In the evalua- tion of heat sink performance, heat transfer across a contact intermediate formed by any two solid surfaces is accompanied by a measurable temperature drop, because there exist interfacial thermal contact resistance (ITCR) to heat flows. The ITCR is caused by the imperfect heat transfer surfaces at the joint, which only give a small fraction of the apparent contact area while rest is filled with air or other media. The complex geometrical and thermo-physical parameters in such a case depend on many factors, such as surface micro-topography, micro- hardness, contact pressure, temperature level of the inter- face, solid thermal conductivities and type of substance in the interstitial gaps. The ITCR between aluminium alloys and copper has received lots of attention due to their wide use in practical applications. Yüncü47 found that the thermal contact conductance of Al–Cu varies between 10.3 and 30.88 kW/(m2 K), depending on the contact pressure and micro-hardness. Shi et al.48 evalu- ated the influence of temperature on ITCR between alu- minium nitride and copper at cryogenic conditions. The ITCR was found to decrease with increasing interface temperature and contact pressure, and ranging between 5 × 10−4 and 3 × 10−3 m2W/K. Depending on the contact surface condition, the thermal conductance of ground and milled Al–Cu was found to be 42-56 kW/(m2 K) and 12-22 kW/(m2 K),49 respectively. In this work, an electrically insulting fatty acid ester- based PCD is used for efficient HVDC converter cooling through a heat exchanger with double-layer oblique- crossed ribs. The PCD was characterized and the heat transfer performance of the heat exchanger was experi- mentally examined, by heating two copper blocks clamped closely to the aluminium heating surfaces. Such a device is able to disrupt thermal and hydraulic bound- ary layers, providing a higher temperature gradient near the wall and rapid removal of thermal energy. A 3D Euler-Euler multiphase approach was then used to model the PCD cooling performance under different heating powers, flowrates and PCD concentrations. The complex geometry of heat sink brings about difficulties in the eval- uation of heat transfer coefficient, thus three analysis methods based on temperatures of the heating wall, the wetted wall and the inlet/outlet fluid were compared. An optimal set of working conditions was proposed and rec- ommended for industrial cooling operations. The present work provides a high-performance and economic cooling method, using a novel coolant inside a heat sink with interrupted hydraulic and thermal boundary layers. Due to the complexity of performing local concentra- tion and velocity measurements of PCD in an insulated pipe, numerical simulations have therefore been widely used for the study. In particular, the Euler-Euler model has been shown to be a fruitful approach to dealing with two- phase mixtures. Göktepe et al.50 compared Eulerian- Eulerian and Eulerian-mixture model to study heat transfer coefficients and friction factors, and recommended the for- mer for description of the two-phase interaction. Wang LI ET AL. 1 \| INTRODUCTION 5820 Under a targeted set of operational conditions, the PCD can be designed to enhance thermal performance of heat sink with a uniform temperature distribution and a small temperature change.19-22 The relative low thermal conductivity of the organic PCD reduces the cooling efficacy, leading to challenges in enhancing and optimizing the thermal management of PCD based heat sinks.35 A great deal of efforts has been devoted to the design of enhanced heat transfer surfaces to lower the convective heat transfer resistance.36 An intro- duction of a large number of fins in the cross-sectional direction can reduce the thermal and hydraulic boundary layers, leading to higher temperature and velocity gradients at the wetted wall for an efficient removal of thermal energy.29 Similarly the use of metal foams could enhance heat transfer.37 However, a significant increase in the pres- sure drop may offset the enhanced heat transfer by an increase in the pumping power. Chai et al.38 studied the heat transfer in microchannel containing offset ribs, with rectangular, backward triangular, isosceles triangular, for- ward triangular and semicircular shapes. Kanargi et al.39 introduced the cross-connected alternating converging- diverging channel heat sink to disrupt boundary layers and 5822 (BCES) of the University of Birmingham. Table 1 summa- rizes the results. Each measurement was repeated at least three times to ensure reproducibility. The solution den- sity was measured using an Anton-Paar DMA 4100 M density meter (UK). The thermal conductivity was derived from laser flash measurements with a Netzsch LFA 427 (Netzsch, German).55 The specific heat capacity, melting point and phase change enthalpy of melting were determined using a DSC (Differential Scanning Calorime- ter, DSC2, Mettler Toledo, USA). The specific heat data are shown in Figure 1 for different PCDs with a PCM volume fraction (ϕ) between 5% and 25%. One can see that the emulsifier is also a PCM, which is therefore multifunctional in this application. The PCD stability against creaming, sedimentation and coalescence was also studied by an optical centrifuge of LUMiSizer, and the results demonstrated the sufficient stability of the proposed PCD even experienced more than 200 cycling times. ensure safety with a long lifespan in a high voltage envi- ronment and a high heat flux over a temperature range of ~320.65-323.15 K. The melting range of the target PCM should align with the temperature range, taking into account the effect of the emulsifier for producing the emulsion (PCD). An extensive screening and formulation exercise was performed to produce the PCD using deionised water and a mixture of commercial fatty acid esters based PCM (50 w/w % Crodatherm 53 + 50 w/w % Crodatherm 47, Croda International PLC, UK). An emul- sifier, consisting of a long chain and a short chain ethoxylated fatty alcohols (75 w/w % Steareth-100 + 25 w/w % Steareth-2, KLK Oleo, Germany), and a small por- tion (1.6 w/w %) of glycerol, were added to the formula- tion to ensure the PCD stability. An optimal PCD was found to be containing 16 w/w % of PCM, 4 w/w % of emulsifier, 0.5 w/w % nucleation agents in continuous phase, which was used in this work. The dispersion was manufactured by dispersing PCM phase into deionized water using a Polytron 10-35 GT lab rotor-stator at the Lucerne University of Applied Sciences and Arts, Switzerland. The rheological behaviour of PCD was measured with a rotational MCR 502 rheometer (Anton Paar, UK). The influences of shear rate and temperature on PCM and emulsifier viscosities are illustrated in Figure 2A. 2.1 \| PCD preparation and properties measurements To meet the cooling requirements of a HVDC converter, the PCD should have low electrical conductivity to LI ET AL. LI ET AL. 5823 0 50 100 150 5 200 10 15 20 200 250 300 350 400 450 μ, mPas γ, 1/s Emulsifier, 328.15 K Emulsifier, 323.15 K Emulsifier, 328.15 K PCM, 328.15 K PCM, 323.15 K 330 332 334 336 338 19 20 21 300 350 400 μ, mPas K T, PCM Emulsifier (A) (B) (C) (D) 0 50 100 150 200 1 10 100 1000 10000 γ, 1/s μ, mPa•s 298.15 K 318.15 K 323.15 K 328.15 K 295 300 305 310 315 325 320 330 335 1 2 3 4 5 6 T, K μ, mPa•s φ=10% φ=20% φ=25% FIGURE 2 A, Rheology behaviour of pure PCM and emulsifier under various shear rates and temperature; B, temperature dependence f viscosity at a shear rate of 100 seconds−1; C, rheological behaviour of PCD under various shear rates and temperatures with ϕ = 20%; nd D, temperature dependence of viscosity of the PCD at γ = 100 s−1 [Colour figure can be viewed at wileyonlinelibrary.com] 0 50 100 150 5 200 10 15 20 200 250 300 350 400 450 μ, mPas γ, 1/s Emulsifier, 328.15 K Emulsifier, 323.15 K Emulsifier, 328.15 K PCM, 328.15 K PCM, 323.15 K (A) (C) 330 332 334 336 338 19 20 21 300 350 400 μ, mPas K T, PCM Emulsifier (B) (D) (A) (B) (D) (C) 0 50 100 150 200 1 10 100 1000 10000 γ, 1/s μ, mPa•s 298.15 K 318.15 K 323.15 K 328.15 K 295 300 305 310 315 325 320 330 335 1 2 3 4 5 6 T, K μ, mPa•s φ=10% φ=20% φ=25% FIGURE 2 A, Rheology behaviour of pure PCM and emulsifier under various shear rates and temperature; B, temperature dependence of viscosity at a shear rate of 100 seconds−1; C, rheological behaviour of PCD under various shear rates and temperatures with ϕ = 20%; and D, temperature dependence of viscosity of the PCD at γ = 100 s−1 [Colour figure can be viewed at wileyonlinelibrary.com] The size distribution of the PCD was measured using the dynamic light scattering method with a Zetasizer Nano-ZS (Malvern, UK) in BCES of University of Birmingham, and the results are shown in Figure 3. The particle size distribu- tion could be fitted by the Rosin-Rammler (R-R) distribution (given in the inset), Yd = e−d=d ð Þ n , with Yd the cumulative PCD volume fraction having a diameter larger than d; n = 1.276 the size distribution parameter; d = 3.0 μm the particle size and size constant; and dpmin = 0.065 μm and dpmax = 6.5 μm, corresponding to 99.9% and 0.1% of the cumulative fraction for the R-R distribution,58 respec- tively. The average deviation between the measurements and fitted correlation is 12.89%. resolution.56,57 An increase in the shear rate to ~25 seconds−1 leads to the viscosity of almost constant for a given temperature, exhibiting the Newtonian behaviour. Such a rheological behaviour is also confirmed by the lin- ear relationship between viscosity and temperature under a constant shear rate of 100 seconds−1 (Figure 2B). Figure 2A and B also indicate that viscosity of the emulsifier is signifi- cantly higher than that of PCM by a factor of 20. resolution.56,57 An increase in the shear rate to ~25 seconds−1 leads to the viscosity of almost constant for a given temperature, exhibiting the Newtonian behaviour. Such a rheological behaviour is also confirmed by the lin- ear relationship between viscosity and temperature under a constant shear rate of 100 seconds−1 (Figure 2B). Figure 2A and B also indicate that viscosity of the emulsifier is signifi- cantly higher than that of PCM by a factor of 20. Unlike the PCM and the emulsifier, the rheological behaviour of the PCD follows a non-Newtonian charac- teristic, as shown in Figure 2A and B. With increasing shear rate, the PCD viscosity decreases regardless of the PCM at the solid or liquid status, and the viscosity of the PCD can reach up to 10 000 mPas at a shear rate below ~ < 25 seconds−1 even when PCD is melted (323.15 K and 328.15 K); see Figure 2C. Increasing the PCM con- centration increases the PCD viscosity (Figure 2D). The PCD viscosity is lower when the PCM is at the liquid sta- tus as shown in Figure 2D, but the decrease is rather smaller compared to the low shear rate values. At a shear rate lower than ~25 seconds−1, an irregular fluctua- tion of the viscosity occurs mainly due to low torque The thermo-physical properties of PCD were mea- sured at the Birmingham Centre for Energy Storage TABLE 1 Thermo-physical properties of water, PCM and PCD under both solid and liquid status of the PCM ρ [kg/m3] k [W/(m K)] cp [kJ/(kg K)] Tmelt [C] Δhpc [kJ/kg] Working Temperature [K] 298.15 328.15 298.15 328.15 298.15 328.15 - - Water 998.2 988.4 0.61 0.651 4.18 4.18 - - PCM 902 827.2 0.231 0.206 2.78 2.38 47.13 193.38 Emulsifier 1011.4 1005.3 0.293 0.301 2.28 2.21 49.72 141.53 PCD 983.9 980.5 0.529 0.561 3.82 3.78 50.95 26.11 300 310 320 330 340 -4 0 4 8 Heat Flow, W/g T, K PCM Emusifier 300 310 320 330 340 3000 6000 9000 12000 T, K Cp, J/kg•K φ = 5% φ = 10% φ = 20% φ = 25% (A) (B) FIGURE 1 A, DSC measurement of PCM and emulsifier, and B, Cp variation with temperature of different PCM concentration, TABLE 1 Thermo-physical properties of water, PCM and PCD under both solid and liquid status of the PCM ρ [kg/m3] k [W/(m K)] cp [kJ/(kg K)] Tmelt [C] Δhpc [kJ/kg] Working Temperature [K] 298.15 328.15 298.15 328.15 298.15 328.15 - - Water 998.2 988.4 0.61 0.651 4.18 4.18 - - PCM 902 827.2 0.231 0.206 2.78 2.38 47.13 193.38 Emulsifier 1011.4 1005.3 0.293 0.301 2.28 2.21 49.72 141.53 PCD 983.9 980.5 0.529 0.561 3.82 3.78 50.95 26.11 300 310 320 330 340 -4 0 4 8 Heat Flow, W/g T, K PCM Emusifier 300 310 320 330 340 3000 6000 9000 12000 T, K Cp, J/kg•K φ = 5% φ = 10% φ = 20% φ = 25% (A) (B) FIGURE 1 A, DSC measurement of PCM and emulsifier, and B, Cp variation with temperature of different PCM concentration, ϕ = 5%, 10%, 20% and 25% [Colour figure can be viewed at wileyonlinelibrary.com] sical properties of water, PCM and PCD under both solid and liquid status of the PCM ABLE 1 Thermo-physical properties of water, PCM and PCD under both solid and liquid status of the PCM 300 310 320 330 340 -4 0 4 8 Heat Flow, W/g T, K PCM Emusifier (A) 300 310 320 330 340 3000 6000 9000 12000 T, K Cp, J/kg•K φ = 5% φ = 10% φ = 20% φ = 25% (B) (B) Heat Flow, W/g FIGURE 1 A, DSC measurement of PCM and emulsifier, and B, Cp variation with temperature of different PCM concentration, ϕ = 5%, 10%, 20% and 25% [Colour figure can be viewed at wileyonlinelibrary.com] LI ET AL. 2.2 \| Experimental rig of cooling performance measurements The performance of the PCD for the HVDC converter cooling was performed by using a heat sink with double- LI ET AL. 5824 PT100 1/3 DIN wire sensors (Roth +Co AG, Switzerland) were placed to measure the temperatures of one of the blocks with a measurement uncertainty less than ±0.1 K; see late for more details. The heating section was thermally insulated by an aluminium foil together with PIR shells (Swisspor, Switzerland). The pressure drop was measured by a Deltabar S PMD75 device (Endress + Hauser, Switzer- land), with a range of 0-40 bar and an accuracy of 0.075%. 2 0 6 4 8 10 0.0 0.2 0.4 0.6 0.8 1.0 1 0.1 10 0 1 2 3 4 5 6 φ, % Yd dp, μm dp, μm Experiment Rosin-Rammler fitting y=exp(-(0.333x)^1.276) 2 0 6 4 8 10 0.0 0.2 0.4 0.6 0.8 1.0 1 0.1 10 0 1 2 3 4 5 6 φ, % Yd dp, μm dp, μm Experiment Rosin-Rammler fitting y=exp(-(0.333x)^1.276) FIGURE 3 Relationship between particle volume fraction and particle diameter. Inset shows the fitting of Rosin–Rammler distribution [Colour figure can be viewed at wileyonlinelibrary.com] Figure 4C illustrates a schematic of the experimental rig. The fluid was driven by an impeller pump (Zuwa NIROSTAR/V 2000-B/PT, Germany) into the system. A Coriolis flow meter (Promass F83, Endress + Hauser AG, Switzerland) and a by-pass valve were used to control the mass flowrate of the flow with an uncertainty of ±0.004 _m, where _m is the mass flow rate. To ensure a stable and safe operation in the electrical environment, the PCD went through a deionization unit filled with a resin, Amberjet UP 6150, purchased from Rohm & Haas, USA. A nylon filter with a mesh size of 200 μm was used to remove any unwanted impurity before the PCD entering the system. To remove the heat added to the fluid in the heat section for recirculation, a thermostat (TYA-201, JUMO Co Ltd, UK) was used to cool the fluid before re- entering the testing rig. All measurements were recorded at the steady state (often reached after ~10 minutes opera- tion). To determine heat loss of the system, the setup was first tested with water under various _m and Pel. 2.2 \| Experimental rig of cooling performance measurements The heat loss, identified as the difference between Pel and water heating from inlet to outletQloss = Pel −_m CpTout−CpTin , was found to be 20 W on average, which was used for the PCD calculations. FIGURE 3 Relationship between particle volume fraction and particle diameter. Inset shows the fitting of Rosin–Rammler distribution [Colour figure can be viewed at wileyonlinelibrary.com] layer oblique-crossed ribs, manufactured by Mersen Co Ltd, Shanghai. The heat sink was made of AlMgSi0.5, which has heat and electrical conductivities of 185 W/ (m K) and 28.6 × 106 S/m, respectively. Figure 4A shows a snapshot of the heat sink. Two identical circular heating surfaces with a diameter of 135 mm were made in close contact with the top and bottom surfaces of the heat sink. The cooling section was formed by overlapping two web layers within the heat sink to increase the heat transfer area and fluid turbulence. The angle between the ribs of upper and bottom layer was 30. The fluid channel was formed between the rib gaps with a cross-section dimension of 3.6 mm (Wc) × 4 mm (Hc) and porosity of 0.576. The cooling fluid was pumped into the circular inlet (internal diameter, ID, = 9 mm), exchanging heat with the cooling section before exiting at the outlet (with the same ID as the inlet). The cooling section had width of 130 mm and length (refers to the distance between inlet and outlet centre) of 118 mm. 2.3 \| Data analysis Temperatures measured by T03 to T06 (Figure 4C) were from the copper block side rather than from the heat sink. To accurately describe the thermal performance of the heat sink, thermal resistance within the copper (Rth1) and ITCR between the aluminium heat sink and copper block (Rth2) should be estimated, as illustrated in Figure 4D. The copper resistance is a function of thermal conductivity, k, distance, s, and heating area, A: The cooling experiment was conducted at Lucerne Uni- versity of Applied Science and Arts (Switzerland). The cir- cular heating surfaces were provided by two copper blocks powered by electrical heating as shown in Figure 4B. The blocks and the heating surfaces of the heat sink were tightly clamped by torque wrenches. Each block was equipped with six heating cartridges so that a constant heat flux boundary condition can be maintained, to simulate the power dissipation of a HVDC converter thyristor. The maximum heating power per cartridge was 200 W, giving a total power of 1.5 kW per side. All 12 cartridges were powered by a thyristor power controller (TYA-201, JUMO Co Ltd, UK), allowing a precise regulation of the heat input with an uncertainty of ±0.02 Pel (Pel is the electrical power). At a distance of 2.5 mm from the interface, four Rth1 = s=kA ð1Þ ð1Þ which yields a constant value of 4.48 × 10−4 K/W. The relationship between ITCR and thermal conductance (hj) is defined as Rth2 = 1 hjA ð2Þ ð2Þ LI ET AL. LI ET AL. 2.3 \| Data analysis 5825 FIGURE 4 A, A snapshot of the heat sink, B, a snapshot of the heating section, C, a schematic diagram of the experimental rig (1 – Deionisation; 2 – Filter; 3 – Converter; 4 – Re-dispersion unit; 5 – Re-cooling unit; 6 – Pump; 7 – By-pass valve; 8 – Throttle valve), and D, schematic of the thermal resistance between heating surface and copper block [Colour figure can be viewed at wileyonlinelibrary.com] FIGURE 4 A, A snapshot of the heat sink, B, a snapshot of the heating section, C, a schematic diagram of the experimental rig (1 – Deionisation; 2 – Filter; 3 – Converter; 4 – Re-dispersion unit; 5 – Re-cooling unit; 6 – Pump; 7 – By-pass valve; 8 – Throttle valve), and D, schematic of the thermal resistance between heating surface and copper block [Colour figure can be viewed at wileyonlinelibrary.com] where ωX is the uncertainty of variable X and ωXn is the uncertainty of parameter xn. The maximum uncertainty of temperature derived is therefore ±0.35 K. Since both metal surfaces were planar and finely polished, the thermal conductance was chosen as hj = 12-22 kW/(m2 K). This gives to Rth2 a range between 3.18 × 10−3 K/W and 5.82 × 10−3 K/W for the present work. The temperature difference between heating sur- face and T3456 (as shown in Figure 4D) can thus be deter- mined by ΔT = P(Rth1 + Rth2), where P is the heating power. 5826 LI ET AL. temperature; and g0 is the radial distribution func- tion defined as64: where the equation of motion is solved for each of particles, the Euler granular model solves only one conservation equation for the solid phase. Thus it can be used for concen- trated fluid in a relative large computational domain. The k-ω turbulence model with shear stress transport (SST) model was adopted, which uses the k-ε model in the core of the flow and switches to the k-ω model in the near-wall regions. The rationale for the use of such a model lies in their proven accuracies in solving mixture problems in the near-wall region.59,60 The detailed turbulence kinetic energy and specific dissipation rate can be referred to Menter.61 g0 = 1−αs=αs,max ð Þ1=3 h i−1 ð9Þ ð9Þ where αs, max is the maximum packing fraction and equals to 0.62 for monodisperse spheres.65 The granular temperature θs is used to describe the fluctuating particle motion, derived from the fluctuating energy balance equation as follows: and specific dissipation rate can be referred to Menter.61 3 2 ∂ ∂t αsρsθs ð Þ + r αsρs v ! sθs = τ = s −Ps I = rv ! s + r kθsrθs ð Þ + φsl −γθs ð10Þ • Continuity equation for ith phase: ð10Þ ∂ ∂t αiρi ð Þ + r αiρi v ! i = 0 ð4Þ ð4Þ where kθs , φsl and γθs represent the diffusion coefficient, interphase energy exchange and collisional dissipation of energy, respectively. The solid pressure, Ps, follows the work of Gidaspow66 and given by: where the subscript i = l, s, representing the liquid phase or solid phase, respectively; α, ρ and v ! denote respectively the volume concentration, density and velocity of the differ- ent phase. Ps = αsρsθs 1 + 2g0αs 1 + ess ð Þ ½ ð11Þ ð11Þ • Momentum conservation equation: • Momentum conservation equation: The granular temperature at the inlet is given by67: Ts = 0:004 Us ð Þ2 ð12Þ ∂ ∂t αiρi v ! i + r αiρi v ! i v ! i = −αirP + r τ = i + αiρi g ! + F ! D + F ! L + F ! td + F ! 3.1 \| Mathematical model The measurement uncertainty is defined as the uncer- tainty of a derived parameter, X, caused by the uncer- tainties of individual measured variables, can be calculated by the following expression43: With the Euler-Euler method, the PCM particles and glycerol-water solution are treated as interpenetrative con- tinua, coupled through the interphase momentum and heat balance. The PCD is assumed to be incompressible and in the turbulent flow regime, and the PCM particles are assumed to be smooth, inelastic and spherical during phase change. Compared to the Eulerian-Lagrangian method ωX = ∂X ∂x1 2 ω2 X1 + ∂X ∂x2 2 ω2 X2 + … + ∂X ∂xn 2 ω2 Xn ð3Þ ð3Þ LI ET AL. 3.2 \| Computational domain and boundary conditions ð16Þ A symmetric computational domain with the same dimension as the experimental heat sink was used and shown in Figure 5A. To avoid poor quality meshed grids, the geometry was slightly simplified by removing the round angles and pilot holes as given in Figure 5B. The simplification has been numerically proved to barely affect the temperature distribution of the heat sink. The fluid domain and the nearby zone were meshed with tet- rahedron grids while the rest solid domain with hexahe- dron grids. Figure 5C gives a side view of the computational grids on cross-section A. The average mesh size of the fluid domain varies between 0.1 and 0.35 mm for a grid-dependence analysis, and the 0.2 mm mesh size was found to be sufficient for an accurate cal- culation without consuming too much computational resource. Mesh sizes larger than 0.25 mm was found to result in temperature deviation higher than 3.83% com- pared to the experimental value thus not recommended. The selected mesh yielded a dimensionless wall distance y + near 1 in the sublayer, and average solid domain size of 0.48 mm. The mesh count was approximately 5 million for the whole solution domain, with 43.19% tetrahedron and 56.81% hexahedron. A periodic meshed fluid domain was given in Figure 5D for easily understand by the readers. where CL is the lift force coefficient and taken as 0.25 for a spherical particle. The particle distribution is domi- nated by the turbulent dispersion force when the size of turbulent eddies are larger than the particle size69: F ! td,1 = Ctdγsl μt,1 σsl rαs αs −rαl αl ð17Þ ð17Þ where Ctd is the dispersion coefficient and determined to be 1. The energy conservation equation for ith phase can be formulated as: ∂ ∂t αiρiHi ð Þ + r αiρi v ! iHi = r λe,irTi ð Þ + τ= i v ! i −hsl Ti −Tq ð18Þ ð18Þ where hsl is the particle–liquid heat transfer coefficient and given by70: hsl = 6αsλl d2 h ð7 + 10αl + 5α2 l Þð1 + 0:7Re0:2 s Pr1=3Þ + ð1:33−2:4αl + 1:2α2 l ÞRe0:7 s Pr1=3i ð19Þ ð19Þ The effective thermal conductivities in the main flow and near-wall regions have been given by Zehner and Schlünder71 and Legawiec and Ziólkowski,72 respectively. 5827 F ! L,l = 2CLv0:5ρdij ρsd dlkdkl ð Þ0:25 v ! s −v ! l ð16Þ VM ð5Þ ð12Þ ð5Þ Among the interfacial forces, the dominate drag force is described by Syamlal and O'Brien63 model which takes the following form: where τ = l represents the stress-strain tensor of the ith phase: F ! D = CDResαl 24v2 s ð13Þ τ = i = αiμi rv ! i + rv ! i T − ζi μi −2 3 r v ! i I ! ð6Þ ð13Þ ð6Þ where μi, ζi and I ! are the shear viscosity, bulk viscosity and unit vector, respectively; F ! D,l , F ! L,l ,F ! td,l , F ! VM are respectively the drag force, lift force, turbulent dispersion force and virtual mass force between the two phases. The solid phase bulk viscosity, ζs, and shear viscosity, μs, are given by Lun et al.62 and Syamlal and O'Brien,63 respectively: where CD is the drag force coefficient expressed as: CD = 0:63 + 4:8 ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Res=vs p !2 ð14Þ ð14Þ A virtual mass would occur when the solid phase accel- erates relative to the carrier phases. The inertia of the carrier-phase mass encountered by the accelerating parti- cles exerts the virtual mass force on particles, given as: ζs = 4αsρsdg0 1 + ess ð Þ ﬃﬃﬃﬃ θs p 3 ﬃﬃﬃπ p ð7Þ ð7Þ 0αs 2 ð8Þ F ! VM = 0:5αsρl dlv ! l−dsvs ! dt ! ð15Þ g0 ﬃﬃﬃﬃ θs p + 10ρsd ﬃﬃﬃﬃﬃﬃﬃ πθs p 96αs 1 + ess ð Þg0 1 + 4 1 + ess ð Þg0αs 5 2 ð8Þ F ! VM = 0:5αsρl dlv ! l−dsvs ! dt ! ð15Þ F ! VM = 0:5αsρl dlv ! l−dsvs ! dt ! ð15Þ μs = 4αsρsd 1 + ess ð Þg0 ﬃﬃﬃﬃ θs p 5 ﬃﬃﬃπ p + 10ρsd ﬃﬃﬃﬃﬃﬃﬃ πθs p 96αs 1 + ess ð Þg0 1 + 4 1 + ess ð Þg0αs 5 2 ð8Þ ð15Þ ð8Þ In a non-uniform or swirling flow, the dispersed par- ticles experience a lift force perpendicular to the relative velocity vector68: where ess is the particle–particle restitution coeffi- cient and chosen to be 0.9; θs is the granular LI ET AL. LI ET AL. 3.2 \| Computational domain and boundary conditions In the near-wall region, no slip boundary is assumed for the liquid phase while a partial slip boundary is assumed for solid phase following Johnson-Jackson equations73 with the shear force given as: The 3D simulations were conducted using a CFD software under the FLUENT 18.2 (ANSYS Inc.) envi- ronment, which employs the Euler-Euler model for two-phase mixtures. The water properties were incor- porated as a polynomial function with coefficients listed in Table 2, so does the measured thermos- physical properties of the PCD. The inlet mass flowrate varied from 4 to 10 L/min, and the atmospheric pres- sure (101 325 Pa) was assigned to the sink outlet. Con- stant velocity of the two phases (Uin = 1.05-2.62 m/s) and solid volume fraction (ϕ = 5-25%) were applied at the heat sink inlet. A uniform heat flux was assigned to the circular heating surface, which ranges from 0.5 to 2.5 kW. The rest walls of the heat sink were set to be adiabatic at the room temperature of 298.15 K. In addi- tion, as mentioned earlier, a no-slip boundary condi- tion was given to the water phase and the Johnson- Jackson partial-slip condition was adopted for the PCM phase at the wall. The governing equations for both fluid and particle phases of the PCD were discretized using the finite-volume method with a second order upwind scheme. The pressure-velocity coupling corre- lation was solved with the phase-coupled SIMPLE (Semi-Implicit Method for Pressure Lined Equation) algorithm.75 The time step was selected to be τsw = − ﬃﬃﬃ 3 p ρsg0αsφ ﬃﬃﬃﬃ θs p ν ! sw 6αs,max ð20Þ ð20Þ and the fluctuating energy expressed by: qsw = ﬃﬃﬃ 3 p ρsg0αsφ ﬃﬃﬃﬃ θs p ν ! sw ν ! sw 6αs,max − ﬃﬃﬃ 3 p ρsg0αsφ 1−e2 w θ3=2 s 4αs,max ð21Þ ð21Þ where φ and ew represent the specularity coefficient and particle–wall restitution coefficient, respectively. The value of φ and ew are adjusted to be between 0.9-0.99 and 0.0001-0.001, respectively, to ensure mass balance for various conditions. The discrepancy of solid distribution and velocity caused by coefficient differences has been proven to be very small.74 LI ET AL. 4.1 \| Pressure drop comparison between PCD and water Figure 6A compares the pressure drop of water and PCD derived from both experiment and simulation for Q = 1-14 L/min. Inlet temperature of 313.15 K and 333.15 K are tested for water. The discrepancy of experimental pressure drop caused by inlet temperature is small, with an average deviation of 5.82% under all flow rates. One can see that the Euler-Euler model agrees well with experimental measure- ments within 6.12%. A change in the particle size (1-7 μm) barely influences the pressure drop (deviation within 0.11%), thus the medium particle size, d = 3.5 μm, was employed for the rest simulations. 5829 0.0001 second and the convergence criterion was set to be 10−4 for all variables. the viscosity of PCD can be 10 times that of water during phase change, the pressure drop only increased by 17.01% on average under the studied conditions. Chen et al.32 showed a PCD (d = 51 μm, 30 wt. %) pressure drop increase by up to 35.71% in a straight pipe with the vis- cosity of the PCD 5.57 times that of water. Similarly Alvarado et al.33 proved that the increased pressure cau- sed by PCD barely affected the pumping work even though the viscosity was 3.5 times higher than water. 828 5828 RE 5 A, Symmetric computational domain and boundaries, B, grid system, C, Side views of computational grids on cross- A, and D, grid of the a periodic web [Colour figure can be viewed at wileyonlinelibrary.com] TABLE 2 Thermal and transport properties of water as a function of temperature A1 A2 A3 A4 A5 /m3] 1.351E+03 −2.422 6.32E-03 −6.989E-06 - /(mK)] −2.946E-01 4.757E-03 −5.712E-06 - - (k K)] 6 515E 01 6 166E 01 2 299E 03 3 755E 06 2 276E 09 LI ET AL 5828 FIGURE 5 A, Symmetric computational domain and boundaries, B, grid system, C, Side views of computational grids on cross- section A, and D, grid of the a periodic web [Colour figure can be viewed at wileyonlinelibrary.com] TABLE 2 Thermal and transport properties of water as a function of temperature TABLE 2 Thermal and transpor properties of water as a function of temperature ; A1 A2 A3 A4 A5 ρ [kg/m3] 1.351E+03 −2.422 6.32E-03 −6.989E-06 - k [W/(mK)] −2.946E-01 4.757E-03 −5.712E-06 - - cp [J/(kgK)] 6.515E+01 −6.166E-01 2.299E-03 −3.755E-06 2.276E-09 μ [kg/(ms)] 3.738E-02 −3.093E-04 9.609E-07 −1.320E-09 6.746E-13 Note: ;(T) = A1 + A2T + A3T2 + A4T3 + A5T4 with T in Kelvin. Note: ;(T) = A1 + A2T + A3T2 + A4T3 + A5T4 with T in Kelvin. Note: ;(T) = A1 + A2T + A3T2 + A4T3 + A5T4 with T in Kelvin. LI ET AL. 830 5830 LI ET AL. LI ET AL. ð23Þ ρfluid,avg = Ð n 0 ρfluid,idvi Ð n 0 dvi ð26Þ Ppump = mfΔP ρavg,fluid ð23Þ ð26Þ ð23Þ where Dhyd is the hydraulic diameter of channel; ρavg, fluid and Uavg represent the volumetric-averaged fluid density and velocity, respectively; N is the channel num- ber of the two-layer web; W is the width of cooling section and α is the angle of web against the side wall (75). Dhyd in Equation (22) is defined by the channel height (Hc = 4.0 mm) and width (Wc = 3.6 mm) as: Figure 6B plots the fFanning and Ppump against the volumetric-averaged velocity at Q = 4-10 L/min. The inlet velocity under such mass flowrate range varies within 1.05-2.62 m/s, which is markedly higher than the Uavg (0.35-0.89 m/s). The Uavg of the PCD is smaller than that of water by 3.04% on average, due to higher viscosity as discussed earlier. Compared to water, the fFanning of the PCD shows a steeper decreasing trend with an aver- age value of 19.40% higher than that of water. Addition- ally, PCD requires a higher pumping power than that of water (by 16.19% on average), especially at a high Uavg. Dhyd = 4A P = 2HCWc HC + Wc ð24Þ ð24Þ and ρavg, fluid and Uavg are given respectively by: and ρavg, fluid and Uavg are given respectively by: The effect of PCM mass fraction on pressure perfor- mance is also evaluated under Q = 8 L/min, as given in Figure 6C. Both the fFanning and Ppump increase with the growth of mass fraction. 4.1 \| Pressure drop comparison between PCD and water It is noteworthy that although To evaluate the Fanning friction fFanning and the pumping power Ppump, the following equations are used: • Averaged Fanning friction factor • Averaged Fanning friction factor fFanning = 2ΔPDhydsinα Nρavg,fluidUavg2W ð22Þ • Pumping power fFanning = 2ΔPDhydsinα Nρavg,fluidUavg2W ð22Þ ð22Þ 2 4 6 8 10 12 14 0.0 0.2 0.4 0.6 0.8 1.0 ∆P, bar Q, L/min PCD, exp, Tin = 323.15 K Water, exp, Tin = 313.15 K 0.3 0.4 0.5 0.6 0.7 0.8 Water, exp, Tin = 333.15 K PCD, sim Water, sim 0.9 0.40 0.45 0.50 0.55 0.60 fFanning Uavg, m/s 0 2 4 6 8 10 12 Ppump, W water, Fanning friction PCD, Fanning friction water, pumping power PCD, pumping power (A) (B) (C) FIGURE 6 A, Experimental and numerical pressure drop as a function of flow rate for water and PCD, Q = 4, 6, 8, 10 L/min; B, Fanning friction factor and pumping power as a function of flow rate, Q = 4, 6, 8, 10 L/min, P = 1.1 kW; C, Pumping power, fanning friction and volumetric average velocity as a function of PCM mass fraction, Q = 8 L/min and P = 1.1 kW [Colour figure can be viewed at wileyonlinelibrary.com] 2 4 6 8 10 12 14 0.0 0.2 0.4 0.6 0.8 1.0 ∆P, bar Q, L/min PCD, exp, Tin = 323.15 K Water, exp, Tin = 313.15 K 0.3 0.4 0.5 0.6 0.7 0.8 Water, exp, Tin = 333.15 K PCD, sim Water, sim 0.9 0.40 0.45 0.50 0.55 0.60 fFanning Uavg, m/s 0 2 4 6 8 10 12 Ppump, W water, Fanning friction PCD, Fanning friction water, pumping power PCD, pumping power (A) (B) (A) (B) (C) (C) FIGURE 6 A, Experimental and numerical pressure drop as a function of flow rate for water and PCD, Q = 4, 6, 8, 10 L/min; B, Fanning friction factor and pumping power as a function of flow rate, Q = 4, 6, 8, 10 L/min, P = 1.1 kW; C, Pumping power, fanning friction and volumetric average velocity as a function of PCM mass fraction, Q = 8 L/min and P = 1.1 kW [Colour figure can be viewed at wileyonlinelibrary.com] 5831 section (L = 118 mm). The experimental heating surface represents the temperature directly derived from copper block, and the deviation caused by thermal resistance and temperature uncertainty is ±1.82 K for P = 1.1 kW. Under flow rates of Q = 4 and 8 L/min (shown in Figure 7A and B), the numerical water temperatures of the inlet and outlet are almost identical to experiment measurement. The fluid temperature difference (ΔTfluid = Tout - Tin) decreases markedly with the increas- ing flow rate, from 8.09 K at Q = 4 L/min to 3.94 K at Q = 8 L/min. On the heating surface, high temperature are spotted near the outlet side from simulation while the experiment found it near the centre. The difference may be attributed to the non-uniform temperature distribu- tion of six heating cartridges in the copper block. Never- theless, the simulation predictions fall into the range of experimental measurement. The surface temperature of becomes 5.52% from ϕ = 5% to ϕ = 25%. The decreasing Uavg at high mass fraction attributed to the increasing PCD viscosity. However, the influence of PCM concentra- tion on Uavg is very limited as the growth is only 3.05% from ϕ = 0 to ϕ = 25%. 830 Taken fFanning as an example, the increment is 13.47% from ϕ = 0 to ϕ = 5% but Ufluid,avg = Ð n 1 Uidvi Ð n 0 dvi ð25Þ ð25Þ 0.0 0.2 0.4 0.6 0.8 320 324 1.0 328 332 336 l/L P = 1.1 kW T, K Q = 4.0 L/min Simulation heating surface Experimental heating surface Simulation water Experimental water .0 0 0.2 .0 4 0.6 .0 8 320 1.0 324 328 332 Q = 8.0 L/min T, K l/L P = 1.1 kW Simulation heating surface Experimental heating surface Simulation water Experimental water .0 0 0.2 .0 4 0.6 .0 8 1.0 320 324 328 332 T, K Q = 4.0 L/min l/L P = 1.1 kW Simulation heating surface Experimental heating surface Simulation PCD Experimental PCD .0 0 0.2 .0 4 0.6 .0 8 320 324 1.0 328 332 T, K Q = 8.0 L/min l/L P = 1.1 kW Simulation heating surface Experimental heating surface Simulation PCD Experimental PCD (C) (D) (A) (B) IGURE 7 Comparison of water temperature on the heating surface and fluid from experiment and simulation, A, Q = 4 L/min, = 1.1 kW; B, Q = 8 L/min, P = 1.1 kW; Comparison of PCD temperature on the heating wall and fluid from experiment and simulation, C, Q = 4 L/min, P = 1.1 kW; D, Q = 8 L/min, P = 1.1 kW [Colour figure can be viewed at wileyonlinelibrary.com] (A) .0 0 0.2 .0 4 0.6 .0 8 320 1.0 324 328 332 Q = 8.0 L/min T, K l/L P = 1.1 kW Simulation heating surface Experimental heating surface Simulation water Experimental water (D) (B) 0.0 0.2 0.4 0.6 0.8 320 324 1.0 328 332 336 l/L P = 1.1 kW T, K Q = 4.0 L/min Simulation heating surface Experimental heating surface Simulation water Experimental water (C) (A) (B) T, K 3 (D) (C) .0 0 0.2 .0 4 0.6 .0 8 320 324 1.0 328 332 T, K Q = 8.0 L/min l/L P = 1.1 kW Simulation heating surface Experimental heating surface Simulation PCD Experimental PCD ( ) .0 0 0.2 .0 4 0.6 .0 8 1.0 320 324 328 332 T, K Q = 4.0 L/min l/L P = 1.1 kW Simulation heating surface Experimental heating surface Simulation PCD Experimental PCD FIGURE 7 Comparison of water temperature on the heating surface and fluid from experiment and simulation, A, Q = 4 L/min, P = 1.1 kW; B, Q = 8 L/min, P = 1.1 kW; Comparison of PCD temperature on the heating wall and fluid from experiment and simulation, C, Q = 4 L/min, P = 1.1 kW; D, Q = 8 L/min, P = 1.1 kW [Colour figure can be viewed at wileyonlinelibrary.com] LI ET AL. 830 LI ET AL. 4.3 \| Parameters influences of PCD Having verified the confidence of numerical model, the influences of flow rate, heating flux and PCM fraction on heat transfer as well as the overall performance were dis- cussed in detail. The inlet temperature of PCD was cho- sen just below the PCM solidus point and kept at Tin = 320.15 K. Figure 8 shows the 3D temperature distri- bution of the heat sink surface under various PCD flow rates (Q = 4-10 L/min), where PCD was pumped into the heat sink from left entrance. High temperature can always be spotted near the outlet region on the heating surface. The highest surface temperature at Q = 4 L/min reaches 353 K but decreases to 339 K at Q = 10 L/min, and a more evenly distributed temperature surface is obtained at high flow rate. The uniform temperature Similar comparison are also conducted for PCD at Q = 4 and 8 L/min and P = 1.1 kW, as shown in Figure 7C and D. The outlet fluid temperature from simulation is slightly lower than the experiment mea- surements, by 0.8 K at Q = 8 L/min. The ΔTfluid of PCD decreases from 3.6 K at Q = 4 L/min to 1.8 K at Q = 8 L/min. The predicted surface temperature fol- lows the trend of experiment, with the highest temper- ature located nearly 1/4 L distance from outlet. The average surface temperature of Q = 8 L/min is 1.73 K lower than that of Q = 4 L/min. 5832 LI ET AL. LI ET AL. features of PCD can now be confidently studied using the Euler-Euler model. Q = 4 L/min is significantly higher than that of Q = 8 L/ min under the same P, by 4.37 K on average from simula- tion. The discrepancies of heat sink and fluid tempera- tures caused by different PCM particle sizes (1-7 μm) are very small (less than ±0.05 K), and the medium particle size d = 3.5 μm will be used for the following thermal analysis. 4.2 \| Temperature comparison between PCM and water Both the measurement uncertainty (± 0.35 K) and ther- mal resistance (3.628 × 10−3 K/ Both the measurement uncertainty (± 0.35 K) and ther- mal resistance (3.628 × 10−3 K/ Both the measurement uncertainty (± 0.35 K) and ther- 3 W < Rth < 6.268× 10−3 K/W) are considered when ana- lysing the temperature heating sink. On Figure 11, the temperatures of heating surface and fluid are plotted against dimensionless distance, which is defined as the ratio of distance from inlet centre and length of cooling FIGURE 8 Comparison of temperature distribution on the heat sink surface under various PCD flow rates with A, Q = 4 L/min, B, Q = 6 L/min, C, Q = 8 L/min and D, Q = 10 L/min at the conditions of P = 2.2 kW, ϕ = 20% and T = 325-353 K [Colour figure can be viewed at wileyonlinelibrary.com] FIGURE 8 Comparison of temperature distribution on the heat sink surface under various PCD flow rates with A, Q = 4 L/min, B, Q = 6 L/min, C, Q = 8 L/min and D, Q = 10 L/min at the conditions of P = 2.2 kW, ϕ = 20% and T = 325-353 K [Colour figure can be viewed at wileyonlinelibrary.com] guarantees the safety working condition of sensitive HVDC converter components. Large temperature gradient can be observed at the surrounding area of heating surface, reaching temperature difference of 6 K within 2-3 mm. Similarly, the temperature on other walls of the heat sink decrease with the growth of flow rates. The non-uniformity of the heat sink temperature, defined as θ = THS,max −THS,min ð Þ q , yields θ = 1.3e-04 at Q = 4 L/min and θ = 7.8e-05 at Q = 10 L/min. The low non-uniformity presents a more evenly distributed temperature field inside the heat sink. temperature. The Tfluid, cal shows the same trend of Tfluid, sim, with average deviation of 0.828%. The temperature of average wetted wall (TWW,cal), which refers to the liquid-solid coupled surface, is employed to cal- culate the heat sink temperature36,76 and given as TWW,cal = THS,surf −Rcond HSQ ð27Þ ð27Þ where the 1D thermal conduction resistance of the heat sink is calculated as The volumetric-averaged fluid/solid temperature can be directly determined by Tfluid,sim = Ð n 0 Tidvi Ð N 0 dvi from simula- The volumetric-averaged fluid/solid temperature can be directly determined by Tfluid,sim = Ð n 0 Tidvi Ð N 0 dvi from simula- tion. However, it is difficult to measure the value from experiment and the connection between Tfluid, sim and Tin/Tout is unclarified for the heat sink with complex geometry. Figure 9A compares the PCD Tfluid, sim and cal- culated temperature, Tfluid,cal = Tin + Tout 2 , under different flow rates and heating powers. Overall, high heat power P and low flow rate Q favour the growth of PCD Rcond HS = HHS kHSAHS = 0:032 m 185 W=mK 3:14 × 0:0675 × 0:0675 ð Þm2 = 0:0121 K W ð28Þ tion. However, it is difficult to measure the value from experiment and the connection between Tfluid, sim and Tin/Tout is unclarified for the heat sink with complex geometry. Figure 9A compares the PCD Tfluid, sim and cal- culated temperature, Tfluid,cal = Tin + Tout 2 , under different flow rates and heating powers. Overall, high heat power P and low flow rate Q favour the growth of PCD ð28Þ where HHS is the average distance between the heating surface and wetted wall and THS, surf is the average tem- perature of the heating surface. 4.3 \| Parameters influences of PCD 5833 4.3 \| Parameters influences of PCD Overall, the proposed model yields accurate results in terms of pressure and temperature, and the hydro- and thermo-dynamic 0.5 1.0 1.5 2.0 2.5 320 322 324 326 T, K P, kW Q = 10 L/min, Tfluid,cal Q = 4 L/min, Tfluid,cal Q = 6 L/min, Tfluid,cal Q = 8 L/min, Tfluid,cal Q = 10 L/min, Tfluid,sim Q = 4 L/min, Tfluid,sim Q = 6 L/min, Tfluid,sim Q = 8 L/min, Tfluid,sim 0.5 1.0 1.5 2.0 2.5 324 328 332 336 340 344 T, K P, kW Q= 4L/min, THS,sim Q= 6L/min, THS,sim Q= 8L/min, THS,sim Q= 10L/min, TWW,cal Q= 10L/min, THS,sim Q= 4L/min, TWW,cal Q= 6L/min, TWW,cal Q= 8L/min, TWW,cal 0.5 1.0 1.5 2.0 319 2.5 320 321 322 323 324 325 T, Κ P, kW water, Q = 8 L/min, Tfluid,sim PCD, Q = 8 L/min, Tfluid,sim water, Q = 8 L/min, Tfluid,cal PCD, Q = 8 L/min, Tfluid,cal 0.5 1.0 1.5 2.0 322 324 2.5 326 328 330 332 334 T, Κ P, kW water, Q = 8 L/min, THS,sim PCD, Q = 8 L/min, THS,sim water, Q = 8 L/min, TWW,cal PCD, Q = 8 L/min, TWW,cal (A) (B) (C) (D) FIGURE 9 Temperature of PCD A, and HS B, from simulation and calculation, Q = 4-10 L/min, P = 0.5-2.5 kW, ϕ = 20%; C, volumetric-averaged fluid temperature; D, volumetric-averaged heat sink and wetting wall from simulation at Q = 8 L/min, P = 0.5-2.5 kW, ϕ = 20% [Colour figure can be viewed at wileyonlinelibrary.com] 0.5 1.0 1.5 2.0 2.5 320 322 324 326 T, K P, kW Q = 10 L/min, Tfluid,cal Q = 4 L/min, Tfluid,cal Q = 6 L/min, Tfluid,cal Q = 8 L/min, Tfluid,cal Q = 10 L/min, Tfluid,sim Q = 4 L/min, Tfluid,sim Q = 6 L/min, Tfluid,sim Q = 8 L/min, Tfluid,sim (A) (C) 0.5 1.0 1.5 2.0 2.5 324 328 332 336 340 344 T, K P, kW Q= 4L/min, THS,sim Q= 6L/min, THS,sim Q= 8L/min, THS,sim Q= 10L/min, TWW,cal Q= 10L/min, THS,sim Q= 4L/min, TWW,cal Q= 6L/min, TWW,cal Q= 8L/min, TWW,cal (B) (D) (A) (B) (C) (D) 0.5 1.0 1.5 2.0 322 324 2.5 326 328 330 332 334 T, Κ P, kW water, Q = 8 L/min, THS,sim PCD, Q = 8 L/min, THS,sim water, Q = 8 L/min, TWW,cal PCD, Q = 8 L/min, TWW,cal 0.5 1.0 1.5 2.0 319 2.5 320 321 322 323 324 325 T, Κ P, kW water, Q = 8 L/min, Tfluid,sim PCD, Q = 8 L/min, Tfluid,sim water, Q = 8 L/min, Tfluid,cal PCD, Q = 8 L/min, Tfluid,cal FIGURE 9 Temperature of PCD A, and HS B, from simulation and calculation, Q = 4-10 L/min, P = 0.5-2.5 kW, ϕ = 20%; C, volumetric-averaged fluid temperature; D, volumetric-averaged heat sink and wetting wall from simulation at Q = 8 L/min, P = 0.5-2.5 kW, ϕ = 20% [Colour figure can be viewed at wileyonlinelibrary.com] LI ET AL. LI ET AL. 5834 PCD is better due to the lower fluid and heat sink tem- perature under the same heat dissipation from HVDC converter. volumetric-averaged heat sink temperature of THS,sim and TWW,cal under the same condition of fluid. Similarly, higher heat sink temperature can be found at high P and low Q. The TWW,cal is slightly higher than that of THS,sim by 1.55% on average. As a result, the temperature of fluid (Tfluid, cal) and heat sink (TWW,cal) derived from outlet and heating surface respectively can be used to evaluate the thermal performance. When it comes to evaluate the thermal performance using heat transfer coefficient (h = q/ΔT), various temp differences (ΔT) driving the convective heat transfer are employed. It is noteworthy that small deviation of ΔT may result in large oscillation of h under high heat flux, leading to a contrary analysis of water and PCD. To ensure reliability of the evaluation, three calculation methods of heat transfer coefficient are compared: hsim based on volumetric-averaged temperature, hcal from temperature of wetted wall and fluid inlet/outlet and hΔT based on log mean temp difference: Figure 9C compare the temperature of fluid and heat sink for water and PCD under the same condition. The temperature of water is higher than that of PCD by 1.42 K for Tfluid, sim and 1.53 K for Tfluid, cal. The temp dif- ference between water and PCD is more obvious at high heating power, reaching 2.25 K at P = 2.5 kW for Tfluid, cal. It is attributed to the latent heat absorption of fatty acid ester, which introduces a more isothermal fluid envi- ronment for the cooling process. Compared to the fluid, the heat sink temp difference between water and PCD is slight larger, with average value of 2.18 K for THS,sim and 1.55 K for TWW,cal. Figure 9B shows the 0.5 1.0 1.5 2.0 2.5 10000 12000 14000 16000 18000 20000 P, kW hcal, W/(m 2•K) 0.5 1.0 1.5 2.0 2.5 10000 12000 14000 16000 18000 20000 22000 Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min P, kW h∆T, W/(m 2 •K) Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min 0.5 1.0 1.5 2.0 2.5 10000 15000 20000 25000 30000 hsim, W/(m 2•K) P, kW Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min (A) (B) (C) FIGURE 10 Heat transfer coefficient using three calculation methods for Q = 4 and 8 L/min, P = 0.5-2.5 kW, ϕ = 20% [Colour figure can be viewed at wileyonlinelibrary.com] 0.5 1.0 1.5 2.0 2.5 10000 12000 14000 16000 18000 20000 P, kW hcal, W/(m 2•K) Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min 0.5 1.0 1.5 2.0 2.5 10000 15000 20000 25000 30000 hsim, W/(m 2•K) P, kW Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min (A) (B) (A) (B) 0.5 1.0 1.5 2.0 2.5 10000 15000 20000 25000 30000 hsim, W/(m 2•K) P, kW Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min (A) 0.5 1.0 1.5 2.0 2.5 10000 12000 14000 16000 18000 20000 P, kW hcal, W/(m 2•K) Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min (B) 0.5 1.0 1.5 2.0 2.5 10000 12000 14000 16000 18000 20000 22000 P, kW h∆T, W/(m 2 •K) Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min (C) 0.5 1.0 1.5 2.0 2.5 10000 12000 14000 16000 18000 20000 22000 P, kW h∆T, W/(m 2 •K) Water, Q = 4 L/min PCD, Q = 4 L/min Water, Q = 8 L/min PCD, Q = 8 L/min (C) (C) FIGURE 10 Heat transfer coefficient using three calculation methods for Q = 4 and 8 L/min, P = 0.5-2.5 kW, ϕ = 20% [Colour figure can be viewed at wileyonlinelibrary.com] Therefore, the cooling performance of hsim = q THS,sim −Tfluid,sim ð29Þ hcal = q TWW,cal −Tfluid,cal ð30Þ ð29Þ ð30Þ 0.5 1.0 1.5 2.0 2.5 0.4 0.8 1.2 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min Q = 10 L/min PEC (h∆T) P, kW 0.5 1.0 1.5 2.0 2.5 0.6 0.8 1.0 1.2 1.4 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min PEC (hcal) Q = 10 L/min P, kW 0.5 1.0 1.5 2.0 2.5 0.8 1.2 1.6 2.0 Q = 4 L/min Q = 6 L/min Q = 8 L/min PEC (hsim) P, kW Q = 10 L/min (A) (B) (C) FIGURE 11 PEC comparison using three calculation methods for Q = 4-10 L/min, P = 0.5-2.5 kW, and ϕ = 20% [Colour figure can be viewed at wileyonlinelibrary.com] 0.5 1.0 1.5 2.0 2.5 0.6 0.8 1.0 1.2 1.4 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min PEC (hcal) Q = 10 L/min P, kW (B) 0.5 1.0 1.5 2.0 2.5 0.6 0.8 1.0 1.2 1.4 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min PEC (hcal) Q = 10 L/min P, kW 0.5 1.0 1.5 2.0 2.5 0.8 1.2 1.6 2.0 Q = 4 L/min Q = 6 L/min Q = 8 L/min PEC (hsim) P, kW Q = 10 L/min (A) (B) 0.5 1.0 1.5 2.0 2.5 0.8 1.2 1.6 2.0 Q = 4 L/min Q = 6 L/min Q = 8 L/min PEC (hsim) P, kW Q = 10 L/min (A) (A) (B) (C) 0.5 1.0 1.5 2.0 2.5 0.4 0.8 1.2 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min Q = 10 L/min PEC (h∆T) P, kW (C) 0.5 1.0 1.5 2.0 2.5 0.4 0.8 1.2 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min Q = 10 L/min PEC (h∆T) P, kW (C) FIGURE 11 PEC i i th l l ti th d f Q 4 10 L/ i P 0 5 2 5 kW d ϕ 20% [C l fi 0.5 1.0 1.5 2.0 2.5 0.4 0.8 1.2 1.6 Q = 4 L/min Q = 6 L/min Q = 8 L/min Q = 10 L/min PEC (h∆T) P, kW (C) FIGURE 11 PEC comparison using three calculation methods for Q = 4-10 L/min, P = 0.5-2.5 kW, and ϕ = 20% [Colour figure can be viewed at wileyonlinelibrary.com] ϕ = 5 % ϕ = 10 % ϕ = 20 % ϕ = 25 % 0 5 10 15 20 25 10000 15000 20000 25000 30000 φ, % h, W/(m 2⋅K) hsim hcal 5 h∆T 0 1 15 20 25 1.0 1.2 1.4 1.6 1.8 2.0 hsim hcal h∆T φ, % PEC (A) (B) (C) (E) (F) (D) E 12 Comparison of temperature distribution on the heating surface under various PCM mass fraction with A, 5%, B, % and D, 25% at the conditions of Q = 8 L/min, P = 2.2 kW, T = 326-343 K; E, heat transfer coefficient and F, PEC as a 5835 LI ET AL. The corresponding P of intersection point between water and PCD at Q = 8 L/min is P = 2.25 kW for hsim, but becomes P = 1.63 kW for hΔT. The favorable work- ing condition of PCD is narrower at low Q, and the cooling performance of PCD is always worse than that of water at Q = 4 L/min for hΔT. To clarify the influence of PCM mass fraction on ther- mal performance, the 2D temperature distribution from a top view are compared under ϕ = 5-25% as given in Figure 12A-D. The PCD enters the heat sink from top left and leaves at a higher temperature on the bottom left. Higher temperature can still be detected on the heating sur- face near outlet side, but no remarkable improvement can be observed under all ϕ. The average heating surface tem- perature of ϕ = 5% is higher than that of ϕ = 25% by 0.71 K. The advantage of PCD is resulted from the latent heat of PCM, but the influence of latent heat reduces when large transport heat applied, where sensible heat becomes dominated. Thus, the superiority of PCD over water is more apparent at small heating power. Mean- while, the high fluid temperature caused by low Q is beyond the latent heat range. For example, Tfluid, sim reaches 326 K at Q = 4 L/min for P = 2.5 kW, but decreases to 322 K at Q = 10 L/min under the same condition. The Cp of PCD at 322 K is nearly 7000 J/ (kg K) but only around 4000 J/(kg K) at 326 K. There- fore, the thermal performance of PCD may be worse than that of water at low flow rates. A quantitative analysis of heat transfer coefficient and PEC under different PCM mass fractions is given in Figure 12E and F. The thermal transfer performance improves with the increasing of PCM concentration. The highest heat transfer coefficient can be found at ϕ = 25%, which is around 1.42 times of water regardless of the calcu- lation method. Taken pressure drop penalty into consider- ation, the best performance can still be observed at the highest PCD concentration, because the enhanced thermal transfer is able to offset the increased pressure drop caused by PCM particles. The PEC results from hcal and hΔT are very close, but lower than that of hsim by 29.88% on average. As a conclusion, the best working condition of the devel- oped fatty acid ester-based PCD is Q = 8 L/min, P = 1.1 kW and ϕ = 25% under the present study range. In order to evaluate the heat transfer enhancement and pressure drop penalty, a performance evaluation criteria (PEC) proposed by Webb77 was used: PEC = Nup,avg=Nuw,avg fp,avg=fw,avg 1=3 ð32Þ ð32Þ LI ET AL. 5835 ϕ = 5 % (A) ϕ = 10 % (B) (A) (B) (C) ϕ = 20 % (C) (E) (D) ϕ = 25 % (F) (D) (E) (F) 0 5 10 15 20 25 10000 15000 20000 25000 30000 φ, % h, W/(m 2⋅K) hsim hcal h∆T (E) 5 0 1 15 20 25 1.0 1.2 1.4 1.6 1.8 2.0 hsim hcal h∆T φ, % PEC FIGURE 12 Comparison of temperature distribution on the heating surface under various PCM mass fraction with A, 5%, B, 10%, C, 20% and D, 25% at the conditions of Q = 8 L/min, P = 2.2 kW, T = 326-343 K; E, heat transfer coefficient and F, PEC as a function of mass fraction at the conditions of Q = 8 L/min and HF = 1.1 kW [Colour figure can be viewed at wileyonlinelibrary.com] LI ET AL. 5836 LI ET AL. Nuavg = havgDhyd λTmean,fluid ð33Þ ð33Þ ð31Þ A comparison of PEC using the three calculation methods are given in Figure 11. The trend predicted by hsim and hcal is very similar but the overall PEC based on hsim is larger. Thus, a wider application range where PCD behaves better than water (PEC > 1) is derived by hsim. The maxi- mum PEC can be obtained at P = 1.1 kW under all flowrates, due to the thermal enhancement. The influence of flow rates on PEC predicted from hcal is less significant than that of hsim. For example, PEC (hsim) and PEC (hcal) are almost identical at Q = 4 L/min, but the difference of the two PECs may reach upto 24.98% at Q = 8 L/min. The peak of PEC (hΔT) can be found at a lower heating power P = 0.5 kW, because the thermal enhancement could not offset the pressure drop penalty at P = 1.1 kW with this method. It is understandable because the hsim increased by 68.99% from P = 0.5 kW to P = 1.1 kW at Q = 8 L/min, while the increment is only 21.30% for hΔT. Considering the representativeness of hsim as a volumetric-averaged parame- ter, the hcal method is recommended for future experimen- tal evaluation due to small deviation. where THS, avg is the average temperature of the heating sur- face. Using expressions (30) and (31), the measured tem- peratures of heating surface and fluid inlet/outlet can be integrated for thermal evaluation. Figure 10 com- pares the heat transfer coefficients of water and PCD using the three methods. The volumetric-averaged tem- perature always yields a higher value (hsim) than the other two methods for both fluids. Increasing Q favours the growth of heat transfer coefficient for water. Under a given Q, the heat transfer coefficient of water shows a nearly linear increasing relationship of heating power P. However, a peak can be spotted at P = 1.1 kW for PCD, and the maximum value increases with increas- ing Q. For P < 1.1 kW, PCD shows better cooling per- formance than water especially at high Q. Exceeding this point, the heat transfer coefficient of PCD decreases until lower than that water, which implies that the working condition is no longer favourable. The corresponding P of intersection point between water and PCD at Q = 8 L/min is P = 2.25 kW for hsim, but becomes P = 1.63 kW for hΔT. The favorable work- ing condition of PCD is narrower at low Q, and the cooling performance of PCD is always worse than that of water at Q = 4 L/min for hΔT. where THS, avg is the average temperature of the heating sur- face. Using expressions (30) and (31), the measured tem- peratures of heating surface and fluid inlet/outlet can be integrated for thermal evaluation. Figure 10 com- pares the heat transfer coefficients of water and PCD using the three methods. The volumetric-averaged tem- perature always yields a higher value (hsim) than the other two methods for both fluids. Increasing Q favours the growth of heat transfer coefficient for water. Under a given Q, the heat transfer coefficient of water shows a nearly linear increasing relationship of heating power P. However, a peak can be spotted at P = 1.1 kW for PCD, and the maximum value increases with increas- ing Q. For P < 1.1 kW, PCD shows better cooling per- formance than water especially at high Q. Exceeding this point, the heat transfer coefficient of PCD decreases until lower than that water, which implies that the working condition is no longer favourable. fatty acid ester-based phase change dispersion (PCD) in a heat sink with double-layer oblique-crossed ribs. Both experimental investigation and numerical modelling were performed and the following conclusions are obtained: CL lift coefficient cp specific heat capacity, (kJ/kg K) Ctd dispersion coefficient d droplet size, (μm) Dhyd hydraulic diameter, (m) ess particle–particle restitution coefficient ew particle–wall restitution coefficient f average fanning friction factor F force, (N) g0 radial distribution function h heat transfer coefficient, (W/m2 K) hsl particle–liquid heat transfer coefficient, (W/m2 K) H height, (m) k thermal conductivity, (W/m K) kθs diffusion coefficient Δh latent heat, (kJ/kg) n size distribution parameter P heating power, (W) Pel electrical power, (W) Ps pressure, (Pa) qsw Johnson–Jackson fluctuating energy Q volume flow rate, (L/min) R thermal resistance, (K/W) s distance, (m) T temperature, (K) v ! velocity vector, (m/s) Uavg average velocity, (m/s) W width, (m) Yd cumulative volume fraction 1. Although the viscosity of the PCD was several times higher than that of water, the increased pumping power was only 17.01% on average. An addition of a small amount of PCM particles significantly increased the friction factor from ϕ = 0% to 5%, but the extent of the increase reduced with further increase in the ϕ from 5% to 25%. 2. The temperature of both heat sink and fluid under water cooling were higher than that of PCD under the same set of working conditions, due to the latent heat of the PCM. Thus, the PCD could achieve a safer and cooler environment for sensitive HVDC converter components like IGBT. 3. The hsim method based on volumetric-averaged tem- perature yielded a larger value than the other two cal- culation methods, which were derived from temperatures of heating surface, wetted wall and fluid inlet/outlet. The thermal performance of PCD was only enhanced at a suitable working range, where heating power, fluid flowrate and PCM concentrations all played important roles. For all three methods, a peak of heat transfer coefficient can be observed at P = 1.1 kW, where average fluid temperature was more close to PCM melting temperature and higher Cp presented. Increasing fluid flowrate favoured the growth of heat transfer coefficient. 4. Considering the fact that the operational condition of industrial heat exchange is usually at turbulent flow regime, the PCD is greatly competitive than water as energy transport HTF. 5 \| CONCLUSIONS where Nuw and fw refers to the Nusselt number and fric- tion factor of water, and the averaged Nu number can be calculated based heat transfer coefficient: The work detailed in this paper concerns with a high- performance cooling method for a HVDC converter using 5837 LI ET AL. ACKNOWLEDGEMENTS The research is partially supported by State Grid Corpo- ration of China and Global Energy Interconnection Research Institute Europe GmbH under Project No. SGRIWLZXQT[2017]882 (Cooling of converters by using phase change materials). SUBSCRIPTS AC alternating current DC direct current HVDC high voltage direct current IGBT insulated gate bipolar transistor l liquid phase PCD phase change dispersion PCM phase change material s solid phase An optimal set of working con- ditions was proposed and a flowrate of 8 L/min under a heating power of 1.1 kW and a PCM concentration of 25% was recommended for industrial cooling opera- tions. Moreover, the hcal method was recommended to evaluate the heat sink with complex cross-distributed multilayer channel, using experimental measurable temperature of TWW, cal and Tfluid, cal. REFERENCES 1. Giampieri A, Ma Z, Ling Chin J, et al. Techno-economic analy- sis of the thermal energy saving options for high-voltage direct current interconnectors. Appl Energy. 2019;247:60-77. 17. Li C, Li Q, Li Y, et al. Heat transfer of composite phase change material modules containing a eutectic carbonate salt for medium and high temperature thermal energy storage applica- tions. Appl Energy. 2019;238:1074-1083. 2. Gomis-Bellmunt O, Junyent-Ferré A, Sumper A, Galceran- Arellano S. Maximum generation power evaluation of variable frequency offshore wind farms when connected to a single power converter. Appl Energy. 2010;87:3103-3109. 18. Li C, Li Q, Ding Y. Investigation on the effective thermal con- ductivity of carbonate salt based composite phase change mate- rials for medium and high temperature thermal energy storage. Energy. 2019;176:728-741. 3. Kabalyk Y. Determination of energy loss in power voltage inverters for power supply of locomotive traction motors. Procedia Eng. 2016;165:1437-1443. 19. Pu L, Xu L, Zhang S, Li Y. Optimization of ground heat exchanger using microencapsulated phase change material slurry based on tree-shaped structure. Appl Energy. 2019;240: 860-869. 4. Lips HP. Technology trends for HVDC thyristor valves. POW- ERCON '98 1998 International Conference on Power System Technology Proceedings (Cat No98EX151); vol 1; August 18-21,1998; Beijing, China:451-455. 20. Ma F, Zhang P, Shi XJ. Investigation of thermo-fluidic perfor- mance of phase change material slurry and energy transport characteristics. Appl Energy. 2018;227:643-654. 5. Wang P, McCluskey P, Bar-Cohen A. Hybrid solid- and liquid- cooling solution for Isothermalization of insulated gate bipolar transistor power electronic devices. IEEE Trans Compon Packag Manuf Technol. 2013;3:601-611. 5. Wang P, McCluskey P, Bar-Cohen A. Hybrid solid- and liquid- cooling solution for Isothermalization of insulated gate bipolar transistor power electronic devices. IEEE Trans Compon Packag Manuf Technol. 2013;3:601-611. 21. Chen J, Zhang P. Preparation and characterization of nano- sized phase change emulsions as thermal energy storage and transport media. Appl Energy. 2017;190:868-879. 6. Guo Y-f, Ma X-l, Shi P. Power losses and temperature varia- tions in a power converter for an electronic power steering sys- tem considering steering profiles. J Instrum Technol. 2014;2: 40-46. 22. Wang F, Zhang C, Liu J, Fang X, Zhang Z. Highly stable graph- ite nanoparticle-dispersed phase change emulsions with little supercooling and high thermal conductivity for cold energy storage. Appl Energy. 2017;188:97-106. 7. Kang SS. Advanced cooling for power electronics. REFERENCES Paper pres- ented at: 2012 7th International Conference on Integrated Power Electronics Systems (CIPS); March 6-8,2012; Nurem- berg, Germany:1-8. 23. Lankveld JMG, Lyklema J. Adsorption of polyvinyl alcohol on the paraffin—water interface. III emulsification of paraffin in aqueous solutions of polyvinyl alcohol and the properties of paraffin-in-water emulsions stabilized by polyvinyl alcohol. J Colloid Interface Sci. 1972;41:475-483. 8. Wrzecionko B, Biela J, Kolar JW. SiC power semiconductors in HEVs: influence of junction temperature on power density, chip utilization and efficiency. Paper presented at: 2009 35th Annual Conference of IEEE Industrial Electronics; November 3-5,2009; Porto, Portugal:3834-3841. 24. Jin Z, Wang Y, Liu J, Yang Z. Synthesis and properties of paraf- fin capsules as phase change materials. Polymer. 2008;49:2903- 2910. 9. Huang AQ. Power semiconductor devices for smart grid and renewable energy systems. Proc IEEE. 2017;105:2019-2047. 25. Xin X, Zhang H, Xu G, Tan Y, Zhang J, Lv X. Influence of CTAB and SDS on the properties of oil-in-water nano-emulsion with paraffin and span 20/tween 20. Colloids Surf A Physicochem Eng Asp. 2013;418:60-67. 10. Haack J, Akyol B, Tenney N, Carpenter B, Pratt R, Carroll T. VOLTTRON™: an agent platform for integrating electric vehi- cles and Smart Grid. Paper presented at: 2013 International Conference on Connected Vehicles and Expo (ICCVE); December 2-6,2013; Las Vegas, NV:81-86. 26. Shchukina E, Graham M, Zheng Z, Shchukin D. Nanoencapsulation of phase change materials for advanced ther- mal energy storage systems. Chem Soc Rev. 2018;47:4156-4175. 11. Faschang M, Cejka S, Stefan M, et al. Provisioning, deploy- ment, and operation of smart grid applications on substation level. Comput Sci Res Develop. 2017;32:117-130. 27. Noël JA, Kahwaji S, Desgrosseilliers L, Groulx D, White MA. Chapter 13 - phase change materials. In: Letcher TM, ed. Storing Energy. Oxford, UK: Elsevier; 2016:249-272. 12. Drofenik U. Theoretical converter power density limits for forced convection cooling. Proceedings of the International PCIM Europe Conference. 2005;2005:608-619. 28. Hasan A, Hejase H, Abdelbaqi S, Assi A, Hamdan M. Compar- ative effectiveness of different phase change materials to improve cooling performance of heat sinks for electronic devices. Appl Sci. 2016;6:226. 13. Kandlikar SG, Hayner CN. Liquid cooled cold plates for indus- trial high-power electronic devices—thermal design and manufacturing considerations. Heat Transf Eng. 2009;30:918-930. 29. Shah RK, London AL. Laminar Flow Forced Convection in Ducts: A Source Book for Compact Heat Exchanger Analytical Data. Cambridge, Massachusetts: Academic Press; 2014. 14. Ali HM. Qi Li https://orcid.org/0000-0002-8209-6637 Qi Li https://orcid.org/0000-0002-8209-6637 Qi Li https://orcid.org/0000-0002-8209-6637 16. Li C, Li Q, Ding Y. Investigation on the thermal performance of a high temperature packed bed thermal energy storage sys- tem containing carbonate salt based composite phase change materials. Appl Energy. 2019;247:374-388. LI ET AL. LI ET AL. 5838 15. Ali H, Abdul W. Qasim Muhammad Arslan, Janjua Muham- mad Mansoor, Ali Muhammad Aon, Ali hafiz Muhammad, Jadoon Tufail Rehman, Ali Ejaz, Raza Ahsan, Javaid Noshairwan. Thermal management and uniform temperature regulation of photovoltaic modules using hybrid phase change materials-nanofluids system. Renew Energy. 2020;145:282-293. NOMENCLATURE A heating area, (m2) CD 5839 30. Huang L, Petermann M. An experimental study on rheological behaviors of paraffin/water phase change emulsion. Int J Heat Mass Transfer. 2015;83:479-486. PCM heat sinks under failure conditions. Appl Energy. 2016; 180:14-26. PCM heat sinks under failure conditions. Appl Energy. 2016; 180:14-26. PCM heat sinks under failure conditions. Appl Energy. 2016; 180:14-26. 47. Yüncü H. Thermal contact conductance of nominaly flat sur- faces. Heat Mass Transfer. 2006;43:1-5. 31. Pollerberg C, Doetsch C. Phase Changing Slurries in cooling and cold supply networks. Paper presented at: 10th Interna- tional Symposium on District Heating and Cooling; September 3-5,2006; Oberhausen, Germany. 48. Shi L, Wu G, Wang H-L, Yu X-M. Interfacial thermal contact resistance between aluminum nitride and copper at cryogenic temperature. Heat Mass Transfer. 2012;48:999-1004. 49. Fletcher L, Gyorog D. Heat Transfer Between Surfaces in Con- tact: An Analytical and Experimental Study of Thermal Contact Resistance of Metallic Interfaces. Mechanical Engineering Department, Arizona State University, Tempe, Arizona, United States. 1971. 32. Chen B, Wang X, Zhang Y, Xu H, Yang R. Experimental research on laminar flow performance of phase change emul- sion. Appl Therm Eng. 2006;26:1238-1245. 33. Alvarado JL, Marsh C, Sohn C, Phetteplace G, Newell T. Ther- mal performance of microencapsulated phase change material slurry in turbulent flow under constant heat flux. Int J Heat Mass Transfer. 2007;50:1938-1952. 50. Göktepe S, Atalık K, Ertürk H. Comparison of single and two-phase models for nanofluid convection at the entrance of a uniformly heated tube. Int J Therm Sci. 2014;80:83-92. 34. Rao Y, Dammel F, Stephan P, Lin G. Flow frictional character- istics of microencapsulated phase change material suspensions flowing through rectangular minichannels. Sci China Ser E: Technol Sci. 2006;49:445-456. 51. Wang J, Wang S, Zhang T, Liang Y. Numerical investigation of ice slurry isothermal flow in various pipes. Int J Refrig. 2013;36: 70-80. 35. Mahmoud S, Tang A, Toh C, Raya A-D, Soo SL. Experimental investigation of inserts configurations and PCM type on the thermal performance of PCM based heat sinks. Appl Energy. 2013;112:1349-1356. 52. Xi W, Cai J, Huai X. Numerical investigation on fluid-solid coupled heat transfer with variable properties in cross-wavy channels using half-wall thickness multi-periodic boundary conditions. Int J Heat Mass Transfer. 2018;122:1040-1052. 36. Kanargi B, Lee PS, Yap C. A numerical and experimental inves- tigation of heat transfer and fluid flow characteristics of an air- cooled oblique-finned heat sink. Int J Heat Mass Transfer. 2018;116:393-416. 53. Fischer L, Maranda S, Stamatiou A, von Arx S, Worlitschek J. Experimental investigation on heat transfer with a phase change dispersion. Appl Therm Eng. 2019;147:61-73. 37. Li WQ, Qu ZG, He YL, Tao WQ. Experimental and numerical studies on melting phase change heat transfer in open-cell metallic foams filled with paraffin. Appl Therm Eng. 2012;37: 1-9. 54. Rajabifar B. Enhancement of the performance of a double lay- ered microchannel heatsink using PCM slurry and nanofluid coolants. Int J Heat Mass Transfer. 2015;88:627-635. 55. Chiu C-P, Maveety JG, Tran QA. Characterization of solder interfaces using laser flash metrology. Microelectron Reliab. 2002;42:93-100. 38. Chai L, Xia GD, Wang HS. Numerical study of laminar flow and heat transfer in microchannel heat sink with offset ribs on sidewalls. Appl Therm Eng. 2016;92:32-41. 56. Xiao X, Zhang G, Ding Y, Wen D. Rheological characteristics of molten salt seeded with Al2O3 Nanopowder and graphene for concentrated solar power. Energies. 2019;12:467. 39. Kanargi OB, Lee PS, Yap C. A numerical and experimental investigation of heat transfer and fluid flow characteristics of a cross-connected alternating converging–diverging channel heat sink. Int J Heat Mass Transfer. 2017;106:449-464. 57. Muñoz-Sánchez B, Nieto-Maestre J, Veca E, et al. Rheology of solar-salt based nanofluids for concentrated solar power. Influ- ence of the salt purity, nanoparticle concentration, temperature and rheometer geometry. Sol Energy Mater Sol Cells. 2018;176: 357-373. 40. Remeljej CW, Hoadley AFA. An exergy analysis of small-scale liquefied natural gas (LNG) liquefaction processes. Energy. 2006;31:2005-2019. 58. Itoh T, Wanibe Y. Derivation of number based size distribution from modified mass based rosin-Rammler distribution and esti- mation of the various mean particle diameters of powder. Trans Jpn Inst Met. 1988;29:671-684. 41. Moein P, Sarmad M, Ebrahimi H, Zare M, Pakseresht S, Vakili SZ. APCI- LNG single mixed refrigerant process for nat- ural gas liquefaction cycle: analysis and optimization. J Nat Gas Sci Eng. 2015;26:470-479. 42. Wang Z, Li Y. Layer pattern thermal design and optimization for multistream plate-fin heat exchangers—a review. Renew Sustain Energy Rev. 2016;53:500-514. 59. Herrmann-Heber R, Reinecke SF, Hampel U. Dynamic aera- tion for improved oxygen mass transfer in the wastewater treat- ment process. Chem Eng J. 2019;122068. https://doi.org/10. 1016/j.cej.2019.122068 43. Radwan A, Ookawara S, Ahmed M. Thermal management of concentrator photovoltaic systems using two-phase flow boiling in double-layer microchannel heat sinks. Appl Energy. 2019; 241:404-419. 60. Bordet A, Poncet S, Poirier M, Galanis N. REFERENCES Recent advancements in PV cooling and efficiency enhancement integrating phase change materials based systems-a comprehensive review. Sol Energy. 2020;197:163-198. LI ET AL. LI ET AL. Advanced numerical modeling of turbulent ice slurry flows in a straight pipe. Int J Therm Sci. 2018;127:294-311. 44. Alhuyi Nazari M, Ahmadi MH, Ghasempour R, et al. A review on pulsating heat pipes: from solar to cryogenic applications. Appl Energy. 2018;222:475-484. 61. Menter F. Zonal two equation kw turbulence models for aero- dynamic flows. Paper presented at: 23rd Fluid Dynamics, Plasma dynamics, and Lasers Conference; July 6-9,1993; Orlando, FL:2906. 45. Srikanth R, Nemani P, Balaji C. Multi-objective geometric opti- mization of a PCM based matrix type composite heat sink. Appl Energy. 2015;156:703-714. 62. Lun C, Savage SB, Jeffrey D, Chepurniy N. Kinetic theories for granular flow: inelastic particles in Couette flow and slightly inelastic particles in a general flowfield. J Fluid Mech. 1984; 140:223-256. 46. Coleman B, Ostanek J, Heinzel J. Reducing cell-to-cell spacing for large-format lithium ion battery modules with aluminum or 5840 LI ET AL. 5840 packed with spherical pellets in tubes. Chem Eng Sci. 1994;49: 2513-2520. 63. Syamlal M, O'Brien T. Computer simulation of bubbles in a flu- idized bed. AIChe Symp Ser. 1989;85:22-31. 64. Ogawa S, Umemura A, Oshima N. On the equations of fully fluidized granular materials. Z Angew Math Phys. 1980;31: 483-493. 73. Johnson PC, Jackson R. Frictional–collisional constitutive rela- tions for granular materials, with application to plane shearing. J Fluid Mech. 1987;176:67-93. 65. Kitanovski A, Poredoš A. Concentration distribution and viscosity of ice-slurry in heterogeneous flow. Int J Refrig. 2002;25:827-835. 74. Jiang YY, Zhang P. Numerical investigation of slush nitrogen flow in a horizontal pipe. Chem Eng Sci. 2012;73:169-180. 75. Patankar S. Numerical Heat Transfer and Fluid Flow. Boca Raton, Florida: CRC Press; 1980. 66. Gidaspow D. Multiphase Flow and Fluidization: Continuum and Kinetic Theory Descriptions. Cambridge, Massachusetts: Academic Press; 1994. 76. Hu S, McFarland AR. Numerical performance simulation of a wetted wall bioaerosol sampling cyclone. Aerosol Sci Tech. 2007;41:160-168. 67. Pu W, Zhao C, Xiong Y, et al. Numerical simulation on dense phase pneumatic conveying of pulverized coal in horizontal pipe at high pressure. Chem Eng Sci. 2010;65:2500-2512. 77. Webb RL. Performance evaluation criteria for use of enhanced heat transfer surfaces in heat exchanger design. Int J Heat Mass Transfer. 1981;24:715-726. 68. Saffman PG. The lift on a small sphere in a slow shear flow. J Fluid Mech. 2006;22:385-400. 69. Simonin C, Viollet P. Predictions of an oxygen droplet pulveri- zation in a compressible subsonic coflowing hydrogen flow. Num Method Multiph Flows. 1990;91:65-82. 70. Gunn DJ. Transfer of heat or mass to particles in fixed and fluidised beds. Int J Heat Mass Transfer. 1978;21: 467-476. 71. Zehner P, Schlünder EU. Wärmeleitfähigkeit von Schüttungen bei mäßigen Temperaturen. Chem Ing Tech. 1970;42:933-941. 72. Legawiec B, Ziólkowski D. Structure, voidage and effective thermal conductivity of solids within near-wall region of beds
https://openalex.org/W3205793267	https://europepmc.org/articles/pmc8551584?pdf=render	English	null	Cost-Effectiveness Analyses on Various Models of The Red Light, Purple Light Self-Regulation Intervention for Young Children	Frontiers in psychology	2,021	cc-by	9,629	ORIGINAL RESEARCH published: 14 October 2021 doi: 10.3389/fpsyg.2021.711578 Cost-Effectiveness Analyses on Various Models of The Red Light, Purple Light Self-Regulation Intervention for Young Children Tao Li 1, Megan M. McClelland 2, Shauna L. Tominey 3 and Alexis Tracy 2 1 Health Management and Policy, Oregon State University, Corvallis, OR, United States, 2 Human Development and Family Sciences and the Hallie E. Ford Center for Healthy Children and Families, Oregon State University, Corvallis, OR, United States, 3 Extension Family and Community Health and the Hallie E. Ford Center for Healthy Children and Families, Oregon State University, Corvallis, OR, United States Early childhood interventions can improve self-regulation, but there are few economic evaluations of such interventions. This study analyzed the cost-effectiveness of an early childhood self-regulation intervention (Red Light Purple Light!; RLPL), comparing three different models of implementation across stages of intervention development: (Model 1) trained research assistants (RAs; graduate students) directly delivered the RLPL intervention to children; (Model 2) RAs trained trainers (e.g., program coaches), who then trained teachers to implement RLPL with children (e.g., train-the-trainer); and (Model 3) program faculty trained teachers to deliver the RLPL intervention to children. We implemented a cost-effectiveness analysis by calculating the incremental cost-effectiveness ratio. We also conducted a series of sensitivity analyses to adjust for parameter uncertainty. Our base- case analysis suggests that Model 2 was the most cost-effective strategy, in that a cost of $23 per child was associated with a one-unit increase of effect size on self-regulation scores. The “train-the-trainer” model remained the optimal strategy across scenarios in our sensitivity analysis. This study fills an important gap in cost-effectiveness analyses on early childhood self-regulation interventions. Our process and results can serve as a model for future cost-effectiveness analyses of early childhood intervention programs and may ultimately inform decisions related to intervention adoption that optimize resource allocation and improve program design. Keywords: cost-effectiveness analysis, economic evaluations, red light purple light, early childhood intervention, self-regulation, school readiness, executive function skills Edited by: Jin Sun, The Education University of Hong Kong, Hong Kong, SAR China , The Education University of Hong Kong, Hong Kong, SAR China Reviewed by: Maria von Salisch, Leuphana University, Germany Marie Catherine White, Gordon College, United States Reviewed by: Maria von Salisch, Leuphana University, Germany Marie Catherine White, Gordon College, United States Reviewed by: Maria von Salisch, Leuphana University, Germany Marie Catherine White, Gordon College, United States Correspondence: Tao Li tao.li2@oregonstate.edu Specialty section: This article was submitted to Developmental Psychology, a section of the journal Frontiers in Psychology Developmental Psychology, a section of the journal Frontiers in Psychology Keywords: cost-effectiveness analysis, economic evaluations, red light purple light, early childhood intervention, self-regulation, school readiness, executive function skills Received: 18 May 2021 Accepted: 06 September 2021 Published: 14 October 2021 Connections Between Self-Regulation and Academic Skills Self-regulation is a key factor supporting academic success across the life span (McClelland et al., 2006, 2013; Blair and Razza, 2007; Moffitt et al., 2011). Self-regulation predicts early achievement (both math and literacy) in preschool, elementary and middle school and even college (Duckworth et al., 2010; McClelland et al., 2013, 2014). Especially important for school readiness skills, self-regulation is related broadly to early aspects of math and emergent literacy (McClelland et al., 2007a, 2014; Schmitt et al., 2017), especially early numeracy skills (counting, cardinality, numeral knowledge), which is most predictive of later mathematics achievement (Nguyen et al., 2016). Moreover, self-regulation interventions have shown significant effects on children’s math and literacy (Blair and Raver, 2014; Schmitt et al., 2015; Pandey et al., 2018) and may be especially predictive of early math skills (Allan et al., 2014; McClelland et al., 2014; Blair et al., 2015; Purpura et al., 2017).f Red Light Purple Light! (RLPL) is an easy-to-implement, short-term intervention designed to promote self-regulation and school readiness (Tominey and McClelland, 2011; McClelland and Tominey, 2015). Although early childhood interventions such as RLPL have been shown to improve self-regulation, there are few economic evaluations of such interventions. This study analyzed the cost-effectiveness of an early childhood self-regulation intervention (RLPL) implemented through three different models: 1) trained research assistants (RAs) implemented the RLPL intervention with children; 2) RAs trained trainers (e.g., early childhood coaches), who then trained teachers to implement the RLPL intervention with children (e.g., train- the-trainer); and 3) program faculty trained teachers to implement RLPL with children. These three models used the same intervention curriculum, but at different phases of testing and iterative development of the implementation process, which allowed us to explore and compare the cost-effectiveness of each implementation model. p In previous studies evaluating the RLPL program, effects have been found for children’s improvements in self-regulation (Schmitt et al., 2015; Duncan et al., 2018), especially those with low baseline levels of self-regulation (Tominey and McClelland, 2011), early math skills (Duncan et al., 2018), particularly for low socioeconomic (McClelland et al., 2019) and low-SES DLL children (Schmitt et al., 2015) and early literacy skills (Tominey and McClelland, 2011; Duncan et al., 2018). These effects are supported by other research showing that classrooms characterized by consistent, organized classroom practices lead to better academic outcomes for children (Cameron et al., 2008; Cameron and Morrison, 2011). Connections Between Self-Regulation and Academic Skills Previous research on RLPL also indicates that participating children demonstrate significantly stronger self-regulation on direct measures and teacher ratings (Schmitt et al., 2015; Keown et al., 2020) suggesting that children are demonstrating improvements in directly assessed self-regulation and observations of children’s behavior in classroom settings. These results are promising but do not provide information related to the costs associated with each model or the corresponding effects, indicating that an evaluation of the cost-effectiveness of the RLPL program is needed. INTRODUCTION Li T, McClelland MM, Tominey SL and Tracy A (2021) Cost-Effectiveness Analyses on Various Models of The Red Light, Purple Light Self- Regulation Intervention for Young Children. Many children, especially those experiencing socio-demographic risks such as poverty, enter formal schooling without key skills needed to thrive in a classroom environment (Blair and Raver, 2015). These skills are included in a construct called school readiness and include self-regulation and early academic skills (Snow, 2006). Unfortunately, children facing socio- demographic risk who struggle with self-regulation and early academic skills are likely to Front. Psychol. 12:711578. doi: 10.3389/fpsyg.2021.711578 Front. Psychol. 12:711578. doi: 10.3389/fpsyg.2021.711578 October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. face achievement gaps that persist and widen over time (Zelazo et al., 2016). Self-regulation is a significant predictor of short- and long-term academic, social, and life outcomes (Moffitt et al., 2011; McClelland et al., 2013; Zelazo et al., 2016; Robson et al., 2020). Importantly, self-regulation can be practiced and improved (Diamond and Ling, 2016), and self-regulation interventions may serve as a mechanism to protect children at risk (Sasser et al., 2017; Pandey et al., 2018). Large-scale classroom curricula that combine self-regulation and academic skills have shown mixed effects (Farran et al., 2013; Blair and Raver, 2014; Morris et al., 2014), but targeted self-regulation interventions that can easily be implemented in the classroom have shown positive effects across both self-regulation and academic domains (Tominey and McClelland, 2011; Schmitt et al., 2015; McClelland et al., 2019). school success (McClelland et al., 2007b, 2014; McClelland and Cameron, 2012). We use the term self-regulation in the present study to capture EF processes in real-world settings. Frontiers in Psychology \| www.frontiersin.org The Importance of Self-Regulation on School Success Self-regulation is a complex construct that includes aspects of emotion, cognition, and behavior (McClelland et al., 2010). This paper focuses on the aspects of self-regulation most relevant for children’s learning in school contexts, which stem from three executive function (EF) processes: working memory, attentional or cognitive flexibility, and inhibitory control (Cameron Ponitz et al., 2009). Working memory refers to the ability to hold and manipulate short-term instructions in mind such as when children need to remember the steps in an art activity (Gathercole, 2008; McClelland et al., 2015); attentional or cognitive flexibility is ability to focus on a task and switch to a new task when needed such as when children need to stop what they are doing and move to a new activity (Rothbart and Posner, 2005); and inhibitory control is the ability to stop a dominant response in favor of a more adaptive one (Blair, 2003). This can be seen when children have to inhibit their impulse to blurt out an answer and raise their hand instead. Although each aspect of EF has been shown to predict academic outcomes, their integration is particularly important for The Present Studyh The Present Study The present study focuses on a cost-effectiveness analysis of the Red Light, Purple Light (RLPL) self-regulation intervention, which has been shown to have significant positive effects on self-regulation (e.g., Tominey and McClelland, 2011; Schmitt et al., 2015; Duncan et al., 2018; McClelland et al., 2019). The RLPL program is an eight-week classroom-based self-regulation intervention. The program consists of large group time music and movement-based classroom games implemented twice a week in 20–30-min sessions (McClelland and Tominey, 2015). Over the course of multiple randomized control trials, the same RLPL curriculum was used, although the training and implementation models varied. This variation allowed for an exploratory cost-effectiveness analysis across each of these models. RLPL was developed through an iterative development process. The intervention was first piloted through a small randomized trial where the sessions were led by a research assistant (RA) who was also a former early childhood classroom teacher (Model 1; Tominey and McClelland, 2011). The intervention was replicated in another randomized trial with a larger sample with RAs leading the games during classroom group time activities (also Model 1; Schmitt et al., 2015). A “train-the- trainer” model was then employed (Model 2; Duncan et al., 2018) where two early childhood professional development coaches were trained in the intervention by the RLPL development team and then delivered the training to additional teachers in their district as part of their existing role supporting professional development efforts. Finally in Model 3, professional development trainings were delivered to teachers by faculty members of the original RLPL development team (McClelland et al., 2019).ff t Interventions focused on the broader construct of social– emotional skills (Diamond et al., 2007; Domitrovich et al., 2007) have emerged in recent years, including programs specifically targeting self-regulation (e.g., Tominey and McClelland, 2011; Schmitt et al., 2015). Numerous programs have demonstrated significant positive effects on children’s self- regulation, although effects are typically modest (Bierman et al., 2008; Raver et al., 2011; McClelland et al., 2017). In some cases, program effects have even been inconsistent across studies (Blair and Raver, 2014; Morris et al., 2014). The resources, time, and skills required to implement early childhood intervention programs to fidelity vary widely and are not often well-documented. Few interventions have conducted cost- effectiveness analyses to identify the costs associated with implementation as well as associated outcomes. The Importance of Cost-Effectiveness Analysis in Early Intervention Programsh Analysis in Early Intervention Programs There is a growing interest among researchers, practitioners, and policymakers in studying the economic impacts of early childhood interventions to inform investment and policy decisions (National Academies of Sciences, Engineering and Medicine, 2016; Cannon et al., 2018; Jones et al., 2019). Although few studies exist for school readiness interventions, researchers have started to examine costs more closely. For example, Jones and colleagues conducted cost analysis and examined the costs of different versions of the Head Start Research-based, Developmentally Informed intervention (Jones et al., 2019). Compared to cost analysis, we expect cost-effectiveness analysis will be a more powerful tool and will provide more important evidence to inform policymakers and stakeholders, because in cost-effectiveness analysis we simultaneously take into account the resources needed for program implementation as well as the outcomes achieved. Thus, to echo this urgent need of economic evaluation evidence, the purpose of this study is to evaluate the cost-effectiveness of the RLPL program. Frontiers in Psychology \| www.frontiersin.org The Potential of Early Interventions y Early childhood is an important developmental period for self- regulation growth because of the rapid development in brain development, especially in the prefrontal cortex, an area that is critical for self-regulation (Diamond, 2002; Montroy et al., 2016). As such, early childhood settings provide an opportune time for interventions aimed at strengthening these skills in young children. Research shows that significant differences in self-regulation are apparent in early childhood (McClelland et al., 2015; Zelazo et al., 2016). Although some of these differences in self-regulation are likely on account of natural October 2021 \| Volume 12 \| Article 711578 2 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. Li et al. variability in children’s developmental ages and stages, these differences relate to children’s abilities to engage in positive classroom behaviors (Day et al., 2015; Zelazo et al., 2016). Research on the high rates of preschool suspensions and expulsions indicates that early childhood teachers could benefit from support managing challenging classroom behaviors likely related to self-regulatory difficulties (Gilliam and Shabar, 2006). Early childhood educators enter the field with a range of educational backgrounds and experiences, making in-service trainings an especially valuable form of professional development for the early childhood field. Professional development is seen as a critical component of high-quality early childhood education settings, and thus, significant time and funds are devoted to supporting these efforts. It is through ongoing professional development that interventions, including self-regulation interventions, are often implemented. The Present Studyh Conducting cost-effectiveness analyses is a critical next step in the field of early childhood intervention to inform practice and policy toward identifying programs that are the most feasible and cost-effective. In the present cost-effectiveness study, cost and effectiveness data are analyzed for each approach to implement and to identify the model(s) that are most cost-effective. Although the RLPL program was implemented across studies using various models, each shared the common goal of improving children’s self-regulation. Therefore, these data provide a unique opportunity to compare cost-effectiveness among different models of the RLPL intervention, as the cost-effectiveness analysis is particularly helpful to compare among closely related interventions (Li et al., 2017, 2019). We expect our findings will help identify the optimal intervention design and strategy to improve children self-regulation in the most cost-effective way, and will provide critical evidence to inform decision making related to intervention selection for schools, districts, and additional stakeholders. Model 1: Trained Research Assistants (RAs; Graduate Students) Deliver the RLPL Games to Children Tominey and McClelland (2011) Sixty-five children (50% low-income as defined as participation in Head Start) from nine classrooms in two preschool centers participated in the study (children were randomized within classrooms). Children were an average of 54.60 months (range: 44–60 months; 60% female). The average maternal education level was about an Associate’s degree (M = 14.6 years) with a range of 6–21 years. Three children spoke Spanish as a first language, were assessed in Spanish, and received the intervention in Spanish. Participants For Model 2 dosage = (25 min per sess dels 1 and 3 dosage = (25 min per session * 16 sessions). For Model 2 the Institutional Review Board at Oregon State University, United States. in the study with a final number of 99 children with data at both time points. Children were an average age of 63.24 months and did not have prior preschool experience. No additional demographic information (i.e., parent education level) was available from the school district. All children received the RLPL intervention and self-regulation assessment in English. Procedure Across the four studies, the RLPL intervention was implemented in three different models as outlined below. In all studies, teachers reported that their normal curricula did not include games or activities that were similar to the RLPL games or emphasized self-regulation. Importantly, for all models in our study, participation in the intervention (versus a delayed control) was randomized. Individuals involved in data collection were blind to the condition of participating children and classrooms where randomization occurred. Children in all delivery models received similar amounts of the intervention. For Models 1 and 3, the duration was twice a week over eight weeks for 16 sessions, and in the Model 2, Schmitt et al. (2015) Two hundred seventy-six children from 14 Head Start classrooms across nine preschool centers participated (7 intervention and 7 control classrooms). Children averaged 51.69 months (range: 37.98 to 66.04 months; 51% female). The average parent education level was less than a high school degree (M = 11.15 years) with a range from 2 to 16 years of education. There were 33% (N = 88) English language learners (ELLs) who were assessed in Spanish and who received the intervention in Spanish. Model 3: Program Faculty Train Teachers, Who Then Deliver the RLPL Games to Children McClelland et al. (2019) One hundred fifty-seven children (52% female) from 13 Head Start classrooms participated in the study (10 intervention, 3 control). Children were an average of 51 months (range of 38–62 months), and parents’ education level ranged from 2 to 17 years (M = 11.27). Fifty-eight percent of the sample identified as Latino and 33% (N = 62) were ELLs. These children were assessed in Spanish but received the intervention in English following best practices for supporting Dual Language Learners (DLL’s; Tominey and O'Bryon, 2018). Participants Data included in this study were collected as part of four intervention studies examining the RLPL intervention in young children. The four studies fit three models of RLPL implementation as described below (see Table 1). Below, we describe the original samples and in the analytic strategy section we detail how sample sizes were analyzed to improve comparability across models. The study was approved by October 2021 \| Volume 12 \| Article 711578 3 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. Li et al. TABLE 1 \| Parameters Sources and Values for Base-case Analysis on Different Models to Implement the RLPL Program. RLPL implementation modela Model 1 Model 2 Model 3 Literature source Tominey and McClelland, 2011; Schmitt et al., 2015 Duncan et al., 2018 McClelland et al., 2019 N (children) 65; 276 125 157 Number of classrooms in interventionb 16 4 10 Child gender (% female) 60%/51% N/Ac 52% Child age (in months) 54.60/51.69 63.24 51 Low-income 50%/100% N/A 100% English language learner 4.6%/33% N/A 33% Average parent education level 14.6 years (Associate’s Degree equivalent)/11.5 years (less than High School) N/A 11.27 years (less than High School equivalent) Total person-hours spent in RLPL implementation 62 6 6 Dosaged 16 sessions (M = 6.67 h) 15 sessions (M = 6.25 h) 16 sessions (M = 6.67 h) Effect size of self-regulation scores 0.34/0.34 0.34 0.31 aModel 1 = trained research assistants (RAs) directly delivered the RLPL games to children; Model 2 = trained RAs trained trainers, who then trained teachers to deliver the RLPL games to children; Model 3 = program faculty trained teachers, who then delivered the RLPL games to children. bThere were 16 classrooms that received RLPL intervention under Model 1 (9 in the Tominey and McClelland, 2011 study, and 7 in the Schmitt et al., 2015 study), and 10 classrooms in intervention under Model 3 (McClelland et al., 2019). In Model 2, RAs trained trainers, who then trained a total of 4 teachers to deliver the RLPL games, so we analyzed cost for 4 classrooms. cN/A = Not available. TABLE 1 \| Parameters Sources and Values for Base-case Analysis on Different Models to Implement the RLPL Program. dFor Models 1 and 3 dosage = (25 min per session * 16 sessions). For Model 2 dosage = (25 min per session * 15 sessions). dFor Models 1 and 3 dosage = (25 min per session * 16 sessions). Duncan et al. (2018) Duncan et al. (2018) In this “train-the-trainer” model, RAs trained trainers who then trained teachers to deliver the RLPL games to children in a summer program. Of the participating schools, three schools were randomly assigned to receive the intervention (N = 67 children) and one school was randomly assigned to serve as a control group (receiving a summer program curricula, which was different from RLPL; N = 32). Teacher fidelity and feasibility surveys indicated that teachers implemented the games as intended and children enjoyed playing them. Children’s self-regulation was directly assessed at pre-program (July) and post-program (August) time points by trained school district staff. Model 2: RAs Train Trainers, Who Then Train Teachers to Deliver the RLPL Games to Children (“Train-the-Trainer”) Duncan et al. (2018) In this implementation model, one hundred twenty-five children from four schools in a large public school district participated October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org 4 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. the duration was five days a week over three weeks for 15 sessions. Each session was 20–30 min across all delivery models. On average, the dosage was 6.67 h for Models 1 and 3 (average of 25 min per session x 16 sessions), and 6.25 h for Model 2 (average of 25 min per session x 15 sessions; see Table 1). Schmitt et al. (2015) Fourteen classrooms were randomly assigned to an intervention group (n = 7; 126 children) or a BAU control group (n = 7; 150 children; N = 276 overall). Trained RAs directly assessed children’s self-regulation in the fall (pre-test) and in the spring (post-test) of the preschool year. The intervention was implemented in the winter by RAs. Parents and teachers completed demographic surveys. Analytic Strategy for the Cost-Effectiveness Analysisf We conducted our cost-effectiveness analyses from the perspective of RLPL program implementation (i.e., the costs associated with conducting research were not included as these expenses were supported by grants and would not be included in future implementation and program costs). In doing so, we reduced parameter uncertainty in our data collection; our analyses were more directly relevant to program implementation; and our findings could better support stakeholders’ decisions on replicating and expanding our intervention. Based on this analytical perspective, we included costs accrued from intervention implementation only, excluding costs accrued for research and evaluation purposes. We applied the concept of opportunity cost when estimating cost, which could better identify the trade-off value of cost items. As recommended by the Panel on Cost-Effectiveness in Health and Medicine, “(a) change in the use of a resource caused by a health intervention should be valued at its opportunity cost, which is the value the resource could have produced if it were spent in its best available alternative use” (Weinstein et al., 1996, p. 1255). We considered program cost as the total cost to implement Outcomes Self-Regulation g To assess children’s self-regulation, the Head-Toes-Knees- Shoulders (HTKS) measure was used in all studies. The HTKS assesses aspects of executive function including attention, working memory, and inhibitory control (McClelland et al., 2014) in children ages 3–8. The task has three parts. In the first part, children are asked to touch their head (or toes) when asked to touch their toes (or head). In the next part, a new rule is introduced where children touch their knees (or shoulders), and then, both rules are included (head/toes opposite and knees/shoulders opposite). In the third part, children are still doing the opposite, but the rules are switched (head goes with knees and shoulders goes with toes). Items are scored 0 for an incorrect response, 1 for a self-corrected response, and 2 for a correct response. The overall scores range from 0 to 60 except in Tominey and McClelland (2011) when the task had two parts and scores ranged from 0 to 40. Tominey and McClelland (2011) Tominey and McClelland (2011) Children within nine classrooms were randomly assigned to receive the intervention (n = 28) or to a Business-As-Usual (BAU) control group (n = 37; overall N = 65). Trained RAs assessed children’s self-regulation individually, and parents and teachers completed demographic surveys. Children were pretested in the fall, the intervention was administered in the winter by RAs, and children were post-tested in the spring of the preschool year. In the McClelland et al. (2019) study (Model 3 above), children received a revised version of the HTKS, called the HTKS-Revised (Gonzales et al., 2021). This version included a downward extension with the same commands where children responded to using verbal cues (e.g., “When I say Toes, you say Head”) rather than gross motor responses. This section was administered to children prior to receiving parts 1–3 of the original HTKS measure and overall scores ranged from 0 to 116. The HTKS-R and HTKS have both demonstrated strong reliability and validity in preschool aged children (McClelland et al., 2014; Gonzales et al., 2021). Previous research has demonstrated that the HTKS is sensitive to intervention effects, where children receiving an intervention showed significant improvement in self-regulation when compared with children in a control group (Tominey and McClelland, 2011; Schmitt et al., 2015; Duncan et al., 2018; Landis et al., 2018; Upshur et al., 2019). Model 1: In this implementation model, RLPL was administered by trained RAs who pulled children out of their classroom to implement the RLPL games and activities. Frontiers in Psychology \| www.frontiersin.org Model 3: McClelland et al. (2019) McClelland et al. (2019) In this model, program faculty trained teachers in the standard three-hour RLPL training, who then delivered the RLPL games to children. Classrooms were randomly assigned to two versions of the RLPL intervention (N = 10 classrooms total; 120 children) or a BAU control group (N = 3 classrooms; 37 children). In the fall (pre-test) and spring (post-test) of the preschool year, direct assessments of self-regulation were administered in English or Spanish. The intervention was implemented in the winter by teachers in either English or Spanish. Teacher fidelity and feasibility surveys demonstrated that teachers implemented the intervention with high fidelity. Parents and teachers completed demographic questionnaires. October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org 5 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. the RLPL program, which include personnel cost and material cost. Regarding personnel cost, we excluded teachers’ time spent in receiving trainings (e.g., trainers’ time in Model 2) and their time spent in delivering the RLPL games in classroom (e.g., teachers’ time in Models 2 and 3) during their regular work hours from our cost estimation, because these teachers would have to spend such time on work even without the RLPL program. For faculty, staff, and graduate students in our RLPL implementation team, we calculated their personnel costs as the product of their wages and their time spent in either training the teachers (e.g., RAs’ and faculty’s time in Models 2 and 3) or directly delivering the intervention (e.g., RA’s cost in Model 1). The RLPL program coordinator recorded the wages and time spent information for faculty, staff, and graduate students in each model. In addition to personnel cost, the intervention cost also includes $100 per classroom materials kit to deliver the RLPL games in each classroom. Classroom material kits included items currently found in early childhood classrooms for use during the intervention games (e.g., different colored circles cut out of construction paper, classroom drum and musical instrument egg shaker set). After estimating total program cost equal to the sum of total personnel cost and total material cost, we divided total program cost by the number of students to get the average intervention cost per student for each model. as the cost of additional investment in our intervention to gain one additional unit of outcome improvement. In addition to base-case analysis, we conducted a series of sensitivity analyses to adjust for parameter uncertainty. RESULTS g Model (3) program faculty trained teachers, who then delivered the RLPL games to children (McClelland et al., 2019). Model (3) program faculty trained teachers, who then delivered the RLPL games to children (McClelland et al., 2019). We extracted data from previous research to assess both cost and effectiveness for each model. There were 16 classrooms that received RLPL intervention under Model 1 (9 in the Tominey and McClelland, 2011 study, and 7 in the Schmitt et al., 2015 study), and 10 classrooms in intervention under Model 3 (McClelland et al., 2019). In the “train-the-trainer” model, RAs trained trainers, who then trained a total of 4 teachers to deliver the RLPL games, so we calculated cost of Model 2 for 4 classrooms (Table 1). As the typical class size for early childhood education settings is 15–20 children (specific to programs included in these studies), in our base-case analysis we standardize our estimates as 18 children in each classroom. This standardization is important to improve comparability across models, and can better support stakeholders’ decisions on replicating and expanding the RLPL programs. We then calculated the average intervention cost per student. Aligned with the goal of the RLPL program to improve children self-regulation, we used the effect size of self-regulation scores to measure effectiveness. Our base-case cost-effectiveness results are shown in Table 2. Although the program effectiveness was similar across the RLPL models, the cost varied from $7.72 per child in Model 2 to $11.15 per child in Model 1. Our base-case analysis suggested that Model 2 was the most cost-effective strategy with an ICER of 22.71. Other RLPL models were dominated, as they were associated with lower or same effectiveness at a higher cost compared to Model 2. Table 3 shows the results under various scenarios in our sensitivity analysis. Model 2 was consistently the most cost-effective strategy across Scenarios 1 to 3, while Models 1 and 3 were dominated. Specifically, the ICER of Model 2 decreased from 27.25 to 20.44 when we adjusted the number of children per classroom from 15 to 20 (Scenario 1), increased from 21.44 to 23.99 when we adjusted the variance of person-hours from 80 to 120% (Scenario 2), and increased from 22.71 to 27.25 when we adjusted the total implementation cost from 100 to 120% (Scenario 3). f Figure 1 illustrates our cost-effectiveness analytical model. RESULTS We evaluated cost-effectiveness among the three models of the RLPL program and compared that to no intervention. To conduct the cost-effectiveness analysis, we calculated the incremental cost-effectiveness ratio (ICER) associated with the RLPL program. Therefore, we would interpret our cost-effectiveness findings Frontiers in Psychology \| www.frontiersin.org McClelland et al. (2019) This would not only test the robustness of our main findings, but also provide critical evidence to inform decision makings on adopting and expanding our interventions at various scales and under different scenarios. Specifically, we conducted sensitivity analyses under three scenarios: Scenario (1) In our base-case analysis, we assumed 18 children in each classroom. The typical class size for early childhood education settings included in the present study is 15–20 children. We ran a sensitivity analysis to adjust the number of children per classroom from 15 to 20 for each model. Scenario (2) We ran a sensitivity analysis to adjust the potential variance in workload to implement the intervention, by increasing and reducing person-hours by 20% for each model. Scenario (3) In our base-case analysis we did not include facility cost, because such cost would still exist even without our RLPL program. Research suggests this cost could range from 10 to 20% of total cost (Jones et al., 2019). We ran a sensitivity analysis to adjust for facility cost, and we increased the total cost by the upper limit of 20% for each model. This allowed us to draw a more conservative cost-effectiveness results, and could comprehensively adjust for various uncertainties of increasing cost.t Table 1 presents the parameters sources and values used in our base-case analysis. According to previous research, the RLPL program had been implemented according to three models: We used TreeAge Pro software (TreeAge Pro 2021, Healthcare R2, 2020) to implement our analyses. We adjusted all monetary values to 2019 dollars by consumer price index (Bureau of Labor Statistics, 2020). Model (1) RAs directly delivered the RLPL games to children (Tominey and McClelland, 2011; Schmitt et al., 2015); Model (2) RAs trained trainers, who then trained teachers to deliver the RLPL games to children (Duncan et al., 2018); and. Model (2) RAs trained trainers, who then trained teachers to deliver the RLPL games to children (Duncan et al., 2018); and. The Importance of Cost-Effectiveness Studies in School Readiness Interventions The Importance of Cost-Effectiveness Studies in School Readiness Interventions In the search to support children’s school readiness and success, a plethora of early childhood interventions have been developed and tested (Blair and Raver, 2014; Sasser et al., 2017; Diamond et al., 2019; Welsh et al., 2020). Although research supports the effectiveness of these interventions, it is unclear if they are cost-effective for early childhood programs to implement and few programs have been evaluated to examine this. Cost- effectiveness analyses can provide important evidence to support decision makings of investing in interventions and optimizing program designs. However, the literature on cost-effectiveness analyses of early childhood self-regulation intervention has been sparse. Corso and colleagues (Corso et al., 2015) conducted cost-effectiveness analyses on a group-based parenting intervention aimed at improving behavioral outcomes among children living in poverty. Their intervention was associated with a cost of $178,000 per child and a cost of $91,100 per child, both in 2008 dollar, for severe behavioral problems avoided and for attention-deficit hyperactivity disorder avoided, respectively. These can be converted into 2019 dollar at around $211,000 and $108,000, which are substantially higher than DISCUSSION The present study conducted cost-effectiveness analyses on three implementation models of the Red Light, Purple Light (RLPL) intervention, a self-regulation intervention for young children. October 2021 \| Volume 12 \| Article 711578 6 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. FIGURE 1 \| Cost-effectiveness Analytical Model. TABLE 2 \| Cost-effectiveness Results in Base-case Analysis. Cost-effectiveness rankinga Cost ($ per child)b Incremental cost Effectiveness (effect size of self-regulation scores) Incremental effectiveness Incremental cost- effectiveness ratio No intervention 0 – 0 − Model 2 7.72 7.72 0.34 0.34 22.71 Model 3 7.85 0.12 0.31 −0.03 dominated Model 1 11.15 3.43 0.34 0 dominated aCost-effectiveness analysis order is from the lowest-cost strategy to the highest-cost strategy. Model 1 = trained research assistants (RAs) directly delivered the RLPL games to children; Model 2 = trained RAs trained trainers, who then trained teachers to deliver the RLPL games to children; Model 3 = program faculty trained teachers, who then delivered the RLPL games to children. bMonetary values were adjusted to 2019 dollars by consumer price index. FIGURE 1 \| Cost-effectiveness Analytical Model. TABLE 2 \| Cost-effectiveness Results in Base-case Analysis. Cost-effectiveness rankinga Cost ($ per child)b Incremental cost Effectiveness (effect size of self-regulation scores) Incremental effectiveness Incremental cost- effectiveness ratio No intervention 0 – 0 − Model 2 7.72 7.72 0.34 0.34 22.71 Model 3 7.85 0.12 0.31 −0.03 dominated Model 1 11.15 3.43 0.34 0 dominated aCost-effectiveness analysis order is from the lowest-cost strategy to the highest-cost strategy. Model 1 = trained research assistants (RAs) directly delivered the RLPL games to children; Model 2 = trained RAs trained trainers, who then trained teachers to deliver the RLPL games to children; Model 3 = program faculty trained teachers, who then delivered the RLPL games to children. bMonetary values were adjusted to 2019 dollars by consumer price index. FIGURE 1 \| Cost-effectiveness Analytical Model. FIGURE 1 \| Cost-effectiveness Analytical Model. FIGURE 1 \| Cost-effectiveness Analytical Model. FIGURE 1 \| Cost-effectiveness Analytical Model. TABLE 2 \| Cost-effectiveness Results in Base-case Analysis. Cost-effectiveness rankinga Cost ($ per child)b Incremental cost Effectiveness (effect size of self-regulation scores) Incremental effectiveness Incremental cost- effectiveness ratio No intervention 0 – 0 − Model 2 7.72 7.72 0.34 0.34 22.71 Model 3 7.85 0.12 0.31 −0.03 dominated Model 1 11.15 3.43 0.34 0 dominated aCost-effectiveness analysis order is from the lowest-cost strategy to the highest-cost strategy. Frontiers in Psychology \| www.frontiersin.org DISCUSSION Model 1 = trained research assistants (RAs) directly delivered the RLPL games to children; Model 2 = trained RAs trained trainers, who then trained teachers to deliver the RLPL games to children; Model 3 = program faculty trained teachers, who then delivered the RLPL games to children. bMonetary values were adjusted to 2019 dollars by consumer price index. aCost-effectiveness analysis order is from the lowest-cost strategy to the highest-cost strategy. Model 1 = trained research assistants (RAs) directly delivered the RLPL games to children; Model 2 = trained RAs trained trainers, who then trained teachers to deliver the RLPL games to children; Model 3 = program faculty trained teachers, who then delivered the RLPL games to children The three implementation models included: (Model 1) trained RAs directly delivered the RLPL intervention to children; (Model 2) RAs trained trainers (e.g., program coaches), who then trained teachers to implement RLPL with children (e.g., train-the-trainer); and (Model 3) program faculty trained teachers to deliver the RLPL intervention to children. Results of base-case analysis indicated that Model 2 was the most cost-effective strategy. We estimated that by implementing the “train-the-trainer” model, a cost of $23 per child was associated with a one-unit increase of the effect size of self-regulation scores. Our base-case finding remains robust to a series of sensitivity analyses, when we adjusted the number of students per classroom, the variance of person- hours, and the total cost to implement the RLPL program. The “train-the-trainer” model remained the optimal strategy across these scenarios, with ICER ranging from 20 to 27. the developer to other trained individuals (research assistants and faculty members with experience in early childhood) and finally to teachers. The second and third models (RAs training trainers who then trained teachers and program faculty training teachers) were only possible in later iterations of the intervention, but also reflected a feasible method of intervention delivery and implementation in current early childhood contexts. Iterative Intervention Developmenth p The present study was unique in several respects. Notably, having access to data from randomized control trials where the same curriculum was tested using different models of implementation allowed for comparison of cost-effectiveness across these models. Each model was implemented at a different point in time and different stage of the iterative development process for the RLPL intervention. Although the intervention curriculum remained the same across implementation methods, the methods of implementation changed over time. In the first study, the intervention was delivered by the curriculum developers. This model is critical as an initial step in testing a new intervention, but not feasible to support long-term scaling efforts. The models that followed shifted the intervention from October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org 7 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. Li et al. TABLE 3 \| Cost and Incremental Cost-effectiveness Ratio (ICER) of Model 2 Under Various Scenarios in Sensitivity Analysis.a Scenario 1: Adjust number of children per classroom Number of Children per classroom Cost per child ($) ICER 15 9.27 27.25 16 8.69 25.55 17 8.18 24.05 18 (Base case) 7.72 22.71 19 7.32 21.52 20 6.95 20.44 Scenario 2: Adjust person-hours to implement interventions Percentage change of person-hours Cost per child ($) ICER −20% 7.29 21.44 −10% 7.51 22.08 Base case 7.72 22.71 +10% 7.94 23.35 +20% 8.16 23.99 Scenario 3: Adjust total cost to implement interventions Percentage change of total cost Cost per child ($) ICER Base case 7.72 22.71 +5% 8.11 23.85 +10% 8.49 24.98 +15% 8.88 26.12 +20% 9.27 27.25 aOnly shows results of Model 2 (the “train-the-trainer” model) compared to no intervention in sensitivity analysis, because other models were dominated. TABLE 3 \| Cost and Incremental Cost-effectiveness Ratio (ICER) of Model 2 Under Various Scenarios in Sensitivity Analysis.a studies to compare with, which we see as the major contribution of our current study. In addition, we collected data from previous interventions of our program and conducted a retrospective cost-effectiveness analysis in this study. In so doing, we could only look back at how the intervention was already implemented in each model, and assess the cost and effectiveness that were already in place. As a result, we encountered some uncertainty and variance in parameters. To address this limitation, we conducted a series of sensitivity analyses aiming to adjust for various possibilities. Conclusion Our economic evaluation of RLPL fills an important gap related to the value of investment in early childhood self-regulation interventions. These results can serve as a model for future cost-effectiveness analyses of early childhood intervention programs and inform policymakers and stakeholders on the cost-effectiveness for the potential impact of an intervention delivered by different models, so that resource allocation and future program design can be optimized. the ICER in our RLPL program. It is worth noting that as the two programs focused on different outcomes, exact comparisons between our study and the one conducted by Corso et al. (2015) cannot be made, but it may still help to illustrate the magnitude of cost-effectiveness, and to aid stakeholders’ investment decisions across programs. In addition, other research with a similar school readiness intervention (e.g., Head Start REDI) has documented the importance of such evaluations along with how initial costs may pay off years later (Bierman et al., 2018).h Limitations Although this study was unique in many respects, we recognize several limitations. First, there is a lack of comparable evidence in the literature that we can use as a reference case to assess our findings. This is mainly because cost-effectiveness analyses on early-childhood self-regulation have been sparse. However, this provides us with a unique opportunity to address this critical knowledge gap, and sets up a model for future cost-effectives DATA AVAILABILITY STATEMENT The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author. The present study extended work in this area by examining cost-effectiveness for the different implementation versions of the RLPL intervention. Results found that costs were low across all implementation methods. In particular, the program could be more cost-effective by using a “train-the-trainer” model, due to its great potential of reaching economies of scale. Future research needs to further investigate the “train-the-trainer” model of implementation as this model is expanded. Iterative Intervention Developmenth Our sensitivity analyses also provide evidence to support stakeholders’ decisions on replicating and expanding our interventions under different scenarios and at various scales. Although fidelity data were collected across each study, each model had different levels of resources to monitor and support fidelity of implementation over time and thus, these analyses were not included as part of the present study. Future studies should consider the relation between cost-effectiveness and fidelity data. Finally, we propose that future studies should analyze cost-effectiveness on early childhood interventions prospectively as part of the development and implementation process. This approach will empower researchers to evaluate interventions alongside intervention implementation, so that cost and effectiveness data will be better aligned with study designs and have greater potential to inform practice and policy. ETHICS STATEMENT The studies involving human participants were reviewed and approved by Oregon State University Institutional Review Board. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin. Frontiers in Psychology \| www.frontiersin.org REFERENCES Diamond, A., Barnett, W. S., Thomas, J., and Munro, S. (2007). Preschool program improves cognitive control. Science 318, 1387–1388. doi: 10.1126/ science.1151148 Allan, N. P., Hume, L. E., Allan, D. M., Farrington, A. L., and Lonigan, C. J. (2014). Relations between inhibitory control and the development of academic skills in preschool and kindergarten: A meta-analysis. Dev. Psychol. 50, 2368–2379. doi: 10.1037/a0037493 Diamond, A., Lee, C., Senften, P., Lam, A., and Abbott, D. (2019). Randomized control trial of tools of the mind: marked benefits to kindergarten children and their teachers. PLoS One 14:e0222447. doi: 10.1371/journal. pone.0222447 Diamond, A., Lee, C., Senften, P., Lam, A., and Abbott, D. (2019). Randomized control trial of tools of the mind: marked benefits to kindergarten children and their teachers. PLoS One 14:e0222447. doi: 10.1371/journal. pone.0222447 Bierman, K. L., Nix, R. L., Greenberg, M. T., Blair, C., and Domitrovich, C. E. (2008). Executive functions and school readiness intervention: impact, moderation, and mediation in the head start REDI program. Dev. Psychopathol. 20, 821–843. doi: 10.1017/S0954579408000394 Diamond, A., and Ling, D. S. (2016). Conclusions about interventions, programs, and approaches for improving executive functions that appear justified and those that, despite much hype, do not. Deve. Cogn. Neurosci. 18, 34–48. doi: 10.1016/j.dcn.2015.11.005 Bierman, K. L., Welsh, J., Heinrichs, B. S., and Nix, R. L. (2018). Effect of preschool home visiting on school readiness and need for services in elementary school: A randomized clinical trial. JAMA Pediatr. 172:e181029. doi: 10.1001/jamapediatrics.2018.1029 Domitrovich, C. E., Cortes, R. C., and Greenberg, M. T. (2007). Improving young children’s social and emotional competence: a randomized trial of the preschool “PATHS” curriculum. J. Prim. Prev. 28, 67–91. doi: 10.1007/ s10935-007-0081-0 Blair, C. (2003). Behavioral inhibition and behavioral activation in young children: relations with self-regulation and adaptation to preschool in children attending head start. Dev. Psychobiol. 42, 301–311. doi: 10.1002/ dev.10103 Duckworth, A. L., Tsukayama, E., and May, H. (2010). Establishing causality using longitudinal hierarchical linear modeling: An illustration predicting achievement from self-control. Soc. Psychol. Personal. Serv. 1, 311–317. doi: 10.1177/1948550609359707 Blair, C., and Raver, C. C. (2014). Closing the achievement gap through modification of neurocognitive and neuroendocrine function: results from a cluster randomized controlled trial of an innovative approach to the education of children in kindergarten. PLoS One 9:e112393. doi: 10.1371/ journal.pone.0112393 Duncan, R. J., Schmitt, S. A., Burke, M., and McClelland, M. M. (2018). REFERENCES Combining a kindergarten readiness summer program with a self-regulation intervention improves school readiness. Early Child Res. Q. 42, 291–300. doi: 10.1016/j.ecresq.2017.10.012 j Blair, C., and Raver, C. C. (2015). School readiness and self-regulation: A developmental psychobiological approach. Annu. Rev. Psychol. 66, 711–731. doi: 10.1146/annurev-psych-010814-015221 Farran, D., Wilson, S. J., and Lipsey, M. (2013). Effects of a Curricular Attempt to Improve Self-Regulation and Achievement in Prekindergarten Children Biennial Meeting for the Society for Research in Child Development, (Seattle, WA). Blair, C., and Razza, R. P. (2007). Relating effortful control, executive function, and false belief understanding to emerging math and literacy ability in kindergarten. Child Dev. 78, 647–663. doi: 10.1111/j.1467- 8624.2007.01019.x Gathercole, S. E. (2008). Working memory in the classroom. The Psychologist 21, 382–385. Gilliam, W. S., and Shabar, G. (2006). Preschool and child care expulsion and suspension: rates and predictors in one state. Infants Young Children: An Interdiscipl. J. Special Care Pract. 19, 228–245. doi: 10.1097/00001163- 200607000-00007 Blair, C., Ursache, A., Greenberg, M., and Vernon-Feagans, L. (2015). Multiple aspects of self-regulation uniquely predict mathematics but not letter–word knowledge in the early elementary grades. Dev. Psychol. 51, 459–472. doi: 10.1037/a0038813 Gonzales, C. R., Bowles, R., Geldhof, G. J., Cameron, C. E., Tracy, A., and McClelland, M. M. (2021). The head-toes-knees-shoulders revised (HTKS-R): development and psychometric properties of a revision to reduce floor effects. Early Child Res. Q. 56, 320–332. doi: 10.1016/j. ecresq.2021.03.008 Bureau of Labor Statistics. (2020). Consumer Price Index. https://www.bls.gov/ cpi/. (Accessed October 29, 2020). Cameron, C. E., Connor, C. M., Morrison, F. J., and Jewkes, A. M. (2008). Effects of classroom organization on letter-word reading in first grade. J. Sch. Psychol. 16, 172–192. doi: 10.1016/j.jsp.2007.03.002 Jones, D. E., Bierman, K. L., Crowley, D. M., Welsh, J. A., and Gest, J. (2019). Important issues in estimating costs of early childhood educational interventions: An example from the REDI program. Child Youth Serv. Rev. 107:104498. doi: 10.1016/j.childyouth.2019.104498 Cameron, C. E., and Morrison, F. J. (2011). Teacher activity orienting predicts preschoolers’ academic and self-regulatory skills. Early Educ. Dev. 22, 620–648. doi: 10.1080/10409280903544405 Cameron Ponitz, C. E., McClelland, M. M., Matthews, J. M., and Morrison, F. J. (2009). A structured observation of behavioral self-regulation and its contribution to kindergarten outcomes. Dev. Psychol. 45, 605–619. doi: 10.1037/a0015365 Keown, L. J., Franke, N., and Triggs, C. M. (2020). An evaluation of a classroom- based intervention to improve executive functions in 4-year old children in New Zealand. AUTHOR CONTRIBUTIONS All authors contributed to the conceptualization and design of the study, data analysis and results, and the writing of the manuscript. October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org 8 Cost-Effectiveness Analyses for Self-Regulation Intervention Li et al. Li et al. The content is the responsibility of the authors and does not necessarily represent the official views of the Institute of Education Sciences, or the U.S. Department of Education. This research was supported by grants from the Ford Family Foundation and Oregon State University. FUNDING The research reported here was supported by the U.S. Department of Education Institute for Education Sciences grants # R305A150196 (PI: McClelland) to Oregon State University. REFERENCES Early Childhood Educ. J. 48, 621–631. doi: 10.1007/s10643- 020-01023-x Cannon, J. S., Kilburn, M. R., Karoly, L. A., Mattox, T., Muchow, A. N., and Buenaventura, M. (2018). Investing early: taking stock of outcomes and economic returns from early childhood programs. RAND health Q. 7:6. doi: 10.7249/RR1993 Landis, T. D., Hart, K. C., and Graziano, P. A. (2018). Targeting self-regulation and academic functioning among preschoolers with behavior problems: are there incremental benefits to including cognitive training as part of a classroom curriculum? Child Neuropsychol. 25, 688–704. doi: 10.1080/09297049. 2018.1526271 Corso, P. S., Visser, S. N., Ingels, J. B., and Perou, R. (2015). Cost-effectiveness of Legacy for Children™ for reducing behavioral problems and risk for ADHD among children living in poverty. J. child adolesc. behavior 3:240. doi: 10.4172/2375-4494.1000240 Li, T., Norcross, M. F., Johnson, S. T., and Koester, M. C. (2019). Cost-benefit of hiring athletic trainers in Oregon high schools From 2011–2014. J. Athl. Train. 54, 165–169. doi: 10.4085/1062-6050-390-17i Day, S. L., Connor, C. M., and McClelland, M. M. (2015). Children’s behavioral regulation and literacy: The impact of the first grade classroom environment. J. Sch. Psychol. 53, 409–428. doi: 10.1016/j.jsp.2015.07.004 Li, T., Waters, T. M., Kaplan, E. K., Kaplan, C. M., Nyarko, K. A., Derefinko, K. J., et al. (2017). Economic analyses of an alcohol misconduct prevention program in a military setting. Mil. Med. 182, e1562–e1567. doi: 10.7205/milmed-d- 16-00098 Diamond, A. (2002). “Normal development of prefrontal cortex from birth to young adulthood: cognitive functions, anatomy, and biochemistry,” in Principles of Frontal Lobe Function. eds. D. T. Stuss and R. T. Knight (Oxford University Press), 466–503. McClelland, M. M., Acock, A. C., and Morrison, F. J. (2006). The impact of kindergarten learning-related skills on academic trajectories at the end of October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org 9 Li et al. Cost-Effectiveness Analyses for Self-Regulation Intervention elementary school. Early Child Res. Q. 21, 471–490. doi: 10.1016/j. ecresq.2006.09.003 elementary school. Early Child Res. Q. 21, 471–490. doi: 10.1016/j. ecresq.2006.09.003 regulation as a mediating mechanism. Child Dev. 82, 362–378. doi: 10.1111/j. 1467-8624.2010.01561.x regulation as a mediating mechanism. Child Dev. 82, 362–378. doi: 10.1111/j. 1467-8624.2010.01561.x Robson, D. A., Allen, M. S., and Howard, S. J. (2020). Self-regulation in childhood as a predictor of future outcomes: A meta-analytic review. Psychol. Bull. 146, 324–354. doi: 10.1037/bul0000227 McClelland, M. M., Acock, A. C., Piccinin, A., Rhea, S. A., and Stallings, M. C. (2013). REFERENCES Relations between preschool attention span-persistence and age 25 educational outcomes. Early Child Res. Q. 28, 314–324. doi: 10.1016/j. ecresq.2012.07.008 Rothbart, M. K., and Posner, M. I. (2005). Genes and experience in the development of executive attention and effortful control. New Dir. Child and Adolesc. Dev. 2005, 101–108. doi: 10.1002/cd.142 McClelland, M. M., and Cameron, C. E. (2012). Self-regulation in early childhood: improving conceptual clarity and developing ecologically valid measures. Child Dev. Perspect. 6, 136–142. doi: 10.1111/j.1750-8606.2011.00191.x Sasser, T. R., Bierman, K. L., Heinrichs, B., and Nix, R. L. (2017). Preschool intervention can promote sustained growth in the executive-function skills of children exhibiting early deficits. Psychol. Sci. 28, 1719–1730. doi: 10.1177/0956797617711640 McClelland, M. M., Cameron, C. E., Connor, C. M., Farris, C. L., Jewkes, A. M., and Morrison, F. J. (2007a). Links between behavioral regulation and preschoolers’ literacy, vocabulary and math skills. Dev. Psychol. 43, 947–959. doi: 10.1037/0012-1649.43.4.947 Schmitt, S. A., Geldhof, G. J., Purpura, D. J., Duncan, R., and McClelland, M. M. (2017). Examining the relations between executive function, math, and literacy during the transition to kindergarten: A multi-analytic approach. J. Educ. Psychol. 109, 1120–1140. doi: 10.1037/edu0000193 McClelland, M. M., Cameron, C. E., Duncan, R., Bowles, R. P., Acock, A. C., Miao, A., et al. (2014). Predictors of early growth in academic achievement: the head-toes-knees-shoulders task. Front. Psychol. 5:599. doi: 10.3389/ fpsyg.2014.00599 Schmitt, S. A., McClelland, M. M., Tominey, S. L., and Acock, A. C. (2015). Strengthening school readiness for head start children: evaluation of a self- regulation intervention. Early Childhood Res. Quarterly 30, 20–31. doi: 10.1016/j.ecresq.2014.08.001 McClelland, M. M., Cameron, C. E., Wanless, S. B., and Murray, A. (2007b). “Executive function, behavioral self-regulation, and social-emotional competence: links to school readiness,” in Contemporary Perspectives on Social Learning in Early Childhood Education. eds. O. N. Saracho and B. Spodek, 83–107. Snow, K. L. (2006). Measuring school readiness: conceptual and practical considerations. Early Educ. Dev. 17, 7–41. doi: 10.1207/s15566935eed1701_2i McClelland, M. M., Geldhof, J., Cameron, C. E., and Wanless, S. B. (2015). “Development and self-regulation,” in Handbook of Child Psychology and Developmental Science. 7th Edn., Vol. 1. eds. W. F. Overton and P. C. M. Molenaar (Wiley), 523–565. Tominey, S. L., and McClelland, M. M. (2011). Red light, purple light: findings from a randomized trial using circle time games to improve behavioral self-regulation in preschool. Early Educ. Dev. 22, 489–519. doi: 10.1080/10409289.2011.574258 McClelland, M. M., Ponitz, C. REFERENCES C., Messersmith, E., and Tominey, S. (2010). “Self-regulation: The integration of cognition and emotion,” in Handbook of Life-Span Development. eds. R. Lerner and B. Overton (Wiley and Sons) Tominey, S., and O'Bryon, E. C. (2018). 45 Strategies that Support Young Dual Language Learners. (Paul H: Brookes Publishing Company).t TreeAge Pro 2021, Healthcare R2 (2020). TreeAge Software. Williamstown, MA; software. Available at. www.treeage.com McClelland, M. M., and Tominey, S. L. (2015). Stop, Think, Act: Integrating Self-Regulation in the Early Childhood Classroom Taylor and Francis. t Upshur, C. C., Wenz-Gross, M., Rhoads, C., Heyman, M., Yoo, Y., and Sawosik, G. (2019). A randomized efficacy trial of the second step early learning (SSEL) curriculum. J. Appl. Dev. Psychol. 62, 145–159. doi: 10.1016/j. appdev.2019.02.008 McClelland, M. M., Tominey, S. L., Schmitt, S. A., and Duncan, R. (2017). Social-emotional learning interventions in early childhood. Futur. Child. 27, 33–48. doi: 10.1353/foc.2017.0002 Weinstein, M. C., Siegel, J. E., Gold, M. R., Kamlet, M. S., and Russell, L. B. (1996). Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 276, 1253–1258. doi: 10.1001/jama.1996.03540 150055031 Weinstein, M. C., Siegel, J. E., Gold, M. R., Kamlet, M. S., and Russell, L. B. (1996). Recommendations of the panel on cost-effectiveness in health McClelland, M. M., Tominey, S. L., Schmitt, S. A., Hatfield, B. E., Purpura, D. J., Gonzales, C. R., et al. (2019). Red light, purple light! Results of an intervention to promote school readiness for children From low-income backgrounds. Front. Psychol. 10:2365. doi: 10.3389/fpsyg.2019.02365 and medicine. JAMA 276, 1253–1258. doi: 10.1001/jama.1996.03540 150055031 Moffitt, T. E., Arseneault, L., Belsky, D., Dickson, N., Hancox, R. J., Harrington, H., et al. (2011). A gradient of childhood self-control predicts health, wealth, and public safety. Proc. Natl. Acad. Sci. 108, 2693–2698. doi: 10.1073/pnas.1010076108 Welsh, J. A., Bierman, K. L., Nix, R. L., and Heinrichs, B. N. (2020). Sustained effects of a school readiness intervention: 5th grade outcomes of the head start REDI program. Early Child Res. Q. 53, 151–160. doi: 10.1016/j. ecresq.2020.03.009 Montroy, J. J., Bowles, R. P., Skibbe, L. E., McClelland, M. M., and Morrison, F. J. (2016). The development of self-regulation Across early childhood. Dev. Psychol. 52, 1744–1764. doi: 10.1037/dev0000159 Zelazo, P. D., Blair, C. B., and Willoughby, M. T. (2016). Executive Function: Implications for Education (NCER 2017–2000). Institute of Education Sciences, U.S. Department of Education. Available at: http://ies.ed.gov/. Morris, P. A., Mattera, S., Castells, N., and Bangser, M. (2014). REFERENCES Improving children’s social and emotional development in Head Start: The Head Start CARES demonstration Head Start Research Conference, Washington: D.C. Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. National Academies of Sciences, Engineering and Medicine (2016). Advancing the Power of Economic Evidence to Inform Investments in Children, Youth, and Families The National Academies Press. Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. h Nguyen, T., Watts, T. W., Duncan, G. J., Clements, D. H., Sarama, J. S., Wolfe, C., et al. (2016). Which preschool mathematics competencies are most predictive of fifth grade achievement? Early Child Res. Q. 36, 550–560. doi: 10.1016/j. ecresq.2016.02.003 Pandey, A., Hale, D., Das, S., Goddings, A., Blakemore, S., and Viner, R. M. (2018). Effectiveness of universal self-regulation–based interventions in children and adolescents: A systematic review and meta-analysis. JAMA Pediatr. 172, 566–575. doi: 10.1001/jamapediatrics.2018.0232 Copyright © 2021 Li, McClelland, Tominey and Tracy. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Purpura, D. J., Schmitt, S. A., and Ganley, C. M. (2017). Foundations of mathematics and literacy: The role of executive functioning components. J. Exp. Child Psychol. 153, 15–34. doi: 10.1016/j.jecp.2016.08.010 Raver, C. C., Jones, S. M., Li-Grining, C., Zhai, F., Bub, K., and Pressler, E. (2011). CSRP’s impact on low-income preschoolers’ preacademic skills: self- October 2021 \| Volume 12 \| Article 711578 Frontiers in Psychology \| www.frontiersin.org 10
https://openalex.org/W3036018319	https://jurnal.unmer.ac.id/index.php/jmdk/article/download/4245/pdf	Indonesian	null	Marketing Intelligence dalam Menumbuhkan Jiwa Entrepreneur pada Generasi Milenial	Jurnal manajemen dan kewirausahaan/Jurnal Manajemen dan Kewirausahaan	2,020	cc-by-sa	4,136	Marketing Intelligence dalam Menumbuhkan Jiwa Entrepreneur pada Generasi Milenial Abdul Waris1, Wahyu Setiyorini2, Chodidjah2 1Jurusan Administrasi Niaga, Politeknik Negeri Malang, Indonesia 2Fakultas Ekonomi dan Bisnis, Universitas Merdeka Malang, Indonesia waris_sip@polinema.ac.id Abstract Indonesia is currently coming into a demographic dividend era occuring due to changes in the age structure of its population because the proportion of productive age is above 2/3 of the total population and around 50.36% of the population of productive age are basically millennial generation. This is an opportunity that must be utilized as well as possible by maximizing the capabilities of the current millennial generations which have open, free, critical, and courageous mindset as the essential things in entrepreneurship. The aim of this research is to find out the effect of marketing intelligence in growing the entrepreneurial spirit in millennial generations by using quantitative approaches and descriptive analysis method. The results show that: 1) marketing intelligence activities performed by millennial generation can help growing the entrepreneurial spirit; and 2) marketing research has the dominant influence of the growing entrepreneurial spirit of the millennial generation in marketing intelligence activities. This marketing intelligence activity does not have to be done in a group, but it can be done independently or alone. Keywords : Entrepreneurial Spirit, Millenial Generations, Marketing Intelligence yang biasa dikenal dengan istilah rasio ketergantungan (dependency rasio). 75 - 84 Marketing Intelligence dalam Menumbuhkan Jiwa Entrepreneur pada Generasi Milenial Abdul Waris1, Wahyu Setiyorini2, Chodidjah2 1Jurusan Administrasi Niaga, Politeknik Negeri Malang, Indonesia 2Fakultas Ekonomi dan Bisnis, Universitas Merdeka Malang, Indonesia waris_sip@polinema.ac.id PENDAHULUAN Indonesia saat ini memasuki era baru yang lebih dikenal dengan era bonus demografi, ini merupakan fenomena yang sangat langka karena dalam perjalanan sejarah suatu Negara hal ini hanya akan terjadi satu kali saja. Salah satu tanda bahwa era bonus demografi ini telah tiba di Indonesia adalah terjadinya kenaikan jumlah populasi penduduk yang produktif (berusia antara 15 tahun s.d. 65 tahun) secara signifikan jika dibandingkan dengan populasi penduduk yang non produktif (yang saat ini berusia dibawah 15 tahun dan berusia 65 tahun ke atas), rasio populasi penduduk yang produktif saat ini di Indonesia mencapai (2/3) dua sepertiga dari total penduduk yang ada atau Era bonus demografi ini merupakan tantangan bagi Indonesia dan harus dijadikan sebuah peluang yang tidak boleh dilewatkan begitu saja, Indonesia harus mampu memanfaatkan potensi penduduk usia produktif yang cukup tinggi ini, apalagi berdasarkan data (Badan Pusat Statistik, 2018) menyebutkan bahwa ada sekitar 50,36% dari total penduduk yang berusia produktif merupakan gerenasi milenial. Dan berdasarkan teori generasi yang dikemukakan oleh (Mannheim, 1952) menyatakan bahwa generasi yang lahir antara tahun 1980-an sampai 2000-an dikategorikan sebagai gerenasi milenial. 75 76 secara optimal apabila dilaksanakan oleh orang yang mengerti tentang teknologi seperti generasi milenial. Dengan kemampuan dibidang teknologi yang melekat dalam diri generasi milenial saat ini juga dapat menjadi modal membuka kesempatan dan peluang untuk berinovasi dalam berbagai bidang (berwirausaha). Di Indonesia para generasi milenial memiliki kesempatan serta peluang yang baik untuk berinovasi dengan dukungan ekosistem digital yang sedang tumbuh dan berkembang, seperti hadirnya bisnis e-commerce yang merupakan karya asli anak bangsa mampu memfasilitasi gerenasi milenial yang memiliki semangat untuk berwirausaha juga dapat berkembang. Selain bisnis e-commerce hadir pula bisnis yang bergerak di bidang transportasi online semisal transportasi online Gojek, Grab, Uber dan masih banyak lagi yang lainnya. Lahirnya bisnis e-commerce dan transportasi online ini merupakan hasil inovasi untuk menjawab tantangan dan kebutuhan dalam mewujudkan kemandirian secara ekonomi bagi generasi milenial di era digital ini. Hal ini sejalan dengan hasil penelitian Ambarwati & Sobari (2020), yang menyatakan bahwa mahasiswa dapat menjadi seorang pengusaha apabila mampu mengubah pola pikir konvensional ke digital dengan memanfaatkan teknologi yang terus berkembang dimulai dari hal yang sederhana seperti membangun pemasaran melalui media social yang dimiliki, bahkan saat ini ada mahasiswa yang mampu menciptakan aplikasi untuk membangun bisnisnya sendiri. Menurut Sebastian dkk. PENDAHULUAN (2016), menyebutkan bahwa generasi milenial memiliki keunggulan yang tidak dimiliki oleh generasi sebelumnya, diantaranya adalah dalam melaksanakan segala aktifitas maupun kegiatannya generasi milenial menginginkan segala sesuatunya dilakukan serba cepat, sulit bertahan dalam pekerjaan yang sama dalam waktu yang relatif lama, sangat dinamis, kreatif, inovatif, sangat friendly dengan media sosial dan perkembangan teknologi. Li (2006) dalam penelitiannya juga memperoleh hal serupa, yakni 68,4% atau lebih dari dua pertiga dari mahasiswa yang diwawancarainnya, mereka lebih memilih untuk mempunyai bisnis sendiri dari pada bekerja kepada orang lain, artinya bahwa semangat untuk berwirausaha dikalangan mahasiswa itu sangat tinggi. Berdasar uraian tersebut, penelitian ini akan mengkaji bagaimana peran marketing intelligence dalam menumbuhkan jiwa entrepreneur generasi milenial. Dunia industri saat ini jika ingin menjangkau segmen generasi milenial harus mulai meningkatkan cara-cara mereka dari sebelumnya, mengingat generasi milenial telah mampu merubah cara berpikir sebagai seorang konsumen yang lebih mengutamakan sisi emosi, keinginan, kebutuhan, dan tuntutannya. Maka dunia industri harus mampu mendeteksi perubahan cara berpikir tersebut. Marketing Intelligence Marketing intelligence (kecerdasan pemasaran) merupakan manifestasi dari sebuah kemampuan dalam mengumpulakn dan menganalisis suatu informasi atau data agar segala sesuatu yang terkait dengan pengambilan keputusan dapat diansipasi dengan baik, terdapat empat bagian yang dibutuhkan dalam kecerdasan ini meliputi product intelligence, competitor intelligence, customer understanding and market understanding (Kurniawan, 2020). Untuk mendeteksi orientasi perilaku konsumen yang efektif maka diperlukan kecerdasan bermarketing atau yang dikenal dengan marketing intelligence, hal ini diperlukan dalam bentuk pengumpulan data dan kemampuan analisis sebagai upaya dalam pengambilan keputusan yang tepat, seperti menganalisis tentang kondisi pesaing, daya saing produk serta memiliki kemampuan untuk memahami pasar dan pelanggan secara menyeluruh. Kemampuan ini akan berjalan Menurut Burns dkk., (2017) terdapat dua unsur penting dalam marketing intelligence yaitu meliputi marketing research dan customer relationship marketing/database marketing, yang menjadi fokus dari marketing research adalah lebih 77 kepada proses perencanaan pemasaran yang dimulai dengan menganalisis situasi dan kondisi serta membangun strategi dan program pemasaran yang diakhiri dengan mengaplikasikan strategi dan program yang sudah di rancang sebelumnya. Sedangkan yang menjadi fokus utama dari database marketing terletak pada pengolahan data dalam sisten database. Cara kerja database marketing ini adalah dengan mengkolaborasikan antara teori pemasaran dengan teori teknologi informasi dan database marketing ini masih tergolong teori yang masih baru (Guoxiang & Zhiheng, 2013). Jiwa Entrepreneur Entrepreneurship selalu berkembang dari generasi ke generasi mengikuti perkembangan zaman, namun ada hal yang tidak akan pernah berubah yaitu kemampuan seorang entrepreneur yang selalu dapat memanfaatkan sumber daya yang ada untuk selalu mengahasilkan produk/jasa yang baru, menciptakan bisnis yang baru atau bahkan mengembangkan bisnis yang sudah ada (Usman dalam Hatta, 2008). Menciptakan nilai tambah dengan memakai cara baru dan berbeda dari sebelumnya merupakan point inti dalam kewirausahaan agar dapat bersaing secara kompetitif dengan para pesainya. Menurut Zimmerer dalam Hatta (2008), terdapat beberapa pendekatan metode dalam penciptaan nilai tambah diantaranya adalah melalui proses penemuan pengetahuan yang baru atau bisa juga memberikan nilai tambah dengan mengembangkan teknologi yang baru atau bisa juga dengan cara memperbaiki produk atau layanan jasa yang sudah ada saat ini atau bahwakan dapat menemukan cara yang lebih efisien dengan sumber daya yang lebih sedikit. Dalam strategi pemasaran modern seringkali mengunakan database marketing ini untuk aktifitas data mining, aktifitas ini dilakukan untuk mendukung dalam pengambilan keputusan guna mempermudah dalam menemukan pengetahuan baru (Luo dkk., 2012). Menariknya dalam aktifitas database marketing ini memanfaatkan teknologi yang ada mulai dari proses pengumpulan data sampai dengan analisis datanya. Salah satu teknik analisis yang bisa digunakan adalah SaaS (Software as a Service) atau juga bisa mengunakan OLAP (Online Analitycal Processing) dan masih banyak lagi tools yang bisa digunakan. Output dari data mining ini sangat penting karena nantinya akan digunakan sebagai dasar dalam pengambilan sebuah keputusan seperti untuk menganalisis target penjualan, membuat forcasting konsumen serta untuk menganalisis pelanggan yang potensial. Setiap orang pada hakekatnya memiliki semangat entrepreneur dalam dirinya, asalkan orang tersebut memiliki kemauan untuk senantiasa berperilaku kreatif, inovatif, selalu siap menghsdapi perubahan dan tantangan serta menyukai kemajuan karena sejatinya rahasia untuk menjadi seorang entrepreneur adalah kreatif dan inovatif. Menurut Li (2006) dari hasil penelitian yang pernah dilakukan memberikan kesimpulan bahwa Perguruan Tinggi memiliki tanggungjawab untuk ikut serta menumbuhkan potensi-potensi yang dimiliki oleh peserta didiknya dengan memberikan pendidikan dan wawasan terkait kewirausahaan karena dengan memahami secara detail tentang kewirausahaan ternyata dapat mempengaruhi minat kaum muda untuk menjadi seorang wirausahawan. Berdasarkan hasil penelitian Amborowati & Suyanto (2015) menyatakan bahwa memanfaatkan teknologi data mining bisa menjadi salah satu opsi yang dapat digunakan dalam melakukan segmentasi pasar/konsumen potensial untuk mendukung kegiatan marketing intelligence. Menurut Kottler & Armstrong (2008), marketing intelligence adalah kumpulan informasi dan analisis secara sistematis terkait informasi yang tersedia tentang pesaing dan perkembangan di pasar. Generasi Milenial Dalam generation theory yang dikemukakan oleh Mannheim (1952) menyebut generasi milenial dengan istilah generasi Y, jika berdasarkan dari karakteriktik tahun kelahirannya generasi Y ini adalah generasi yang lahir dalam rasio tahun 1980-an sampai tahun 2000-an, dan melalui editorial dalam sebuah surat kabar utama yang terbit di Amerika Serikat tahun 1993 berperan sangat besar dalam mempopulerkan istilah ini. Berikut ini adalah hipotesis yang digunakan : H1: Marketing research berpengaruh positif dalam menumbuhkan jiwa entrepreneur pada generasi milenial H2: Database marketing berpengaruh positif dalam menumbuhkan jiwa entrepreneur pada generasi milenial Carlson (2008) menyatakan hal yang sama dengan Mannheim terkait dengan karakteristik generasi milenial berdasarkan tahu lahirnya, Carlson menyebutkan periode tahun 1983 sampai tahun 2001 merupakan rentang tahun kelahiran para generasi milenial. Sedangkan Howe & Strauss (2000) memberikan ilustrasi terkait dengan generasi milenial adalah mereka yang lahir pada tahun 1982 sudah mulai masuk prasekolah dan mereka yang pada tahun 2000 lulus dari Sekolah Menengah Atas (SMA) dan istilah ini mulai populer pada tahun 1987. H3: Marketing research berpengaruh dominan dalam menumbuhkan jiwa entrepreneur pada generasi milenial Jiwa Entrepreneur 78 informasi yang disajikan harus yang berkualitas (no hoax) dengan tampilan informasi yang menarik (Hidayatullah dkk., 2020). Generasi Milenial METODE Pendekatan metode yang digunakan adalah deskriftif kuantitatif, yaitu untuk menganalisis dan mengetahui bagaimana peran marketing intelligence dalam menumbuhkan jiwa entrepreneur pada generasi milenial saat ini. Pendekatan ini akan berupaya untuk mengkaji dan mengurai keterkaitan antar variabel yang digunakan dalam penelitian dengan menggunakan tools descriptive analysis dan multiple linear regression analysis. Hasil penelitian dari Lancaster & Stillman (2003) menghasilkan karakteristik yang melekat dalam diri generasi milenial, Lancaster & Stillman menyatakan bahwa generasi milenial memiliki sikap yang realistis, sangat menghormati perbedaan yang ada, lebih senang bekerjasama dalam tim serta memiliki sikap optimis dan percaya diri yang tinggi dalam setiap hal yang dilakukan dan ternyata generasi milenial memiliki kepercayaan terhadap nilai-nilai moral dan sosial, namun sangat pragmatis dalam menyelesaikan masalah. Penelitian ini mengunakan lokus Perguruan Tinggi yang berada di Kota Malang dengan jumlah 150 mahasiswa sebagai sampel yang tersebar di seluruh Perguruan Tinggi di Kota Malang dengan mengunakan google form, sedangkan pengambilan sampelnya mengunakan teknik multistage random sampling. Instrumen yang digunakan ini sudah melalui uji kelayakan mengunakan uji validitas dan reliabilitas dengan merujuk pada pendapat Sugiono P.D, (2011) yang mengatakan bawa apabila ingin mengetahui sejauh mana instrument yang digunakan dapat mengukur sesuatu yang seharusnya diukur maka dapat mengunakan uji validitas. Dari semua item (24) yang telah di uji menggunakan uji validitas, menyatakan bahwan semua item dalam kuesioner adalah valid karena nilai probabilitas yang dimiliki lebih kecil dari pada 0,05. Sedangkan uji reliabilitas jika merujuk pada pendapat Sedangkan berdasarkan hasil penelitian Hidayatullah dkk. (2018) dengan judul perilaku generasi milenial dalam mengunakan aplikasi go-food menyatakan bahwa salah satu karakterikstik generasi milenial yang berpengaruh secara signifikan adalah kemampuan generasi milenial dalam memanfaatkan teknologi dan informasi dalam setiap aktifitas yang dilakukan sehari-hari dimulai dari mereka bangun tidur sampai kembali tidur. Disamping itu untuk menjamin kepuasan kelompok generasi milenial, 79 Nurgiyantoro & Gunawan (2000) yang mengatakan bahwa nilai koefisien 0,60 menjadi syarat minimum sebuah instrument akan dinyatakan reliable. Dari semua variabel yang telah di uji mengunakan uji reliabilitas telah dinyatakan reliable. Sehingga instrument dinyatakan layak untuk digunakan. Untuk sebaran responden berdasarkan Perguruan Tingginya, maka sebaran responden tertinggi untuk penelitian marketing intelligence dalam menumbuhkan jiwa entrepreneur pada generasi milenial yang dilakukan di Kota Malang berada di Politeknik Negeri Malang dengan jumlah sebarannya sebanyak 18,67%, kemudian sebaran terbanyak berikutnya berada di Universitas Brawijaya sebanya 14,00%, sedangkan sebaran responden yang paling sedikit berada di Universitas Widyagama Malang yaitu sebanyak 2,67%. METODE Dari sebaran responden menunjukkan bahwa sebagian besar berada di Politeknik Negeri Malang. HASIL DAN PEMBAHASAN Deskripsi Responden Gambar 1 : Jesnis Kelamin Responden Sumber: data primer (diolah), 2020 HASIL DAN PEMBAHASAN Deskripsi Responden Gambar 1 : Jesnis Kelamin Responden umber: data primer (diolah), 2020 Marketing Research Gambar 3 : Deskripsi Marketing Research berdasarkan Indikatornya Sumber: data primer (diolah), 2020 Gambar 1 : Jesnis Kelamin Responden Sumber: data primer (diolah), 2020 Gambar 1 : Jesnis Kelamin Responden Sumber: data primer (diolah), 2020 Responden dalam penelitian marketing intelligence dalam menumbuhkan jiwa entrepreneur pada generasi milenial yang dilakukan di Kota Malang ini didominasi oleh responden yang berjenis kelamin laki-laki sebanyak 58,67%, sedangkan sisanya adalah responden perempuan yang sebanyak 41,33% responden. Gambar 3 : Deskripsi Marketing Research berdasarkan Indikatornya Sumber: data primer (diolah), 2020 Nilai rata-rata untuk indikator marketing research dalam penelitian marketing intelligence dalam menumbuhkan jiwa entrepreneur pada generasi milenial yang dilakukan di Kota Malang adalah sebesar 4,12. Nilai ini memberi gambaran bahwa pada saat responden menerapkan prinsip marketing intelligence yang dalam hal ini adalah marketing research dapat membantu dalam menumbuhkan jiwa entrepreneur mereka, hal ini dikarenakan responden akan mendapatkan informasi tentang gambaran situasi pasar sehingga dapat membagun strategi dan program pemasaran yang cocok untuk diimplementasikan dalam memasarakan produknya kepada konsumennya maupun calon konsumennya. Kegiatan marketing research dapat dilakukan secara tim maupun seorang diri hal ini Gambar 2 : Deskripsi Responden Berdasarkan sebaran di Perguruan Tinggi Sumber: data primer (diolah), 2020 Gambar 2 : Deskripsi Responden Berdasarkan sebaran di Perguruan Tinggi Sumber: data primer (diolah), 2020 80 80 tergantung tingkat kebutuhan dan biaya yang dimilikinya. Menumbuhkan Jiwa Entrepreneu tergantung tingkat kebutuhan dan biaya yang dimilikinya. Menumbuhkan Jiwa Entrepreneur Menumbuhkan Jiwa Entrepreneur Gambar 5 : Deskripsi dalam Menumbuhkan Jiwa Entrepreneur berdasarkan Indikatornya Sumber: data primer (diolah), 2020 Database Marketing Analisis Regresi Linier Berganda Analisis Regresi Linier Berganda Tabel 2. Rekapitulasi Hasil Analisis Regresi Linier Berganda Sumber: data primer (diolah), 2020 B β X1 Marketing research 0,469 0,565 7,127 0,000 X2 Database marketing 0,174 0,191 2,417 0,017 1,511 0,495 74,086 0 150 1,976 3,06 Jiwa entrepreneur Variabel Keterangan Y Ttabel Ftabel Fhitung Sig. F N Konstanta R2 Koefisien Regresi thitung Sig. Tabel 2. Rekapitulasi Hasil Analisis Regresi Linier Berganda Gambar 6 : Uji Heterokedastisitas Sumber: data primer (diolah), 2020 Uji heterokedastisitas pada gambar 6 di atas menunjukkan jika data observasi yang digunakan tidak terjadi problem heterokedastisitas karena penyebaran titik- titik yang ada menyebar secara sempurna baik diatas maupun di bawah sumbu Y serta pola penyebaranya tidak teratur. Uji Asumsi Klasik Normalitas Uji Asumsi Klasik Normalitas Gambar 7 : Uji Normalitas Sumber: data primer (diolah), 2020 Gambar 7 : Uji Normalitas Sumber: data primer (diolah), 2020 Dalam regresi linier berganda diketahui tidak terjadi gejala multikolinieritas setalah dilakukan uji multikolinieritas seperti yang terdapat pada tabel 1 di atas. Uji Asumsi Klasik Heterokedastisitas Gambar 6 : Uji Heterokedastisitas Sumber: data primer (diolah), 2020 Gambar 6 : Uji Heterokedastisitas Sumber: data primer (diolah), 2020 Gambar 7 : Uji Normalitas Gambar 7 : Uji Normalitas Sumber: data primer (diolah), 2020 Gambar 7 : Uji Normalitas Sumber: data primer (diolah), 2020 j Sumber: data primer (diolah), 2020 Data dalam penelitian ini berdistribusi secara normal, hal ini dapat dilihat dari norma probability plots dimana titik-titik tersebut mengikuti garis diagonal yang terdapat pada garis kurva normal pada gambar 7 di atas. Analisis Regresi Linier Berganda Database Marketing Database Marketing Gambar 4 : Deskripsi Database Marketing berdasarkan Indikatornya Sumber: data primer (diolah), 2020 Gambar 5 : Deskripsi dalam Menumbuhkan Jiwa Entrepreneur berdasarkan Indikatornya Sumber: data primer (diolah), 2020 Gambar 4 : Deskripsi Database Marketing berdasarkan Indikatornya Sumber: data primer (diolah), 2020 Nilai rata-rata untuk indikator upaya dalam menumbuhkan jiwa entrepreneur dalam penelitian marketing intelligence dalam menumbuhkan jiwa entrepreneur pada generasi milenial yang dilakukan di Kota Malang adalah sebesar 4,15. Nilai ini memberi gambaran bahwa responden dalam penelitian ini dapat menumbuhkan jiwa entrepreneur mereka bisa melalui berbagai cara diantaranya adalah dengan mengembangkan teknologi yang sudah ada saat ini, kadangkala jiwa entrepreneur mereka tumbuh karena mereka mendapatkan pengetahuan yang baru tentang entrepreneurship di bangku perkuliahan, sehingga apabila responden telah melakukan aktifitas marketing intelligence yang terdiri dari marketing research dan database marketing akan menjadi bekal yang sangat berarti bagi responden untuk memulai usahanya atau bisa juga digunakan untuk memperbaiki layanan jasa ataupun produk yang sudah ada dengan memberikan terobosan-terobosan yang baru sehingga dapat menghemat waktu dan biaya. Namun tumbuh dan berkembangnya jiwa entrepreneur seseorang bisa datang dari mana saja baik faktor internal maupun eksternal dari yang bersangkutan Nilai rata-rata untuk indikator database marketing dalam penelitian marketing intelligence dalam menumbuhkan jiwa entrepreneur pada generasi milenial yang dilakukan di Kota Malang adalah sebesar 4,07. Nilai ini memberi gambaran bahwa pada saat responden menerapkan prinsip marketing intelligence yang dalam hal ini adalah database marketing dapat membantu dalam menumbuhkan jiwa entrepreneur-nya, karena setelah responden melakukan pengumpulan data melalui marketing research, maka data tersebut akan dijadikan dasar dalam melakukan prediksi tentang prilaku konsumen sehingga nantinya hal tersebut akan menjadi dasar untuk menganalisis konsumen yang potensial terhadap produk atau jasa yang akan dilakukan, selain untuk memperoleh konsumen potensial. Hal ini juga dilakukan untuk menganalisis target penjualan dimasa yang akan datang. Database marketing ini menjadi sangat penting sebagai dasar dalam pengambilan keputusan, dan kegiatan database marketing dapat memanfaatkan aplikasi dan teknologi yang ada saat ini, diantaranya adalah SaaS (Software as a Service) atau OLAP (Online Analytical Processing) dll. 81 81 Uji Asumsi Klasik Multikolinieritas Tabel 1. Uji Multikolinieritas Sumber: data primer (diolah), 2020 Variable VIF Hasil Keterangan Marketing research (X1) 1,852 < 10 Tidak terjadi multikolinieritas Database marketing (X2) 1,852 < 10 Tidak terjadi multikolinieritas Uji Asumsi Klasik Multikolinieritas Uji Kelayakan Model Uji F dalam penelitian ini digunakan untuk menagalisis uji model dengan menggukanan nilai standar probabilitas di bawah 0,05. Dari hasil uji F yang dilakukan maka diperoleh nilai probabilitasnya sebasesar 0,000 sehingga diperoleh perbandingan (0,000<0,05), dengan demikian maka model regresi linier yang diestimasi layak untuk digunakan karena derajat signifikansi yang ditentukan lebih besar dari nilai probabilitas yang diperoleh. Pada tabel 2 dapat diketehui bahwa uji t untuk variabel database marketing memiliki pengaruh positif dan pengaruhnya signifikan dalam menumbuhkan jiwa entrepreneur di kalangan generasi milenial. Oleh karena itu, maka hipotesis yang kedua teruji atau dapat diterima karena nilai probabilitas dari database marketing sebesar 0,017 (< 0,05) dari nilai yang ditentukan. Hasil ini tidak dapat dipisahkan dari kemampuan responden yang merupakan generasi milenial dalam melakukan aktivitas database marketing, dimana mereka tumbuh dan berkembang dalam ekosistem digital yang sangat friendly dengan perkembangan teknologi informasi membuat para generasi milenial menjadi lebih kreatif dan dinamis dalam melakukan aktivitas database marketing, sehingga data/informasi yang sudah terkumpul dapat dijadikan dasar menganalisis konsumen yang potensial untuk mencapai target penjualan dan memenangkan persaingan dimasa yang akan datang. Uji Asumsi Klasik Autokorelasi Metode Durbin Waston (DW) digunakan untuk menguji ada tidaknya gejala autokorelasi dengan mengunakan rumus: du<DW<(4-du). Dengan mengunakan pendekatan Durbin Waston (DW) ini diperoleh bahwa dalam model regresi yang digunakan bebas dari gejala autokorelasi dengan nilai : 1,760 <1,927< 2,240. Persamaan regresi linier berganda yang diperoleh adalah Y = 1,511 + 0,469X1 + 0,174X2 dengan nilai R Square-nya (koefisien determinasi) dalam penelitian ini sebesar 82 82 yang nantinya akan dijadikan dasar untuk membagun strategi dan program pemasaran yang cocok untuk diimplementasikan dalam kegiatan memasarkan produk atau jasanya kepada konsumennya maupun calon konsumennya, sehingga akan mampu bersaing secara kompetitif dengan para pesaingnya. Hal ini akan menjadi stimulus yang luar biasa bagi generasi milenial untuk berani dalam berwirausaha, karena dengan bekal informasi dan data yang diperoleh dari proses marketing research yang dilakukan akan menjadi bekal untuk memberikan kepercayaan diri yang tinggi bagi generasi milenial dalam berwirausaha. 0,495. Ini memberikan pengertian bahwa pada saat variabel independennya yang terdiri dari marketing research (X1) dan database marketing (X2) jika bersamaan akan mampu memberikan kontribusi dalam menumbuhkan jiwa entrepreneur pada generasi milenial dengan nilai 49,5%, sehingga dapat dikatakan bahwa kontribusi ini berada dalam kategori sedang. 0,495. Ini memberikan pengertian bahwa pada saat variabel independennya yang terdiri dari marketing research (X1) dan database marketing (X2) jika bersamaan akan mampu memberikan kontribusi dalam menumbuhkan jiwa entrepreneur pada generasi milenial dengan nilai 49,5%, sehingga dapat dikatakan bahwa kontribusi ini berada dalam kategori sedang. Pengujian Hipotesis Uji T dalam penelitian ini digunakan untuk melakukan pengujian hipotesis dengan mengunakan nilai probabilitas di bawah 0,05, uji t ini untuk mengetahui apakah terdapat pengaruh yang signifikan atau tidak dari independent variable terhadap dependent variable. Output uji t yang terdapat pada tabel 2 dapat dilihat bahwa marketing research memiliki pengaruh positif dan pengaruhnya signifikan dalam menumbuhkan jiwa entrepreneur di kalangan generasi milenial. Oleh karena itu, maka hipotesis yang pertama teruji atau dapat diterima karena nilai probabilitas dari marketing research adalah sebesar 0,000 (<0,05) dari nilai yang ditentukan. Hasil ini tidak dapat dipisahkan dari kemampuan responden yang dalam hal ini adalah generasi milenial dalam melakukan aktivitas marketing research karena mereka adalah geranasi yang sangat fleksibel dan melek teknologi informasi, dimana mereka dengan kemampuan fleksibelitasnya tersebut dapat dengan mudah melakukan marketing research secara mandiri untuk mendapatkan informasi tentang gambaran situasi pasar Berdasarkan nilai koefisien beta (β) dapat diketahui bahawa variabel yang paling dominan pengaruhnya dalam menumbuhkan jiwa entrepreneur pada generasi milenial adalah variabel marketing research dibandingkan dengan variabel database marketing. Oleh karena itu, maka hipotesis yang ketiga dalam penelitian ini teruji atau dapat diterima karena marketing research memiliki nilai koefisien beta (β) paling tinggi sebesar 0,565 jika dibandingkan dengan nilai koefisien beta (β) database marketing yang 83 Dukungan Marketing Intelligence Pada Strategi Pemasaran Seminar Nasional Informatika (SEMNASIF). http://jurnal.upnyk.ac.id/index.php/sem nasif/article/view/1365 hanya memiliki nilai sebesar 0,191. Selain itu argumentasi yang mendasari marketing research dapat berpengaruh dominan dalam menumbuhkan jiwa entrepreneur pada generasi milenial ini tidak terlepas dari tingkat urgensi dari marketing research dalam aktifitas marketing intelligence karena marketing research merupakan langkah awal dalam rangkaian marketing intelligence berupa pengumpulan informasi dan data yang valid dengan cara yang sistematis, kreatif dan inovatif, karena informasi dan data yang diperoleh ini nantinya akan dijadikan dasar analisa untuk menetukan strategi dan program dalam rangka memenangkan persaingan dalam bidang usahanya. Badan Pusat Statistik. (2018). Statistik Gender Tematik: Profil Generasi Milenial Indonesia. Kementerian Pemberdayaan Perempuan dan Perlindungan Anak. Burns, A. C., Veeck, A., & Bush, R. F. (2017). Marketing Research (Ninth edit). New York : Pearson Education. Carlson, E. (2008). Who are the Lucky Few? In The Lucky Few (pp. 1–10). Springer. https://link.springer.com/chapter/10.100 7%2F978-1-4020-8541-3_1 Kesimpulan Guoxiang, L., & Zhiheng, Q. (2013). Data Mining Applications in Marketing Strategy. 2013 Third International Conference on Intelligent System Design and Engineering Applications, 518–520. https://doi.org/10.1109/ISDEA.2012.12 6 Berikut ini adalah kesimpulan yang diperoleh dari hasil analisa: 1) marketing research memiliki pengaruh yang signifikan dan pengaruh tersebut positif dalam menumbuhkan jiwa entrepreneur pada generasi milenial; 2) database marketing memiliki pengaruh yang signifikan dan pengaruh tersebut positif dalam menumbuhkan jiwa entrepreneur pada generasi milenial; dan 3) faktor utama bagi generasi milenial dalam menumbuhkan jiwa entrepreneur-nya adalah aktivitas marketing research. Sampai saat ini masih tidak banyak para generasi milenial maupun pengusaha menegah kebawah yang melakukan marketing intelligence baik pada saat memulai usaha maupun untuk mengembangkan usahanya, padahal aktifitas marketing intelligence ini dapat dilakukan secara mandiri (sendiri) sehingga biaya yang dibutuhkan relatif kecil. 6 Hatta, G. (2008). Pedoman Manajemen Kesehatan di sarana pelayanan kesehatan. Revisi, Jakarta Universitas Indonesia. Hidayatullah, S., Rachmawati, I. K., Aristanto, E., Waris, A., & Patalo, R. G. (2020). Peran Sistem Informasi Pemasaran, Kualitas Pelayanan dan Entrepreneurial marketing serta Kepuasan Terhadap Loyalitas Generasi Milenial Berkunjung ke Tempat Wisata. Jurnal Ilmiah Bisnis dan Ekonomi Asia, 14(1), 74–83. https://jurnal.stie.asia.ac.id/index.php/ji beka/article/view/184 Hidayatullah, S., Waris, A., Devianti, R. C., Sari, S. R., Wibowo, I. A., & PW, P. M. (2018). Perilaku Generasi Milenial dalam Menggunakan Aplikasi Go-Food. Jurnal Manajemen dan Kewirausahaan, 6(2), 240–249. https://doi.org/10.26905/jmdk.v6i2.256 0 DAFTAR PUSTAKA Ambarwati, A., & Sobari, I. S. (2020). Membangun Jiwa Kewirausahaan Di Era Milenial Bagi ahasiswa Institut Stiami Kampus Tangerang Selatan. Jurnal Komunitas : Jurnal Pengabdian Kepada Masyarakat, 2(2), 140–144. https://doi.org/10.31334/jks.v2i2.736 Amborowati, A., & Suyanto, M. (2015). Studi Howe, N., & Strauss, W. (2000). Millennials rising: The next great generation. Vintage. https://doi.org/10.31334/jks.v2i2.736 Amborowati, A., & Suyanto, M. (2015). Studi 84 Kurniawan V. (2020). Definisi dan Sistem Kerja Marketing Intelligence \| BINUS UNIVERSITY BANDUNG - Kampus Teknologi Kreatif. https://binus.ac.id/bandung/2020/04/def inisi-dan-sistem-kerja-marketing- intelligence/ Kottler, P., & Armstrong, G. (2008). Prinsip- Prinsip Pemasaran. Ed. 12, Jilid 1. Jakarta: Erlangga. Lancaster, L. C., & Stillman, D. (2003). When generations collide: Who they are, why they clash, how to solve the generational puzzle at work. HarperBusiness New York, NY. https://pdfs.semanticscholar.org/bd24/8f 640f066d6a81b470f93aa3fe49295c852 b.pdf p Li, W. (2006). Entrepreneurial Intention Among International Students: Testing a Model of entrepreneurial intention. Paper Presented at the USASBE Small Business Advancent National Center. Luo, Y., Cai, Q. R., Xi, H. X., Liu, Y. J., & Yu, Z. M. (2012). Telecom Customer Segmentation With K-means Clustering. ICCSE 2012 - Proceedings of 2012 7th International Conference on Computer Science and Education. https://doi.org/10.1109/ICCSE.2012.62 95158 Mannheim, K. (1952). The Sociological Problem of Generation. Essays on the Sociology of Knowledge. Nurgiyantoro, B., & Gunawan, M. (2000). Statistik Terapan untuk Ilmu-Ilmu Sosial. Gadjah Mada University Press, Yogyakarta. Sebastian, Y., Amran, D., & Lab, Y. (2016). Generasi Langgas Millenials Indonesia. Gagas Media. Sugiono P.D. (2011). Metode Penelitian Pendidikan (pendekatan kuwantitatif, kuwalitatif,R&D). In Alfabbeta Pres.
https://openalex.org/W2000007305	https://repub.eur.nl/pub/61379/OABMC_1471-2407-14-429.pdf	English	null	MicroRNA response to hypoxic stress in soft tissue sarcoma cells: microRNA mediated regulation of HIF3α	BMC cancer	2,014	cc-by	10,283	RESEARCH ARTICLE Open Access Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 * Correspondence: e.wiemer@erasmusmc.nl 1Department of Medical Oncology, Erasmus University Medical Center – Erasmus MC Cancer Institute, Rotterdam, the Netherlands 3Department of Medical Oncology, Erasmus MC Cancer Institute, Erasmus University Medical Center, Rm Be422, Wytemaweg 80, 3015 CN Rotterdam, the Netherlands Full list of author information is available at the end of the article MicroRNA response to hypoxic stress in soft tissue sarcoma cells: microRNA mediated regulation of HIF3α Caroline MM Gits1, Patricia F van Kuijk1, Jonneke CWM de Rijck1, Nikky Muskens1, Moniek BE Jonkers1, Wilfred F van IJcken2, Ron HJ Mathijssen1, Jaap Verweij1, Stefan Sleijfer1 and Erik AC Wiemer1,3* Abstract Background: Hypoxia is often encountered in solid tumors and known to contribute to aggressive tumor behavior, radiation- and chemotherapy resistance resulting in a poor prognosis for the cancer patient. MicroRNAs (miRNAs) play a role in the regulation of the tumor cell response to hypoxia, however, not much is known about the involvement of miRNAs in hypoxic signalling pathways in soft tissue sarcomas (STS). Method: A panel of twelve STS cell lines was exposed to atmospheric oxygen concentrations (normoxia) or 1% oxygen (hypoxia) for up to 48 h. Hypoxic conditions were verified and miRNA expression profiles were assessed by LNA™oligonucleotide microarrays and RT-PCR after 24 h. The expression of target genes regulated by hypoxia responsive miRNAs is examined by end-point PCR and validated by luciferase reporter constructs. Results: Exposure of STS cell lines to hypoxic conditions gave rise to upregulation of Hypoxia Inducible Factor (HIF) 1α protein levels and increased mRNA expression of HIF1 target genes CA9 and VEGFA. Deregulation of miRNA expression after 24 h of hypoxia was observed. The most differentially expressed miRNAs (p < 0.001) in response to hypoxia were miR-185-3p, miR-485-5p, miR-216a-5p (upregulated) and miR-625-5p (downregulated). The well-known hypoxia responsive miR-210-3p could not be reliably detected by the microarray platform most likely for technical reasons, however, its upregulation upon hypoxic stress was apparent by qPCR. Target prediction algorithms identified 11 potential binding sites for miR-485-5p and a single putative miR-210-3p binding site in the 3’UTR of HIF3α, the least studied member of the HIF family. We showed that HIF3α transcripts, expressing a 3’UTR containing the miR-485-5p and miR-210-3p target sites, are expressed in all sarcoma cell lines and upregulated upon hypoxia. Additionally, luciferase reporter constructs containing the 3’UTR of HIF3α were used to demonstrate regulation of HIF3α by miR-210-3p and miR-485-5p. Conclusion: Here we provide evidence for the miRNA mediated regulation of HIF3α by hypoxia responsive miRNAs in STS, which may help to tightly regulate and fine-tune the hypoxic response. This provides a better insight into the mechanisms underlying the hypoxic response in STS and may ultimately yield information on novel prognostic and predictive markers or targets for treatment. Keywords: miRNA, Hypoxia, HIF3α, Soft tissue sarcomas, miR-210-3p, miR-485-5p © 2014 Gits et al.; licensee BioMed Central Ltd. Abstract This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Page 2 of 13 Cell culture A panel of 12 soft tissue sarcoma cell lines was used, consisting of the fibrosarcoma cell lines HT1080, SW684 and BG-8, the liposarcoma cell line SW872, the leiomyo- sarcoma cell lines SK-UT-1 and SK-LMS-1, the synovial sarcoma cell line SW982 and the rhabdomyosarcoma cell lines RH30, A204, A673, RD and SJCRH30. Note that the A673 (ATCC®CRL-1598) cell line is now listed on the ATCC website as Ewing sarcoma. The myxoid fibrosarcoma cell line BG-8 was a kind gift from Dr. A. Carnero (Centro Nacional de Investigaciones Oncológicas, Madrid, Spain); HT1080, SK-UT-1, RD and RH-30 were acquired from Dr. M. Debiec-Rychter (KU Leuven, Leu- ven, Belgium) and SJCRH30 was obtained from Dr. L. Alberti (Université de Lyon, Centre Léon Bérard, Lyon, France). All further cell lines were obtained from the ATCC, Rockville, MD, USA. The cell lines were cultured in RPMI 1640/GlutaMAX (Invitrogen, Bleiswijk, The Netherlands) supplemented with 10% fetal bovine serum, at 37°C, 5% CO2. Tumor hypoxia and the accompanying biochemical and cellular changes in the tumor cells contribute to ag- gressive tumor behavior, and radiation- and chemother- apy resistance in many tumor types, including STS [12,13]. STS is a group of relatively rare tumors of mes- enchymal origin in which currently more than 50 differ- ent histological subtypes are recognized. STS have been reported to present with hypoxic areas resulting in in- creased tumor proliferation, distant metastases and shorter overall disease survival [14,15]. Given the poor prognosis STS patients face with, a median overall sur- vival of only 1 year for patients presenting with meta- static disease, more insight is needed into the effects of hypoxia in STS and how hypoxia is regulated in STS. Background oncomiR, having either a tumor suppressive or an onco- genic role [17-20]. MiRNA expression can be influenced by changes in the tumor microenvironment [21,22] in- cluding conditions like tumor hypoxia [23-26]. The miRNA response to hypoxia, which is often cell-type specific, suggests miRNAs are intimately involved in the cellular reaction to hypoxia [23,24,26] and some are in turn involved in feedback mechanisms that regulate HIF [27-32]. Interestingly, many hypoxia responsive miRNAs (HRMs) are also upregulated in cancer suggesting they also function in tumorigenesis and/or tumor progression [26]. As opposed to the role of HRMs in the major can- cer types relatively little is known about the role of hyp- oxic miRNAs in STS. In this study, we describe the miRNA response to hypoxic stress in sarcoma cell lines revealing the involvement of miRNAs in the hypoxic signalling in STS. Hypoxia is a condition of reduced oxygen tension that is often encountered in solid tumors including soft tissue sarcomas (STS) when they outgrow their blood supply. The tumor cells respond to the hypoxic conditions by inducing genes that regulate various biological processes including cellular proliferation, apoptosis, metabolism, angiogenesis and migration [1]. A key element in the re- sponse to hypoxia is the upregulation of HIF1, a member of the family of hypoxia inducible factors (HIFs) also comprising HIF2 and HIF3 [2]. These transcription fac- tors consist of a tightly controlled alpha (HIFα), and a constitutively activated beta subunit (HIFβ or ARNT family). Upon hypoxia, the alpha subunit is stabilized, translocated to the nucleus, where it dimerizes to the beta subunit. The HIF complex binds to hypoxia respon- sive elements (HREs) in gene promoters, thereby in- ducing the transcription of genes important for the adaption to low oxygen concentrations. HIF1α and HIF2α have similar protein domain structures, are regu- lated in a similar fashion and share several target genes [3], although their expression patterns differ in some tis- sues and some unique target genes have been identified [4]. The function(s) of HIF3α and the way it is regulated are still largely unknown. Complicating the study of HIF3α is the fact that alternative splice variants are rec- ognized coding for at least six isoforms, with structural and functional differences [5-7]. HIF3α is regulated by HIF1 at the transcriptional level and exerts an inhibitory effect on HIF1α or HIF-dependent gene regulation, in a cell-type specific manner [5-11]. cDNA-synthesis Total RNA was isolated using RNAbee (Tel Test Inc., Friendswood, TX, USA) according to the manufacturer’s recommendation. The RNA concentration and quality were determined on a Nanodrop-1000 (Nanodrop Tech- nologies, Wilmington, DE, USA). cDNA synthesis RNA (1 μg) was reversed transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Bio- systems) according to the manufacturer’s protocol. The resulting cDNA was used to perform qPCR and end- point PCR analysis. qPCR analysis of miRNA qPCR analysis of miRNA MiR-210-3p expression was examined using TaqManW MicroRNA Assays (Applied Biosystems, Nieuwerkerk aan den IJssel, The Netherlands). In brief, total RNA (50 ng) was reverse transcribed using specific miRNA primers and the TaqManW MicroRNA Reverse Tran- scription Kit (Applied Biosystems). The resulting cDNA was used as input in a quantitative real-time PCR (qPCR) using the miRNA specific primer/probe mix to- gether with the TaqManW Universal PCR Master Mix No AmpErase® UNG (Applied Biosystems) according to manufacturer’s protocol. qPCR data were analyzed with Hypoxic experiments h ll l The 12 cell lines comprising the soft tissue sarcoma panel were cultured under standard conditions and seeded in 6 well culture plates or 9 cm cell culture dishes (miRNA profiling experiment) in duplicate at such a cell density that 50% confluency was reached after 48 h. Subsequently half of the culture plates was kept under standard, normoxic (21% O2) culture condi- tions whereas the other half was transferred to a hypoxic incubator and cultured at 37°C, 5% CO2, 1% O2 for 6 h to 48 h as indicated. After the appropriate normoxic/ hypoxic incubation period cells were rapidly washed with ice-cold PBS after which total RNA was isolated MicroRNAs (miRNAs) are small (~23 nt) non-protein coding RNA molecules that negatively regulate gene ex- pression. MiRNAs play essential roles in a wide variety of cellular processes, such as differentiation, cell cycle regulation, metabolism, apoptosis and stem cell main- tenance and are associated with various diseases includ- ing cancer [16]. Most cancers display a deregulated miRNA expression profile when compared to relevant normal tissues. Some miRNAs may function as an Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Page 3 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 (three separate wells/9 cm dish) and total protein lysates (three separate wells) were prepared. (three separate wells/9 cm dish) and total protein lysates (three separate wells) were prepared. SDS software (version 2.4, Applied Biosystems). MiRNA expression was normalized using RNU43 expression and the comparative CT-method [34]. End-point PCR analysis cDNA (50 ng) was used to perform endpoint PCR using 1 × Green Go Taq Flexi Buffer, 1.25U Go Taq Flexi poly- merase, 200 μM dNTP mix, 1.5 mM MgCl2 and 300 nM forward and reverse primer of HIF3α (Additional file 1: Table S1a) on a thermal cycler using the following PCR- program: 2 min 95°C, 30-40 cycles of (45 sec 95°C, 45 sec 60-65°C, 45 sec 72°C), 5 min 72°C. PCR-products were analyzed on a 1.5% agarose gel in 0.5 × TBE buffer with 0.5 μg/ml ethidium bromide. qPCR analysis of mRNAs MiRNA profiling was performed essentially as previously described [33]. In brief, 1 μg total RNA was fluorescently labeled with Cy3 using the ULS™aRNA Labeling Kit (Kreatech Diagnostics, Amsterdam, The Netherlands). Labeled RNA was hybridized with locked nucleic acid (LNA™) modified oligonucleotide capture probes (Exi- qon) spotted in duplicate on Nexterion E slides. The capture probe set (based on miRBase version 10, annota- tion version 13) contains 1344 probes of which 725 are capable of detecting human miRNAs. Hybridized slides were scanned and median spot intensity was determined using ImaGene software (BioDiscovery Inc., Hawthorne, CA, USA). After background subtraction, expression values were quantile normalized using R software (http://cran.r-project.org; version 2.15.2, release oct. 2012), low expressed miRNAs and obvious outliers/bad spots were removed, and duplicate spots were averaged. Outliers/bad spots were identified by determining the average expression value of each miRNA and examining whether duplicate spots differ from the average (>2 × greater or >5 × smaller). Flagged spots were checked by eye to determine why the signal is aberrant e.g. a speck of dust or printing error. Only in case of an apparent technical error, noted in less than 1% of the spots, a spot was discarded. For each expression value, the ratio to the geometric mean of the samples was log2 transformed. These values were used to determine differentially expressed miRNAs. Hierarchical clustering analyses were performed in Spotfire (Spotfire DecisionSite 9.1, Tibco Software, Somerville, MA. USA). cDNA (45 ng) was used to perform qPCR using the pri- mer/probe mix from the TaqManW Gene Expression As- says of human VEGFA (assay ID Hs00900055_m1) and CA9 (assay ID Hs00154208_m1) with exon spanning probes, and TaqManW Universal PCR Master Mix using the 7500 Fast Real-Time PCR system (all Applied Bio- systems) according to the manufacturer’s protocol. HPRT was used as a housekeeper gene for normalization purposes using the comparative CT-method. qPCR data was analyzed with SDS software (Applied Biosystems). Protein extraction C ll h Cells were harvested in MCB lysis buffer (50 mM Tris– HCl pH 7.5, 50 mM NaCl, 10% glycerol, 1% NP-40, 0.5% Na-deoxycholate, 20 mM NaF) supplemented with a cocktail of protease and phosphatase inhibitors. Lysates were thoroughly vortexed and further lysed by two sub- sequent freeze-thaw cycles using liquid nitrogen. Cell debris was spun down and protein concentration was determined by a Bradford assay (BioRad, Veenendaal, The Netherlands). Western blotting To determine whether the putative promoter regions of miR-216a-5p, miR-185-3p and miR-625-5p contain hyp- oxia responsive elements (HRE) we downloaded 600 bp of 5’ flanking sequences for each of the three miRNA genes from the Ensembl Genome Browser release 72 (http://www.ensembl.org). We scanned the flanking se- quences for the presence of consensus HREs (A/GCGTG) and imperfect HREs. Mimic transfections MiRIDIAN microRNA mimics (Thermo Scientific, Etten- Leur, Netherlands) of hsa-miR-210-3p and hsa-miR-485- 5p and miRIDIAN microRNA Mimic Negative Control #1 (Thermo Scientific) were transfected in a final concen- tration of 50 nM using DharmaFECT1 transfection reagent (Thermo Scientific) 24 h after cell seeding. Transfection efficiency was optimized to >95% using fluorescent mimics. Western blotting Twenty μg of total protein was subjected to SDS-PAGE. Proteins were transferred to a PVDF membrane followed by blocking of the membrane in 5% non-fat dry milk in PBS-Tween (Phosphate buffered saline, 0.05% (v/v) Tween 20) to prevent non-specific antibody binding. Primary and secondary antibody incubations were carried out in the same buffer using anti-HIF1α (610958, mouse monoclo- nal, 1:500, BD Biosciences, Breda, The Netherlands), anti- Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Page 4 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 HIF3α (ab10134, rabbit polyclonal, 1:500, Abcam, Cam- bridge, UK) and anti-β-Actin (AC-15, mouse monoclonal, 1:5000, Sigma-Aldrich, Zwijndrecht, The Netherlands) as a loading control. As secondary antibody HRP- conjugated goat-anti-mouse (1:10000, Santa Cruz Bio- technology, Heidelberg, Germany) or goat-anti-rabbit (1:10000, Jackson Immunoresearch, Suffolk, UK) anti- bodies were used. Antibody incubations were followed by enhanced chemoluminescence (Supersignal West Pico Chemiluminescent Substrate, Thermo Scientific) and visualized on film (Amersham Hyperfilm ECL, GE Healthcare, Diegem, Belgium). Reporter Assay System, Promega) was added, immedi- ately followed by quantitation of the firefly luciferase activity on a luminometer (Aspekt Fluoroskan). Subse- quently, Stop & GloW Reagent (Dual LuciferaseW Reporter Assay System, Promega) was added to the mixture, imme- diately followed by quantitation of the Renilla luciferase activity. Relative luciferase signals of duplicates were aver- aged, values (n = 3) were normalized, and average and standard deviations were calculated. Significant differ- ences in luciferase activity were determined by two- sample t-tests. HIF3α (ab10134, rabbit polyclonal, 1:500, Abcam, Cam- bridge, UK) and anti-β-Actin (AC-15, mouse monoclonal, 1:5000, Sigma-Aldrich, Zwijndrecht, The Netherlands) as a loading control. As secondary antibody HRP- conjugated goat-anti-mouse (1:10000, Santa Cruz Bio- technology, Heidelberg, Germany) or goat-anti-rabbit (1:10000, Jackson Immunoresearch, Suffolk, UK) anti- bodies were used. Antibody incubations were followed by enhanced chemoluminescence (Supersignal West Pico Chemiluminescent Substrate, Thermo Scientific) and visualized on film (Amersham Hyperfilm ECL, GE Healthcare, Diegem, Belgium). Reporter Assay System, Promega) was added, immedi- ately followed by quantitation of the firefly luciferase activity on a luminometer (Aspekt Fluoroskan). Subse- quently, Stop & GloW Reagent (Dual LuciferaseW Reporter Assay System, Promega) was added to the mixture, imme- diately followed by quantitation of the Renilla luciferase activity. Relative luciferase signals of duplicates were aver- aged, values (n = 3) were normalized, and average and standard deviations were calculated. Significant differ- ences in luciferase activity were determined by two- sample t-tests. Cloning Hypoxia responsive miRNAs in soft tissue sarcoma cells In order to examine the miRNA response to hypoxia in sarcomas, a panel of 12 soft tissue sarcoma cell lines was cultured for 0 h, 6 h, 24 h and 48 h under hypoxic (1% oxygen) or normoxic (21% oxygen) conditions. Exposure to hypoxic conditions was verified by the upregulation of HIF1α protein levels, which is one of the earliest markers of hypoxia. HIF1α protein, being degraded under normoxic conditions, is stabilized during hypoxia. As expected, HIF1α protein levels were significantly in- creased upon hypoxic stress (Figure 1A). In all 12 cell lines examined peak HIF1α levels were observed at the 6 h time point which decreased after 24 h and 48 h of hypoxia as shown for 5 representative cell lines (Figure 1A). As a consequence of HIF1α upregulation, the transcription of specific HIF1 target genes like CA9 and VEGFA was slightly induced after 6 h and further increased after 24-48 h of hypoxia, as shown for 5 repre- sentative cell lines (Figure 1B, C). Note that in particular the increase in CA9 varies considerably between cell lines. The carbonic anhydrase CA9 is a member of a zinc metalloenzyme family that catalyses the reversible hydration of carbon dioxide to carbonic acid [35,36]. CA9 is a transmembrane protein which is found overex- pressed in numerous tumor types and induced under hypoxic conditions. Similarly HIF1α stimulates the ex- pression of the vascular endothelial growth factor (VEGFA) which functions as an signal mediating vascu- logenesis and angiogenesis [37]. Fragments of the 3’UTR of HIF3α (HIF3α-short: 817 bp fragment, HIF3α-long: 3807 bp fragment) were PCR amplified from human genomic DNA (Promega, Leiden, The Netherlands) introducing a XhoI (5’-end) and a NotI site (3’-end). The PCR products were cloned in PCR®-Blunt (Invitrogen), followed by a XhoI and NotI restriction and ligation in the psiCHECK™-2 vector (Pro- mega) behind a Renilla luciferase gene. The psiCHECK- 2 vector also contains a firefly luciferase gene, which was used for normalization. The resulting constructs are psi- CHECK2/HIF3α-short and psiCHECK2/HIF3α-long. Site directed mutagenesis (QuickChange II XL, site directed mutagenesis kit, Agilent Technologies, Amsterdam, The Netherlands) was performed to mutate miR-210-3p and miR-485-5p target sites in the 3’UTR of HIF3α in psi- CHECK2/HIF3α-short (Additional file 1: Figure S1). Pri- mer sequences used for cloning and mutagenesis are listed in Additional file 1: Table S1b and S1c. Luciferase assay PsiCHECK-2/HIF3α-short and psiCHECK2/HIF3α-long constructs were transfected using Fugene HD transfec- tion reagent (Promega) 24 h after cell seeding, or when mimics were used, 24 h after miRNA mimic transfec- tion, according to recommendations by the manufac- turer. After 24 hours the cells were washed with PBS and lysed by Passive Lysis Buffer (Dual LuciferaseW Re- porter Assay System, Promega) for 30 min on a shaker platform. Lysates were transferred to white 96-wells mi- croplates. Luciferase Assay Reagent II (Dual LuciferaseW Next, the miRNA expression profile was analyzed using microarrays containing LNA™oligonucleotide capture Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Page 5 of 13 Page 5 of 13 Figure 1 Hypoxia induces HIF1α protein levels and CA9 and VEGFA transcription in soft tissue sarcoma cell lines. (A) HIF1α protein levels are stabilized and increase during hypoxia. Cell lines were cultured under hypoxia (H) or normoxia (N), after the indicated times cell lysates were prepared and analysed by Western blotting for HIF1α expression. HIF1α levels peak at 6 hours of hypoxia. Upon prolonged hypoxia (24 and 48 hours) HIF1α protein levels decrease, but remain increased compared to the HIF1α levels detected in cells cultured under normoxic conditions. β-Actin was used as a loading control. (B,C) mRNA levels of HIF1 target genes CA9 (B) and VEGFA (C) are upregulated during hypoxia (grey bars) compared to normoxia (black bars). CA9 and VEGFA expression was determined by RT-PCR and normalized to the expression of HPRT. Bars indicate average fold change of mRNA expression ± S.D.(n = 2) compared to the levels detected at 0 h which are arbitrarily set at 1. Statistical significance between CA9 and VEGF expression in normoxic and hypoxic samples at the various time points was determined by two-sample t-tests: * = p < 0.05, = p < 0.005, * = p < 0.0005. The complete panel of 12 soft tissue sarcoma cell lines was examined, the results obtained with five representative cell lines are shown. Figure 1 Hypoxia induces HIF1α protein levels and CA9 and VEGFA transcription in soft tissue sarcoma cell lines. (A) HIF1α protein levels are stabilized and increase during hypoxia Cell lines were cultured under hypoxia (H) or normoxia (N) after the indicated times cell lysates a induces HIF1α protein levels and CA9 and VEGFA transcription in soft tissue sarcoma cell lines. Luciferase assay (A) HIF1α protein Figure 1 Hypoxia induces HIF1α protein levels and CA9 and VEGFA transcription in soft tissue sarcoma cell line levels are stabilized and increase during hypoxia. Cell lines were cultured under hypoxia (H) or normoxia (N), after the indicated times cell lysates were prepared and analysed by Western blotting for HIF1α expression. HIF1α levels peak at 6 hours of hypoxia. Upon prolonged hypoxia (24 and 48 hours) HIF1α protein levels decrease, but remain increased compared to the HIF1α levels detected in cells cultured under normoxic conditions. β-Actin was used as a loading control. (B,C) mRNA levels of HIF1 target genes CA9 (B) and VEGFA (C) are upregulated during hypoxia (grey bars) compared to normoxia (black bars). CA9 and VEGFA expression was determined by RT-PCR and normalized to the expression of HPRT. Bars indicate average fold change of mRNA expression ± S.D.(n = 2) compared to the levels detected at 0 h which are arbitrarily set at 1. Statistical significance between CA9 and VEGF expression in normoxic and hypoxic samples at the various time points was determined by two-sample t-tests: * = p < 0.05, = p < 0.005, * = p < 0.0005. The complete panel of 12 soft tissue sarcoma cell lines was examined, the results obtained with five representative cell lines are shown. were compared. Unsupervised hierarchical clustering based on the miRNA expression profiles of 407 miRNAs grouped the cell lines on cell type rather than on exposure to hypoxia (Figure 2A), which implies that the probes capable of detecting 725 human miRNAs. Since HIF1 target genes were clearly induced after 24 h of hypoxia, miRNA expression levels of cell lines that were cultured for 24 h under normoxic or hypoxic conditions Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Page 6 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Figure 2 (See legend on next page.) Figure 2 (See legend on next page.) Figure 2 (See legend on next page.) Figure 2 (See legend on next page.) Page 7 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 ( g p p g ) Figure 2 Hypoxia induces changes in miRNA expression. (A) Unsupervised hierarchical clustering using miRNA expression data (407 miRNAs) of hypoxic (H) and normoxic (N) cell line samples does not discriminate the hypoxic and normoxic samples. The hypoxic samples cluster together with their normoxic counterparts. Luciferase assay (B) The most significant differentially expressed miRNAs (two-sample t-test, p < 0.05) between cell lines that were cultured under normoxic and hypoxic conditions are used for a supervised hierarchical clustering. The expression of these 32 miRNAs can distinguish hypoxic from normoxic sarcoma cell line samples. (C) The four most significant (P < 0.001) differentially expressed miRNAs (miR-485-5p, miR-216a-5p, miR-185-3p and miR-625-5p) discriminate sarcoma cell lines that were cultured under normoxic conditions from those cultured under hypoxic conditions in a supervised hierarchical clustering. (D) MiR-210-3p is upregulated during hypoxia. RT-PCR was used to determine miR-210-3p levels in normoxic (black bars) and hypoxic (grey bars) samples from five representative sarcoma cell lines. Expression is normalized against RNU43 expression. Bars indicate average expression fold change ± S.D. (n = 2) of miR-210-3p compared to the expression at 0 h. Statistical significance between expression in normoxic and hypoxic samples at a time point were determined by two-sample t-tests: * = p < 0.05, = p < 0.005, * = p < 0.0005. The different cell lines are indicated by a color code that identifies their tumor of origin: with synovial sarcoma (yellow); Ewing sarcoma (green); rhabdomyosarcoma (purple); leiomyosarcoma (blue); liposarcoma (orange) and fibrosarcoma (red). the Ensembl Genome Browser release 72 (http://www. ensembl.org) the HIF3α transcript variants 003 and 201 (consistent with HIF3α transcription variant 1, 2 and 3 according to GenBank (http://www.ncbi.nlm.nih.gov/ genbank)) contain a long 3’UTR of 3811 bp which in- cludes the potential miR-210-3p and miR-485-5p binding sites. In order to determine whether these transcripts are expressed in our sarcoma cell line panel we used an adapted RT-PCR procedure involving an end-point PCR and specific primers located in the 3’UTR (primer sites are indicated in Figure 3B, primer sequences are listed in Additional file 1: Table S1a). HIF3α transcripts containing a long 3’UTR harboring the mentioned miRNA binding sites could be detected in all cell lines and were upregu- lated in time under hypoxic conditions (Figure 3D). changes in miRNA expression due to hypoxia are rela- tively minor. One major cluster branch contains the rhabdomyosarcoma cell lines and the synovial sarcoma cell line, the other branch contains the other cell lines (leiomyosarcoma, liposarcoma and fibrosarcoma cell lines). MiR-485-5p and miR-210-3p target HIF3α MiR 485 5p and miR 210 3p target HIF3α The effects of miR-210-3p and miR-485-5p overexpression on HIF3α, protein levels were examined by immunoblot- ting. The sarcoma cell lines SW872 (liposarcoma) and SK- UT-1 (leiomyosarcoma) were transfected with mimics of miR-210-3p, miR-485-5p or a scrambled negative con- trol. 48 h post-transfection the cells were exposed to hyp- oxia for 24 h to induce HIF3α. Figure 4A clearly shows that overexpression of the miRNA mimics resulted in a reduced induction of HIF3α when compared to the HIF3α protein levels observed in cells transfected with a scrambled control mimic (mneg). The detected HIF3α protein has a relative molecular weight of approximately 70 kDa which is in line with the predicted molecular weight of the HIF3α proteins encoded by the 003 or 201 HIF3α transcript variants. Luciferase assay Supervised hierarchical clustering using the 32 most differentially expressed miRNAs (p < 0.05; Additional file 1: Table S2) between cell lines exposed to hypoxia or nor- moxia clearly separated the hypoxic from the normoxic samples (Figure 2B). Fifteen miRNAs were upregulated during hypoxia, while 17 miRNAs were downregulated. The SJCRH30 rhabdomyosarcoma cell line cultures under hypoxia was the only sample that misclustered, which in- dicates that the hypoxia induced miRNA response in this cell line was apparently weak. The top four differentially expressed miRNAs (p < 0.001) consisted of miR-185-3p, miR-485-5p, miR-216a-5p and miR-625-5p (Figure 2C; Additional file 1: Table S2). Only miR-625-5p was down- regulated in most hypoxic samples, whereas the other three miRNAs were predominantly upregulated. MiR- 210-3p, a commonly found hypoxia responsive miRNA (HRM), was not reliably detected on our microarray plat- form. Therefore the miR-210-3p expression in normoxic and hypoxic sarcoma cell line samples was separately de- termined by qPCR, which revealed a significant upregula- tion of miR-210-3p in all cell lines after 24 h of hypoxia (Figure 2D). HIF3α transcripts are expressed and upregulated under hypoxia in sarcoma cell lines Luciferase ac- tivity in SW872 hypoxic lysates was reduced to 57% and 65% in the HIF3α-short and HIF3α-long transfectants, respectively, when compared to the levels seen at nor- moxic circumstances (Figure 4B, left panel). In the hyp- oxic SK-UT-1 lysates the luciferase activity was decreased to 79% and 67% in the HIF3α-short and HIF3α-long transfectants, respectively (Figure 4B, right panel). These results indicate that during 24 h hypoxia the 3’UTR of HIF3α is targeted, most likely by hypoxia responsive miRNAs, causing a reduction of the luciferase activity. (Figure 4C, left panel). Overexpression of miR-485-5p de- creased luciferase activity to 52% in SK-UT-1 cells trans- fected with psiCHECK2/HIF3α-short and further reduced luciferase activity to 24% in the HIF3α-long transfectants (Figure 4C, right panel). To prove whether the regulation of HIF3α is due to direct binding of miR-210-3p and miR-485-5p to target sites in the 3’UTR, the predicted binding sites were mutated to prevent miRNA binding (Additional file 1: Figure S1). Since the regulatory effect of the single miR- 485-5p site present in the HIF3α-short construct was greater than the effect of the additional miR-485-5p binding sites present in the HIF3α-long construct, only the miR-485-5p target site in the HIF3α-short construct was mutated. The presence of a mutated miR-210-3p binding site significantly increased (SW872) or restored (SK-UT-1) luciferase activity (Figure 4D). Mutation of the miR-485-5p target site did not affect luciferase activ- ity in SW872 cells and restored luciferase activity to some extent in the SK-UT-1 cells. We conclude that miR-210-3p directly regulates HIF3α whereas the regula- tion observed with miR-485-5p is primarily due to an in- direct effect. However, we cannot rule out that direct regulation of HIF3α does occur mediated by one or more of the additional miR-485-5p binding sites in the HIF3α 3’UTR. To examine whether downregulation of HIF3α is medi- ated by miR-210-3p and/or miR-485-5p we transfected miRNA mimics for miR-210-3p and miR-485-5p in SW872 and SK-UT-1 cells, after 24 h followed by the transfection of the psiCHECK2/HIF3α 3’UTR luciferase reporter constructs. Compared to the scrambled negative control (mneg), miR-210-3p overexpression decreased the luciferase activity to 36% and 41% in SW872 (Figure 4C, left panel) and 37% and 43% in SK-UT-1 (Figure 4C, right panel) in the HIF3α-short and HIF3α-long transfectants respectively. MiR-485-5p overexpression reduced the luciferase activity to 61% in the HIF3α-short construct in the SW872 cell line. HIF3α transcripts are expressed and upregulated under hypoxia in sarcoma cell lines Figure 3 The 3’UTR of HIF3α contains putative miR-210-3p and miR-485-5p binding sites and is expressed in sarcoma cell lines. (A) Schematic representation of the 3’UTR of the HIF3α transcript variants 003 and 201 (http://www.ensembl.org, release 72). Depicted are predicted target sites of miR-210-3p (open box) and miR-485-5p (black boxes). (B) Arrows indicate location of primers used to detect the presence of HIF3α 3’UTR transcripts. A 212 bp and 313 bp amplification product indicate expression of the 3’UTR of HIF3α transcription variants 003 and 201. (C) Lines designate the HIF3α 3’UTR fragments (HIF3α-short, HIF3α-long) which were cloned into the psiCHECK-2 luciferase reporter to verify regulation of HIF3α by miRNAs. (D) End-point RT-PCR was used to determine the presence of HIF3α 3’UTR transcripts and their induction upon hypoxia. Depicted are EtBr stained PCR amplification fragments of 212 and 313 bp derived from HIF3α 3’UTR cDNA and an amplified 336 bp HPRT fragment as input control. The results obtained with five representative cell lines are shown. Figure 3 The 3 UTR of HIF3α contains putative miR-210-3p and miR-485-5p binding sites and is expressed in sarcoma cell lines. (A) Schematic representation of the 3’UTR of the HIF3α transcript variants 003 and 201 (http://www.ensembl.org, release 72). Depicted are predicted target sites of miR-210-3p (open box) and miR-485-5p (black boxes). (B) Arrows indicate location of primers used to detect the presence of HIF3α 3’UTR transcripts. A 212 bp and 313 bp amplification product indicate expression of the 3’UTR of HIF3α transcription variants 003 and 201. (C) Lines designate the HIF3α 3’UTR fragments (HIF3α-short, HIF3α-long) which were cloned into the psiCHECK-2 luciferase reporter to verify regulation of HIF3α by miRNAs. (D) End-point RT-PCR was used to determine the presence of HIF3α 3’UTR transcripts and their induction upon hypoxia. Depicted are EtBr stained PCR amplification fragments of 212 and 313 bp derived from HIF3α 3’UTR cDNA and an amplified 336 bp HPRT fragment as input control. The results obtained with five representative cell lines are shown. miR-485-5p. The SW872 an SK-UT-1 cell lines were transfected with the psiCHECK2/HIF3α-short and psi- CHECK2/HIF3α-long constructs. The next day the cells were exposed to hypoxia for 24 h, after which the lucifer- ase activity in cell lysates was determined. HIF3α transcripts are expressed and upregulated under hypoxia in sarcoma cell lines When ten additional putative bind- ing sites for miR-485-5p were introduced in the HIF3α- long transfectant, the luciferase activity dropped to 45% HIF3α transcripts are expressed and upregulated under hypoxia in sarcoma cell lines Candidate target genes for the most differentially expressed miRNAs identified in the microarray analysis were predicted using TargetScanHuman 6.2 (http://www. targetscan.org/) [38]. When we focused on genes known to play a role in the hypoxic response it was noted that miR-485-5p has as much as 11 potential binding sites in the 3’UTR of HIF3α, the less studied member of the family of hypoxia inducible factors. Also miR-210-3p has a single putative binding site in this transcript (Figure 3A). Several human HIF3α transcripts exist due to alternative splicing, coding for different HIF3α isoforms [5-7]. According to To demonstrate that during hypoxia HIF3α is regulated by hypoxia responsive miRNAs, a short (HIF3α-short, 817 bp) and a long (HIF3α-long, 3807 bp) fragment of the HIF3α 3’UTR (variants 003/201) (Figure 3C) were cloned in a psiCHECK2 luciferase reporter. The psi- CHECK2/HIF3α-short construct encompassed the puta- tive miR-210-3p binding site and the first miR-485-5p binding site. The psiCHECK2/HIF3α-long construct included an additional 10 potential binding sites for Gits et al. BMC Cancer 2014, 14:429 Page 8 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 http://www.biomedcentral.com/1471-2407/14/429 Figure 3 The 3’UTR of HIF3α contains putative miR-210-3p and miR-485-5p binding sites and is expressed in sarcoma cell lines. (A) Schematic representation of the 3’UTR of the HIF3α transcript variants 003 and 201 (http://www.ensembl.org, release 72). Depicted are predicted target sites of miR-210-3p (open box) and miR-485-5p (black boxes). (B) Arrows indicate location of primers used to detect the presence of HIF3α 3’UTR transcripts. A 212 bp and 313 bp amplification product indicate expression of the 3’UTR of HIF3α transcription variants 003 and 201. (C) Lines designate the HIF3α 3’UTR fragments (HIF3α-short, HIF3α-long) which were cloned into the psiCHECK-2 luciferase reporter to verify regulation of HIF3α by miRNAs. (D) End-point RT-PCR was used to determine the presence of HIF3α 3’UTR transcripts and their induction upon hypoxia. Depicted are EtBr stained PCR amplification fragments of 212 and 313 bp derived from HIF3α 3’UTR cDNA and an amplified 336 bp HPRT fragment as input control. The results obtained with five representative cell lines are shown. Figure 3 The 3’UTR of HIF3α contains putative miR-210-3p and miR-485-5p binding sites and is expressed in sa UTR of HIF3α contains putative miR-210-3p and miR-485-5p binding sites and is expressed in sarcoma cell lines. Discussion STS, like all solid tumors, can present with hypoxic areas, a state which is associated with disease progression and Page 9 of 13 Page 9 of 13 Gits et al. BMC Cancer 2014, 14:429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 http://www.biomedcentral.com/1471-2407/14/429 Figure 4 (See legend on next page.) Figure 4 (See legend on next page.) Page 10 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 (See figure on previous page.) Figure 4 HIF3α is regulated by miR-210-3p and miR-485-5p. (A) MiR-210-3p and miR-485-5p mimic overexpression reduce HIF3α protein induction under hypoxic conditions. The sarcoma cell lines SW872 and SK-UT-1 were transfected with mimics of miR-210-3p (m210), miR-485-5p (m485) and a scrambled negative control mimic (mneg). 48 h post-transfection cells were cultured under hypoxic conditions (1% O2) for 24 h. Subsequently total protein lysates were prepared and analysed by Western blotting for HIF3α protein expression. β-Actin is used as a loading control. (B) Hypoxia responsive miRNAs target HIF3α 3’UTR. SW872 and SK-UT-1 cell lines were transfected with psiCHECK 2 constructs containing short and long 3’UTR fragments of HIF3α. 24 h later the cells were exposed to hypoxia for 24 h. Bars indicate average luciferase activity ± SD (n = 3) measured in hypoxic cell lysates relative to the luciferase activity in normoxic cell lysates which is arbitrarily set at 100. (C) MiR-210-3p and miR-485-5p regulate HIF3α. SW872 and SK-UT-1 cell lines were transfected with mimics of miR-210-3p (m210), miR-485-5p (m485) or a scrambled control mimic (mneg) followed after 24 h by a transfection with psiCHECK 2 constructs containing short and long fragments of the HIF3α 3’UTR. Bars indicate average luciferase activity ± SD (n = 3) measured in cell lysates after 24 h. (D) MiR-210-3p and miR-485-5p regulate HIF3α. SW872 and SK-UT-1 cell lines were transfected with mimics of miR-210-3p (m210), miR-485-5p (m485) or a scrambled control mimic (mneg) followed, after 24 h, by a transfection with a psiCHECK 2-HIF3α-short construct containing either wild-type (WT) or mutated (mut) miR-210-3p and miR-485-5p binding sites. Bars indicate average luciferase activity ± SD (n = 3) measured in cell lysates. Statistical significance was determined by two-sample t-tests: * = p < 0.05, = p < 0.005, * = p < 0.0005. independent mechanisms (reviewed in [44,45]). Discussion Some HRMs, such as miR-210-3p, contain hypoxia responsive elements (HREs) in their promoter region and are dir- ectly regulated by the HIF1 transcription factor in re- sponse to hypoxia [26,46,47]. Although HIF1 is generally known as a transcriptional activator, it can also function as a transcriptional repressor [48], which could account for the downregulated miRNAs under hypoxic condi- tions. Another HIF-dependent mechanism by which hypoxia responsive miRNAs are regulated is via a HIF1 induced transcription factor, such as TWIST, which in- duces miR-10b expression [49]. A HIF-independent mechanism involved in the regulation of HRMs is e.g. the induction of Akt2 upon hypoxia, which in turn upre- gulates miR-21 [50]. Also deregulation of the miRNA biogenesis during hypoxia could contribute to altered miRNA expression levels. Recently it was reported that hypoxia enhanced the association between EGFR and Ago2, resulting in increased Ago2-phosphorylation and reduced binding of Dicer to Ago2, thereby inhibiting miRNA processing [51]. In contrast, in another study hypoxia did not alter the expression of key miRNA ma- chinery proteins (i.e. Drosha, Exp5, Dicer, Ago2 and DP103) in human trophoblasts [52]. In order to deter- mine whether our top hypoxia responsive miRNAs (miR-185-3p; miR-485-5p; miR-216a-5p and miR-625- 5p) contain HREs in their putative promoter region we screened 600 bp flanking sequences upstream of the transcription start site of the primary miRNA. With the exception of miR-485-5p of which the gene resides in a densely packed miRNA cluster on chromosome 14, the genes for miR-185-3p, miR-216a-5p and miR-625-5p are not clustered. No consensus HRE (A/GCGTG) could be detected upstream of miR-625-5p (which was downregu- lated during hypoxia), whereas miR-216a-5p contains a ‘ACGTGC’(position -42 to -37) and miR-185-3p contains a ‘CCGTG’(position -305 to -301). Although the latter HRE does not perfectly match the consensus HRE se- quence, imperfect HREs have been found to be functional bad prognosis [14,15]. To date, little is known about the involvement of miRNAs in the hypoxic response of STS. Greither et al. reported that the expression of miR-210- 3p, a well-known hypoxia responsive miRNA, is associ- ated with poor survival in STS [39]. We identified 32 miRNAs that are deregulated upon hypoxia in a panel of 12 soft tissue sarcoma cell lines. Particularly, miR-185- 3p, miR-485-5p, miR-216a-5p were significantly upregu- lated, while miR-625-5p was significantly downregulated. In addition we detected induction of miR-210-3p upon hypoxia using a RT-PCR procedure. Discussion Besides transcriptional regulation of HIF3α by HIF1 and oxygen dependent degradation of HIF3α proteins, we provide evidence for an additional layer of regulation of HIF3α, i.e. miRNAs that act on the translational level. These different mechanisms can com- plement one another in fine tuning HIF3α expression. This may be critical because of the versatile role HIF3α plays in the hypoxic response. [11,53,54]. Therefore it is possible that miR-216a-5p and miR-185-3p are regulated by HIF1. Since miR-210-3p is considered a general responder to hypoxia irrespective of cell-type and exact hypoxic con- ditions, its function is thought to be universal. MiR-210- 3p has been demonstrated to have a role in e.g. cell proliferation, angiogenesis, apoptosis, DNA repair and mitochondrial metabolism (reviewed in [55-57]). The function of other, less commonly deregulated HRMs is poorly understood, but these miRNAs are likely to have a cell-type specific function. Not much is known about miR-485-5p. This miRNA was downregulated in ependy- momas [58] and suppressed dendritic spine development in rats [59]. Furthermore miR-485-5p was downregu- lated in Alzheimers disease [60] and malignant serous ovarian tumors, and correlated with FIGO grade in the latter [61]. Computational analysis predicted as much as eleven putative miR-485-5p binding sites in the 3’UTR of HIF3α. Also miR-210-3p has a potential binding site in the 3’UTR of this member of the family of hypoxia inducible factors. Note that there are only two HIF3α transcripts (variants 003 and 201) listed (http://www. ensembl.org; release 72) that habor a 3’UTR long enough to encompass all miR-485-5p binding sites in addition to the miR-210-3p site. At least two other HIF3α transcripts that contain a smaller 3’UTR (478 bp in variant 006 and 287 bp in variant 007) also contain the miR-210-3p binding site suggesting that the HIF3α isoforms they encode may be controlled by miR-210-3p. We demonstrated that HIF3α is regulated by miR-210- 3p and in an indirect fashion by miR-485-5p. We also showed that HIF3α mRNA levels are induced upon hyp- oxia as expected of a hypoxia inducible factor. Under the same conditions, however, the expression of hypoxia responsive miRNAs like miR-210-3p and miR-485-5p is upregulated that affect HIF3α protein expression. We postulate that miRNAs help to fine-tune HIF3α protein expression and in this way regulate the cellular hypoxic response. Discussion In all cell lines ex- amined, our array platform did not detect a significant upregulation of miR-210-3p expression upon hypoxia. In fact, expression levels detected were too low to be con- sidered for subsequent analyses. This is most likely due to the fact that the LNA capture probe for miR-210-3p was not optimally designed. Exiqon varies both length and LNA contents of the capture probes to obtain a Tm normalized probe with appropriate affinity for its target miRNA. Although the design and composition of the capture probes are proprietary information we learned that Exiqon has redesigned the miR-210-3p capture probe in subsequent versions of the capture probe set. A subset of the deregulated miRNAs, i.e. miR-185-3p, miR- 191, miR-210-3p, miR-373 (upregulated), miR-148b, miR-181b and miR-342-5p (downregulated), was previ- ously identified as hypoxia responsive miRNAs (HRM) in other studies involving different cancer types [40-43]. These miRNAs may be regarded as general responders to hypoxia. However the majority of miRNAs we detected has not been previously associated with hypoxia suggest- ing a cell-type specific miRNA response, as has been re- ported before [25,44]. The molecular mechanisms responsible for miRNA deregulation in response to hypoxia are for a large part still unclear as is the precise role of each of the HRM in the hypoxia response. As the key response to hypoxic conditions is the stabilization of HIF1, HRMs can be regulated by HIF-dependent or, alternatively, by HIF- Page 11 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 promoter region [5,7,11]. This probably accounts for the increased HIF3α mRNA expression upon hypoxia. In addition HIF3α can be regulated at the protein level. HIF1α and HIF2α contain an oxygen dependent degrad- ation domain (ODDD), which harbors two conserved prolines that can be hydroxylated by prolyl hydroxylases (PHDs) in presence of oxygen. This marks the proteins for proteasomal degradation via the von Hippel-Lindau (VHL) E3 ubiquitin ligase complex. It was postulated that HIF3α, in contrast with HIF1α and HIF2α, is not regulated at the level of protein stability, since the longer HIF3α transcripts miss part of the ODDD and the shorter isoforms lack the entire ODDD. This renders them less prone to hydroxylation and subsequent pro- teasomal degradation [11]. However, other studies showed that the remaining proline residue in the ODDD of HIF3α transcripts is targeted by VHL and subjected to proteasomal degradation in an oxygen dependent manner [5,6]. Discussion They can do so by affecting the translation of specific HIF3α transcripts, however, we cannot rule out that part of the regulation takes place via HIF3α mRNA degradation. Although HRMs targeting HIF3α has never been documented before, miRNA regulation of HIFs is not uncommon: miR-155 [46], miR-424 [28], and the miR-17-92 cluster [30] are known to target HIF1α. Thus HIFs can induce the expression of miRNAs, and in turn, miRNAs can also target HIFs. Our understanding of the regulation of HIF3α and its None of the HIF3α variants are likely to act as potent transcription factors, since they lack a C-terminal trans- activation domain. However, they can have an effect on HRE-driven gene expression, depending on the level of HIFβ [5,8,10]. When HIFβ is not limiting, it will dimerize with HIF1α, HIF2α and HIF3α to induce a sub- set of hypoxia regulated genes. HIF1α/HIFβ and HIF2α/ HIFβ complexes will associate with HRE sequences, while the HIF3α/HIFβ complexes likely bind a response element different from the canonical HRE, to maximize hypoxia induced gene expression [5]. When HIFβ is lim- iting, HIF3α will either compete with HIF1α and HIF2α for binding with HIFβ [8], or HIF3α will associate with HIF1α and HIF2α [5,10], both resulting in decreased capability to bind HRE sequences and diminished tran- scription of HIF target genes. As such, HIF3α splice vari- ants will not act as global regulators of gene expressions, but may modulate specific genes in a cell type dependent manner [5,11]. Additional file 5. Heikkila M, Pasanen A, Kivirikko KI, Myllyharju J: Roles of the human hypoxia-inducible factor (HIF)-3alpha variants in the hypoxia response. Cell Mol Life Sci 2011, 68(23):3885–3901. Additional file 1: Table S1. Primers used for PCR and cloning. (A) Primers used for end-point RT-PCRs of HIF3α 3’UTR fragments, resulting in amplification products of 212 bp (Fw/Rv1) and 313 bp (Fw/Rv2). Amplification of HPRT (product of 335 bp) was used as input control. (B) Primers used for cloning HIF3α-short and HIF3α-long 3’UTR constructs. (C) Primers used for site mutagenesis of predicted miR-485-5p and miR-210-3p binding sites in HIF3α-short 3’UTR constructs. Table S2 Differentially expressed miRNAs (p<0.05) between cell lines that were cultured under hypoxic and normoxic conditions. P-values of two-sample t-test as well as fold change in miRNA expression and miRNA genomic locations are indicated. False Discovery Rate (FDR) for the top four miRNAs are: hsa-miR-185-3p (FDR 0.002893); hsa-miR-485-5p (FDR 0.004466); hsa-miR-216a-5p (FDR 0.068687) and hsa-miR-625-5p (FDR 0.112324). Figure S1 Predicted 3’UTR target sites in HIF3α for miR-210-3p and miR-485-5p and the mutations that have been generated in the target site sequence where the miRNA seed sequence (bold) binds. The wild-type (WT) and mutated (mut) sites in HIF3α-short are shown. The vertical lines represent possible base pairing between miRNA and 3’UTR target site, and the x’s indicate abrogated base pairing where nucleotides are mutated (red). The resulting mutated 3’UTR fragments were cloned into the psiCHECK-2 luciferase reporter. 6. Maynard MA, Qi H, Chung J, Lee EH, Kondo Y, Hara S, Conaway RC, Conaway JW, Ohh M: Multiple splice variants of the human HIF-3 alpha locus are targets of the von Hippel-Lindau E3 ubiquitin ligase complex. J Biol Chem 2003, 278(13):11032–11040. 7. Pasanen A, Heikkila M, Rautavuoma K, Hirsila M, Kivirikko KI, Myllyharju J: Hypoxia-inducible factor (HIF)-3alpha is subject to extensive alternative splicing in human tissues and cancer cells and is regulated by HIF-1 but not HIF-2. Int J Biochem Cell Biol 2010, 42(7):1189–1200. 8. Hara S, Hamada J, Kobayashi C, Kondo Y, Imura N: Expression and characterization of hypoxia-inducible factor (HIF)-3alpha in human kidney: suppression of HIF-mediated gene expression by HIF-3alpha. Biochem Biophys Res Commun 2001, 287(4):808–813. 9. Maynard MA, Evans AJ, Hosomi T, Hara S, Jewett MA, Ohh M: Human HIF-3alpha4 is a dominant-negative regulator of HIF-1 and is down-regulated in renal cell carcinoma. FASEB J 2005, 19(11):1396–1406. 10. Conclusion Our understanding of the regulation of HIF3α and its role in the hypoxic response is still rather limited. Re- ports in the literature are sometimes contradictory or difficult to compare as they focus on different HIF3α splice variants in different cell types. It is clear that HIF3α can be regulated at different levels. Transcription of HIF3α can be induced by HIF via HREs in the In conclusion, this study describes, in addition to miR- 210-3p, a new panel of HRMs in hypoxic sarcoma cells. In turn, two of these HRMs, i.e. miR-210-3p and miR- 485-5p, regulate HIF3α in a direct or indirect fashion, respectively. Fine-tuning of HIF3α expression is of great importance for the hypoxic response, since HIF3α affects HIF1/2-induced gene expression. Keeping these negative Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Page 12 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 feedback loops under tight control enables the cells to adapt to hypoxic conditions. Deregulation of these mech- anisms, e.g. by therapeutic modulation of levels of miR- NAs that are important for the hypoxic response, may inhibit the adaptive potential of the cells, and reduce their resistance against radiation and systemic agents. 2. Wang GL, Semenza GL: General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. Proc Natl Acad Sci U S A 1993, 90(9):4304–4308. 3. Ema M, Taya S, Yokotani N, Sogawa K, Matsuda Y, Fujii-Kuriyama Y: A novel bHLH-PAS factor with close sequence similarity to hypoxia-inducible factor 1alpha regulates the VEGF expression and is potentially involved in lung and vascular development. Proc Natl Acad Sci U S A 1997, 94(9):4273–4278. 4. Hu CJ, Wang LY, Chodosh LA, Keith B, Simon MC: Differential roles of hypoxia-inducible factor 1alpha (HIF-1alpha) and HIF-2alpha in hypoxic gene regulation. Mol Cell Biol 2003, 23(24):9361–9374. Authors’ contributions CG, SS, JV, EW conceived the study. CG, EW, SS, RM, JV were involved in the design and coordination of the study. CG, PK, JR, MJ and NM performed qPCR, end-point PCR and Western blotting experiments. CG, WIJ, MJ carried out miRNA expression profiling CG, NM, EW participated in cloning experiments and luciferase reporter assays. CG, PK, WIJ, SS, RM, JV and EW drafted the manuscript. All authors read and approved the final version of the manuscript. 14. Brizel DM, Scully SP, Harrelson JM, Layfield LJ, Bean JM, Prosnitz LR, Dewhirst MW: Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma. Cancer Res 1996, 56(5):941–943. 15. Nordsmark M, Hoyer M, Keller J, Nielsen OS, Jensen OM, Overgaard J: The relationship between tumor oxygenation and cell proliferation in human soft tissue sarcomas. Int J Radiat Oncol Biol Phys 1996, 35(4):701–708. 16. Wiemer EA: The role of microRNAs in cancer: no small matter. Eur J Cancer 2007, 43(10):1529–1544. Author details 1 1Department of Medical Oncology, Erasmus University Medical Center – Erasmus MC Cancer Institute, Rotterdam, the Netherlands. 2Center for Biomics, Erasmus University Medical Center, Rotterdam, the Netherlands. 3Department of Medical Oncology, Erasmus MC Cancer Institute, Erasmus University Medical Center, Rm Be422, Wytemaweg 80, 3015 CN Rotterdam, the Netherlands. 20. Zhang B, Pan X, Cobb GP, Anderson TA: microRNAs as oncogenes and tumor suppressors. Dev Biol 2007, 302(1):1–12. 21. Li X, Wu Z, Fu X, Han W: A microRNA component of the neoplastic microenvironment: microregulators with far-reaching impact. Biomed Res Int 2013, 2013:762183. 22. Wentz-Hunter KK, Potashkin JA: The role of miRNAs as key regulators in the neoplastic microenvironment. Mol Biol Int 2011, 2011:839872. Received: 2 September 2013 Accepted: 29 May 2014 Published: 13 June 2014 Received: 2 September 2013 Accepted: 29 May 2014 Published: 13 June 2014 Received: 2 September 2013 Accepted: 29 May 2014 Published: 13 June 2014 23. Crosby ME, Devlin CM, Glazer PM, Calin GA, Ivan M: Emerging roles of microRNAs in the molecular responses to hypoxia. Curr Pharm Des 2009, 15(33):3861–3866. Competing interests The authors declare they have no competing interests. 13. Hockel M, Schlenger K, Aral B, Mitze M, Schaffer U, Vaupel P: Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res 1996, 56(19):4509–4515. Acknowledgements 17. Chen CZ: MicroRNAs as oncogenes and tumor suppressors. N Engl J Med 2005, 353(17):1768–1771. We thank dr. A. Carnero, dr. M. Debiec-Rychter and dr. L. Alberti for providing cell lines. This study was funded by EC FP6 CONTICANET network of excellence (LSHC-CT-2005-018806) from the European Commission. We thank dr. A. Carnero, dr. M. Debiec-Rychter and dr. L. Alberti for providing cell lines. This study was funded by EC FP6 CONTICANET network of excellence (LSHC-CT-2005-018806) from the European Commission. 18. Di Leva G, Croce CM: Roles of small RNAs in tumor formation. Trends Mol Med 2010, 16(6):257–267. 19. Lujambio A, Lowe SW: The microcosmos of cancer. Nature 2012, 482(7385):347–355. Additional file Maynard MA, Evans AJ, Shi W, Kim WY, Liu FF, Ohh M: Dominant-negative HIF-3 alpha 4 suppresses VHL-null renal cell carcinoma progression. Cell Cycle 2007, 6(22):2810–2816. 11. Tanaka T, Wiesener M, Bernhardt W, Eckardt KU, Warnecke C: The human HIF (hypoxia-inducible factor)-3alpha gene is a HIF-1 target gene and may modulate hypoxic gene induction. Biochem J 2009, 424(1):143–151. Abbreviations HIF h d HIF: hypoxia inducible factor; HRE: hypoxia responsive elements; HRM: hypoxia responsive miRNA; LNA: locked nucleic acid; miRNA: microRNA; qPCR: quantitative polymerase chain reaction; STS: soft tissue sarcomas; 3’UTR: 3’untranslated region. 12. Airley RE, Phillips RM, Evans AE, Double J, Burger AM, Feibig HH, West CM, Stratford IJ: Hypoxia-regulated glucose transporter glut-1 may influence chemosensitivity to some alkylating agents: results of EORTC (first translational award) study of the relevance of tumour hypoxia to the outcome of chemotherapy in human tumour-derived xenografts. Int J Oncol 2005, 26(6):1477–1484. Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 MicroRNA-155 promotes resolution of hypoxia-inducible factor 1alpha activity during prolonged hypoxia. Mol Cell Biol 2011, 31(19):4087–4096. MicroRNA-155 promotes resolution of hypoxia-inducible factor 1alpha activity during prolonged hypoxia. Mol Cell Biol 2011, 31(19):4087–4096. 25. Kulshreshtha R, Ferracin M, Negrini M, Calin GA, Davuluri RV, Ivan M: Regulation of microRNA expression: the hypoxic component. Cell Cycle 2007, 6(12):1426–1431. 47. Huang X, Ding L, Bennewith KL, Tong RT, Welford SM, Ang KK, Story M, Le QT, Giaccia AJ: Hypoxia-inducible mir-210 regulates normoxic gene expression involved in tumor initiation. Mol Cell 2009, 35(6):856–867. 26. Kulshreshtha R, Ferracin M, Wojcik SE, Garzon R, Alder H, Agosto-Perez FJ, Davuluri R, Liu CG, Croce CM, Negrini M, Calin GA, Ivan M: A microRNA signature of hypoxia. Mol Cell Biol 2007, 27(5):1859–1867. 48. Chen KF, Lai YY, Sun HS, Tsai SJ: Transcriptional repression of human cad gene by hypoxia inducible factor-1alpha. Nucleic Acids Res 2005, 33(16):5190–5198. 27. Cha ST, Chen PS, Johansson G, Chu CY, Wang MY, Jeng YM, Yu SL, Chen JS, Chang KJ, Jee SH, Tan C, Lin MT, Kuo ML: MicroRNA-519c suppresses hypoxia-inducible factor-1alpha expression and tumor angiogenesis. Cancer Res 2010, 70(7):2675–2685. 49. Haque I, Banerjee S, Mehta S, De A, Majumder M, Mayo MS, Kambhampati S, Campbell DR, Banerjee SK: Cysteine-rich 61-connective tissue growth factor-nephroblastoma-overexpressed 5 (CCN5)/Wnt-1-induced signaling protein-2 (WISP-2) regulates microRNA-10b via hypoxia-inducible factor-1alpha-TWIST signaling networks in human breast cancer cells. J Biol Chem 2011, 286(50):43475–43485. 28. Ghosh G, Subramanian IV, Adhikari N, Zhang X, Joshi HP, Basi D, Chandrashekhar YS, Hall JL, Roy S, Zeng Y, Ramakrishnan S: Hypoxia-induced microRNA-424 expression in human endothelial cells regulates HIF-alpha isoforms and promotes angiogenesis. J Clin Invest 2010, 120(11):4141–4154. 50. Polytarchou C, Iliopoulos D, Hatziapostolou M, Kottakis F, Maroulakou I, Struhl K, Tsichlis PN: Akt2 regulates all Akt isoforms and promotes resistance to hypoxia through induction of miR-21 upon oxygen deprivation. Cancer Res 2011, 71(13):4720–4731. 29. Kelly TJ, Souza AL, Clish CB, Puigserver P: A hypoxia-induced positive feedback loop promotes hypoxia-inducible factor 1alpha stability through miR-210 suppression of glycerol-3-phosphate dehydrogenase 1-like. Mol Cell Biol 2011, 31(13):2696–2706. 51. Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Devlin C, Greco S, Martelli F, Ivan M: miR-210: more than a silent player in hypoxia. IUBMB Life 2011, 63(2):94–100. 36. Wykoff CC, Beasley NJ, Watson PH, Turner KJ, Pastorek J, Sibtain A, Wilson GD, Turley H, Talks KL, Maxwell PH, Pugh CW, Ratcliffe PJ, Harris AL: Hypoxia-inducible expression of tumor-associated carbonic anhydrases. Cancer Res 2000, 60(24):7075–7083. 57. Ivan M, Harris AL, Martelli F, Kulshreshtha R: Hypoxia response and microRNAs: no longer two separate worlds. J Cell Mol Med 2008, 12(5A):1426–1431. 58. Costa FF, Bischof JM, Vanin EF, Lulla RR, Wang M, Sredni ST, Rajaram V, Bonaldo Mde F, Wang D, Goldman S, Tomita T, Soares MB: Identification of microRNAs as potential prognostic markers in ependymoma. PLoS One 2011, 6(10):e25114. 37. Pugh CW, Ratcliffe PJ: Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med 2003, 9(6):677–684. 38. Lewis BP, Burge CB, Bartel DP: Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 2005, 120(1):15–20. 59. Cohen JE, Lee PR, Chen S, Li W, Fields RD: MicroRNA regulation of homeostatic synaptic plasticity. Proc Natl Acad Sci U S A 2011, 108(28):11650–11655. 39. Greither T, Wurl P, Grochola L, Bond G, Bache M, Kappler M, Lautenschlager C, Holzhausen HJ, Wach S, Eckert AW, et al: Expression of microRNA 210 associates with poor survival and age of tumor onset of soft-tissue sarcoma patients. Int J Cancer 2012, 130(5):1230–1235. 60. Faghihi MA, Zhang M, Huang J, Modarresi F, Van der Brug MP, Nalls MA, Cookson MR, St-Laurent G 3rd, Wahlestedt C: Evidence for natural antisense transcript-mediated inhibition of microRNA function. Genome Biol 2010, 11(5):R56. 40. Crosby ME, Kulshreshtha R, Ivan M, Glazer PM: MicroRNA regulation of DNA repair gene expression in hypoxic stress. Cancer Res 2009, 69 (3):1221–1229. 61. Kim TH, Kim YK, Kwon Y, Heo JH, Kang H, Kim G, An HJ: Deregulation of miR-519a, 153, and 485-5p and its clinicopathological relevance in ovarian epithelial tumours. Histopathology 2010, 57(5):734–743. 41. Guimbellot JS, Erickson SW, Mehta T, Wen H, Page GP, Sorscher EJ, Hong JS: Correlation of microRNA levels during hypoxia with predicted target mRNAs through genome-wide microarray analysis. BMC Med Genomics 2009, 2:15. doi:10.1186/1471-2407-14-429 Cite this article as: Gits et al.: MicroRNA response to hypoxic stress in soft tissue sarcoma cells: microRNA mediated regulation of HIF3α. BMC Cancer 2014 14:429. 42. Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Shen J, Xia W, Khotskaya YB, Huo L, Nakanishi K, Lim SO, Du Y, Wang Y, Chang WC, Chen CH, Hsu JL, Wu Y, Lam YC, James BP, Liu X, Liu CG, Patel DJ, Hung MC: EGFR modulates microRNA maturation in response to hypoxia through phosphorylation of AGO2. Nature 2013, 497(7449):383–387. 30. Taguchi A, Yanagisawa K, Tanaka M, Cao K, Matsuyama Y, Goto H, Takahashi T: Identification of hypoxia-inducible factor-1 alpha as a novel target for miR-17-92 microRNA cluster. Cancer Res 2008, 68(14):5540–5545. 31. Xu Q, Liu LZ, Qian X, Chen Q, Jiang Y, Li D, Lai L, Jiang BH: MiR-145 directly targets p70S6K1 in cancer cells to inhibit tumor growth and angiogenesis. Nucleic Acids Res 2012, 40(2):761–774. 52. Donker RB, Mouillet JF, Nelson DM, Sadovsky Y: The expression of Argonaute2 and related microRNA biogenesis proteins in normal and hypoxic trophoblasts. Mol Hum Reprod 2007, 13(4):273–279. 32. Yamakuchi M, Lotterman CD, Bao C, Hruban RH, Karim B, Mendell JT, Huso D, Lowenstein CJ: P53-induced microRNA-107 inhibits HIF-1 and tumor angiogenesis. Proc Natl Acad Sci U S A 2010, 107(14):6334–6339. 53. Augstein A, Poitz DM, Braun-Dullaeus RC, Strasser RH, Schmeisser A: Cell-specific and hypoxia-dependent regulation of human HIF-3alpha: inhibition of the expression of HIF target genes in vascular cells. Cell Mol Life Sci 2011, 68(15):2627–2642. 33. Pothof J, Verkaik NS, van IJcken W, Wiemer EA, TA VT, van der Horst GT, Jaspers NG, van Gent DC, Hoeijmakers JH, Persengiev SP: MicroRNA-mediated gene silencing modulates the UV-induced DNA-damage response. EMBO J 2009, 28(14):2090–2099. 54. Makino Y, Uenishi R, Okamoto K, Isoe T, Hosono O, Tanaka H, Kanopka A, Poellinger L, Haneda M, Morimoto C: Transcriptional up-regulation of inhibitory PAS domain protein gene expression by hypoxia-inducible factor 1 (HIF-1): a negative feedback regulatory circuit in HIF-1-mediated signaling in hypoxic cells. J Biol Chem 2007, 282(19):14073–14082. 34. Schmittgen TD, Livak KJ: Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 2008, 3(6):1101–1108. 35. Beasley NJ, Wykoff CC, Watson PH, Leek R, Turley H, Gatter K, Pastorek J, Cox GJ, Ratcliffe P, Harris AL: Carbonic anhydrase IX, an endogenous hypoxia marker, expression in head and neck squamous cell carcinoma and its relationship to hypoxia, necrosis, and microvessel density. Cancer Res 2001, 61(13):5262–5267. 55. Chan YC, Banerjee J, Choi SY, Sen CK: miR-210: the master hypoxamir. Microcirculation 2012, 19(3):215–223. 56. References 1. Harris AL: Hypoxia–a key regulatory factor in tumour growth. Nat Rev Cancer 2002, 2(1):38–47. 24. Kulshreshtha R, Davuluri RV, Calin GA, Ivan M: A microRNA component of the hypoxic response. Cell Death Differ 2008, 15(4):667–671. 24. Kulshreshtha R, Davuluri RV, Calin GA, Ivan M: A microRNA component of the hypoxic response. Cell Death Differ 2008, 15(4):667–671. 1. Harris AL: Hypoxia–a key regulatory factor in tumour growth. Nat Rev Cancer 2002, 2(1):38–47. 1. Harris AL: Hypoxia–a key regulatory factor in tumour growth. Nat Rev Cancer 2002, 2(1):38–47. Page 13 of 13 Gits et al. BMC Cancer 2014, 14:429 http://www.biomedcentral.com/1471-2407/14/429 Hebert C, Norris K, Scheper MA, Nikitakis N, Sauk JJ: High mobility group A2 is a target for miRNA-98 in head and neck squamous cell carcinoma. Mol Cancer 2007, 6:5. 43. Xu X, Jia R, Zhou Y, Song X, Wang J, Qian G, Ge S, Fan X: Microarray-based analysis: identification of hypoxia-regulated microRNAs in retinoblastoma cells. Int J Oncol 2011, 38(5):1385–1393. 44. Shen G, Li X, Jia YF, Piazza GA, Xi Y: Hypoxia-regulated microRNAs in human cancer. Acta Pharmacol Sin 2013, 34(3):336–341. 45. Nallamshetty S, Chan SY, Loscalzo J: Hypoxia: a master regulator of microRNA biogenesis and activity. Free Radic Biol Med 2013, 64:20–30. 45. Nallamshetty S, Chan SY, Loscalzo J: Hypoxia: a master regulator of microRNA biogenesis and activity. Free Radic Biol Med 2013, 64:20–30. 46. Bruning U, Cerone L, Neufeld Z, Fitzpatrick SF, Cheong A, Scholz CC, Simpson DA, Leonard MO, Tambuwala MM, Cummins EP, Taylor CT: 46. Bruning U, Cerone L, Neufeld Z, Fitzpatrick SF, Cheong A, Scholz CC, Simpson DA, Leonard MO, Tambuwala MM, Cummins EP, Taylor CT:
https://openalex.org/W4381511905	https://tc.copernicus.org/preprints/tc-2023-43/tc-2023-43.pdf	English	null	Reply on RC1	null	2,023	cc-by	27,363	Reconciling ice dynamics and bed topography with a versatile and fast ice thickness inversion Th F k1 W d J J P lt1 d J k K hl 2 Finally, we apply the approach to the tidewater glacier Kronebreen on as the misfit functional, iterative pointwise updates to an initial guess of bed topography are made, while mismatches between 5 observed and modelled velocities are used to simultaneously infer basal friction. The final product of the inversion is not only a map of ice thickness, but a fully spun-up glacier model representing the dynamic state of a given glacier. We here present the method, and use an artificial ice-cap built inside a numerical model to test it and conduct sensitivity experiments. Even under a range of perturbations, the method is stable and fast. Finally, we apply the approach to the tidewater glacier Kronebreen on 5 Svalbard. Ultimately, our method shown here represents a fast way of inferring ice thickness where the final output forms a 10 consistent picture of model physics, input observations and bed topography. Reconciling ice dynamics and bed topography with a versatile and fast ice thickness inversion Th F k1 W d J J P lt1 d J k K hl 2 Thomas Frank1, Ward J. J. van Pelt1, and Jack Kohler2 1Department of Earth Sciences, Uppsala University, Uppsala, Sweden 2Norwegian Polar Institute, Fram Centre, Tromsø, Norway Correspondence: Thomas Frank (thomas.frank@geo.uu.se) Abstract. We present a novel thickness inversion approach that leverages globally available satellite products and state-of-the- art ice flow models to produce distributed maps of subglacial topography independent of bed observations. While the method can use any complexity of ice physics as represented in ice dynamical models, it is computationally cheap, enabling applications both on local and large scales. Using the mismatch between observed and modelled rates of surface elevation change (dh/dt) Abstract. We present a novel thickness inversion approach that leverages globally available satellite products and state-of-the- art ice flow models to produce distributed maps of subglacial topography independent of bed observations. While the method can use any complexity of ice physics as represented in ice dynamical models, it is computationally cheap, enabling applications both on local and large scales. Using the mismatch between observed and modelled rates of surface elevation change (dh/dt) as the misfit functional, iterative pointwise updates to an initial guess of bed topography are made, while mismatches between 5 observed and modelled velocities are used to simultaneously infer basal friction. The final product of the inversion is not only a map of ice thickness, but a fully spun-up glacier model representing the dynamic state of a given glacier. We here present the method, and use an artificial ice-cap built inside a numerical model to test it and conduct sensitivity experiments. Even under a range of perturbations, the method is stable and fast. Finally, we apply the approach to the tidewater glacier Kronebreen oni g U g w v v g ( / ) as the misfit functional, iterative pointwise updates to an initial guess of bed topography are made, while mismatches between 5 observed and modelled velocities are used to simultaneously infer basal friction. The final product of the inversion is not only a map of ice thickness, but a fully spun-up glacier model representing the dynamic state of a given glacier. We here present the method, and use an artificial ice-cap built inside a numerical model to test it and conduct sensitivity experiments. Even under a range of perturbations, the method is stable and fast. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. 1 Introduction Early works include the use of scaling relationships that allow the computation of 25 mean bed elevation from surface area (i.e. volume-area scaling, Chen and Ohmura (1990); Bahr et al. (1997, 2014)). More ad- vanced approaches apply computational methods relying on a physical understanding of how ice flows, often (when available) in conjunction with measured ice thicknesses. The underlying rationale is that easily observable surface characteristics, such as the surface elevation height, are a product of the external climate forcing (e.g. climatic mass balance), ice dynamics and the subglacial topography, and therefore an inference about the latter can be made from knowledge of the previous. The recent 30 years have seen an increasing amount of such new ice thickness inversion approaches; an overview of widely used methods is given in Farinotti et al. (2017, 2021). The large majority of approaches follow at least one of the two strategies: in the first, the physical calculations of ice dynamics underpinning the inversion are conducted along flow-lines or at randomly selected points on a glacier, and subsequently the results are extrapolated to surrounding areas, often while taking into account some subglacial topography, and therefore an inference about the latter can be made from knowledge of the previous. The recent 30 years have seen an increasing amount of such new ice thickness inversion approaches; an overview of widely used methods is given in Farinotti et al. (2017, 2021). The large majority of approaches follow at least one of the two strategies: in the first, the physical calculations of ice dynamics underpinning the inversion are conducted along flow-lines or at randomly selected points on a glacier, and subsequently the results are extrapolated to surrounding areas, often while taking into account some first-order glaciological principles, e.g. the strong link between surface slope and ice thickness (Farinotti et al., 2009; Linsbauer 35 et al., 2009; Huss and Farinotti, 2012; Li et al., 2012; Linsbauer et al., 2012; Paul and Linsbauer, 2012; Clarke et al., 2013; Frey et al., 2014; Brinkerhoff et al., 2016; Rabatel et al., 2018; Maussion et al., 2019; Werder et al., 2020). This is computation- ally efficient as compared to calculating distributed ice thickness fields, but comes with the limitation that the physics-based inversion is only conducted at specific locations. In the second approach, a simple model considering only internal shear (i.e. 1 Introduction Knowledge of the ice thickness of glaciers is crucial for a range of applications, from field work planning, to water management Knowledge of the ice thickness of glaciers is crucial for a range of applications, from field work Knowledge of the ice thickness of glaciers is crucial for a range of applications, from field work planning, to water management and ultimately for projections of expected sea level rise in the face of climate warming (Oppenheimer et al., 2019; Rounce et al., 15 2023). Ice thickness not only determines the ice volume of a glacier, but the subglacial topography is furthermore a key control on future retreat of glaciers. As such, accurate bed maps are instrumental to initialize prognostic simulations of numerical ice flow models with the correct boundary conditions. and ultimately for projections of expected sea level rise in the face of climate warming (Oppenheimer et al., 2019; Rounce et al., 15 2023). Ice thickness not only determines the ice volume of a glacier, but the subglacial topography is furthermore a key control on future retreat of glaciers. As such, accurate bed maps are instrumental to initialize prognostic simulations of numerical ice flow models with the correct boundary conditions. Ice thickness can be established using geophysical methods such as ground-penetrating radar (Bogorodsky et al., 1985). However, observations are only dense along their acquisition lines, while considerable spacing between lines is the norm. 20 To produce distributed maps of ice thickness, statistical techniques have been developed to interpolate between observations However, observations are only dense along their acquisition lines, while considerable spacing between lines is the norm. 20 To produce distributed maps of ice thickness, statistical techniques have been developed to interpolate between observations 1 1 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. (Flowers and Clarke, 1999; Fretwell et al., 2013; Neven et al., 2021). However, only a small fraction of glaciers worldwide have any thickness observations at all (Welty et al., 2020). (Flowers and Clarke, 1999; Fretwell et al., 2013; Neven et al., 2021). However, only a small fraction of glaciers worldwide have any thickness observations at all (Welty et al., 2020). As an alternative to field measurements, inversion techniques have been introduced that allow to derive a bed estimate from easily-obtainable surface observations alone. 1 Introduction the shallow ice approximation, SIA) is applied to turn a modelled or observed quantity (e.g. depth-integrated ice flux, surface 40 velocities) into ice thickness, sometimes complemented with a parameterization for basal sliding (Farinotti et al., 2009; Huss and Farinotti, 2012; Li et al., 2012; Linsbauer et al., 2012; Paul and Linsbauer, 2012; Clarke et al., 2013; Frey et al., 2014; Gantayat et al., 2014; Fürst et al., 2017; Rabatel et al., 2018; Langhammer et al., 2019; Maussion et al., 2019; Werder et al., 2020; Zorzut et al., 2020; Millan et al., 2022). These approaches are also computationally cheap compared to calculations based on a higher-order or Full-Stokes model, and they exploit the simple mathematical nature of the depth-integrated SIA. 45 However, the simplifying assumptions on ice dynamics that are at the heart of the SIA translate into errors in the calculated ice thicknesses at locations where the conditions are not met. Although different in their complexity, their required inputs and their strengths and limitations (for more details the reader is referred to the respective publications), the thickness inversion approaches discussed here share limitations that arise from following one or both of the mentioned strategies. For instance, based on a higher-order or Full-Stokes model, and they exploit the simple mathematical nature of the depth-integrated SIA. 45 However, the simplifying assumptions on ice dynamics that are at the heart of the SIA translate into errors in the calculated ice thicknesses at locations where the conditions are not met. Although different in their complexity, their required inputs and their strengths and limitations (for more details the reader is referred to the respective publications), the thickness inversion approaches discussed here share limitations that arise from following one or both of the mentioned strategies. For instance, using a bed derived with such methods for prognostic simulations with map-plane higher-order or Full-Stokes models requires 50 initial model relaxation that likely results in a glacier state significantly different from observations. A different approach was presented by van Pelt et al. (2013), building on previous work in the context of inversion problems in fluid mechanics (Heining, 2011). Using historical observations of external forcing, and starting from an initial guess for bed topography, they use the ice flow model PISM (Bueler and Brown, 2009) to simulate centuries of glacier evolution up until A different approach was presented by van Pelt et al. 1 Introduction (2013) is based on the idea that the present-day dynamic state of a glacier is best reproduced by simulating that glacier for centuries up until today using historical forcing. However, such forcing data is rarely available and subject to considerable uncertainties. Furthermore, due to the long time spans modelled the approach is computationally costly. Here, we exploit the fact that the instantaneous rate of surface elevation change dh/dt, which also represents the dynamic state of a glacier, is much faster to model allowing one to apply a similar methodology as in van Pelt et al. (2013) 65 while only requiring a fraction of the computational resources. In our new approach, we furthermore extend the capabilities of the thickness inversion in that we simultaneously infer basal friction using observed surface velocities akin to a previously published approach (Pollard and DeConto, 2012). Ultimately, we establish a framework that is able to provide a comprehensive view of the state of a glacier consistent with its external forcing, its ice dynamics and its bed topography while being at low computational cost. Allowing to choose between different levels of complexity for the ice-flow physics and whether or not to 70 simultaneously infer basal friction, our approach is suitable both for detailed inversions at the local scale as well as for fast large-scale applications. In the following, we describe our new fast ice thickness inversion approach using PISM as a forward model, we demonstrate its capabilities with an example of an artificial ice cap grown inside a numerical model, and apply and test it on the tidewater glacier Kronebreen on Svalbard. 75 2.1 Method The general functional principle derives from the idea that a numerical model of a glacier initialized with the correct boundary conditions should behave similarly to a real glacier; if not, one needs to conclude that the model setup is erroneous in some part. In a simple case without basal sliding and neglecting thermodynamic processes, the first-order controls on ice dynamics 0 are the climatic mass balance, the glacier surface shape and the bed. By assuming that it is possible to set up a numerical ice-flow model where all of these inputs are sufficiently well represented except for bed topography (because they are surface observations), we may attribute any differences in modelled vs. observed ice geometry and dynamics to originate from errors The general functional principle derives from the idea that a numerical model of a glacier initialized with the correct boundary conditions should behave similarly to a real glacier; if not, one needs to conclude that the model setup is erroneous in some part. In a simple case without basal sliding and neglecting thermodynamic processes, the first-order controls on ice dynamics 0 The general functional principle derives from the idea that a numerical model of a glacier initialized with the correct boundary conditions should behave similarly to a real glacier; if not, one needs to conclude that the model setup is erroneous in some part. In a simple case without basal sliding and neglecting thermodynamic processes, the first-order controls on ice dynamics 80 are the climatic mass balance, the glacier surface shape and the bed. By assuming that it is possible to set up a numerical ice-flow model where all of these inputs are sufficiently well represented except for bed topography (because they are surface observations), we may attribute any differences in modelled vs. observed ice geometry and dynamics to originate from errors in the bed. part. In a simple case without basal sliding and neglecting thermodynamic processes, the first-order controls on ice dynamics 80 are the climatic mass balance, the glacier surface shape and the bed. By assuming that it is possible to set up a numerical ice-flow model where all of these inputs are sufficiently well represented except for bed topography (because they are surface observations), we may attribute any differences in modelled vs. observed ice geometry and dynamics to originate from errors in the bed. 1 Introduction (2013), building on previous work in the context of inversion problems in fluid mechanics (Heining, 2011). Using historical observations of external forcing, and starting from an initial guess for bed topography, they use the ice flow model PISM (Bueler and Brown, 2009) to simulate centuries of glacier evolution up until A different approach was presented by van Pelt et al. (2013), building on previous work in the context of inversion problems in fluid mechanics (Heining, 2011). Using historical observations of external forcing, and starting from an initial guess for bed topography, they use the ice flow model PISM (Bueler and Brown, 2009) to simulate centuries of glacier evolution up until present day, when a comparison between modelled and observed surface elevation is made. Based on the assumption that any 55 deviations between the two are due to errors in the bed, this mismatch is then used to update the subglacial topography. These present day, when a comparison between modelled and observed surface elevation is made. Based on the assumption that any 55 deviations between the two are due to errors in the bed, this mismatch is then used to update the subglacial topography. These 2 2 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. two steps are repeated iteratively until a certain stopping criterion is reached. This approach allows to include any physics that state-of-the-art ice flow models feature and it results in a distributed map of bed topography that is in balance with the present-day dynamic state of the glacier and the external forcing. For ensuing prognostic simulations, no relaxation is needed as all input parameters and physics used for the forward simulations can also be used in the thickness inversion. 60 two steps are repeated iteratively until a certain stopping criterion is reached. This approach allows to include any physics that state-of-the-art ice flow models feature and it results in a distributed map of bed topography that is in balance with the present-day dynamic state of the glacier and the external forcing. For ensuing prognostic simulations, no relaxation is needed as all input parameters and physics used for the forward simulations can also be used in the thickness inversion. 0 60 The approach by van Pelt et al. 2.1 Method To turn this framework into an inversion approach, the following workflow is applied (Fig. 1): First, a numerical representa- 85 tion of a glacier is set up inside an ice flow model with observations of surface shape from a digital elevation model (DEM) and climatic mass balance. Furthermore, some arbitrary bed shape is assumed. For instance, it is straightforward to derive a rough bed estimate from a DEM with the perfect-plasticity method (Nye, 1952; Haeberli and Hoelzle, 1995). As Habermann et al. To turn this framework into an inversion approach, the following workflow is applied (Fig. 1): First, a numerical representa- 85 tion of a glacier is set up inside an ice flow model with observations of surface shape from a digital elevation model (DEM) and climatic mass balance. Furthermore, some arbitrary bed shape is assumed. For instance, it is straightforward to derive a rough bed estimate from a DEM with the perfect-plasticity method (Nye, 1952; Haeberli and Hoelzle, 1995). As Habermann et al. 85 3 3 Figure 1. Schematic representation of bed corrections during thickness inversion workflow, assuming a steady-state glacier (dh/dtobs = 0) with observed surface Sobs. Without the bed bump at iteration i, the modelled surface Smod subsides during a short forward simulation, rendering dh/dtmod negative. As given by eq. (1), this negative misfit between modelled and observed dh/dt results in an uplift of the bed. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure 1. Schematic representation of bed corrections during thickness inversion workflow, assuming a steady-state glacier (dh/dtobs = 0) with observed surface Sobs. Without the bed bump at iteration i, the modelled surface Smod subsides during a short forward simulation, rendering dh/dtmod negative. As given by eq. (1), this negative misfit between modelled and observed dh/dt results in an uplift of the bed. (2012) points out, though, an initial bed guess should not contain roughness not justified by the inp (2012) points out, though, an initial bed guess should not contain roughness not justified by the input data, so some smoothing, e.g., with a gaussian filter, may be in order. 0 e.g., with a gaussian filter, may be in order. 90 Then, the ice-flow model is run forward with the goal of comparing whether modelled and observed ice dynamics agree. 2.1 Method As described before, the target variable here is the rate of surface elevation change dh/dt as a measure for the dynamic state of the glacier. From the model side, this requires only a stress balance simulation or a simulation with a short timescale length dt to compute the instantaneous dh/dtmod. From the observational side, it requires either an already existing dh/dt product (e.g. Hugonnet et al., 2021), or two DEMs from different years, where it is assumed that the rate of surface elevation change between 95 them is constant, allowing the derivation of the quantity dh/dtobs. If dh/dtmod = dh/dtobs, it is reasonable to conclude that the bed topography is correct, because the model correctly reproduces the observations. However, in the more likely event that they are not equal, the differences between the two are ascribed to originate from errors in the bed, following the logic discussed above. If there is a mismatch between a modelled and observed quantity, previous work has shown point-wise updates can be 100 applied to the unknown parameter (in this case bed topography) as a simple function of that mismatch (van Pelt et al., 2013; Michel et al., 2013; Heining, 2011; Pollard and DeConto, 2012). Here, a new bed Bi+1 is hence produced by subtracting, at any point in space in the domain, the misfit in modelled vs. observed dh/dt at that point from the present bed Bi such that If there is a mismatch between a modelled and observed quantity, previous work has shown point-wise updates can be 100 applied to the unknown parameter (in this case bed topography) as a simple function of that mismatch (van Pelt et al., 2013; Michel et al., 2013; Heining, 2011; Pollard and DeConto, 2012). Here, a new bed Bi+1 is hence produced by subtracting, at any point in space in the domain, the misfit in modelled vs. observed dh/dt at that point from the present bed Bi such that Bi+1 = Bi −β dhi mod dt −dhobs dt (1) Bi+1 = Bi −β dhi mod dt −dhobs dt (1) where β is a scalar that controls the magnitude of bed corrections. These simple steps described here (running the model 105 forward, computing the misfit in modelled vs. observed dh/dt and updating the bed) are repeated iteratively until a satisfactory solution for the bed is found (Fig 1). Note that van Pelt et al. 2.1 Method Here, we present a novel way of handling this issue: every time the bed topography is updated, a small proportion θ of 120 the dh/dt misfit is used to also update the glacier surface S in the opposite direction of the bed such that a new surface Si+1 is given by Si+1 = Si + θβ dhi mod dt −dhobs dt . (2) (2) Doing so results in a ’squeezing’ effect where an upwards correction of the bed (i.e. the ice thickness becomes smaller) results Doing so results in a ’squeezing’ effect where an upwards correction of the bed (i.e. the ice thickness becomes smaller) results in a lowered glacier surface at that point, and vice versa for a downwards bed correction. This leads to locally steeper surface 125 slopes and therefore increased driving stresses which gently induce further bed corrections in surrounding areas, thus evening out small bed undulations. At the same time, the small changes in the surface shape allow the model to accommodate observa- tions which it otherwise cannot reproduce, be it because of input errors or imperfect model physics. Indeed, our experiments show that the bed and surface updates interact to finally yield a configuration where dh/dtmod = dh/dtobs. in a lowered glacier surface at that point, and vice versa for a downwards bed correction. This leads to locally steeper surface 125 slopes and therefore increased driving stresses which gently induce further bed corrections in surrounding areas, thus evening out small bed undulations. At the same time, the small changes in the surface shape allow the model to accommodate observa- tions which it otherwise cannot reproduce, be it because of input errors or imperfect model physics. Indeed, our experiments show that the bed and surface updates interact to finally yield a configuration where dh/dtmod = dh/dtobs. Finally, our method works well under the assumption that sliding is not significant. Where this simplification does not hold, 130 one solution is to prescribe some friction field. However, this makes the modelled ice thickness dependent on an often poorly constrained parameter which could lead to considerable errors in the recovered bed. Here, we address this issue by inverting for basal friction following a similar methodology as in Pollard and DeConto (2012). 2.1 Method (2013) used a similar equation, although instead of dh/dt, 105 4 4 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. the modelled surface elevation Smod after century-long time-dependent runs was used to calculate a misfit with the observed surface Sobs, requiring much greater computational resources. the modelled surface elevation Smod after century-long time-dependent runs was used to calculate a misfit with the observed surface Sobs, requiring much greater computational resources. Our experiments show that this simple approach alone is capable of recovering subglacial topography for a non-sliding 110 glacier yielding an overall good match with the true bed. However, previous research has shown that ice thickness inversion is non-unique. Small scale bed undulations leave no expressions on the glacier surface and hence an inversion, regardless of the method used, can make no statement on the existence or non-existence of such features (Gudmundsson, 2003; Bahr et al., 2014). For our approach, this implies that the derived bed may contain such small-scale undulations that are not justified by 110 the input data if the solution is not biased towards outcomes that take these limitations into account. Furthermore, real-world 115 problems imply that input data and model physics are never perfect, meaning that not all deviations between modelled and observed dynamics can be attributed to originate from errors in the bed. problems imply that input data and model physics are never perfect, meaning that not all deviations between modelled and observed dynamics can be attributed to originate from errors in the bed. Previous work in the spirit of traditional inversion theory has tackled this issue by introducing so-called regularization terms which force the modelled bed to be smooth and which account for a degree of imperfection in the model (Habermann et al., Previous work in the spirit of traditional inversion theory has tackled this issue by introducing so-called regularization terms which force the modelled bed to be smooth and which account for a degree of imperfection in the model (Habermann et al., 2012). Here, we present a novel way of handling this issue: every time the bed topography is updated, a small proportion θ of 120 the dh/dt misfit is used to also update the glacier surface S in the opposite direction of the bed such that a new surface Si+1 is given by 2012). 2.2 Model While the method is not bound to a specific ice-flow model, we use the Parallel Ice Sheet Model (PISM v.2.0.4, Bueler and Brown (2009)) for the ice-flow simulations. Below, we briefly present relevant aspects of PISM with respect to our thickness inversion method, but refer the reader to the literature for a more in-depth description (Bueler and Brown, 2009; Winkelmann 155 et al., 2011). et al., 2011). PISM allows the glaciological stress-balance equations to be solved using one of four approximations to the Full-Stokes equations: the shallow ice approximation (SIA), the shallow shelf approximation (SSA), a hybrid SSA+SIA scheme or the Blatter-Pattyn higher-order model (Blatter, 1995; Pattyn, 2003). Any of them is suitable for the thickness inversion, but we PISM allows the glaciological stress-balance equations to be solved using one of four approximations to the Full-Stokes equations: the shallow ice approximation (SIA), the shallow shelf approximation (SSA), a hybrid SSA+SIA scheme or the Blatter-Pattyn higher-order model (Blatter, 1995; Pattyn, 2003). Any of them is suitable for the thickness inversion, but we here apply the hybrid SSA+SIA scheme. To simulate sliding, different laws are available, some of which allow interaction with 160 modelled thermodynamics to yield values of subglacial water availability which in turn determines basal friction. In this work, we do not include thermodynamic processes (sec. Discussion), and rely instead on a simple pseudo-plastic power law where basal shear stress τ b is given by here apply the hybrid SSA+SIA scheme. To simulate sliding, different laws are available, some of which allow interaction with 160 modelled thermodynamics to yield values of subglacial water availability which in turn determines basal friction. In this work, we do not include thermodynamic processes (sec. Discussion), and rely instead on a simple pseudo-plastic power law where basal shear stress τ b is given by τ b = −τc u uq th\|u\|1−q (4) (4) where uth is the threshold velocity where τb has the same magnitude as the yield stress τc and q ∈[0,1] is an exponent 165 determining the plasticity of the sliding law (q = 1 represents a linear sliding law). τc is a spatially variable constant which reflects the strength of the subglacial bed and hence determines the degree of sliding. This is the quantity that we invert for in eq. (3). 2.1 Method Ultimately, this alternation between many bed updates and one 150 friction update is repeated until a combination of a bed and friction field is found that explains both dh/dtobs and uobs. velocity mismatch is evaluated and if needed, F is updated. Ultimately, this alternation between many bed updates and one 150 friction update is repeated until a combination of a bed and friction field is found that explains both dh/dtobs and uobs. 2.1 Method Most sliding laws have a parameter (often termed friction coefficient, yield stress, etc.) that varies in space but is constant in time which controls the strength of the subglacial material, and hence the amount of sliding. We invert for this parameter F in a similar fashion as shown 135 before. Specifically, again an initial guess for F is made which is subsequently updated, however, this time using a mismatch in modelled vs. observed ice flow speed u, such that F i+1 =        F i −λF i, if ui mod−uobs uobs < −λ F i + λF i, if ui mod−uobs uobs > λ F i + F i ui mod−uobs uobs , otherwise. (3) 5 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. The parameter λ serves to prevent too strong updates to the friction field. Importantly, bed and friction are not updated in the same iteration. This is because the dh/dt misfit and the velocity misfit both represent a mismatch in modelled vs. observed 140 flow dynamics and it is not clear how to disentangle which part of that mismatch is due to an erroneous friction field vs. an erroneous bed. So instead, F is only updated once after the model has converged to a solution for ice thickness. At that point, a model configuration that can explain dh/dtobs has been found. However, only if that model also explains observed surface speed is it reasonable to conclude that the imposed friction field is correct, assuming that the friction field is the largest remaining unknown control on ice dynamics. If there is no agreement in modelled vs. observed surface speed, F is likely not 145 correct (again under that same assumption) and hence the modelled bed elevation is erroneous. A change in the friction field will trigger a response of the ice dynamics, and so dh/dtmod will no longer match dh/dtobs after having updated F. Following a friction update, a new bed thus needs to be found that can explain dh/dtobs with that new friction field. Therefore, again bed and surface updates according to eq. (1) and (2) need to be applied until any dh/dt mismatch is resolved. Subsequently, the velocity mismatch is evaluated and if needed, F is updated. 2.1 Method Ultimately, this alternation between many bed updates and one 150 friction update is repeated until a combination of a bed and friction field is found that explains both dh/dtobs and uobs. The parameter λ serves to prevent too strong updates to the friction field. Importantly, bed and friction are not updated in the same iteration. This is because the dh/dt misfit and the velocity misfit both represent a mismatch in modelled vs. observed 140 flow dynamics and it is not clear how to disentangle which part of that mismatch is due to an erroneous friction field vs. an erroneous bed. So instead, F is only updated once after the model has converged to a solution for ice thickness. At that point, a model configuration that can explain dh/dtobs has been found. However, only if that model also explains observed surface speed is it reasonable to conclude that the imposed friction field is correct, assuming that the friction field is the largest remaining unknown control on ice dynamics. If there is no agreement in modelled vs. observed surface speed, F is likely not 145 correct (again under that same assumption) and hence the modelled bed elevation is erroneous. A change in the friction field will trigger a response of the ice dynamics, and so dh/dtmod will no longer match dh/dtobs after having updated F. Following a friction update, a new bed thus needs to be found that can explain dh/dtobs with that new friction field. Therefore, again bed and surface updates according to eq. (1) and (2) need to be applied until any dh/dt mismatch is resolved. Subsequently, the remaining unknown control on ice dynamics. If there is no agreement in modelled vs. observed surface speed, F is likely not 145 correct (again under that same assumption) and hence the modelled bed elevation is erroneous. A change in the friction field will trigger a response of the ice dynamics, and so dh/dtmod will no longer match dh/dtobs after having updated F. Following a friction update, a new bed thus needs to be found that can explain dh/dtobs with that new friction field. Therefore, again bed and surface updates according to eq. (1) and (2) need to be applied until any dh/dt mismatch is resolved. Subsequently, the velocity mismatch is evaluated and if needed, F is updated. 3.1 The synthetic ice cap 170 We start by using our new method on an artificial ice cap. We first design a domain with a 50 x 50 grid and a resolution of 2 km where we define a bed topography consisting of a flat bed including two large (500 m tall) and two small (350 m tall) bumps with a prescribed mass balance field (Fig. 2a). Furthermore, we design a friction field by setting the parameter τc such that the left half of the domain is slippery, allowing ice to slide, while the right half is sticky, preventing any basal motion. We use a non-linear sliding law as in the MISMIP experiments (Pattyn et al., 2012) by setting the sliding law parameters q = 1/3 and 175 uth = 1m s−1 in eq. (4) (Tab. 1). Subsequently, we grow the ice cap inside PISM for 10 kyrs using the isothermal SSA+SIA scheme allowing us to obtain a steady-state configuration with a maximum ice thickness of ∼1000 m. Then, we ’forget’ the bed and friction field and try to recover them. For the thickness inversion, we use as inputs the same mass balance as applied when growing the ice cap, the surface of the grown ice cap, a reference dh/dtobs field that is zero everywhere (since the ice non-linear sliding law as in the MISMIP experiments (Pattyn et al., 2012) by setting the sliding law parameters q = 1/3 and 175 uth = 1m s−1 in eq. (4) (Tab. 1). Subsequently, we grow the ice cap inside PISM for 10 kyrs using the isothermal SSA+SIA scheme allowing us to obtain a steady-state configuration with a maximum ice thickness of ∼1000 m. Then, we ’forget’ the bed and friction field and try to recover them. For the thickness inversion, we use as inputs the same mass balance as applied when growing the ice cap, the surface of the grown ice cap, a reference dh/dtobs field that is zero everywhere (since the ice cap is in steady-state), and make a first bed guess assuming a completely flat bed without bumps (Fig. 2a). For the friction 180 inversion, our first guess is that the domain is sticky everywhere, and the input required to later update this guess is the velocity field of the grown ice cap. 2.2 Model where uth is the threshold velocity where τb has the same magnitude as the yield stress τc and q ∈[0,1] is an exponent 165 determining the plasticity of the sliding law (q = 1 represents a linear sliding law). τc is a spatially variable constant which reflects the strength of the subglacial bed and hence determines the degree of sliding. This is the quantity that we invert for in eq. (3). 165 6 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Table 1. Parameter values used for building the ice cap and the subsequent inversion. Parameter Value Ice temperature 268 K Ice density 910 kg m−3 q 0.333 uth 1 m s−1 β 1 θ 0.025 dt 0.1 yr λ 0.5 3.1 The synthetic ice cap 170 Subsequently, we simulate a short forward run with dt = 0.1 yrs and update the bed and the ice cap surface according to eq. (1) and (2). To prevent irregularities at the ice cap margins where spurious boundary effects can occur, we do not apply any updates in a band of two cells width at the ice cap margin, but set the bed elevation to zero there. cap is in steady-state), and make a first bed guess assuming a completely flat bed without bumps (Fig. 2a). For the friction 180 inversion, our first guess is that the domain is sticky everywhere, and the input required to later update this guess is the velocity field of the grown ice cap. Subsequently, we simulate a short forward run with dt = 0.1 yrs and update the bed and the ice cap surface according to eq. (1) and (2). To prevent irregularities at the ice cap margins where spurious boundary effects can occur, we do not apply any updates in a band of two cells width at the ice cap margin, but set the bed elevation to zero there. As we iteratively update the bed and the surface in that way, we calculate for all n grid points inside the ice-covered domain 185 the median dh/dt misfit ∆dh defined as As we iteratively update the bed and the surface in that way, we calculate for all n grid points inside the ice-covered domain 185 the median dh/dt misfit ∆dh defined as ∆dh = med dhmod dt −dhobs dt , (5) (5) and the mean absolute bed misfit \|∆B\|, given by 7 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure 2. Example of the thickness inversion for a ‘home-grown’ ice cap (i.e. ice cap is created, then the bed and friction field is ‘forgotten’ before the inversion is run with the same physics as the build-up). a) Bed and friction field used to build the ice cap plus the surface elevation and velocity of the artificial ice cap. b) dh/dt and bed misfit per iteration step i with sharp peaks and distinct steps indicating times at which friction field was updated (top panel). 3.1 The synthetic ice cap 170 Note that this amount of iterations between friction 190 updates is rather arbitrary; we here choose it simply because this number is large enough to ensure that the dh/dt misfit is sufficiently small at that point so that no more bed updates are occurring. While \|∆B\| is generally reduced as the iterative inversion progresses, each friction update induces a short response in the ice dynamics that briefly increases the ∆dh.As seen in Fig. 2b, the steep decline of the bed misfit at these instances highlights that those friction updates are crucial for finding the After 1000 iterations, we apply a friction update according to eq. (3). Note that this amount of iterations between friction 190 updates is rather arbitrary; we here choose it simply because this number is large enough to ensure that the dh/dt misfit is sufficiently small at that point so that no more bed updates are occurring. While \|∆B\| is generally reduced as the iterative inversion progresses, each friction update induces a short response in the ice dynamics that briefly increases the ∆dh.As seen in Fig. 2b, the steep decline of the bed misfit at these instances highlights that those friction updates are crucial for finding the After 1000 iterations, we apply a friction update according to eq. (3). Note that this amount of iterations between friction 190 updates is rather arbitrary; we here choose it simply because this number is large enough to ensure that the dh/dt misfit is sufficiently small at that point so that no more bed updates are occurring. While \|∆B\| is generally reduced as the iterative inversion progresses, each friction update induces a short response in the ice dynamics that briefly increases the ∆dh.As seen in Fig. 2b, the steep decline of the bed misfit at these instances highlights that those friction updates are crucial for finding the 190 correct bed. Indeed, a cross-section through the icecap reveals that the bed in the right (sticky) half of the domain is correctly 195 identified already before the first friction update. In the left (slippery) half where the initial friction field does not allow any sliding, however, ice thickness is greatly overestimated at that point. This is because in the absence of sliding, the ice thickness needs to be much greater to be able to achieve an ice flux that is in line with dh/dtobs and mass balance. 3.1 The synthetic ice cap 170 Cross-section through bed along y=23 km at each iteration step (lower left) and final bed elevation after 12000 iterations (lower right panel) Figure 2. Example of the thickness inversion for a ‘home-grown’ ice cap (i.e. ice cap is created, then the bed and friction field is ‘forgotten’ before the inversion is run with the same physics as the build-up). a) Bed and friction field used to build the ice cap plus the surface elevation and velocity of the artificial ice cap. b) dh/dt and bed misfit per iteration step i with sharp peaks and distinct steps indicating times at which friction field was updated (top panel). Cross-section through bed along y=23 km at each iteration step (lower left) and final bed elevation after 12000 iterations (lower right panel) Figure 2. Example of the thickness inversion for a ‘home-grown’ ice cap (i.e. ice cap is created, then the bed and friction field is ‘forgotten’ before the inversion is run with the same physics as the build-up). a) Bed and friction field used to build the ice cap plus the surface elevation and velocity of the artificial ice cap. b) dh/dt and bed misfit per iteration step i with sharp peaks and distinct steps indicating times at which friction field was updated (top panel). Cross-section through bed along y=23 km at each iteration step (lower left) and final bed elevation after 12000 iterations (lower right panel) Figure 2. Example of the thickness inversion for a ‘home-grown’ ice cap (i.e. ice cap is created, then the bed and friction field is ‘forgotten’ before the inversion is run with the same physics as the build-up). a) Bed and friction field used to build the ice cap plus the surface elevation and velocity of the artificial ice cap. b) dh/dt and bed misfit per iteration step i with sharp peaks and distinct steps indicating times at which friction field was updated (top panel). Cross-section through bed along y=23 km at each iteration step (lower left) and final bed elevation after 12000 iterations (lower right panel) \|∆B\| = 1 n n X 1 \|Bmod −Bref\|. (6) \|∆B\| = 1 n n X 1 \|Bmod −Bref\|. (6) (6) After 1000 iterations, we apply a friction update according to eq. (3). 3.1 The synthetic ice cap 170 With each friction update, the bed is made more slippery which increases flow velocities, thins the ice and lifts up the bed until eventually the correct bed is found here as well The final bed is in good agreement with the original one (Fig 2) 200 8 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. 3.2 Ice cap sensitivity Next, we explore how sensitive the inversion is to errors in the input data and to modelling choices. To that end, we use the synthetic ice cap where we have full knowledge of the ’true’ parameters. In the following, we provide an overview of the results, while detailed figures and more elaborations can be found in the appendix. 3.2.1 Input errors 205 However, this overestimation of ice thickness induces too-large ice-deformational velocities, leading to an even more sticky bed when eq. (3) is applied, thus causing further thickening. Again, the opposite is the case for a positive velocity error. Errors in the observed ice velocity induce similar mechanisms. A negative error here means that the bed is made more sticky 220 than it should be, resulting in too thick ice. However, this overestimation of ice thickness induces too-large ice-deformational velocities, leading to an even more sticky bed when eq. (3) is applied, thus causing further thickening. Again, the opposite is the case for a positive velocity error. In the case of ice temperature errors, friction updates have a compensating effect. For instance, a negative ice temperature error causes too-stiff ice and hence too-slow flow, which leads to too-thick ice without friction updates. With friction updates, 225 however, too-slow flow results in a negative velocity mismatch (eq. (3)) leading to a more slippery bed which, in turn, thins the ice again. Overall, the exact magnitude of amplifying or damping effects of friction updates in the presence of errors in ice temperature, mass balance, dh/dt and velocity depends on several characteristics that vary between glaciers, e.g. sliding magnitude and extent, and is thus hard to generalize. error causes too-stiff ice and hence too-slow flow, which leads to too-thick ice without friction updates. With friction updates, 225 however, too-slow flow results in a negative velocity mismatch (eq. (3)) leading to a more slippery bed which, in turn, thins the ice again. Overall, the exact magnitude of amplifying or damping effects of friction updates in the presence of errors in ice temperature, mass balance, dh/dt and velocity depends on several characteristics that vary between glaciers, e.g. sliding magnitude and extent, and is thus hard to generalize. 230 For surface noise, we find that it changes the bed shape, with already small surface perturbations causing large bed re- 230 sponses, as is expected from theory (Raymond and Gudmundsson, 2005; Bahr et al., 2014). However, increasing the regular- ization parameter θ serves to mitigate these issues, although it makes the final bed shape more dependent on initial conditions. Furthermore, we find that the calculated final ice volume is largely insensitive to surface noise. 3.2.1 Input errors 205 We test the sensitivity to changes in the input mass balance (which is equal to modifying dh/dt, since it can be shown that those two parameters may be understood as one through the concept of the apparent mass balance (Farinotti et al., 2009)) and velocity in experiments where their magnitude is reduced or increased by 25%, 50% and 75%, respectively, relative to their true value. Furthermore, we conduct experiments where we introduce random surface elevation noise with a standard deviation We test the sensitivity to changes in the input mass balance (which is equal to modifying dh/dt, since it can be shown that those two parameters may be understood as one through the concept of the apparent mass balance (Farinotti et al., 2009)) and velocity in experiments where their magnitude is reduced or increased by 25%, 50% and 75%, respectively, relative to their true value. Furthermore, we conduct experiments where we introduce random surface elevation noise with a standard deviation of 5 m, 10 m, 15 m, 25 m and 50 m. Finally, we modify the ice temperature from its true value of 268 K to values of 264 K, 210 266 K, 270 K and 272 K. For the sensitivity tests of mass balance and ice temperature, we investigate two scenarios: in one, we assume that the true friction field is known, and hence no friction updates are applied; in the other, we assume no prior knowledge of the friction field, requiring inversion for this parameter. With a known friction field, mass balance errors translate to ice volume errors as the 5th root, as is expected from theory for shear-dominated flow (Cuffey and Paterson, 2010). With friction updates, however, mass balance (and dh/dt) errors are 215 reinforced. Consider, for instance, a negative mass balance error: this yields overall smaller ice thicknesses and therefore reduced driving stresses, which in turn results in slower ice flow. To match the observed velocities, sliding must be increased when friction updates (eq.(3)) are applied, resulting in further thinning of the ice. The same effect but with opposite sign follows from a positive mass balance errors. Errors in the observed ice velocity induce similar mechanisms. A negative error here means that the bed is made more sticky 220 than it should be, resulting in too thick ice. 3.2.2 Inversion parameters We now investigate how different parameter choices for the inversion influence results. Specifically, we investigate the length 235 dt of the ice-flow model forward simulation, the relaxation parameter β (eq. (1)), the regularization parameter θ (eq. (2)) and the role of initial conditions. We find that our results show virtually no dependence on dt. Likewise, the recovered bed and friction field is not sensitive to their initial guess. For β, larger values lead to faster convergence, but lower detail in the recovered bed as compared to smaller values. This is because a large β can cause the bed to oscillate around its true value. The largest influence on the recovered bed 240 is θ which serves to balance observations and model physics in the presence of input errors as well as to regularize the solution. Larger expected errors in input data require larger θ to allow the model to accommodate these ’unreproducible’ characteristics of the data. As mentioned previously, however, this increases the dependence on initial conditions since it means that also parts of the non-erroneous observations are accommodated by surface updates rather than bed adjustments. Regardless of the quality of the input, we find that choosing θ > 0 is beneficial to avoid small bed irregularities in the solution which are not justified by 245 of the input, we find that choosing θ > 0 is beneficial to avoid small bed irregularities in the solution which are not justified by 245 the input data. 3.2.1 Input errors 205 For surface noise, we find that it changes the bed shape, with already small surface perturbations causing large bed re- 230 sponses, as is expected from theory (Raymond and Gudmundsson, 2005; Bahr et al., 2014). However, increasing the regular- ization parameter θ serves to mitigate these issues, although it makes the final bed shape more dependent on initial conditions. Furthermore, we find that the calculated final ice volume is largely insensitive to surface noise. 9 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. 4 Application to Kronebreen Next, we apply our method to the tidewater glacier Kronebreen on Svalbard. With Kronebreen we here refer to a combined glacier system consisting of the ice cap Holtedahlfonna, feeding the outlet glacier Kronebreen. The glacier system covers an area of ∼380 km2 and an elevation range from sea level to around 1400 m. Highest flow velocities of up to >1000 m a−1 250 are found on the highly crevassed glacier tongue (Schellenberger et al., 2015). Due to the large availability of in situ and remote sensing data, Kronebreen has in recent years been subject of numerous glaciological studies on ice dynamics (e.g. Schellenberger et al., 2015; Vallot et al., 2017), mass balance (e.g. van Pelt and Kohler, 2015; Deschamps-Berger et al., 2019), subglacial hydrology (How et al., 2017), calving (Luckman et al., 2015; Vallot et al., 2018), perennial firn aquifers (Christianson et al., 2015), seismicity (Köhler et al., 2019) and basal topography (Lindbäck et al., 2018). 255 4.1 Observational data The input data for the thickness inversion of Kronebreen comprises two DEMs based on Pleiades imagery from 2014 (Deschamps- Berger et al., 2019) and 2020 which we use to generate a dh/dt field dh dt obs = S2020−S2014 6 (Fig. 3). The 2020 DEM has some holes at higher elevations in the southern part of Kronebreen which we fill using 2014 data. Since dh/dt hence would be zero there, we generate a dh/dt field as a function of surface elevation which we use to patch those holes. Whether to use the 2014 260 DEM, the 2020 DEM or a mean of the two as input surface for the ice dynamics model is not an obvious choice. From a the- oretical point, the surface that gave rise to (or was the product of) the aggregate ice dynamics in the period 2014-2020 should be used. While this would favor using the mean of the two DEMs, it would imply using a surface that may not have existed ever. We therefore choose to use the 2020 DEM instead as this avoids complications at the terminus where Kronebreen has The input data for the thickness inversion of Kronebreen comprises two DEMs based on Pleiades imagery from 2014 (Deschamps- Berger et al., 2019) and 2020 which we use to generate a dh/dt field dh dt obs = S2020−S2014 6 (Fig. 3). The 2020 DEM has some holes at higher elevations in the southern part of Kronebreen which we fill using 2014 data. Since dh/dt hence would be zero holes at higher elevations in the southern part of Kronebreen which we fill using 2014 data. Since dh/dt hence would be zero there, we generate a dh/dt field as a function of surface elevation which we use to patch those holes. Whether to use the 2014 260 DEM, the 2020 DEM or a mean of the two as input surface for the ice dynamics model is not an obvious choice. From a the- oretical point, the surface that gave rise to (or was the product of) the aggregate ice dynamics in the period 2014-2020 should be used. While this would favor using the mean of the two DEMs, it would imply using a surface that may not have existed ever. 4.1 Observational data We therefore choose to use the 2020 DEM instead as this avoids complications at the terminus where Kronebreen has there, we generate a dh/dt field as a function of surface elevation which we use to patch those holes. Whether to use the 2014 260 DEM, the 2020 DEM or a mean of the two as input surface for the ice dynamics model is not an obvious choice. From a the- oretical point, the surface that gave rise to (or was the product of) the aggregate ice dynamics in the period 2014-2020 should be used. While this would favor using the mean of the two DEMs, it would imply using a surface that may not have existed ever. We therefore choose to use the 2020 DEM instead as this avoids complications at the terminus where Kronebreen has 260 10 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. retreated. There, differencing or taking a mean of the two DEMs would not yield a representative dh/dt field or glacier surface 265 because the difference and mean between the 2014 surface elevation and sea level (which is what the 2020 DEM shows) does not reflect the true ice loss or a realistic glacier surface shape. We also smooth the input DEM with a gaussian kernel prior to feeding it to the inversion algorithm. In terms of mass balance, we use observations from ten stakes placed approximately along the central flow line of Kronebreen and covering an altitude range from 505 m (mean over 2014-2020: -0.53 m w. eq. retreated. There, differencing or taking a mean of the two DEMs would not yield a representative dh/dt field or glacier surface 265 because the difference and mean between the 2014 surface elevation and sea level (which is what the 2020 DEM shows) does not reflect the true ice loss or a realistic glacier surface shape. We also smooth the input DEM with a gaussian kernel prior to feeding it to the inversion algorithm. In terms of mass balance, we use observations from ten stakes placed approximately along the central flow line of Kronebreen and covering an altitude range from 505 m (mean over 2014-2020: -0.53 m w. eq. yr−1) to 1116 m (mean: 0.96 m w. eq. yr−1) to generate a linear mass balance gradient which we apply over the whole altitu- 270 dinal range of Kronebreen. 4.1 Observational data Following general observations made at glaciers in Svalbard, we cap the maximum mass balance at 1 m yr−1 (similar to van Pelt and Kohler (2015)). Internal and basal mass balance are neglected. Finally, we use surface velocity observations from Millan et al. (2022) with a nominal date 2017-2018. Their performance is validated using measured stake velocities (mean difference: 2.4 m yr−1 / 3.5%). For validation of the modelled thicknesses, we use an extensive data set of measured bed elevations acquired using ground- 275 penetrating radar (Lindbäck et al., 2018). The errors associated with that data are on the order of 10 m where the ice is thin (mainly along the fast-flowing lower reaches of Kronebreen) and up to 30 m in the upper thick parts of Kronebreen. We resample the radar tracks to a grid of 400 m resolution to match it with the model results. For validation of the modelled thicknesses, we use an extensive data set of measured bed elevations acquired using ground- 275 penetrating radar (Lindbäck et al., 2018). The errors associated with that data are on the order of 10 m where the ice is thin (mainly along the fast-flowing lower reaches of Kronebreen) and up to 30 m in the upper thick parts of Kronebreen. We resample the radar tracks to a grid of 400 m resolution to match it with the model results. For validation of the modelled thicknesses, we use an extensive data set of measured bed elevations acquired using ground- 275 penetrating radar (Lindbäck et al., 2018). The errors associated with that data are on the order of 10 m where the ice is thin (mainly along the fast-flowing lower reaches of Kronebreen) and up to 30 m in the upper thick parts of Kronebreen. We resample the radar tracks to a grid of 400 m resolution to match it with the model results. 4.2 Modelling choices We set up a rectangular modelling domain with a regular grid spacing of 400 m. We define a glacier mask based on manually 280 drawn outlines of Kronebreen using satellite imagery and ice flow observations. Inside the mask, we allow ice to exist while in all other parts of the domain, the ice thickness is forced to zero by removing any ice existing there after each iteration. The mask also covers some small parts of what is believed to belong to the neighboring glacier Kongsbreen where we have a surface DEM from 2020. Since the boundary between Kronebreen and Kongsbreen is not perfectly known, some ice exchange We set up a rectangular modelling domain with a regular grid spacing of 400 m. We define a glacier mask based on manually 280 drawn outlines of Kronebreen using satellite imagery and ice flow observations. Inside the mask, we allow ice to exist while in all other parts of the domain, the ice thickness is forced to zero by removing any ice existing there after each iteration. The mask also covers some small parts of what is believed to belong to the neighboring glacier Kongsbreen where we have a surface DEM from 2020. Since the boundary between Kronebreen and Kongsbreen is not perfectly known, some ice exchange We set up a rectangular modelling domain with a regular grid spacing of 400 m. We define a glacier mask based on manually 280 drawn outlines of Kronebreen using satellite imagery and ice flow observations. Inside the mask, we allow ice to exist while in all other parts of the domain, the ice thickness is forced to zero by removing any ice existing there after each iteration. The mask also covers some small parts of what is believed to belong to the neighboring glacier Kongsbreen where we have a surface DEM from 2020. Since the boundary between Kronebreen and Kongsbreen is not perfectly known, some ice exchange here cannot be excluded. While our experiments indicate that this is a better approach than cutting the mask at the assumed 285 Kronebreen/Kongsbreen border (not shown), some boundary effects can still not be ruled out since we only model a small part of Kongsbreen, thus not accounting for the entire mass flux potentially reaching the Kronebreen/Kongsbreen border in reality. 4.2 Modelling choices At all other mask boundaries, we do not apply any flux constraints or lateral friction since the friction inversion will account for lateral drag if needed. here cannot be excluded. While our experiments indicate that this is a better approach than cutting the mask at the assumed 285 Kronebreen/Kongsbreen border (not shown), some boundary effects can still not be ruled out since we only model a small part of Kongsbreen, thus not accounting for the entire mass flux potentially reaching the Kronebreen/Kongsbreen border in reality. At all other mask boundaries, we do not apply any flux constraints or lateral friction since the friction inversion will account for lateral drag if needed. At the glacier terminus, we apply the hydrostatic water pressure. We also remove all floating ice and do not allow calving; 290 this is because we cut off all ice outside the mask after each iteration anyway, hence artificially ’calving’ all ice that has advanced beyond the known glacier extent. Furthermore, it needs to be ensured that in each forward simulation, the ice can advance into the ocean unblocked by the bathymetry in front of the glacier (which is unknown). We therefore apply a zero bed-slope in the flow direction at the ice-ocean boundary. Ice flow is described by the SSA+SIA hybrid scheme available in PISM (Bueler and Brown, 2009). Note that while the SSA 295 accounts for membrane stresses, the SIA does not and so ice flows strictly downhill in the absence of sliding (see sec. 4.5). Fur- thermore, we set an isothermal flow law with a fixed ice temperature of 0°C where the ice softness A = 3.96×10−24 s−1 Pa−3. While this ice temperature assumption is a simplification not valid everywhere, it follows from observations and modelling at 11 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure 3. Input data for Kronebreen inversion (coordinate system is UTM zone 33N). 4.2 Modelling choices a) Kronebreen outline in black and border to neigh- boring glacier Kongsbreen (KK border) in green; b) Location of Kronebreen in Svalbard (marked by red dot); c) observed rate of surface elevation change dh/dt as computed by subtracting two digital elevation models; d) climatic mass balance derived from extrapolating stake observations (locations shown by circles in main plot, and values by orange dots in inset) to the entire glacier using an elevation dependent linear model (inset plot); e) observed velocities from remote sensing (Millan et al., 2022) validated with stake observations (inset plot) Figure 3. Input data for Kronebreen inversion (coordinate system is UTM zone 33N). a) Kronebreen outline in black and border to neigh- boring glacier Kongsbreen (KK border) in green; b) Location of Kronebreen in Svalbard (marked by red dot); c) observed rate of surface elevation change dh/dt as computed by subtracting two digital elevation models; d) climatic mass balance derived from extrapolating stake observations (locations shown by circles in main plot, and values by orange dots in inset) to the entire glacier using an elevation dependent linear model (inset plot); e) observed velocities from remote sensing (Millan et al., 2022) validated with stake observations (inset plot) Figure 3. Input data for Kronebreen inversion (coordinate system is UTM zone 33N). a) Kronebreen outline in black and border to neigh- boring glacier Kongsbreen (KK border) in green; b) Location of Kronebreen in Svalbard (marked by red dot); c) observed rate of surface elevation change dh/dt as computed by subtracting two digital elevation models; d) climatic mass balance derived from extrapolating stake observations (locations shown by circles in main plot, and values by orange dots in inset) to the entire glacier using an elevation dependent linear model (inset plot); e) observed velocities from remote sensing (Millan et al., 2022) validated with stake observations (inset plot) many Svalbard glaciers where extensive meltwater refreezing in the firn keeps the ice close to the melting point (e.g. van Pelt et al., 2019). In the bare ice areas of the ablation zone and near the glacier edges, the largest deviations from this assumption 300 can be expected. We choose a linear sliding law by setting the parameters q = 1 and uth = 1 ms−1 in eq. (4). This causes the sliding velocity to be directly proportional to the velocity misfit (eq. (3)), thus stabilizing the inversion. et al., 2019). 4.2 Modelling choices In the bare ice areas of the ablation zone and near the glacier edges, the largest deviations from this assumption 300 can be expected. We choose a linear sliding law by setting the parameters q = 1 and uth = 1 ms−1 in eq. (4). This causes the sliding velocity to be directly proportional to the velocity misfit (eq. (3)), thus stabilizing the inversion. 300 4.4 Inversion parameters We set the forward-modelling time step dt = 0.01 yrs. This small value allows a fast inversion while our experiments show 315 that the results are not sensitive to this choice. For the scaling parameter β (eq. (1)), we apply a ’ramp-up’ procedure where we let β increase as a function of the iteration count c, such that β = −10·β0 c+10 + β0, where β0 = 0.5. After each friction update, c is reset to one. We do this since we know that the dh/dt misfit, and hence the bed corrections, are generally largest at the start of the inversion and after friction updates; to stabilize the inversion, we find that it is beneficial to scale the bed updates with a small β at those points. Since we assume that we have high-quality input data, we apply only weak regularization by 320 setting the parameter θ = 0.05, meaning that 5% of bed updates are changing the surface in the opposite direction. Based on our experiments, we see that after ∼8000 iterations, no more changes to bed or friction field occur, and so we stop the inversion there. We also see that it takes at most ∼1000 iterations after a friction update to converge to a solution, making this our interval for updates of τc (eq. (3)). with a small β at those points. Since we assume that we have high-quality input data, we apply only weak regularization by 320 setting the parameter θ = 0.05, meaning that 5% of bed updates are changing the surface in the opposite direction. Based on our experiments, we see that after ∼8000 iterations, no more changes to bed or friction field occur, and so we stop the inversion there. We also see that it takes at most ∼1000 iterations after a friction update to converge to a solution, making this our interval for updates of τc (eq. (3)). 4.3 Initial conditions We derive an initial bed guess by subtracting, from the input 2020 DEM, an initial ice thickness Hinit calculated with the perfect plasticity assumption (Nye, 1952), such that 305 (7) Hinit = τb ρgsinα, Hinit = τb ρgsinα, 12 12 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. where ρ is the ice density, g is gravitational acceleration, α is surface slope as computed from the input DEM and τb is the basal shear stress which we arbitrarily set to τb = 120 kPa. Based on the sensitivity experiments with the artificial ice cap, we do not expect the choice of τb to impact our results significantly. We iteratively smooth the derived thickness field over four times the ice thickness to account for longitudinal stress gradients (Kamb and Echelmeyer, 1986), and we apply a gaussian 310 kernel to remove small-scale variability not justified by the input data (Habermann et al., 2012). For basal friction, we derive an initial guess of τc in eq. (4) as τc = τduth uobs where τd is the driving stress at the start of the inversion calculated as τd = ρgHinit tanα (Vallot et al., 2017). where ρ is the ice density, g is gravitational acceleration, α is surface slope as computed from the input DEM and τb is the basal shear stress which we arbitrarily set to τb = 120 kPa. Based on the sensitivity experiments with the artificial ice cap, we do not expect the choice of τb to impact our results significantly. We iteratively smooth the derived thickness field over four times the ice thickness to account for longitudinal stress gradients (Kamb and Echelmeyer, 1986), and we apply a gaussian 10 kernel to remove small-scale variability not justified by the input data (Habermann et al., 2012). For basal friction, we derive an initial guess of τc in eq. (4) as τc = τduth uobs where τd is the driving stress at the start of the inversion calculated as τd = ρgHinit tanα (Vallot et al., 2017). 4.5 Inversion results 325 We find that the iterations consistently reduce the dh/dt misfit (eq. (1)), interrupted by friction updates, as expected (Fig. 4a). Owing to the surface updates, the misfit approaches zero in each interval between friction updates, showing that with our regularization method, no stopping criterion is needed. It turns out that our initial bed guess overestimated ice thickness in most parts of the domain as the bed generally becomes shallower. This is particularly true at the fast-flowing outlet glacier where the friction updates decrease basal drag and thus are instrumental in ensuring that ice thickness there is not overestimated. 330 The final mismatch between modelled and observed velocity magnitudes is generally on the order of a few percent where the glacier is sliding, while they can be higher in the non-sliding parts of the accumulation area (Fig. 4b). Overall, this indicates that the friction updates have helped to align model dynamics with reality. When comparing with observed bed elevations from radar measurements, we see that our ice thickness generally matches the observations well with deviations on the order of a few 10s of meters (Fig. 4). However, there is one larger area where our 335 ice thickness is greatly overestimated by more than 250 m (Fig. 4c). This area is a flat and slow-flowing part which, as shown by the observed flow-velocities, is only weakly connected to the main flow and hence only receives little ice from upstream (Fig. 4d). However, in our model, we see that there is a distinct ’flow channel’ diverging from the main flow through which ice is discharged into that area. Where the ’flow channel’ diverges in the model, basal velocities are close to zero, meaning that the observations well with deviations on the order of a few 10s of meters (Fig. 4). However, there is one larger area where our 335 ice thickness is greatly overestimated by more than 250 m (Fig. 4c). This area is a flat and slow-flowing part which, as shown by the observed flow-velocities, is only weakly connected to the main flow and hence only receives little ice from upstream (Fig. 4d). However, in our model, we see that there is a distinct ’flow channel’ diverging from the main flow through which ice is discharged into that area. 4.5 Inversion results 325 Where the ’flow channel’ diverges in the model, basal velocities are close to zero, meaning that 13 Fi 4 R f K b i i ) dh/dt ( (1)) d b d i fi ( (6)) i i b) V l i i fi i h l l https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure 4. Raw output of Kronebreen inversion. a) dh/dt (eq. (1)) and bed misfit (eq. (6)) per iteration. b) Velocity misfit with purple color indicating where modelled velocities exceed observed velocities by more than 50%. c) Difference between radar-derived and modelled bed elevation. d) Zoom into area with largest bed deviations (green box in middle row) for observed velocities, modelled velocities and surface aspect (with arrows denoting downhill direction, and colors showing into which sector the ice surface is sloping) Figure 4. Raw output of Kronebreen inversion. a) dh/dt (eq. (1)) and bed misfit (eq. (6)) per iteration. b) Velocity misfit with purple color indicating where modelled velocities exceed observed velocities by more than 50%. c) Difference between radar-derived and modelled bed elevation. d) Zoom into area with largest bed deviations (green box in middle row) for observed velocities, modelled velocities and surface aspect (with arrows denoting downhill direction, and colors showing into which sector the ice surface is sloping) internal shear controls ice flux. As mentioned before, in the absence of sliding, the shear-dominated flow calculated by the SIA 340 is strictly downhill. And indeed, we see that our input DEM is sloping towards the ’flow channel’ in that area (Fig. 4d). We hence conclude that in reality, there are likely membrane stresses forcing the ice to not flow downhill which are not resolved by our model physics, thus causing a mismatch in flow directions. Because of that, our model discharges ice into a slow and flat area where, as is known from theory (Gudmundsson, 2003; Raymond and Gudmundsson, 2005), errors in the flux will lead to internal shear controls ice flux. As mentioned before, in the absence of sliding, the shear-dominated flow calculated by the SIA 340 is strictly downhill. And indeed, we see that our input DEM is sloping towards the ’flow channel’ in that area (Fig. 4d). 4.5 Inversion results 325 Furthermore, note that us+ here refers to us + 0.5us to account for the uncertainty in velocities. 360 With this post-processing step, we are able to greatly reduce the errors in the modelled bed. Where we have identified and corrected the bed based on the mismatch between ice-deformational and observed velocities, errors have decreased by several hundreds of meters. We derive a final bed product by smoothing the computed bed with a gaussian kernel and so arrive at a mean absolute deviation from observations (eq. (6)) of 78 m (as compared to 108 m for the raw output prior to post-processing) infinity as α approaches zero. Furthermore, note that us+ here refers to us + 0.5us to account for the uncertainty in velocities. 360 With this post-processing step, we are able to greatly reduce the errors in the modelled bed. Where we have identified and corrected the bed based on the mismatch between ice-deformational and observed velocities, errors have decreased by several hundreds of meters. We derive a final bed product by smoothing the computed bed with a gaussian kernel and so arrive at a mean absolute deviation from observations (eq. (6)) of 78 m (as compared to 108 m for the raw output prior to post-processing) (Fig. 5). 365 From our input datasets, the least constrained one is arguably surface mass balance, since we rely on a few stakes that are extrapolated to the entire glacier. An ice core drilled in 2005 on an ice divide in the upper accumulation area (1200 m a.s.l.) of Kronebreen indeed shows considerably lower accumulation (0.5 m w.e. yr−1 during 1963-2005, (van der Wel et al., 2011)) than the values we use here (1 m w.e. yr−1), demonstrating that there is large spatial inhomogeneity, likely as a result of variability (Fig. 5). 365 From our input datasets, the least constrained one is arguably surface mass balance, since we rely on a few stakes that are extrapolated to the entire glacier. An ice core drilled in 2005 on an ice divide in the upper accumulation area (1200 m a.s.l.) of Kronebreen indeed shows considerably lower accumulation (0.5 m w.e. yr−1 during 1963-2005, (van der Wel et al., 2011)) than the values we use here (1 m w.e. yr−1), demonstrating that there is large spatial inhomogeneity, likely as a result of variability in wind exposure and associated snow redistribution. 4.5 Inversion results 325 We hence conclude that in reality, there are likely membrane stresses forcing the ice to not flow downhill which are not resolved by our model physics, thus causing a mismatch in flow directions. Because of that, our model discharges ice into a slow and flat area where, as is known from theory (Gudmundsson, 2003; Raymond and Gudmundsson, 2005), errors in the flux will lead to internal shear controls ice flux. As mentioned before, in the absence of sliding, the shear-dominated flow calculated by the SIA 340 is strictly downhill. And indeed, we see that our input DEM is sloping towards the ’flow channel’ in that area (Fig. 4d). We hence conclude that in reality, there are likely membrane stresses forcing the ice to not flow downhill which are not resolved by our model physics, thus causing a mismatch in flow directions. Because of that, our model discharges ice into a slow and flat area where, as is known from theory (Gudmundsson, 2003; Raymond and Gudmundsson, 2005), errors in the flux will lead to l i th b d N t th t f i ti d t f h l h i th i i t thi k i th i l iti i th 345 large errors in the bed. Note that friction updates are of no help here: since the ice is too thick in the area, ice velocities in the 345 14 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. model are larger than in reality due to enhanced internal deformation, which means that a friction update makes the bed more sticky (even if there already is no sliding) - while the opposite would be needed to lift up the bed. model are larger than in reality due to enhanced internal deformation, which means that a friction update makes the bed more sticky (even if there already is no sliding) - while the opposite would be needed to lift up the bed. This being said, we can still use the velocity observations to identify areas that are likely too thick. We do so by scanning our model domain for areas where the ice velocity through internal deformation exceeds the observed velocities. Assuming that our model domain for areas where the ice velocity through internal deformation exceeds the observed velocities. 4.5 Inversion results 325 Assuming that our ice viscosity estimate is correct, a too-large ice thickness is the only first-order control on ice dynamics not constrained by the 350 inputs that could induce higher velocities than observed in the absence of basal sliding. We apply this logic and conservatively only mark areas where the observed ice velocity is exceeded by at least 50% to account for errors in the velocity observations (purple areas in Fig. 4b). And indeed, this reliably identifies the area where the mismatch in flow directions induces the largest errors in the modelled bed, as well as picking out other locations where the ice thickness is overestimated (as shown by the ice viscosity estimate is correct, a too-large ice thickness is the only first-order control on ice dynamics not constrained by the 350 inputs that could induce higher velocities than observed in the absence of basal sliding. We apply this logic and conservatively only mark areas where the observed ice velocity is exceeded by at least 50% to account for errors in the velocity observations (purple areas in Fig. 4b). And indeed, this reliably identifies the area where the mismatch in flow directions induces the largest errors in the modelled bed, as well as picking out other locations where the ice thickness is overestimated (as shown by the 350 radar validation). To update the bed elevation estimate in this area, we turn towards an analytical ice thickness equation based 355 on the SIA (e.g. Millan et al., 2022) in which ice thickness H can be inferred from observed velocities such that radar validation). To update the bed elevation estimate in this area, we turn towards an analytical ice thickness equation based 355 on the SIA (e.g. Millan et al., 2022) in which ice thickness H can be inferred from observed velocities such that H = (us+ −ub)(n + 1) 2A(ρgα)n 1 n+1 (8) H = (us+ −ub)(n + 1) 2A(ρgα)n 1 n+1 (8) with n = 3 being Glen´s flow law exponent (Glen, 1955) and us and ub surface and basal velocity, respectively. Note that this equation requires a minimum slope α i which we set to 0 025 (e g van Pelt et al 2013) to prevent ice thickness going to with n = 3 being Glen´s flow law exponent (Glen, 1955) and us and ub surface and basal velocity, respectively. Note that infinity as α approaches zero. 4.5 Inversion results 325 This is also known from other Svalbard glaciers (e.g. van Pelt et al., 370 2014). The velocity misfit map of our raw output indicates that in large parts in the accumulation area the ice is flowing too fast despite no or very limited basal sliding. While this could also be the result of a too-high ice viscosity (note, however, that we already use an ice temperature of 0°C), a likely explanation is a too positive mass balance which induces too thick ice and hence too fast surface motion by internal deformation. We therefore repeat our thickness inversion after multiplying the mass in wind exposure and associated snow redistribution. This is also known from other Svalbard glaciers (e.g. van Pelt et al., 370 2014). The velocity misfit map of our raw output indicates that in large parts in the accumulation area the ice is flowing too fast despite no or very limited basal sliding. While this could also be the result of a too-high ice viscosity (note, however, that we already use an ice temperature of 0°C), a likely explanation is a too positive mass balance which induces too thick ice and hence too fast surface motion by internal deformation. We therefore repeat our thickness inversion after multiplying the mass balance by 0.8. Indeed, this yields a better result with a mean absolute deviation of 100 m (was 108 m) before post-processing. 375 balance by 0.8. Indeed, this yields a better result with a mean absolute deviation of 100 m (was 108 m) before post-processing. 375 The velocity misfit in the accumulation area is also reduced. 15 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure 5. Final output after ’post-processing’. a) Modelled ice thickness. b) bed elevation error as compared to radar observations, with mean absolute error (eq.(6)) of 78 m. c) Correlation between radar-derived and modelled ice thickness. Figure 5. Final output after ’post-processing’. a) Modelled ice thickness. b) bed elevation error as compared to radar observations, with mean absolute error (eq.(6)) of 78 m. c) Correlation between radar-derived and modelled ice thickness. 5 Discussion and conclusions We here presented a new thickness inversion method and tested it on both a synthetic ice cap as well as the tidewater glacier Kronebreen in Svalbard. The input datasets are easily-obtainable surface observations where ready-to-use products already exist on a global scale (dh/dtobs by Hugonnet et al. (2021), surface velocities by e.g. Millan et al. (2022), DEMs from e.g. 380 NASA JPL (2020), glacier outlines from RGI Consortium (2017), mass balance by Rounce et al. (2023)). Observations of bed height are not required. The key strength of the method lies in the fact that it produces distributed maps of ice thickness and basal friction using any level of complexity in ice physics, while being computationally cheap. In essence, the inversion output is not only a bed and friction map, but a fully spun-up glacier model that reproduces the present dynamic state of the modelled glacier as constrained by observations. Therefore, using the output for prognostic simulations is possible without requiring any 385 f h d l l i We here presented a new thickness inversion method and tested it on both a synthetic ice cap as well as the tidewater glacier Kronebreen in Svalbard. The input datasets are easily-obtainable surface observations where ready-to-use products already exist on a global scale (dh/dtobs by Hugonnet et al. (2021), surface velocities by e.g. Millan et al. (2022), DEMs from e.g. 380 NASA JPL (2020), glacier outlines from RGI Consortium (2017), mass balance by Rounce et al. (2023)). Observations of bed height are not required. The key strength of the method lies in the fact that it produces distributed maps of ice thickness and basal friction using any level of complexity in ice physics, while being computationally cheap. In essence, the inversion output is not only a bed and friction map, but a fully spun-up glacier model that reproduces the present dynamic state of the modelled glacier as constrained by observations. Therefore, using the output for prognostic simulations is possible without requiring any 385 further model relaxation. Although always converging to a bed solution, the experiments with the synthetic ice cap showed that errors in input data can have varying impacts on the modelled bed depending on the inversion setup. With friction updates, the method is less sensitive to errors in ice temperature than without friction updates, since the modelled basal friction field can compensate for ice viscosity errors. 5 Discussion and conclusions However, the method is more sensitive to errors in mass balance and dh/dt with friction updates since 390 an underestimation (overestimation) of mass flux and hence ice thickness reduces (increases) internal deformation, prompting 380 Although always converging to a bed solution, the experiments with the synthetic ice cap showed that errors in input data can have varying impacts on the modelled bed depending on the inversion setup. With friction updates, the method is less sensitive to errors in ice temperature than without friction updates, since the modelled basal friction field can compensate for Although always converging to a bed solution, the experiments with the synthetic ice cap showed that errors in input data can have varying impacts on the modelled bed depending on the inversion setup. With friction updates, the method is less sensitive to errors in ice temperature than without friction updates, since the modelled basal friction field can compensate for ice viscosity errors. However, the method is more sensitive to errors in mass balance and dh/dt with friction updates since 390l Although always converging to a bed solution, the experiments with the synthetic ice cap showed that errors in input data can have varying impacts on the modelled bed depending on the inversion setup. With friction updates, the method is less sensitive to errors in ice temperature than without friction updates, since the modelled basal friction field can compensate for ice viscosity errors. However, the method is more sensitive to errors in mass balance and dh/dt with friction updates since 390 an underestimation (overestimation) of mass flux and hence ice thickness reduces (increases) internal deformation, prompting sensitive to errors in ice temperature than without friction updates, since the modelled basal friction field can compensate for ice viscosity errors. However, the method is more sensitive to errors in mass balance and dh/dt with friction updates since 390 an underestimation (overestimation) of mass flux and hence ice thickness reduces (increases) internal deformation, prompting ice viscosity errors. However, the method is more sensitive to errors in mass balance and dh/dt with friction updates since 390 an underestimation (overestimation) of mass flux and hence ice thickness reduces (increases) internal deformation, prompting 16 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. changes in the friction field that further exacerbate the error. Likewise, errors in velocities reinforce themselves. 5 Discussion and conclusions Because of that, we argue that friction updates should be used with caution, i.e. only when it is likely that sliding is important for a specific glacier and when the input data is of sufficient quality. Indeed, for both the synthetic ice cap and Kronebreen, not doing friction updates would have required us to assume a friction field without any knowledge of whether it is reflective of 395 the ’true’ basal conditions or not, hence resulting in a somewhat arbitrary bed shape. For slow-flowing mountain glaciers with only poor velocity observations, however, it may be more meaningful to assume a friction field or simply not allow any sliding. In the case of non-sliding mountain glaciers with high-quality velocity data, velocity observations could be used to tune the ice temperature. Regarding errors in an input DEM, we showed that our regularization methodology involving small adjustments to changes in the friction field that further exacerbate the error. Likewise, errors in velocities reinforce themselves. Because of that, we argue that friction updates should be used with caution, i.e. only when it is likely that sliding is important for a specific glacier and when the input data is of sufficient quality. Indeed, for both the synthetic ice cap and Kronebreen, not doing friction updates would have required us to assume a friction field without any knowledge of whether it is reflective of 395 the ’true’ basal conditions or not, hence resulting in a somewhat arbitrary bed shape. For slow-flowing mountain glaciers with only poor velocity observations, however, it may be more meaningful to assume a friction field or simply not allow any sliding. In the case of non-sliding mountain glaciers with high-quality velocity data, velocity observations could be used to tune the ice temperature. Regarding errors in an input DEM, we showed that our regularization methodology involving small adjustments to the surface in each iteration is capable of producing a reasonable bed shape even in the presence of considerable surface errors, 400 albeit with less spatial detail and stronger dependence on initial conditions for larger values of the regularization parameter θ. Generally, the comparatively large sensitivity of a bed inversion to the input surface is an inherent problem that follows from the ill-posed nature of all thickness inversions, and is thus largely unavoidable (Bahr et al., 2014). Code and data availability. PISM v.2.0.4 is freely available for download at http://www.pism.io/. Note that PISM needs to be built including PISM´s Python bindings for optimal computational performance. Key scripts needed to perform the simulations done in this study will be made publicly available on GitHub once the manuscript is published. Radar data for Kronebreen are available at https://doi.org/10.21334/ npolar.2017.702ca4a7(Lindbäck et al., 2018), and ice velocity at https://doi.org/10.6096/1007 (Millan et al., 2022) 5 Discussion and conclusions Since global DEMs are generally quite accurate though (Chen et al., 2022), we anticipate that surface elevation may be a smaller problem for regional shape. As such, unregularized data and model errors are subsumed in the final bed as seen for the Kronebreen experiments. 410 While it is possible to estimate how errors in the input data are propagated to bed errors, this is unfortunately not true for model errors which are difficult to quantify. Using our ’post-processing’ approach as outlined in sec. 4.5, however, allows alleviation of some of the possible errors. Such an approach represents an interference with model physics in favor of a more accurate bed map, and should therefore be applied when the modelled bed likely contains significant errors. This may be assessed based on shape. As such, unregularized data and model errors are subsumed in the final bed as seen for the Kronebreen experiments. 410 While it is possible to estimate how errors in the input data are propagated to bed errors, this is unfortunately not true for model errors which are difficult to quantify. Using our ’post-processing’ approach as outlined in sec. 4.5, however, allows alleviation of some of the possible errors. Such an approach represents an interference with model physics in favor of a more accurate bed map, and should therefore be applied when the modelled bed likely contains significant errors. This may be assessed based on i the mismatch in observed and modelled velocities as shown here. If only minor deviations are found, we propose to refrain 415 from any post-processing to conserve the consistency between bed, model physics and input observations. We do not consider thermodynamics in this study, but use a constant ice viscosity parameter. This could account for some errors that we see at Kronebreen such that, for instance, a locally stiffer ice body might have resulted in slightly different modelled flow directions in those places where we find the largest errors. How to implement thermodynamics in our inversion methodology should hence be explored in future work. 420 the mismatch in observed and modelled velocities as shown here. If only minor deviations are found, we propose to refrain 415 from any post-processing to conserve the consistency between bed, model physics and input observations. We do not consider thermodynamics in this study, but use a constant ice viscosity parameter. 5 Discussion and conclusions This could account for some errors that we see at Kronebreen such that, for instance, a locally stiffer ice body might have resulted in slightly different modelled flow directions in those places where we find the largest errors. How to implement thermodynamics in our inversion methodology should hence be explored in future work. 420 We do not consider thermodynamics in this study, but use a constant ice viscosity parameter. This could account for some errors that we see at Kronebreen such that, for instance, a locally stiffer ice body might have resulted in slightly different modelled flow directions in those places where we find the largest errors. How to implement thermodynamics in our inversion methodology should hence be explored in future work. 420 Code and data availability. PISM v.2.0.4 is freely available for download at http://www.pism.io/. Note that PISM needs to be built including PISM´s Python bindings for optimal computational performance. Key scripts needed to perform the simulations done in this study will be made publicly available on GitHub once the manuscript is published. Radar data for Kronebreen are available at https://doi.org/10.21334/ npolar.2017.702ca4a7(Lindbäck et al., 2018), and ice velocity at https://doi.org/10.6096/1007 (Millan et al., 2022) 17 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Appendix A 425 Below, we elaborate further and present figures on the sensitivity experiments described in sec. 3.2. A1 Input errors This effect, however, is only present in those areas where the glacier is sliding. In non-sliding areas, a 445 further increase in the friction field will not lead to a thickness increase, putting an upper bound on this positive feedback effect. To quantify the effects of flux errors when friction updates are applied is, therefore, more difficult than when no friction updates are applied. Both the ratio between sliding and non-sliding areas in the glacier as well as the sliding magnitude control the final errors. Generally though, due to the above, an underestimation of ice flux introduces larger errors than an overestimation, and end of the inversion. This effect, however, is only present in those areas where the glacier is sliding. In non-sliding areas, a 445 further increase in the friction field will not lead to a thickness increase, putting an upper bound on this positive feedback effect. To quantify the effects of flux errors when friction updates are applied is, therefore, more difficult than when no friction updates are applied. Both the ratio between sliding and non-sliding areas in the glacier as well as the sliding magnitude control the final errors. Generally though, due to the above, an underestimation of ice flux introduces larger errors than an overestimation, and fi d th t f tifi i l i ti b l t l t b t li l i t l hil iti 450 end of the inversion. This effect, however, is only present in those areas where the glacier is sliding. In non-sliding areas, a 445 further increase in the friction field will not lead to a thickness increase, putting an upper bound on this positive feedback effect. To quantify the effects of flux errors when friction updates are applied is, therefore, more difficult than when no friction updates are applied. Both the ratio between sliding and non-sliding areas in the glacier as well as the sliding magnitude control the final errors. Generally though, due to the above, an underestimation of ice flux introduces larger errors than an overestimation, and Next, we assume errors in the velocity field by increasing and reducing the reference velocities by 25%, 50% and 75% uniformly throughout our domain (Fig. A2). A1 Input errors We start by perturbing the mass balance such that it is reduced and increased uniformly in the domain by 25%, 50% and 75%. Mass balance and dh/dt jointly control the mass flux through the glacier (and can even be lumped together into one variable, following the concept of the apparent mass balance (Farinotti et al., 2009)), and so errors in either of them have the same effect. 430 Therefore, we do not investigate them separately here, but focus on mass balance alone. A reduction in flux should lead to a We start by perturbing the mass balance such that it is reduced and increased uniformly in the domain by 25%, 50% and 75%. M b l d dh/dt j i tl t l th fl th h th l i ( d b l d t th i t i bl We start by perturbing the mass balance such that it is reduced and increased uniformly in the Mass balance and dh/dt jointly control the mass flux through the glacier (and can even be lum Mass balance and dh/dt jointly control the mass flux through the glacier (and can even be lumped together into one variable, following the concept of the apparent mass balance (Farinotti et al., 2009)), and so errors in either of them have the same effect. 430 Therefore, we do not investigate them separately here, but focus on mass balance alone. A reduction in flux should lead to a thinner glacier and shallower bed while the opposite should be the case for an increase in flux, as observed in our experiments (Fig. A1). From the integrated form of Glen´s flow law (Cuffey and Paterson, 2010) for shear-dominated flow following the concept of the apparent mass balance (Farinotti et al., 2009)), and so errors in either of them have the same effect. 430 Therefore, we do not investigate them separately here, but focus on mass balance alone. A reduction in flux should lead to a thinner glacier and shallower bed while the opposite should be the case for an increase in flux, as observed in our experiments (Fig. A1). From the integrated form of Glen´s flow law (Cuffey and Paterson, 2010) for shear-dominated flow H = n+2 s q 2A · n + 2 (ρgsinα)n , (A1) it can be seen that flux (denoted by q) errors translate into thickness errors as the 5th root. A1 Input errors If we prescribe the true friction 435 field at the start of the inversion and do not apply any updates to it, the computed ice volumes using the perturbed mass balances follow this relationship accurately. However, with friction updates, there are positive feedback effects which induce larger errors. Since a reduction in mass flux means that the calculated ice thickness is smaller than its true value, the driving stresses are smaller, resulting in slower ice flow. To match the observed velocities, sliding must hence be increased when it can be seen that flux (denoted by q) errors translate into thickness errors as the 5th root. If we prescribe the true friction 435 field at the start of the inversion and do not apply any updates to it, the computed ice volumes using the perturbed mass balances follow this relationship accurately. However, with friction updates, there are positive feedback effects which induce larger errors. Since a reduction in mass flux means that the calculated ice thickness is smaller than its true value, the driving stresses are smaller, resulting in slower ice flow. To match the observed velocities, sliding must hence be increased when friction updates (eq. (3)) are applied, resulting in a further thinning of the ice through associated bed uplift. Therefore, with 440 friction updates the final ice volume is smaller than what might be expected from the mass balance underestimation alone. The opposite is the case when the mass flux through the glacier is overestimated. In this case, the overestimated ice thickness induces larger driving stresses and hence an overestimation in surface velocities, prompting the friction field to be made more sticky during a friction update. This induces further bed lowering, resulting in a stronger overestimation of ice thickness at the friction updates (eq. (3)) are applied, resulting in a further thinning of the ice through associated bed uplift. Therefore, with 440 friction updates the final ice volume is smaller than what might be expected from the mass balance underestimation alone. The opposite is the case when the mass flux through the glacier is overestimated. In this case, the overestimated ice thickness induces larger driving stresses and hence an overestimation in surface velocities, prompting the friction field to be made more sticky during a friction update. This induces further bed lowering, resulting in a stronger overestimation of ice thickness at the end of the inversion. A1 Input errors a) Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of mass balance perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. Triangles represent runs where the true friction field was prescribed from the start of the inversion, meaning that no friction updates were applied. Dots denote runs where the friction field was not known a priori, and therefore friction updates according to eq. (3) were applied. The dashed line is the 5th root of the mass balance errors, i.e. the theoretical relationship between mass balance and ice volume errors in the absence of friction updates; final inverted bed topography for the largest positive and negative (i.e. +-75%) mass balance perturbation when friction updates are applied shown in b) and c), respectively. stresses, inducing smaller ice deformational velocities and hence a larger negative velocity misfit as compared to a situation with the correct friction field. This results in further updates that reduce the basal friction, thin the ice and lift up the bed. For a negative error in the ’observed’ velocity, however, again an upper bound in final thickness errors is found. This is because lower observed velocities increase friction, but this is only relevant to the point until no more sliding at the base occurs. Further stresses, inducing smaller ice deformational velocities and hence a larger negative velocity misfit as compared to a situation with the correct friction field. This results in further updates that reduce the basal friction, thin the ice and lift up the bed. For a negative error in the ’observed’ velocity, however, again an upper bound in final thickness errors is found. This is because lower observed velocities increase friction, but this is only relevant to the point until no more sliding at the base occurs. Further 460 strengthening of the bed has no consequences, as is particularly well visible in the non-sliding (right) half of the domain where 460 the correct bed is found even when ’observed’ velocities are underestimated. As discussed previously, it is difficult to generalize the impact of these complicating effects on ice volume estimates as they depend on sliding extent and magnitude. A1 Input errors Generally, a positive error in ’observed’ velocities should lead to a shallower bed and hence a thinner ice cap because faster ’observed’ velocities mean that the bed must be more slippery under the same mass flux. However, again a positive feedback mechanism acts on top of that: a thinning of the ice leads to a reduction in driving 455 18 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure A1. a) Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of mass balance perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. Triangles represent runs where the true friction field was prescribed from the start of the inversion, meaning that no friction updates were applied. Dots denote runs where the friction field was not known a priori, and therefore friction updates according to eq. (3) were applied. The dashed line is the 5th root of the mass balance errors, i.e. the theoretical relationship between mass balance and ice volume errors in the absence of friction updates; final inverted bed topography for the largest positive and negative (i.e. +-75%) mass balance perturbation when friction updates are applied shown in b) and c), respectively. Figure A1. a) Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of mass balance perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. Triangles represent runs where the true friction field was prescribed from the start of the inversion, meaning that no friction updates were applied. Dots denote runs where the friction field was not known a priori, and therefore friction updates according to eq. (3) were applied. The dashed line is the 5th root of the mass balance errors, i.e. the theoretical relationship between mass balance and ice volume errors in the absence of friction updates; final inverted bed topography for the largest positive and negative (i.e. +-75%) mass balance perturbation when friction updates are applied shown in b) and c), respectively. Figure A1. A1 Input errors To show the unmitigated effects of surface noise for the purpose of this sensitivity analysis, we do not smooth the input data, and only apply the same weak regularization as in all other experiments shown here by setting the parameter θ = 2.5% in eq. (2). Indeed, Fig. A3 demonstrates that already with little surface noise, considerable bed noise is introduced, although the general bed shape can still be seen. As predicted by thicker and non-sliding ice, while thin sliding ice is affected to a lesser degree. Smoothing of the input data or regularization 470 tools (e.g. eq. (2)) can mitigate problems arising from surface noise. To show the unmitigated effects of surface noise for the purpose of this sensitivity analysis, we do not smooth the input data, and only apply the same weak regularization as in all other experiments shown here by setting the parameter θ = 2.5% in eq. (2). Indeed, Fig. A3 demonstrates that already with little surface noise, considerable bed noise is introduced, although the general bed shape can still be seen. As predicted by theory, the sliding (left) half of the domain is less affected than the non-sliding right half. For large surface noise, the bed shape 475 is essentially unrecognizable, as is exemplified by a mean absolute bed deviation above 200. However, Fig. A3 also shows that the modelled ice cap volume is largely unaffected by surface noise, as is expected when the mass flux through the domain and the reference velocity field are unaltered. Next, we consider ice temperature errors (Fig. A4). Specifically, we change the ice temperature from its true value (used to Next, we consider ice temperature errors (Fig. A4). Specifically, we change the ice temperature from its true value (used to build the ice cap) of 268 K to values of 264 K, 266 K, 270 K and 272 K. With the isothermal flow law used here, this means 480 that the ice is uniformly made stiffer or softer as compared to its true viscosity. Colder ice deforms less, and so a thicker ice Next, we consider ice temperature errors (Fig. A4). Specifically, we change the ice temperature from its true value (used to build the ice cap) of 268 K to values of 264 K, 266 K, 270 K and 272 K. A1 Input errors In the case of the ice cap, we find that even a small underestimation of, say 25% of the flow velocities already results in no more sliding, thus reaching the upper ice volume error limit of about 25% volume overestimation (Fig. A2). Furthermore, we simulate how errors in the glacier surface shape translate into bed errors. To that end, we add random 465 noise with a standard deviation of 5 m, 10 m, 15 m, 25 m and 50 m to the ice cap surface. From theory, it is well-known that surface shape errors can lead to large bed elevation errors since bed undulations typically leave a surface expression that is only Furthermore, we simulate how errors in the glacier surface shape translate into bed errors. To that end, we add random 465 noise with a standard deviation of 5 m, 10 m, 15 m, 25 m and 50 m to the ice cap surface. From theory, it is well-known that surface shape errors can lead to large bed elevation errors since bed undulations typically leave a surface expression that is only 19 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure A2. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of velocity perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. The dashed line denotes the ice volume if the inversion is run without allowing any sliding anywhere; the final inverted bed topography for the largest positive and negative (i.e. +-75%) velocity perturbation shown in b) and c), respectively. Figure A2. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of velocity perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. The dashed line denotes the ice volume if the inversion is run without allowing any sliding anywhere; the final inverted bed topography for the largest positive and negative (i.e. +-75%) velocity perturbation shown in b) and c), respectively. a fraction of their magnitude (Raymond and Gudmundsson, 2005; Habermann et al., 2012; Bahr et al., 2014). Even a small surface bump, therefore, will be interpreted as a large bed undulation by the inversion. A1 Input errors Generally, this effect is stronger for thicker and non-sliding ice, while thin sliding ice is affected to a lesser degree. Smoothing of the input data or regularization 470 tools (e.g. eq. (2)) can mitigate problems arising from surface noise. To show the unmitigated effects of surface noise for the purpose of this sensitivity analysis, we do not smooth the input data, and only apply the same weak regularization as in all other experiments shown here by setting the parameter θ = 2.5% in eq. (2). Indeed, Fig. A3 demonstrates that already with little surface noise, considerable bed noise is introduced, although the general bed shape can still be seen. As predicted by theory, the sliding (left) half of the domain is less affected than the non-sliding right half. For large surface noise, the bed shape 475 is essentially unrecognizable, as is exemplified by a mean absolute bed deviation above 200. However, Fig. A3 also shows that th d ll d i l i l l ff t d b f i i t d h th fl th h th d i d a fraction of their magnitude (Raymond and Gudmundsson, 2005; Habermann et al., 2012; Bahr et al., 2014). Even a small surface bump, therefore, will be interpreted as a large bed undulation by the inversion. Generally, this effect is stronger for thicker and non sliding ice while thin sliding ice is affected to a lesser degree Smoothing of the input data or regularization 0 a fraction of their magnitude (Raymond and Gudmundsson, 2005; Habermann et al., 2012; Bahr et al., 2014). Even a small surface bump, therefore, will be interpreted as a large bed undulation by the inversion. Generally, this effect is stronger for a fraction of their magnitude (Raymond and Gudmundsson, 2005; Habermann et al., 2012; Bahr et al., 2014). Even a small surface bump, therefore, will be interpreted as a large bed undulation by the inversion. Generally, this effect is stronger for thicker and non-sliding ice, while thin sliding ice is affected to a lesser degree. Smoothing of the input data or regularization 470 tools (e.g. eq. (2)) can mitigate problems arising from surface noise. A1 Input errors This is because the friction updates compensate for too stiff or too soft ice by reducing the subglacial friction if the ice is too slow, thinning the glacier, or by increasing friction if the ice is too cap can be expected from a negative ice temperature error, whereas the opposite is the case for a positive ice temperature error. However, the results of the perturbation analysis suggest that this is not the case. Rather, only small errors in final ice volume below 10% are introduced by perturbing ice temperature. This is because the friction updates compensate for too stiff or too cap can be expected from a negative ice temperature error, whereas the opposite is the case for a positive ice temperature error. However, the results of the perturbation analysis suggest that this is not the case. Rather, only small errors in final ice volume below 10% are introduced by perturbing ice temperature. This is because the friction updates compensate for too stiff or too soft ice by reducing the subglacial friction if the ice is too slow, thinning the glacier, or by increasing friction if the ice is too 485 fast (again, however, if there is no sliding at all, this does not have any effect). Indeed, when prescribing the true friction field at the start of the inversion and not applying any updates to it, larger errors in final ice volume are apparent for both too cold and too warm ice, as compared to when friction updates were enabled (Fig. A4). A1 Input errors With the isothermal flow law used here, this means 480 that the ice is uniformly made stiffer or softer as compared to its true viscosity. Colder ice deforms less, and so a thicker ice 20 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure A3. Modelled ice volume after the inversion relative to true ice volume as a function of surface shape perturbations with random noise. Colors indicate the mean absolute deviation between the inverted bed and the true bed; final inverted bed topography for the smallest and largest surface shape perturbation shown in b) and c), respectively. Figure A3. Modelled ice volume after the inversion relative to true ice volume as a function of surface shape perturbations with random noise. Colors indicate the mean absolute deviation between the inverted bed and the true bed; final inverted bed topography for the smallest and largest surface shape perturbation shown in b) and c), respectively. cap can be expected from a negative ice temperature error, whereas the opposite is the case for a positive ice temperature error. However, the results of the perturbation analysis suggest that this is not the case. Rather, only small errors in final ice volume below 10% are introduced by perturbing ice temperature. This is because the friction updates compensate for too stiff or too soft ice by reducing the subglacial friction if the ice is too slow, thinning the glacier, or by increasing friction if the ice is too 5 fast (again, however, if there is no sliding at all, this does not have any effect). Indeed, when prescribing the true friction field at the start of the inversion and not applying any updates to it, larger errors in final ice volume are apparent for both too cold and too warm ice, as compared to when friction updates were enabled (Fig. A4). cap can be expected from a negative ice temperature error, whereas the opposite is the case for a positive ice temperature error. However, the results of the perturbation analysis suggest that this is not the case. Rather, only small errors in final ice volume below 10% are introduced by perturbing ice temperature. A2 Inversion parameters 264 K and 272 K, respectively) ice temperature perturbation when friction updates are applied is shown in b) and c), respectively. Figure A4. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of ice temperature perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. Triangles represent runs where the true friction field was prescribed from the start of the inversion, meaning that no friction updates were applied. Dots denote runs where the friction field was not known a priori, and therefore friction updates according to eq. (3) were applied; the final inverted bed topography for the largest positive and negative (i.e. 264 K and 272 K, respectively) ice temperature perturbation when friction updates are applied is shown in b) and c), respectively. they are overcompensated, resulting in a bed oscillating around its true value during the iterations. Conversely, a small β slows down the inversion, but results in a better resolved bed. Our sensitivity experiments confirm this. Indeed, with a larger β the bed misfit (eq. (6)) is quick to converge towards a constant value after each friction update while this is substantially slower for small β. However, the final inverted bed has a larger remaining misfit for a large than for a small β (Fig. A6). they are overcompensated, resulting in a bed oscillating around its true value during the iterations. Conversely, a small β slows down the inversion, but results in a better resolved bed. Our sensitivity experiments confirm this. Indeed, with a larger β the bed misfit (eq. (6)) is quick to converge towards a constant value after each friction update while this is substantially slower for small β. However, the final inverted bed has a larger remaining misfit for a large than for a small β (Fig. A6). Now, we investigate different choices of the regularization parameter θ in eq. (2) (Fig. A7). As discussed previously, θ 500 serves the purpose of both regularizing the solution as well as allowing the inversion to converge when input data and model physics otherwise cannot balance, e.g. due to errors in the input data that are not reproducible with ice-flow physics. In the presence of larger errors, a larger θ will be required, because this gives the model flexibility to accommodate these erroneous observations. A2 Inversion parameters We now investigate how different parameter choices for the inversion influence results. We first look at different values for the 490 length dt of the ice-flow forward simulation. We explore values of 1 yr, 0.5 yr, 0.25 yr, 0.15 yr, 0.05 yr, 0.025 yr and 0.01 yr. Our results show virtually no difference in the final bed shape or remaining bed misfit (fig. A5), such that using small values for dt can be used to save computational time. Furthermore, we investigate different values for the relaxation parameter β in eq. (1) (Fig. A6). β scales the dh/dt misfit into bed corrections. If β is large, the inversion should progress faster, but smaller bed features might not be resolved since 495 21 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure A4. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of ice temperature perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. Triangles represent runs where the true friction field was prescribed from the start of the inversion, meaning that no friction updates were applied. Dots denote runs where the friction field was not known a priori, and therefore friction updates according to eq. (3) were applied; the final inverted bed topography for the largest positive and negative (i.e. 264 K and 272 K, respectively) ice temperature perturbation when friction updates are applied is shown in b) and c), respectively. Figure A4. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of ice temperature perturbations. Figure A4. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of ice temperature perturbations. Colors indicate the mean absolute deviation between the inverted bed and the true bed. Triangles represent runs where the true friction field was prescribed from the start of the inversion, meaning that no friction updates were applied. Dots denote runs where the friction field was not known a priori, and therefore friction updates according to eq. (3) were applied; the final inverted bed topography for the largest positive and negative (i.e. A2 Inversion parameters The choice of θ may thus be regarded as a measure for how well the ice-flow model is deemed to be able to reproduce observations. At the same time, however, a larger θ also increases the dependence on initial conditions because it 505 means that a larger part of the dh/dt misfit is not accommodated with bed updates, but with surface adjustments. We start by Now, we investigate different choices of the regularization parameter θ in eq. (2) (Fig. A7). As discussed previously, θ 500 serves the purpose of both regularizing the solution as well as allowing the inversion to converge when input data and model physics otherwise cannot balance, e.g. due to errors in the input data that are not reproducible with ice-flow physics. In the presence of larger errors, a larger θ will be required, because this gives the model flexibility to accommodate these erroneous observations. The choice of θ may thus be regarded as a measure for how well the ice-flow model is deemed to be able to 500 reproduce observations. At the same time, however, a larger θ also increases the dependence on initial conditions because it 505 means that a larger part of the dh/dt misfit is not accommodated with bed updates, but with surface adjustments. We start by testing the regularization effect of θ by varying it without perturbing the input data. We test values of 0, 0.015, 0.05, 0.075, 22 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure A5. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of different values for the forward modelling time step dt. Colors indicate the mean absolute deviation between the inverted bed and the true bed; the final inverted bed topography for the smallest and largest dt (i.e. 0.01 yrs and 1 yr, respectively) shown in b) and c), respectively. Figure A5. Modelled ice volume after the inversion relative to true ice volume of the ice cap as a function of different values for the forward modelling time step dt. Colors indicate the mean absolute deviation between the inverted bed and the true bed; the final inverted bed topography for the smallest and largest dt (i.e. 0.01 yrs and 1 yr, respectively) shown in b) and c), respectively. A2 Inversion parameters In this 515 case, the forward model crashes since it cannot accommodate the perturbed surface shape if θ = 0, whereas increasing θ helps eliminating some of the surface noise. Again, though, when θ is too large, the bed is smooth, but also largely devoid of its characteristic topographic highs (Fig. A7). Note that the computed ice volume is always within 1% of the true ice volume regardless of θ. 515 Finally, we also test whether a different initial bed guess influences the results. To that end, we assume that the ice thickness 520 is zero everywhere in the domain at the start of the inversion. We find that this has no significant impact on the recovered bed (not shown). Likewise, a different initial guess for the friction field (namely, we assume that it is slippery everywhere), does not change the bed solution. This suggests that our inversion is relatively stable towards initial conditions, although this is dependent on the choice of θ as described above. Finally, we also test whether a different initial bed guess influences the results. To that end, we assume that the ice thickness 520 is zero everywhere in the domain at the start of the inversion. We find that this has no significant impact on the recovered bed (not shown). Likewise, a different initial guess for the friction field (namely, we assume that it is slippery everywhere), does not change the bed solution. This suggests that our inversion is relatively stable towards initial conditions, although this is dependent on the choice of θ as described above. Finally, we also test whether a different initial bed guess influences the results. To that end, we assume that the ice thickness 520 is zero everywhere in the domain at the start of the inversion. We find that this has no significant impact on the recovered bed (not shown). Likewise, a different initial guess for the friction field (namely, we assume that it is slippery everywhere), does not change the bed solution. This suggests that our inversion is relatively stable towards initial conditions, although this is dependent on the choice of θ as described above. Author contributions. TF and WvP developed the inversion methodology. TF conducted the experiments with significant input from WvP. 525 JK provided and compiled data for the Kronebreen experiments, including processing of Pléiades imagery for generation of the 2020 DEM. A2 Inversion parameters We investigate this here by applying the different values for θ on our ice cap where the surface shape was perturbed by adding random noise with a standard deviation of 5 m. In this 515 case, the forward model crashes since it cannot accommodate the perturbed surface shape if θ = 0, whereas increasing θ helps eliminating some of the surface noise. Again, though, when θ is too large, the bed is smooth, but also largely devoid of its characteristic topographic highs (Fig. A7). Note that the computed ice volume is always within 1% of the true ice volume regardless of θ. or not. However, it is generally undesirable to include features in the solution if they are not justified by the input data, and so regularization is in order. With increasing θ, the computed bed becomes smoother, however, this eventually comes at the cost of detail in the solution. For θ = 0.3, the bed highs are no longer sharply visible. Next, we test how different choices of θ influence the inverted bed in the presence of input data errors. We investigate this here by applying the different values for or not. However, it is generally undesirable to include features in the solution if they are not justified by the input data, and so regularization is in order. With increasing θ, the computed bed becomes smoother, however, this eventually comes at the cost of detail in the solution. For θ = 0.3, the bed highs are no longer sharply visible. Next, we test how different choices of θ influence the inverted bed in the presence of input data errors. We investigate this here by applying the different values for θ on our ice cap where the surface shape was perturbed by adding random noise with a standard deviation of 5 m. In this 515 case, the forward model crashes since it cannot accommodate the perturbed surface shape if θ = 0, whereas increasing θ helps eliminating some of the surface noise. Again, though, when θ is too large, the bed is smooth, but also largely devoid of its characteristic topographic highs (Fig. A7). Note that the computed ice volume is always within 1% of the true ice volume regardless of θ. θ on our ice cap where the surface shape was perturbed by adding random noise with a standard deviation of 5 m. A2 Inversion parameters Figure A6. Sensitivity of bed inversion to the parameter β (eq. (1)). a) bed misfit per iteration step; b) dh/dt misfit per iteration step. Inset plots show zoom to iterations 8000 to 10000. Sharp steps (a) or peaks (b) occur when friction updates are applied. Figure A6. Sensitivity of bed inversion to the parameter β (eq. (1)). a) bed misfit per iteration step; b) dh/dt misfit per iteration step. Inset plots show zoom to iterations 8000 to 10000. Sharp steps (a) or peaks (b) occur when friction updates are applied. 0.1 and 0.3 for θ. The results show that without any surface updates (i.e. θ = 0), small-scale bed irregularities are introduced to the solution. They are too small to leave any noticeable impression on ice dynamics or surface shape, and so the inversion is not wrong in producing them; from the available surface ’observations’, it is impossible to tell whether these features exist 0 510 23 Figure A7. Final bed shape as a function of changing the regularization parameter θ. Upper row: setup without perturbed surface. Lower row: surface perturbed by adding random noise with a standard deviation of 5 m. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Figure A7. Final bed shape as a function of changing the regularization parameter θ. Upper row: setup without perturbed surface. Lower row: surface perturbed by adding random noise with a standard deviation of 5 m. Figure A7. Final bed shape as a function of changing the regularization parameter θ. Upper row: setup without perturbed surface. Lower row: surface perturbed by adding random noise with a standard deviation of 5 m. or not. However, it is generally undesirable to include features in the solution if they are not justified by the input data, and so regularization is in order. With increasing θ, the computed bed becomes smoother, however, this eventually comes at the cost of detail in the solution. For θ = 0.3, the bed highs are no longer sharply visible. Next, we test how different choices of θ influence the inverted bed in the presence of input data errors. Acknowledgements. We thank the PISM developers, in particular Constantine Khrulev, but also Ed Bueler and Andy Aschwanden for col- 530 laboration on changes to the PISM source code to speed up the inversion, for fruitful discussions and for their hospitality. TF was supported by the Swedish Research Council (VR grant 2020-04319). WvP acknowledges funding from the Swedish National Space Agency (grant 189/18). The Pléiades stereo-pairs used to generate the 2014 and 2020 DEMs used in this study were provided by the Pléiades Glacier Observatory initiative of the French Space Agency (CNES). References 535 J., Aschwanden, A., and Truffer, M.: Bayesian Inference of Subglacial Topography Using Mass Conservation, Frontiers in Earth Science, 4, 8, https://doi.org/10.3389/feart.2016.00008, 2016. Bueler, E. and Brown, J.: Shallow shelf approximation as a “sliding law” in a thermomechanically coupled ice sheet model, Journal of 545 Geophysical Research: Earth Surface, 114, https://doi.org/10.1029/2008JF001179, 2009. Chen, J. and Ohmura, A.: Estimation of Alpine glacier water resources and their change since the 1870s, IAHS publ, 193, 127–135, 1990. Chen, W., Yao, T., Zhang, G., Li, F., Zheng, G., Zhou, Y., and Xu, F.: Towards ice-thickness inversion: an evaluation of global digital elevation models (DEMs) in the glacierized Tibetan Plateau, The Cryosphere, 16, 197–218, https://doi.org/10.5194/tc-16-197-2022, 2022. Bueler, E. and Brown, J.: Shallow shelf approximation as a “sliding law” in a thermomechanically coupled ice sheet model, Journal of 545 Geophysical Research: Earth Surface, 114, https://doi.org/10.1029/2008JF001179, 2009. Chen, J. and Ohmura, A.: Estimation of Alpine glacier water resources and their change since the 1870s, IAHS publ, 193, 127–135, 1990. Bueler, E. and Brown, J.: Shallow shelf approximation as a “sliding law” in a thermomechanically coupled ice sheet model, Journal of 545 Geophysical Research: Earth Surface, 114, https://doi.org/10.1029/2008JF001179, 2009. Geophysical Research: Earth Surface, 114, https://doi.org/10.1029/2008JF001179, 2009. Chen, J. and Ohmura, A.: Estimation of Alpine glacier water resources and their change since the 1870s, IAHS publ, 193, 127–135, 1990. Chen, W., Yao, T., Zhang, G., Li, F., Zheng, G., Zhou, Y., and Xu, F.: Towards ice-thickness inversion: an evaluation of global digital elevation d l ( ) i h l i i d ib l h C h 16 19 218 h //d i /10 5194/ 16 19 2022 2022 Chen, J. and Ohmura, A.: Estimation of Alpine glacier water resources and their change since the 1870s, IAHS publ, 193, 127–135, 1990. Chen, W., Yao, T., Zhang, G., Li, F., Zheng, G., Zhou, Y., and Xu, F.: Towards ice-thickness inversion: an evaluation of global digital elevation models (DEMs) in the glacierized Tibetan Plateau, The Cryosphere, 16, 197–218, https://doi.org/10.5194/tc-16-197-2022, 2022. Christianson, K., Kohler, J., Alley, R. B., Nuth, C., and Van Pelt, W. J.: Dynamic perennial firn aquifer on an Arctic glacier, Geophysical 550 Research Letters, 42, 1418–1426, 2015. References 535 Cl k G K C A l F S J h A H R di´ V M B B l h T d B thi E I V l d S b l i l T h Christianson, K., Kohler, J., Alley, R. B., Nuth, C., and Van Pelt, W. J.: Dynamic perennial firn aquifer on an Arctic glacier, Geophysical 550 Research Letters, 42, 1418–1426, 2015. Clarke, G. K. C., Anslow, F. S., Jarosch, A. H., Radi´c, V., Menounos, B., Bolch, T., and Berthier, E.: Ice Volume and Subglacial Topography for Western Canadian Glaciers from Mass Balance Fields, Thinning Rates, and a Bed Stress Model, Journal of Climate, 26, 4282–4303, https://doi.org/10.1175/JCLI-D-12-00513.1, 2013. Clarke, G. K. C., Anslow, F. S., Jarosch, A. H., Radi´c, V., Menounos, B., Bolch, T., and Berthier, E.: Ice Volume and Subglacial Topography for Western Canadian Glaciers from Mass Balance Fields, Thinning Rates, and a Bed Stress Model, Journal of Climate, 26, 4282–4303, https://doi.org/10.1175/JCLI-D-12-00513.1, 2013. Cuffey, K. M. and Paterson, W. S. B.: The physics of glaciers, Butterworth-Heinemann, Amsterdam, 2010. 555 Deschamps-Berger, C., Nuth, C., Pelt, W. V., Berthier, E., Kohler, J., and Altena, B.: Closing the mass budget of a tidewater glacier: the example of Kronebreen, Svalbard, Journal of Glaciology, 65, 136–148, https://doi.org/10.1017/jog.2018.98, 2019. Farinotti, D., Huss, M., Bauder, A., Funk, M., and Truffer, M.: A method to estimate the ice volume and ice-thickness distribution of alpine glaciers, Journal of Glaciology, 55, 422–430, https://doi.org/10.3189/002214309788816759, 2009. Deschamps-Berger, C., Nuth, C., Pelt, W. V., Berthier, E., Kohler, J., and Altena, B.: Closing the mass budget of a tidewater glacier: the example of Kronebreen, Svalbard, Journal of Glaciology, 65, 136–148, https://doi.org/10.1017/jog.2018.98, 2019. Farinotti, D., Huss, M., Bauder, A., Funk, M., and Truffer, M.: A method to estimate the ice volume and ice-thickness distribution of alpine glaciers, Journal of Glaciology, 55, 422–430, https://doi.org/10.3189/002214309788816759, 2009. Farinotti, D., Brinkerhoff, D. J., Clarke, G. K. C., Fürst, J. J., Frey, H., Gantayat, P., Gillet-Chaulet, F., Girard, C., Huss, M., Leclercq, 560 P. W., Linsbauer, A., Machguth, H., Martin, C., Maussion, F., Morlighem, M., Mosbeux, C., Pandit, A., Portmann, A., Rabatel, A., Ramsankaran, R., Reerink, T. J., Sanchez, O., Stentoft, P. A., Singh Kumari, S., van Pelt, W. J. J., Anderson, B., Benham, T., Binder, D., Dowdeswell, J. A., Fischer, A., Helfricht, K., Kutuzov, S., Lavrentiev, I., McNabb, R., Gudmundsson, G. H., Li, H., and Andreassen, L. A2 Inversion parameters TF wrote the manuscript and created the figures with significant input from WvP and JK. Author contributions. TF and WvP developed the inversion methodology. TF conducted the experiments with significant input from WvP. 525 JK provided and compiled data for the Kronebreen experiments, including processing of Pléiades imagery for generation of the 2020 DEM. TF wrote the manuscript and created the figures with significant input from WvP and JK. Competing interests. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 24 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Acknowledgements. We thank the PISM developers, in particular Constantine Khrulev, but also Ed Bueler and Andy Aschwanden for col- 530 laboration on changes to the PISM source code to speed up the inversion, for fruitful discussions and for their hospitality. TF was supported by the Swedish Research Council (VR grant 2020-04319). WvP acknowledges funding from the Swedish National Space Agency (grant 189/18). The Pléiades stereo-pairs used to generate the 2014 and 2020 DEMs used in this study were provided by the Pléiades Glacier Observatory initiative of the French Space Agency (CNES). 25 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. References 535 Bahr, D. B., Meier, M. F., and Peckham, S. D.: The physical basis of glacier volume-area scaling, Journal of Geophysical Research: Solid Earth, 102, 20 355–20 362, https://doi.org/10.1029/97JB01696, 1997. Bahr, D. B., Pfeffer, W. T., and Kaser, G.: Glacier volume estimation as an ill-posed inversion, Journal of Glaciology, 60, 922–934, https://doi.org/10.3189/2014JoG14J062, 2014. Bahr, D. B., Meier, M. F., and Peckham, S. D.: The physical basis of glacier volume area scaling, Journal of Geophysical Research: Solid Earth, 102, 20 355–20 362, https://doi.org/10.1029/97JB01696, 1997. Bahr, D. B., Pfeffer, W. T., and Kaser, G.: Glacier volume estimation as an ill-posed inversion, Journal of Glaciology, 60, 922–934, https://doi.org/10.3189/2014JoG14J062, 2014. Bahr, D. B., Pfeffer, W. T., and Kaser, G.: Glacier volume estimation as an ill-posed inversion, Journal of Glaciology, 60, 922–934, https://doi.org/10.3189/2014JoG14J062, 2014. Blatter, H.: Velocity and stress fields in grounded glaciers: a simple algorithm for including deviatoric stress gradients, Journal of Glaciology, 540 41, 333–344, https://doi.org/10.3189/S002214300001621X, 1995. Bogorodsky, V. V., Bogorodski˘ı, V. V., Bentley, C. R., and Gudmandsen, P. E.: Radioglaciology, Springer Science & Business Media, 1985. Brinkerhoff, D. J., Aschwanden, A., and Truffer, M.: Bayesian Inference of Subglacial Topography Using Mass Conservation, Frontiers in Earth Science, 4, 8, https://doi.org/10.3389/feart.2016.00008, 2016. Blatter, H.: Velocity and stress fields in grounded glaciers: a simple algorithm for including deviatoric stress gradients, Journal of Glaciology, 540 41, 333–344, https://doi.org/10.3189/S002214300001621X, 1995. Bogorodsky, V. V., Bogorodski˘ı, V. V., Bentley, C. R., and Gudmandsen, P. E.: Radioglaciology, Springer Science & Business Media, 1985. Brinkerhoff, D. J., Aschwanden, A., and Truffer, M.: Bayesian Inference of Subglacial Topography Using Mass Conservation, Frontiers in Earth Science, 4, 8, https://doi.org/10.3389/feart.2016.00008, 2016. Blatter, H.: Velocity and stress fields in grounded glaciers: a simple algorithm for including deviatoric stress gradients, Journal of Glaciology, 540 41, 333–344, https://doi.org/10.3189/S002214300001621X, 1995. Blatter, H.: Velocity and stress fields in grounded glaciers: a simple algorithm for including deviatoric stress gradients, Journal of Glaciology, 540 41, 333–344, https://doi.org/10.3189/S002214300001621X, 1995. Bogorodsky, V. V., Bogorodski˘ı, V. V., Bentley, C. R., and Gudmandsen, P. E.: Radioglaciology, Springer Science & Business Media, 1985. Blatter, H.: Velocity and stress fields in grounded glaciers: a simple algorithm for including deviatoric stress gradients, Journal of Glaciology, 540 41, 333–344, https://doi.org/10.3189/S002214300001621X, 1995. Bogorodsky, V. V., Bogorodski˘ı, V. V., Bentley, C. R., and Gudmandsen, P. E.: Radioglaciology, Springer Science & Business Media, 1985. Brinkerhoff, D. Flowers, G. E. and Clarke, G. K. C.: Surface and bed topography of Trapridge Glacier, Yukon Territory, Canada: digital elevation models 570 and derived hydraulic geometry, Journal of Glaciology, 45, 165–174, https://doi.org/10.3189/S0022143000003142, 1999. References 535 M.: How accurate are estimates of glacier ice thickness? Results from ITMIX, the Ice Thickness Models Intercomparison eXperiment, The Cryosphere, 11, 949–970, https://doi.org/10.5194/tc-11-949-2017, 2017. 565 Farinotti, D., Brinkerhoff, D. J., Fürst, J. J., Gantayat, P., Gillet-Chaulet, F., Huss, M., Leclercq, P. W., Maurer, H., Morlighem, M., Pandit, A., Rabatel, A., Ramsankaran, R., Reerink, T. J., Robo, E., Rouges, E., Tamre, E., van Pelt, W. J. J., Werder, M. A., Azam, M. F., Li, H., and Andreassen, L. M.: Results from the Ice Thickness Models Intercomparison eXperiment Phase 2 (ITMIX2), Frontiers in Earth Science, 8, https://doi.org/10.3389/feart.2020.571923, 2021. Flowers, G. E. and Clarke, G. K. C.: Surface and bed topography of Trapridge Glacier, Yukon Territory, Canada: digital elevation models 570 and derived hydraulic geometry, Journal of Glaciology, 45, 165–174, https://doi.org/10.3189/S0022143000003142, 1999. 26 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Fretwell, P., Pritchard, H. D., Vaughan, D. G., Bamber, J. L., Barrand, N. E., Bell, R., Bianchi, C., Bingham, R. G., Blankenship, D. D., Casassa, G., Catania, G., Callens, D., Conway, H., Cook, A. J., Corr, H. F. J., Damaske, D., Damm, V., Ferraccioli, F., Forsberg, R., Fujita, S., Gim, Y., Gogineni, P., Griggs, J. A., Hindmarsh, R. C. A., Holmlund, P., Holt, J. W., Jacobel, R. W., Jenkins, A., Jokat, W., Jordan, T., King, E. C., Kohler, J., Krabill, W., Riger-Kusk, M., Langley, K. A., Leitchenkov, G., Leuschen, C., Luyendyk, B. P., Matsuoka, K., 575 Mouginot, J., Nitsche, F. O., Nogi, Y., Nost, O. A., Popov, S. V., Rignot, E., Rippin, D. M., Rivera, A., Roberts, J., Ross, N., Siegert, M. J., Smith, A. M., Steinhage, D., Studinger, M., Sun, B., Tinto, B. K., Welch, B. C., Wilson, D., Young, D. A., Xiangbin, C., and Zirizzotti, A.: Bedmap2: improved ice bed, surface and thickness datasets for Antarctica, The Cryosphere, 7, 375–393, https://doi.org/10.5194/tc-7- 375-2013, 2013. 575 Frey, H., Machguth, H., Huss, M., Huggel, C., Bajracharya, S., Bolch, T., Kulkarni, A., Linsbauer, A., Salzmann, N., and Stoffel, 580 M.: Estimating the volume of glaciers in the Himalayan-Karakoram region using different methods, The Cryosphere, 8, 2313–2333, https://doi.org/10.5194/tc-8-2313-2014, 2014. Fürst, J. J., Gillet-Chaulet, F., Benham, T. J., Dowdeswell, J. References 535 Habermann, M., Maxwell, D., and Truffer, M.: Reconstruction of basal properties in ice sheets using iterative inverse methods, Journal of Gudmundsson, G. H.: Transmission of basal variability to a glacier surface, Journal of Geophysical Research: Solid Earth, 108, 590 https://doi.org/10.1029/2002JB002107, 2003. Habermann, M., Maxwell, D., and Truffer, M.: Reconstruction of basal properties in ice sheets using iterative inverse methods, Journal of Glaciology, 58, 795–808, https://doi.org/10.3189/2012JoG11J168, 2012. Haeberli W and Hoelzle M : Application of inventory data for estimating characteristics of and regional climate-change effects on mountain https://doi.org/10.1029/2002JB002107, 2003. Habermann, M., Maxwell, D., and Truffer, M.: Reconstruction of basal properties in ice sheets using iterative inverse methods, Journal of Glaciology, 58, 795–808, https://doi.org/10.3189/2012JoG11J168, 2012. Habermann, M., Maxwell, D., and Truffer, M.: Reconstruction of basal properties in ice sheets using iterative inverse methods, Journal of Glaciology, 58, 795–808, https://doi.org/10.3189/2012JoG11J168, 2012. Haeberli, W. and Hoelzle, M.: Application of inventory data for estimating characteristics of and regional clim Haeberli, W. and Hoelzle, M.: Application of inventory data for estimating characteristics of and regional climate-change effects on mountain glaciers: a pilot study with the European Alps, Annals of Glaciology, 21, 206–212, https://doi.org/10.3189/S0260305500015834, 1995. 595 Haeberli, W. and Hoelzle, M.: Application of inventory data for estimating characteristics of and regional climate-change effects on mountain glaciers: a pilot study with the European Alps, Annals of Glaciology, 21, 206–212, https://doi.org/10.3189/S0260305500015834, 1995. 595 Heining, C.: Velocity field reconstruction in gravity-driven flow over unknown topography, Physics of Fluids, 23, 032 101, https://doi.org/10.1063/1.3559144, 2011. glaciers: a pilot study with the European Alps, Annals of Glaciology, 21, 206–212, https://doi.org/10.3189/S0260305500015834, 1995. 595 Heining, C.: Velocity field reconstruction in gravity-driven flow over unknown topography, Physics of Fluids, 23, 032 101, https://doi.org/10.1063/1.3559144, 2011. How, P., Benn, D. I., Hulton, N. R. J., Hubbard, B., Luckman, A., Sevestre, H., van Pelt, W. J. J., Lindbäck, K., Kohler, J., and Boot, W.: Rapidly changing subglacial hydrological pathways at a tidewater glacier revealed through simultaneous observations of water pressure, p g How, P., Benn, D. I., Hulton, N. R. J., Hubbard, B., Luckman, A., Sevestre, H., van Pelt, W. J. J., Lindbäck, K., Kohler, J., and Boot, W.: Rapidly changing subglacial hydrological pathways at a tidewater glacier revealed through simultaneous observations of water pressure, supraglacial lakes, meltwater plumes and surface velocities, The Cryosphere, 11, 2691–2710, https://doi.org/10.5194/tc-11-2691-2017, 600 2017. References 535 A., Grabiec, M., Navarro, F., Pettersson, R., Moholdt, G., Nuth, C., Sass, B., Aas K Fettweis X Lang C Seehaus T and Braun M : Application of a two step approach for mapping ice thickness to various Frey, H., Machguth, H., Huss, M., Huggel, C., Bajracharya, S., Bolch, T., Kulkarni, A., Linsbauer, A., Salzmann, N., and Stoffel, 580 M.: Estimating the volume of glaciers in the Himalayan-Karakoram region using different methods, The Cryosphere, 8, 2313–2333, https://doi.org/10.5194/tc-8-2313-2014, 2014. Fürst, J. J., Gillet-Chaulet, F., Benham, T. J., Dowdeswell, J. A., Grabiec, M., Navarro, F., Pettersson, R., Moholdt, G., Nuth, C., Sass, B., Aas, K., Fettweis, X., Lang, C., Seehaus, T., and Braun, M.: Application of a two-step approach for mapping ice thickness to various Fürst, J. J., Gillet-Chaulet, F., Benham, T. J., Dowdeswell, J. A., Grabiec, M., Navarro, F., Pettersson, R., Moholdt, G., Nuth, C., Sass, B., Aas, K., Fettweis, X., Lang, C., Seehaus, T., and Braun, M.: Application of a two-step approach for mapping ice thickness to various glacier types on Svalbard, The Cryosphere, 11, 2003–2032, https://doi.org/10.5194/tc-11-2003-2017, 2017. 585 glacier types on Svalbard, The Cryosphere, 11, 2003–2032, https://doi.org/10.5194/tc-11-2003-2017, 2017. 585 Gantayat, P., Kulkarni, A. V., and Srinivasan, J.: Estimation of ice thickness using surface velocities and slope: case study at Gangotri Glacier, India, Journal of Glaciology, 60, 277–282, https://doi.org/10.3189/2014JoG13J078, 2014. Glen, J. W.: The creep of polycrystalline ice, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 228, 519–538, 1955. Gantayat, P., Kulkarni, A. V., and Srinivasan, J.: Estimation of ice thickness using surface velocities and slope: case study at Gangotri Glacier, India, Journal of Glaciology, 60, 277–282, https://doi.org/10.3189/2014JoG13J078, 2014. Gantayat, P., Kulkarni, A. V., and Srinivasan, J.: Estimation of ice thickness using surface velocities and slope: case study at Gangotri Glacier, India, Journal of Glaciology, 60, 277–282, https://doi.org/10.3189/2014JoG13J078, 2014. Glen, J. W.: The creep of polycrystalline ice, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 228, 519–538, 1955. Glen, J. W.: The creep of polycrystalline ice, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 228, 519–538, 1955. Gudmundsson, G. H.: Transmission of basal variability to a glacier surface, Journal of Geophysical Research: Solid Earth, 108, 590 https://doi.org/10.1029/2002JB002107, 2003. Gudmundsson, G. H.: Transmission of basal variability to a glacier surface, Journal of Geophysical Research: Solid Earth, 108, 590 https://doi.org/10.1029/2002JB002107, 2003. References 535 Michel, L., Picasso, M., Farinotti, D., Bauder, A., Funk, M., and Blatter, H.: Estimating the ice thickness of m , , , , , , , , , , , g g approach using surface topography and mass-balance, Inverse Problems, 29, 035 002, https://doi.org/10.1088/0266-5611/29/3/035002, 630 2013. Millan, R., Mouginot, J., Rabatel, A., and Morlighem, M.: Ice velocity and thickness of the world’s glaciers, Nature Geoscience, 15, 124–129, https://doi org/10 1038/s41561-021-00885-z 2022 2013. Millan, R., Mouginot, J., Rabatel, A., and Morlighem, M.: Ice velocity and thickness of the world’s glaciers, Nature Geoscience, 15, 124–129, https://doi.org/10.1038/s41561-021-00885-z, 2022. Millan, R., Mouginot, J., Rabatel, A., and Morlighem, M.: Ice velocity and thickness of the world’s glaciers, Nature Geoscience, 15, 124–129, https://doi.org/10.1038/s41561-021-00885-z, 2022. NASA JPL: NASADEM Merged DEM Global 1 arc second V001. 2020, distributed by NASA EOSDIS Land Processes DAAC, https: //lpdaac.usgs.gov/products/nasadem_hgtv001, 2020. 635 NASA JPL: NASADEM Merged DEM Global 1 arc second V001. 2020, distributed by NASA EOSDIS Land Processes DAAC, https: //lpdaac.usgs.gov/products/nasadem_hgtv001, 2020. 635 Neven, A., Dall’Alba, V., Juda, P., Straubhaar, J., and Renard, P.: Ice volume and basal topography estimation using geostatistical methods NASA JPL: NASADEM Merged DEM Global 1 arc second V001. 2020, distributed by NASA EOSDIS Land Processes DAAC, https: //lpdaac.usgs.gov/products/nasadem_hgtv001, 2020. 635 Neven, A., Dall’Alba, V., Juda, P., Straubhaar, J., and Renard, P.: Ice volume and basal topography estimation using geostatistical methods and ground-penetrating radar measurements: application to the Tsanfleuron and Scex Rouge glaciers, Swiss Alps, The Cryosphere, 15, 5169–5186, https://doi.org/10.5194/tc-15-5169-2021, 2021. Nye J F : A Method of Calculating the Thicknesses of the Ice Sheets Nature 169 529 530 https://doi org/10 1038/169529a0 1952 //lpdaac.usgs.gov/products/nasadem_hgtv001, 2020. 635 Neven, A., Dall’Alba, V., Juda, P., Straubhaar, J., and Renard, P.: Ice volume and basal topography estimation using geostatistical methods and ground-penetrating radar measurements: application to the Tsanfleuron and Scex Rouge glaciers, Swiss Alps, The Cryosphere, 15, 5169–5186, https://doi.org/10.5194/tc-15-5169-2021, 2021. Calculating the Thicknesses of the Ice-Sheets, Nature, 169, 529–530, https://doi.org/10.1038/169529a0, 1952. Nye, J. F.: A Method of Calculating the Thicknesses of the Ice-Sheets, Nature, 169, 529–530, https://doi.org/ Oppenheimer, M., Glavovic, B., Hinkel, J., van de Wal, R., Magnan, A., Abd-Elgawad, A., Cai, R., Cifuentes-Jara, M., DeConto, R., Ghosh, 640 T., Hay, J., Isla, F., Marzeion, B., and Sebesvari, Z.: Sea Level Rise and Implications for Low-Lying Islands, Coasts and Communities, in: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, 2019. References 535 Li, H., Ng, F., Li, Z., Qin, D., and Cheng, G.: An extended “perfect-plasticity” method for estimating ice thickness along the flow line of mountain glaciers, Journal of Geophysical Research: Earth Surface, 117, https://doi.org/https://doi.org/10.1029/2011JF002104, 2012. 615 615 mountain glaciers, Journal of Geophysical Research: Earth Surface, 117, https://doi.org/https://doi.org/10.1029/2011JF002104, 2012. 615 Lindbäck, K., Kohler, J., Pettersson, R., Nuth, C., Langley, K., Messerli, A., Vallot, D., Matsuoka, K., and Brandt, O.: Subglacial topography, ice thickness, and bathymetry of Kongsfjorden, northwestern Svalbard, Earth System Science Data, 10, 1769–1781, https://doi.org/10.5194/essd-10-1769-2018, 2018. Linsbauer, A., Paul, F., Hoelzle, M., Frey, H., and Haeberli, W.: The Swiss Alps without glaciers – a GIS-based modelling approach for Linsbauer, A., Paul, F., Hoelzle, M., Frey, H., and Haeberli, W.: The Swiss Alps without glaciers – a GIS-based modelling approach for reconstruction of glacier beds, Proceedings of Geomorphometry, pp. 243–247, https://doi.org/10.5167/uzh-27834, 2009. 620 reconstruction of glacier beds, Proceedings of Geomorphometry, pp. 243–247, https://doi.org/10.5167/uzh-27834, 2009. 620 Linsbauer, A., Paul, F., and Haeberli, W.: Modeling glacier thickness distribution and bed topography over entire mountain ranges with GlabTop: Application of a fast and robust approach, Journal of Geophysical Research: Earth Surface, 117, https://doi.org/10.1029/2011JF002313, 2012. Luckman A Benn D I Cottier F Bevan S Nilsen F and Inall M : Calving rates at tidewater glaciers vary strongly with ocean Linsbauer, A., Paul, F., and Haeberli, W.: Modeling glacier thickness distribution and bed topography over entire mountain ranges with GlabTop: Application of a fast and robust approach, Journal of Geophysical Research: Earth Surface, 117, https://doi.org/10.1029/2011JF002313, 2012. Luckman, A., Benn, D. I., Cottier, F., Bevan, S., Nilsen, F., and Inall, M.: Calving rates at tidewater glaciers vary strongly with ocean temperature, Nature Communications, 6, 8566, https://doi.org/10.1038/ncomms9566, 2015. 625 temperature, Nature Communications, 6, 8566, https://doi.org/10.1038/ncomms9566, 2015. 625 Maussion, F., Butenko, A., Champollion, N., Dusch, M., Eis, J., Fourteau, K., Gregor, P., Jarosch, A. H., Landmann, J., Oesterle, F., Recinos, B., Rothenpieler, T., Vlug, A., Wild, C. T., and Marzeion, B.: The Open Global Glacier Model (OGGM) v1.1, Geoscientific Model Development, 12, 909–931, https://doi.org/10.5194/gmd-12-909-2019, 2019. Michel, L., Picasso, M., Farinotti, D., Bauder, A., Funk, M., and Blatter, H.: Estimating the ice thickness of mountain glaciers with an inverse approach using surface topography and mass-balance, Inverse Problems, 29, 035 002, https://doi.org/10.1088/0266-5611/29/3/035002, 630 2013. References 535 Hugonnet, R., McNabb, R., Berthier, E., Menounos, B., Nuth, C., Girod, L., Farinotti, D., Huss, M., Dussaillant, I., Brun, F., and Kääb, A.: Accelerated global glacier mass loss in the early twenty-first century, Nature, 592, 726–731, https://doi.org/10.1038/s41586-021-03436-z, 2021. Hugonnet, R., McNabb, R., Berthier, E., Menounos, B., Nuth, C., Girod, L., Farinotti, D., Huss, M., Dussaillant, I., Brun, F., and Kääb, A.: Accelerated global glacier mass loss in the early twenty-first century, Nature, 592, 726–731, https://doi.org/10.1038/s41586-021-03436-z, 2021. Huss, M. and Farinotti, D.: Distributed ice thickness and volume of all glaciers around the globe, Journal of Geophysical Research: Earth Surface, 117, https://doi.org/10.1029/2012JF002523, 2012. Huss, M. and Farinotti, D.: Distributed ice thickness and volume of all glaciers around the globe, Journal of Geophysical Research: Earth 605 Surface, 117, https://doi.org/10.1029/2012JF002523, 2012. Kamb, B. and Echelmeyer, K. A.: Stress-Gradient Coupling in Glacier Flow: I. Longitudinal Averaging of the Influence of Ice Thickness and Surface Slope, Journal of Glaciology, 32, 267–284, https://doi.org/10.3189/S0022143000015604, 1986. Kamb, B. and Echelmeyer, K. A.: Stress-Gradient Coupling in Glacier Flow: I. Longitudinal Averaging of the Influence of Ice Thickness and Surface Slope, Journal of Glaciology, 32, 267–284, https://doi.org/10.3189/S0022143000015604, 1986. 27 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Köhler, A., Petlicki, M., Lefeuvre, P.-M., Buscaino, G., Nuth, C., and Weidle, C.: Contribution of calving to frontal ablation quan- tified from seismic and hydroacoustic observations calibrated with lidar volume measurements, The Cryosphere, 13, 3117–3137, 610 https://doi.org/10.5194/tc-13-3117-2019, 2019. L h L G b M B d A d M H Gl i thi k ti ti f l i l i i d t d d li t i t Köhler, A., Petlicki, M., Lefeuvre, P.-M., Buscaino, G., Nuth, C., and Weidle, C.: Contribution of calving to frontal ablation quan- tified from seismic and hydroacoustic observations calibrated with lidar volume measurements, The Cryosphere, 13, 3117–3137, 610 https://doi.org/10.5194/tc-13-3117-2019, 2019. Langhammer, L., Grab, M., Bauder, A., and Maurer, H.: Glacier thickness estimations of alpine glaciers using data and modeling constraints, The Cryosphere, 13, 2189–2202, https://doi.org/10.5194/tc-13-2189-2019, 2019. Li, H., Ng, F., Li, Z., Qin, D., and Cheng, G.: An extended “perfect-plasticity” method for estimating ice thickness along the flow line of 610 Langhammer, L., Grab, M., Bauder, A., and Maurer, H.: Glacier thickness estimations of alpine glaciers using data and modeling constraints, The Cryosphere, 13, 2189–2202, https://doi.org/10.5194/tc-13-2189-2019, 2019. References 535 M.: A simple inverse method for the distribution of basal sliding coefficients under ice sheets, applied to 650 Antarctica, The Cryosphere, 6, 953–971, https://doi.org/10.5194/tc-6-953-2012, 2012. Rabatel, A., Sanchez, O., Vincent, C., and Six, D.: Estimation of Glacier Thickness From Surface Mass Balance and Ice Flow Velocities: A Case Study on Argentière Glacier, France, Frontiers in Earth Science, 6, 112, https://doi.org/10.3389/feart.2018.00112, 2018. y g , , , , , p g , Raymond, M. J. and Gudmundsson, G. H.: On the relationship between surface and basal properties on glaciers, ice sheets, and ice streams, J l f G h i l R h S lid E h 110 h //d i /10 1029/2005JB003681 2005 655 Raymond, M. J. and Gudmundsson, G. H.: On the relationship between surface and basal properties on glaciers, ice sheets, and ice streams, Journal of Geophysical Research: Solid Earth, 110, https://doi.org/10.1029/2005JB003681, 2005. 655 Raymond, M. J. and Gudmundsson, G. H.: On the relationship between surface and basal properties on glaci Journal of Geophysical Research: Solid Earth, 110, https://doi.org/10.1029/2005JB003681, 2005. 655 RGI Consortium: Randolph Glacier Inventory – A Dataset of Global Glacier Outlines: Version 6.0, NSIDC: National Snow and Ice Data Center, Boulder, Colorado USA, https://doi.org/10.7265/N5-RGI-60, 2017. Rounce, D. R., Hock, R., Maussion, F., Hugonnet, R., Kochtitzky, W., Huss, M., Berthier, E., Brinkerhoff, D., Compagno, L., Copland, L., Farinotti, D., Menounos, B., and McNabb, R. W.: Global glacier change in the 21st century: Every increase in temperature matters, Journal of Geophysical Research: Solid Earth, 110, https://doi.org/10.1029/2005JB003681, 2005. 655 RGI Consortium: Randolph Glacier Inventory – A Dataset of Global Glacier Outlines: Version 6.0, NSIDC: National Snow and Ice Data Center, Boulder, Colorado USA, https://doi.org/10.7265/N5-RGI-60, 2017. Rounce, D. R., Hock, R., Maussion, F., Hugonnet, R., Kochtitzky, W., Huss, M., Berthier, E., Brinkerhoff, D., Compagno, L., Copland, L., Farinotti, D., Menounos, B., and McNabb, R. W.: Global glacier change in the 21st century: Every increase in temperature matters, Science, 379, 78–83, https://doi.org/10.1126/science.abo1324, 2023. 660 Science, 379, 78–83, https://doi.org/10.1126/science.abo1324, 2023. 660 Schellenberger, T., Dunse, T., Kääb, A., Kohler, J., and Reijmer, C. H.: Surface speed and frontal ablation of Kronebreen and Kongsbreen, NW Svalbard, from SAR offset tracking, The Cryosphere, 9, 2339–2355, https://doi.org/10.5194/tc-9-2339-2015, 2015. Vallot, D., Pettersson, R., Luckman, A., Benn, D. I., Zwinger, T., Pelt, W. J. J. V., Kohler, J., Schäfer, M., Claremar, B., and Hulton, N. R. References 535 J.: Basal dynamics of Kronebreen, a fast-flowing tidewater glacier in Svalbard: non-local spatio-temporal response to water input, Journal of Schellenberger, T., Dunse, T., Kääb, A., Kohler, J., and Reijmer, C. H.: Surface speed and frontal ablation of Kronebreen and Kongsbreen, NW Svalbard, from SAR offset tracking, The Cryosphere, 9, 2339–2355, https://doi.org/10.5194/tc-9-2339-2015, 2015. Vallot, D., Pettersson, R., Luckman, A., Benn, D. I., Zwinger, T., Pelt, W. J. J. V., Kohler, J., Schäfer, M., Claremar, B., and Hulton, N. R. J.: from SAR offset tracking, The Cryosphere, 9, 2339–2355, https://doi.org/10.5194/tc-9-2339-2015, 2015. Vallot, D., Pettersson, R., Luckman, A., Benn, D. I., Zwinger, T., Pelt, W. J. J. V., Kohler, J., Schäfer, M., Claremar, B., and Hulton, N. R. J.: Basal dynamics of Kronebreen, a fast-flowing tidewater glacier in Svalbard: non-local spatio-temporal response to water input, Journal of Glaciology, 63, 1012–1024, https://doi.org/10.1017/jog.2017.69, 2017. 665 Glaciology, 63, 1012 1024, https://doi.org/10.1017/jog.2017.69, 2017. 665 Vallot, D., Åström, J., Zwinger, T., Pettersson, R., Everett, A., Benn, D. I., Luckman, A., van Pelt, W. J. J., Nick, F., and Kohler, J.: Effects of undercutting and sliding on calving: a global approach applied to Kronebreen, Svalbard, The Cryosphere, 12, 609–625, https://doi.org/10.5194/tc-12-609-2018, publisher: Copernicus GmbH, 2018. van der Wel, L. G., Streurman, H. J., Isaksson, E., Helsen, M. M., Wal, R. S. W. V. D., Martma, T., Pohjola, V van der Wel, L. G., Streurman, H. J., Isaksson, E., Helsen, M. M., Wal, R. S. W. V. D., Martma, T., Pohjola, V. A., Moore, J. C., and Meijer,i van der Wel, L. G., Streurman, H. J., Isaksson, E., Helsen, M. M., Wal, R. S. W. V. D., Martma, T., Pohjola, V. A., Moore, J. C., and Meijer, H. a. J.: Using high-resolution tritium profiles to quantify the effects of melt on two Spitsbergen ice cores, Journal of Glaciology, 57, 670 108 109 h //d i /10 3189/002214311 98843368 2011 H. a. J.: Using high-resolution tritium profiles to quantify the effects of melt on two Spitsbergen ice cores, Journal of Glaciology, 57, 670 1087–1097, https://doi.org/10.3189/002214311798843368, 2011. van Pelt, W. J. J. and Kohler, J.: Modelling the long-term mass balance and firn evolution of glaciers around Kongsfjorden, Svalbard, Journal of Glaciology, 61, 731–744, https://doi.org/10.3189/2015JoG14J223, 2015. van Pelt, W. J. J., Oerlemans, J., Reijmer, C. H., Pettersson, R., Pohjola, V. References 535 Oppenheimer, M., Glavovic, B., Hinkel, J., van de Wal, R., Magnan, A., Abd-Elgawad, A., Cai, R., Cifuentes-Jara, M., DeConto, R., Ghosh, 640 T., Hay, J., Isla, F., Marzeion, B., and Sebesvari, Z.: Sea Level Rise and Implications for Low-Lying Islands, Coasts and Communities, in: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, 2019. Pattyn, F.: A new three-dimensional higher-order thermomechanical ice sheet model: Basic sensitivity, ice stream development, and ice flow across subglacial lakes Journal of Geophysical Research: Solid Earth 108 https://doi org/10 1029/2002JB002329 2003 T., Hay, J., Isla, F., Marzeion, B., and Sebesvari, Z.: Sea Level Rise and Implications for Low-Lying Islands, Coasts and Communities, in: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, 2019. Pattyn, F.: A new three-dimensional higher-order thermomechanical ice sheet model: Basic sensitivity, ice stream development, and ice flow across subglacial lakes, Journal of Geophysical Research: Solid Earth, 108, https://doi.org/10.1029/2002JB002329, 2003. Pattyn, F.: A new three-dimensional higher-order thermomechanical ice sheet model: Basic sensitivity, ice stream development, and ice flow across subglacial lakes, Journal of Geophysical Research: Solid Earth, 108, https://doi.org/10.1029/2002JB002329, 2003. 28 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Pattyn, F., Schoof, C., Perichon, L., Hindmarsh, R. C. A., Bueler, E., de Fleurian, B., Durand, G., Gagliardini, O., Gladstone, R., Goldberg, 645 D., Gudmundsson, G. H., Huybrechts, P., Lee, V., Nick, F. M., Payne, A. J., Pollard, D., Rybak, O., Saito, F., and Vieli, A.: Results of the Marine Ice Sheet Model Intercomparison Project, MISMIP, The Cryosphere, 6, 573–588, https://doi.org/10.5194/tc-6-573-2012, 2012. Paul, F. and Linsbauer, A.: Modeling of glacier bed topography from glacier outlines, central branch lines, and a DEM, International Journal of Geographical Information Science, 26, 1173–1190, https://doi.org/10.1080/13658816.2011.627859, 2012. Pollard, D. and DeConto, R. M.: A simple inverse method for the distribution of basal sliding coefficients under ice sheets, applied to 650 Antarctica, The Cryosphere, 6, 953–971, https://doi.org/10.5194/tc-6-953-2012, 2012. Rabatel, A., Sanchez, O., Vincent, C., and Six, D.: Estimation of Glacier Thickness From Surface Mass Balance and Ice Flow Velocities: A Case Study on Argentière Glacier, France, Frontiers in Earth Science, 6, 112, https://doi.org/10.3389/feart.2018.00112, 2018. Raymond, M. J. and Gudmundsson, G. H.: On the relationship between surface and basal properties on glaciers, ice sheets, and ice streams, Pollard, D. and DeConto, R. References 535 A., Isaksson, E., and Divine, D.: An iterative inverse method to estimate basal topography and initialize ice flow models, The Cryosphere, 7, 987–1006, https://doi.org/10.5194/tc-7-987-2013, 2013. 675 675 van Pelt, W. J. J., Pettersson, R., Pohjola, V. A., Marchenko, S., Claremar, B., and Oerlemans, J.: Inverse estimation of snow accu- mulation along a radar transect on Nordenskiöldbreen, Svalbard, Journal of Geophysical Research: Earth Surface, 119, 816–835, https://doi.org/10.1002/2013JF003040, 2014. van Pelt, W. J. J., Pohjola, V., Pettersson, R., Marchenko, S., Kohler, J., Luks, B., Hagen, J. O., Schuler, T. V., Dunse, T., Noël, B., and Reijmer, C.: A long-term dataset of climatic mass balance, snow conditions, and runoff in Svalbard (1957–2018), The Cryosphere, 13, 680 2259–2280, https://doi.org/10.5194/tc-13-2259-2019, 2019. 680 29 https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. https://doi.org/10.5194/tc-2023-43 Preprint. Discussion started: 4 April 2023 c⃝Author(s) 2023. CC BY 4.0 License. Welty, E., Zemp, M., Navarro, F., Huss, M., Fürst, J. J., Gärtner-Roer, I., Landmann, J., Machguth, H., Naegeli, K., Andreassen, L. M., Farinotti, D., Li, H., and GlaThiDa Contributors: Worldwide version-controlled database of glacier thickness observations, Earth System Science Data, 12, 3039–3055, https://doi.org/10.5194/essd-12-3039-2020, 2020. Welty, E., Zemp, M., Navarro, F., Huss, M., Fürst, J. J., Gärtner-Roer, I., Landmann, J., Machguth, H., Naegeli, K., Andreassen, L. M., Farinotti, D., Li, H., and GlaThiDa Contributors: Worldwide version-controlled database of glacier thickness observations, Earth System Science Data, 12, 3039–3055, https://doi.org/10.5194/essd-12-3039-2020, 2020. Werder, M. A., Huss, M., Paul, F., Dehecq, A., and Farinotti, D.: A Bayesian ice thickness estimation model for large-scale applications, 85 Journal of Glaciology, 66, 137–152, https://doi.org/10.1017/jog.2019.93, 2020. Werder, M. A., Huss, M., Paul, F., Dehecq, A., and Farinotti, D.: A Bayesian ice thickness estimation model for large-scale applications, 685 Journal of Glaciology, 66, 137–152, https://doi.org/10.1017/jog.2019.93, 2020. Winkelmann, R., Martin, M. A., Haseloff, M., Albrecht, T., Bueler, E., Khroulev, C., and Levermann, A.: The Potsdam Parallel Ice Sheet Model (PISM-PIK) – Part 1: Model description, The Cryosphere, 5, 715–726, https://doi.org/10.5194/tc-5-715-2011, 2011. Zorzut, V., Ruiz, L., Rivera, A., Pitte, P., Villalba, R., and Medrzycka, D.: Slope estimation influences on ice thickness in- version models: a case study for Monte Tronador glaciers, North Patagonian Andes, Journal of Glaciology, 66, 996–1005, 690 https://doi.org/10.1017/jog.2020.64, 2020. Journal of Glaciology, 66, 137 152, https://doi.org/10.1017/jog.2019.93, 2020. Winkelmann, R., Martin, M. References 535 A., Haseloff, M., Albrecht, T., Bueler, E., Khroulev, C., and Levermann, A.: The Potsdam Parallel Ice Sheet Model (PISM-PIK) – Part 1: Model description, The Cryosphere, 5, 715–726, https://doi.org/10.5194/tc-5-715-2011, 2011. Z t V R i L Ri A Pitt P Vill lb R d M d k D Sl ti ti i fl i thi k i Winkelmann, R., Martin, M. A., Haseloff, M., Albrecht, T., Bueler, E., Khroulev, C., and Levermann, A.: The Potsdam Parallel Ice Sheet Model (PISM-PIK) – Part 1: Model description, The Cryosphere, 5, 715–726, https://doi.org/10.5194/tc-5-715-2011, 2011. Zorzut, V., Ruiz, L., Rivera, A., Pitte, P., Villalba, R., and Medrzycka, D.: Slope estimation influences on ice thickness in- version models: a case study for Monte Tronador glaciers, North Patagonian Andes, Journal of Glaciology, 66, 996–1005, 690 https://doi.org/10.1017/jog.2020.64, 2020. 30
https://openalex.org/W2022702793	https://zenodo.org/records/2505882/files/article.pdf	English	null	Samplers and sampling of vegetable oils	Journal of oil & fat industries	1,921	public-domain	4,157	sAMPLERS AND SAMPLING OF VEGETABLE OILS and is the most defective in construction of any of these samplers, as will be noted from the following table. By P. W. TompMns, Curtis & Tompkins, San IEi'ranc~sco, Cal. p g Either sampling tube No. 3 is not drawn to scale or a very obvious objection exists in the valve. We assume the interior diameter of the tube to be 2 inches, and if so represents an area of 3.14 square inches, while the valve at its widest point represents an area of 2.62 square inches. This leaves but 0.52 square inch for the by-pass around the valve, or an opening of only 16.6 per cent of the total area through which the liquid must flow to fill the 2-inch tube. This constitutes a serious objection, since the tube, if not lowered very slowly, will largely fill when the descent is stopped at the bottom of the car. The sampler submitted by Mr. Fash (Fig. 5, May, page 55) is the only one of the three capable of a free and unobstructed flow and of taking a Z-inch core throughout, and has but a single defect, although that is at the most vital point, the bottom. The ~-inch plate and holding nut underneath permits a sample be- ing drawn only to within about ,;~ inch of the bottom, where deposits are at their maximum concentration. If this arrangement were in some way altered to permit the tube reaching to within ~ or ~ inch of the bottom, it would then take practically a full 2-inch core. This reference naturally has only to do with procuring a full 2-inch core, but not to the multiple of error caused by taking a sample in this way, where settlings are present in a tank car. The following remarks concerning the proportions each one of these triers would represent with a given amount of water in a car are understood at the outset to be only approximations, with no greater accuracy attached to the figures than computations from drawings of this kind would permit. In this we take a standard 8,000-gallon tank car containing 5 inches of water at the bottom, the remainder dry oil: p The tube with the standard 2-inch two-way stop-cock and side arm for opening (No. 4--A. O. C. S. Sampler) has the defect of feeding a 2-inch pipe (3.1416 sq. in. sAMPLERS AND SAMPLING OF VEGETABLE OILS area) with an opening in the stop-cock of approximately 1.5 square inch, or an area about 51 per cent less than the 2-inch pipe which it supplies. This necessitates that the tube be submerged no faster than the oil in the tube can reach the level of the outside, as otherwise the back fill will unequally represent some particular part of the car and usually when the descent is stopped at the bottom. The stop-cock also has the disadvantage of a restricted volume and one not in keeping with the 2-inch core. Five inct~es welter in the car, representing Per ce~t 243.5 gallons (2,028 tbs.) ...... 3.3% excess error A true 2-in. core taken to within in. of bottom--4~ in. water in trier ....................... 6.6% 100 Fig. No. I--~ in. above bottom--4~ in. water in trier .............. 6.8% 106 Fig. No. 2--9/10 in. above bottom-- 4 1/10 in. water in trier ........ 7.1% 115 (If trap on side of valve, cage fills.) Fig. No. 3--7/10 in. above bottom-- 4 3/10 in. water in trier ........ 5.5% 67 Fig. No. Z~-about 1 in. above bot- tom-4 in. water in trier ....... 5.3% 60 Fig. No. 5--3~ in. above bottom-- 4~ in. water in trier ........... 5.9% 78 Five inct~es welter in the car, representing Per ce~t 243.5 gallons (2,028 tbs.) ...... 3.3% excess error In the Barrow-Agee Sampler (Fig. 1) a 2-inch intake feeds a 2-inch tube, and the valve at its widest point is 2N inches, leaving a by-pass of only 3~ inch, or a re- striction at that point of about 54 per cent of the intake area supplying the 2-inch tube. Obviously when this instrument is submerged, it should be with the knowl- edge of this restricted by-pass, in order that sufficient time be allowed to permit the liquid to flow in only as fast as it can equalize the outside level. If this pre- caution were not observed and the tube submerged too rapidly and only retarded or stopped when the bottom of the car was reached, the back fill would probably draw several inches (if not more) from the bottom of the car before equalization took place. The idea of the valve is excellent, but without a readjustment to permit a more rapid intake, there is danger of exaggerating settlings more than would be encountered by the geo- metric error of the 2-inch core alone. sAMPLERS AND SAMPLING OF VEGETABLE OILS The several sketches representing tank-car sampling euipment developed by us are merely submitted as sug- gestions for some ingenious mind to perfect and over- come the mechanical defects as they appear to us. Out of this a sampler might be developed that would embody oil desirable elements with none of their disadvantages. g Fig. "A" is our original type of 10-foot sampling tube used many years ago. The valve is ground to a close seat and works satisfactorily as far as this clumsy equipment permits. It is made of 1/16-inch steel or brass tubing, the brass valve lifting against the springs and is caught by a sliding slot on the top guide. The tube is arranged to unscrew in the center, making 5-foot lengths, facilitating cleaning and transportation. It has, however, the same disadvantage of most of this type of equipment, including the difficulty of cleaning and the volumetric error due to the valve cage arrangement making any deposit appear even greater than the geo- metric error of a 2-inch core alone. If a by-pass equal to the area of a 2-inch intake were provided, it would only tend to increase the error of representation, and no matter how it is arranged, the error cannot be avoided by maintaining the standard measurements. Naturally, if a 2-inch intake supplies a 2-inch tube and enough space is allowed for an equiva- lent flow past a 288 valve, there must be a volu- metric error created in the valve cage that cannot be compensated by these measurements and still correspond to the proportions of a 2-inch core. The error caused by the valve cage arrangement in Fig. "A" over that of the 2-inch core was the reason of its being replaced by the type of tube in Fig. "B." The valve is ground in the same manner as in Fig. "A." Later a valve of the type represented in Fig. "C" was employed, which eliminated much of the complicated mechanism of the valve arrangement in Fig. "B." A lead or leather washer produced a sufficiently tight seat to insure against leakage. With this tube, as with all others, handles were used, though not indicated in sketches. p p The sketch referred to in the May issue of the Co~o~ OIL PRESS shows two sampling tubes not for- merly considered (No. 2 P. C. O. R. G. and No. 3 P. C. O. R. Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers" Association (CHEMISTS' SECTION) LERS AND SAMPLING OF VEGETABLE and is the most defective in construction of any 29 Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers" Association (CHEMISTS' SECTION) Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers" Association (CHEMISTS' SECTION) Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers' Association (CE~.mSTS' S~.CTmm Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers' Association (CE~.mSTS' S~.CTmm 3I of the tank. In this way, also, the geometric error in tank-car sampling can be overcome to a large extent. These plungers are made in various lengths and for large storage tanks 2 feet is best. washer was substituted for the ground valve of Fig. "B." Fig. "B," when used as a 10-foot sampling tube, did not of course require the top stationary lid. The arrangement of the tube suggested the possibility of taking independent samples at any depth by decreasing the length and merely utilizing a cover the same as the bottom valve. Fig. "C" has been drawn in the latter form, though without the lid it represents the 10-foot sampler as it was originally used. Its awkward nature, either as a 10-foot tube or as a 10-inch zone sampler, was not overcome. washer was substituted for the ground valve of Fig. "B." Fig. "B," when used as a 10-foot sampling tube, did not of course require the top stationary lid. The arrangement of the tube suggested the possibility of taking independent samples at any depth by decreasing the length and merely utilizing a cover the same as the bottom valve. Fig. "C" has been drawn in the latter form, though without the lid it represents the 10-foot sampler as it was originally used. Its awkward nature, either as a 10-foot tube or as a 10-inch zone sampler, was not overcome. g g As has been brought out in the committee report al- ready printed, it is not possible to make or utilize a sampler that will take a correct sample from tank cars containing sediment. As Mr. D. C. Morris aptly sets forth in the July, 1920, issue of the Cotton OIL PRESS, the manner in which the sediment collects at the bottom of a car, and particularly if banked at both ends, ren- ders any of the usual sampling methods inadequate. The results of Mr. A. W. Putland's observations of moisture and insoluble impurities, determined, we pre- sume, from samples drawn by the 2-inch core trier, in comparison with the kettle samples, clearly illustrates one of the principal defects in the present system, which, however, can be largely rectified if the geometric con- tour of the ear is considered. As an outgrowth of Fig. Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers' Association (CE~.mSTS' S~.CTmm "C" as a 10-inch zone sam- pier, and the latitude it presented under varying condi- tions, the automatic plungers, Figs. "E" and "F," made possible many features of sampling precluded in the 10-foot equipment. However, the first of these plungers were arranged with the flapper valve, but these were too easily disarranged. As a substitute, the valves indicated in Figs. "E" and "F" were employed most successfully. The type of plunger in Fig. "E" has the defect that a l~-inch intake (2.4 sq. in. area) feeds a 2M-inch tube with an area of 3.97 sq. in., or a restricted supply of a little less than 40 per cent. This necessitates that the plunger be lowered slowly enough to allow the fill to equalize the outside (which is a defect), though the rate at which the plunger wouId normally be lowered by the knotted cord is ample for the purpose. In order to permit a valve by-pass equal in capacity to the 13~- inch intake, the valve cage had to be somewhat larger than the rest Of the tube, and to compensate for this increased volume, the balance of the tube must be 2M inches, to equalize the composition o~ the 10-inch sample as a whole. Differences in analyses are rarely traceable to the laboratory, but are the result of incorrectly taken sam- ples. We find in all rules governing tank-car sampling the proverbial "with a trier, of uniform diameter of 2 inches, and which trier must remain open as it passes through the oil and closed only when the bottom of the tank is reached," etc. But how often does the instrument employed even comply with this requirement, to say nothing of the geometric error ? g Let us consider a standard tank car with 1~ inches of water at the bottom, and for convenience the re- mainder dry oil. This would represent about 52 gallons of water, or about 0.68 per cent in the ear as a whole. Assuming that all instruments mentioned below would take samples to the bottom, the Samples procured, though intended to represent the average composition of the car, would give the following results, according to their respective errors : To overcome the defect in this plunger (Fig. "E"), the one indicated by Fig. "F" proved a most satisfactory substitute. This pipe valve is properly weighted and moves as readily as the valve in Fig. Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers' Association (CE~.mSTS' S~.CTmm "C"; in fact, it has a quicker response. The intake of 2 inches has a by-pass of about the same capacity and holds the equiva- lent of a 2-inch core, with the exception of a small section at the top, which, however, is well above the usual sediment strata and therefore not a factor. Per cent. ACTUAL WATER IX Ca~ 1~ I~rCI~ES ............ 0.68 True 2-inch core (Fig. No. 5--Fash Sampler).. 2.1 Tube with the 2-inch two-way stop-cock vaIve (No. 4--A. O. C. S. Sampler) approximately. 1.0 The 2-inch tube of the Barrow-Agee valve type, (Fig. No. 1) approximately .................. 2.3 Tube with valve cage represented by our Fig. "A" (discarded years ago) .................. 3.7 Per cent. ACTUAL WATER IX Ca~ 1~ I~rCI~ES ............ 0.68 True 2-inch core (Fig. No. 5--Fash Sampler).. 2.1 Tube with the 2-inch two-way stop-cock vaIve (No. 4--A. O. C. S. Sampler) approximately. 1.0 The 2-inch tube of the Barrow-Agee valve type, (Fig. No. 1) approximately .................. 2.3 Tube with valve cage represented by our Fig. "A" (discarded years ago) .................. 3.7 Both these plungers are so simple and easily handled and accessible fo.r cleaning that their use has become quite general here, and for sampling large tanks are of great assistance. In tank cars, the plunger is lowered to the bottom, when the bottom valve and lid open and the liquid passes through as it descends, closing only when the instrument is brought to a stationary position. The sample is consequently taken from the particular section where the instrument is stopped in its descent and the valve and lid closed. The depth to which the sampler is desired to descend is regulated by a cord attached to the handle (knots every 10 inches the length of the plunger), thus enabling every 10 inches of the tank to be sampled (complete core), or as many samples from any particular zone as desired, or that may be necessary to compensate for geometric errors of the container. We have in mind several other sampling devices of this kind, where the valve cage or other construction would cause the sample to appear as if the car con- tained much more water than in reality, yet the parties using them had every confidence in the accuracy and authentic character of their samples. sAMPLERS AND SAMPLING OF VEGETABLE OILS G.) that are apparently quite defective in their construction. If in Figure No. 2 the projecting cage space on left side of valve entirely fills, a serious objection is found, and, with the excess volume from the convex right side of cage, would together constitute a considerable error The tube represented by Fig. "C" was the successful outgrowth by which the valve with leather or lead THE COTTON Of L PRESS (CIIEI~ISTS' SECTION 3 ~ ( THE COTTON OIL PRESS (CHEMISTS' SECTION) 7 p g, ] , I wish to call attention to the very splendid work done by the Oil Sampling Sub-Committee under Mr. Perry as chairman. y MR. P.va~Y: We decided to recommend to the Society for its consideration a change in the rule on sampling (see May issue, page 54). Then we decided, too, that a great many samplers were being used, and to collect and present for your consideration the different types that were being used, with the idea that the discussion which you would have over these samples would stand- ardize our ideas as to the type of sampler to be used. Messrs. Curtis and Tompkins sent us some drawings which show the different types of samplers. The sam- pler A (see Tompkins' paper following) has the objec- tion that in order to make a full opening he has enlarged the chamber just at the point where it would hold more settlings than the rest of the sampler. That has been corrected in B, and still another type comes in C. This is very similar to the type suggested by Mr. Fash, except that the valve is pushing down into the fools and would, I think, not be quite so satisfactory as the other type. While it is not possible to use the "Official Sampling Tube" or the "Automatic Plunger" altogether correctly, without previous knowledge of the depth of water in the car, and that is not always practically possible at the time of sampling, even with the "test paper" used to learn the approximate depth of water in gasoline tanks, yet it is practical and certainly more logical to use the sampling tubes with some idea of maintaining an aver- age geometric ratio that will come much closer to the facts than is possible by the present practice, as will be noted in Column 6. Mr. Tompkins brings up a new type of sampler, a zone sampler (Figs. D, E, F). It has a valve with seats in the bottom and it is lowered down open, and wherever you want to take a sample in the car the core is closed and brought out. It keeps the oil from mixing in the top. That, he says, is used a great deal where they have large storage tanks in San Francisco and want to ascertain the condition of the oil in the tank at different strata. THE COTTON OIL PRESS (CHEMISTS' SECTION) As a further suggestion on sampling tank cars, where a refining kettle is not accessible or the trier not ade- quate, the pipe-line bleeders represented in Figs. "G" and "H" have been found very useful. However, Bleeder "G" has the objection that it restricts the pipe flow, and though it is intended to represent the pipe in its geometric proportions, it has the added disad- vantage of collecting sediment and fiber, stopping some of the openings, and acts more as a strainer than is desired. This does not occur, however, where a pipe- line strainer is placed in the line. He gives us another suggestion of what they are do- ing out there, in this drawing showing the bleeder sam- pler (Figs. G and H). In the first one there is an inside pipe connected with a small valve over the bottom. The perforations are around the diameter of the pipe. p p p f would like to have discussion as to the suggestion we have made about permitting the refiner to unload the oil where settlffigs are present, and also a discussion as to the type of sampler. p As a rest/lt of this occasional obstruction, the bleeder represented by Fig. "H" was developed and found most successful, and is made up entirely of standard plumb- ing equipment. Occasionally one of these openings will clog, but not often, as the holes widen on the outward flow and if fragments pass in at all they usually go through. It can be coupled into any line or taken out in a minute or two and cleaned and replaced without trouble. The usual type of bleeder which consists of a small faucet on the under side of a pipe is wholly un- reliable, as any water passing along the bottom of the line is represented in greater proportions than exists and misrepresents the car. The official sampler used at New Orleans, it seems to me, is open to very serious criticism in that the core of oil is not a full core; it is cut off right in the center and you are taking just approximately one-half of the settlings that would be present in that section of the tank. DR. THE COTTON OIL PRESS (CHEMISTS' SECTION) 32 SAMPLING COMMITTEE REPORT By Ed. R. Barrow, Barrow-Agee Laboratories, Memphis, Tenn. and if triers or plungers are to be used at all, let it be with a view of proportionate ratios. In any event, the most logical way in which the trier could be used is always to accompany the composite with an independent bottom sample (10 inches or 1 foot), and with this information at hand some idea of the actual deposit in the car can be obtained. and if triers or plungers are to be used at all, let it be with a view of proportionate ratios. In any event, the most logical way in which the trier could be used is always to accompany the composite with an independent bottom sample (10 inches or 1 foot), and with this information at hand some idea of the actual deposit in the car can be obtained. EDITOR'S NOTE: The report of the Oil Sampling Sub- Committee appeared in our May issue, on Page 54. What follows is subsequent discussion c~t the conven~iovr and revision of the report Mr. Tompki~,s submitted to Mr. Perry, which appears only in abstract in the convention ~otes. Even if sufficient attention was paid to the construc- tion of the official sampling tube to insure a true core, whether it be 2 inches or any other dimension, it is use- less to use it as described in the "Official Rules," which disregard any attempt to correct the geometric ratio. The sub-committees under this Sampling Committee are on Meal Sampling, of which R. F. Montsatvat~ i~ chairman; Cake Sampling, N C. Hamner, chairm~2 ~ an Hulls and Lint Sampling, H. G. Lewis, chairman; Oil Samphng," R. W. Perry, chairman; Seed Sampli-a~.,o, J.r o.r~ Evans, chairman; Soap Stock and Grease Sampling, Archibald Campbell, chairman; Other Oil Seed Prod- ucts Sampling, W. ]3. Gilmer, chairman. g y p g The following table is a striking illustration of a car containing deposited water, in the following amounts, and sampled by the "Official Method" (Column 5) : y-,, ~ < ~ < Gallons Pounds 78" 66" Down % % % i 32.39 ~7 o i 2 37" 0-44 1.4 0.55 2 70.27 586 1 2 I9" 0-95 2.8 I.IO 3 I2I-3~ IOlO I I 54" 1-64 4,I 1.64 4 178.56 1485 I I 34" 2-41 5.5 2.2 5 ~43.5 z 2030 1 1 2o" 3.3 ~ 6. 9 2. Ot~cial Monthly Bulletin Interstate Cotton Seed Crushers' Association (CE~.mSTS' S~.CTmm In spite of this, any variation in results is considered by the trade, and often by the profession, as an error in the analysis. If tank cars are to be sampled by triers at all, why not arrange and use them as accurately as they permit? If an official sampling tube or the automatic plunger, capa- ble of taking a true core, was employed in the following manner, considering the above example, a reasonably cor- rect approximation to the 0.68 per cent would be obtained. By taking the official tube once the full length twice to within 1 foot of the bottom and once 30 inches down from the top, and by the 10-inch plunger once from the bottom and 23 times from above this point, the geometric error of the sample is largely corrected. Of course, it is very convenient to work out a remedy and correct errors when the amount of water is known in advance, but a uniform system can be employed that will greatly reduce errors in samples caused by deposits, One instance where these plungers were used satis- factorily was for a firm dealing exclusively in cocoanut oil, with 14 storage tanks, each having a capacity of about 317,000 gallons and nearly 30 feet high. Here "zone sampling" frequently became a necessity, to de- termine if all parts of the tank were similar, or at what points loots or sediment would be encountered. If enzymatic action set in, it would be necessary to deter- mine to what extent the storage tank was affected, or rather how far down the oil might be drawn before the affected zone would be reached, as with these pIungers it is possible to take individual samples from any one zone without interference from any other part THE COTTON OIL PRESS (CHEMISTS' SECTION) WESSON: I believe it is absolutely impossible to get a fair sample of a tank car by putting a tube down into it if loots are present, on account of the geometric shape of the tank. You can't get any more out of a job than what you put into it. p MR. PERRY: That was commented upon in the report. At the same time, we know there is a tendency of the settlings to go to the end of the car. The motion of the tank seems to throw them back and forth, so there is no sampler that can take this into consideration. If you don't think cooperation is necessary, notice what happens when a wheel comes off a wagon.
https://openalex.org/W4313397892	https://dirros.openscience.si/Dokument.php?id=21948&dn=	Pashto, Pushto	null	Redrawing HER Volcano with Interfacial Processes—The Role of Hydrogen Spillover in Boosting H2 Evolution in Alkaline Media	Catalysts	2,023	cc-by	31,600	WXYXZ[\Y\ ]^8_`9a bcdLeMfgbhijeNkcLljimLjcnncnopNcbicl fKdLMcqJiircLsnMf<grituivicwcd qxyz{t\| j} ~wcLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ~ sxcd\| jhnixcr\| j> dkMLuzm>~ 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, $)# 0R/ # 3 EFFFP%$% $ # 00 $ $3# # Y\ bhijeNkcLljimLjcnncnopNcbicl ircLsnMf<grituivicwcd cLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ~ dkMLuzm>~ E 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, 4 $)# 0R/ # 3 EFFFP%$% $ G $ 30'/3# $R/ # 3 # 30&$,! EEFFF&$,! V 0R/ # 3 $- ,/.! !#/)# 0/ R$$,0 #E4II . R#!C ,00/,# unLj2/ #0 /0 $ 00## $0$ !+ / /,+ , ,3 # # $#0#/3!,! .0 ! # 00 $ $3# # !!0-4! !$# , #0 / ! #3# !$!- + $ Q! R$SRU# $ # !# / /3!,$ S-.U# ! /3!,# $$ /# +/ / $3# #0 #$,$3 Y\ bhijeNkc iircLsnMf< cLJ~uicvndLzyrr\| j dkMLuzm E 0R/ # 3 $ 30 4 $)# 0R/ # 3 EFFF G $ 30'/3# $R/ # 3 V 0R/ # 3 $ #E4II . R#!C ,00/,# unLj2/ #0 / ,3 # # $#0#/3! !!0-4! ! $ Q! R$SR S-.U# ! /3!, ! 3!#!/3!,! Y\ bhijeNkcLljimLjcnncnopNcbicl ircLsnMf grituivicwcd 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, $)# 0R/ # 3 EFFFP%$% $ G $ 30'/3# $R/ # 3 # 30&$,! EEFFF&$,! V 0R/ # 3 $- ,/.! !#/)# 0/ R$$,0 #E4II . R#!C ,00/,# unLj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cKeLdn/3!,$ /3!,# $$ 0 $#### ! $3# # !!,/ !5 R$$/3!,!# }zkLdtj -3!,.$ * S-.U/#$+3#/!# $$ // #0 $ / # #-+ # $ ! 3 5 $E 0 /F3# 0 #/ # , /$, $$#/ !! /+ !#!#0+ $ $3# #0/3!,! )# 0+/ / # -. ! * !C4I * #!C4A !C4A '$ #/!CE JKLMNO P #') 2/ # $ # ! # ! ! #0 / * !C4I * #!C4A !C4A '$ #/!CE JKLMNO 4 P #') 2/ # $ # ! # ! ! ! #0 / $ / # #-+ # $ ! 3 5 $E 0 /F3# 0 #/ # , /$, $$#/ !! /+ !#!#0+ $ $3# #0/3!,! )# 0+/ / # -.* ! $5$ ! 4 G / /#/ $ ! /53 ###$ ! /+ !#! #0$ $3# #0-4! CSEU$+,30 3 S4U# $ 30 / $ !#! ,! # !SGU$ 0$ $# +/ / /# $3# #0 / ##E-+ / ,3 # #$ # /#/0+#$ #00 -4! 2 / #!3 ! 00 $###0 $#/#!0-.2/# $! $$ /##S$!$ $# + / , / !# $# !!0-. $5$ ! U $$3#S #!$$3# , /# 5+ #/# U # $3/3! !# !# #$!# ! /#VA /! 00 $#$### $$ !# ! , $3 !# #0, R# ! , /-.* 3 $3 #0! 00 $##$$3 $ Q! /00 / $$ $$3!! 32# B +/ /+#! Q!# ,-.* # ! ! //3!,!# , #0 /2$$ #3 #5 ,DEFGGBFD $EGFEFFAB / #CDD+++! D%$D $3# # * #!C4A4F44 !C4A4F44 '$ #/!CE34F4G JKLMNO 4F4G3 / /# P #') &#$ + Q$! 2/ # $ ### $ ! # ! ! / # ! ! #0 /R R# SRR&TU$ #S/ #CDD #,D$ ##D3D VFDU 67879:;8;<=<> ?@ AB/ #CDD! VA /! 00 $#$### R# ! , /-.* #$$3 $ Q! WXYXZ[\Y\ ]^8_`9a bcdLeMfgbhijeNkcLljimLjcnncnopNcbicl fKdLMcqJiircLsnMf<grituivicwcd qxyz{t\| j} ~wcLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ~ sxcd\| jhnixcr\| j> dkMLuzm>~ 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, $)# 0R/ # 3 EFFFP%$% $ R$SRU# $ # !# / /3!,$ S-.U# ! /3!,# $$ /# +/ / $3# #0 #$,$3 ! 3!#!/3!,! ,! #&#! / # ,/ #0R + #3 /# Q!# #0!!,/ !# ! $3# #! ! / # +!#-. $5$ ! + / / 0#! , $#, ! / # 00 #, 0' '! R ! /#0 /# /# -.* 32/-.$ # !# ,$$ !/3!, ! ,, #! #!-. # #/+## #/ 0 0 /# , ! ,/2/ #+5!# # /## $ #0 $# 0, ,0! ,00 -.* $3# #! !#,$, !$ #0/# ,$ $3# # #0$ $3 /3!,! cKeLdn/3!,$ /3!,# $$ 0 $#### ! $3# # !!,/ !5 R$$/3!,!# }zkLdtj -3!,.$ * S-.U/#$+3#/!# $$ // #0 WXYXZ[\Y\ ]^8_`9a bcdLeMfgbhijeNkcLljimLjcnncnopNcbicl fKdLMcqJiircLsnMf<grituivicwcd qxyz{t\| j} ~wcLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ~ sxcd\| jhnixcr\| j> dkMLuzm>~ 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, $)# 0R/ # 3 EFFFP%$% $ YXZ[\Y\ MfgbhijeNkcLljimLjcnncnopNcbicl qJiircLsnMf<grituivicwcd ~wcLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ixcr\| j> dkMLuzm>~ E 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, 4 $)# 0R/ # 3 EFFFP%$% $ G $ 30'/3# $R/ # 3 # 30&$,! EEFFF&$,! 0R/ # 3 $- ,/.! !#/)# 0/ R$$,0 #E4II . Y\ bhijeNkcLljimLjcnncnopNcbicl iircLsnMf<grituivicwcd cLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ~ dkMLuzm>~ E 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, 4 $)# 0R/ # 3 EFFFP%$% $ G $ 30'/3# $R/ # 3 # 30&$,! EEFFF&$,! V 0R/ # 3 $- ,/.! !#/)# 0/ R$$,0 #E4II . R#!C ,00/,# unLj2/ #0 /0 $ 00## $0$ !+ / /,+ , ,3 # # $#0#/3!,! .0 ! # 00 $ $3# # Y\ bhijeNkcLljimLjcnncnopNcbicl ircLsnMf<grituivicwcd cLJ<~uicvndLzyrr\| j> ~{z{cLcL >~ ~uqzL> ~ dkMLuzm>~ E 0R/ # 3 $ 30 # 30% EFFF% &# !-Q, 4 $)# 0R/ # 3 EFFFP%$% $ G $ 30'/3# $R/ # 3 # 30&$,! EEFFF&$,! V 0R/ # 3 $- ,/.! !#/)# 0/ R$$,0 #E4II . R#!C ,00/,# unLj2/ #0 /0 $ 00## $0$ !+ / /,+ , ,3 # # $#0#/3!,! .0 ! WXYXZ[\Y\ ]^8_`9a bcdLeMfgbh fKdLMcqJiir qxyz{t\| j ~wcL WXYXZ[\Y\ ]^8_`9a bcdLeMfgbh fKdLMcqJiir qxyz{t\| j ~wcL WXYXZ[\Y\ ]^8_`9a bcdLeMfgb fKdLMcqJiir qxyz{t\| j ~wc WXYXZ[\Y\ ]^8_`9a bcdLeMfgbh fKdLMcqJiir qxyz{t\| j ~wcL WXYXZ[\Y\ dLeMfgbh dLMcqJiir yz{t\| j ~wcL icl cd L ~ /3!,! cKeLdn/3!, !!,/ }zkLdtj -3!,. $ / # # 0 #/ $ $3# #0/ $5$ ! 4 #0$ $3# $ !#! +/ / /# cKeLdn/3!,$ /3!,# $$ 0 $#### ! $3# # !!,/ !5 R$$/3!,!# }zkLdtj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`abcadef_dgh . ,+ &")$1- ,&+& ,&&, + 5- !&")&+" $+7i-+:j . =klk7mG:n . k=.l 7: +& )+" * $55+ 5)+ 7H=: +" $556 B M:, 55!" &5 +&H ,&&5+-FE +!+& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &" + 5- ,&mG & 5 -& $+&!"% 5)+ " 7H=: +" $556 B M:, 55!" &5 +&H= ,&&5+-FEG +! +& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &") + 5- . ,&mG & 5 -& $+&!"% 5)+ " 7H=: +" $556 B M:, 55!" &5 +&H= ,&&5+-FEG +!+& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &") + 5- . ,&mG & 5 -& $+&!"% 7H=: +" $556 > EIH=+" 7E JK) K$ L ( G B B M:, 55!"+5 $./0+"- $- + &5 +&H= $55- &! . )" )- ,&&5+-FEG &$ +&))&5 & )"! )./0 +! +& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &")$1- ,&+& ,&&, + 5- !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: ,&mG & 5 &+" $+D&8 5 . - -& $+&!"%7/3$7::." 1"+7/3$76::j . k. n . k7mG: 7: . =k k. n . k7mG:k=. 76: ./0 , :j 7: 7: 76: FEG &$ +&))&5 & )"! )./0 V dgh )$1- ,&+& ,&&, !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: &+" $+D&8 5 . - %7/3$7::." 1"+7/3$76::j . k. n .k7mG: 7: . =klk. n . k7mG:k=.l 76: +! +& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &")$1- ,&+& ,&&, + 5- !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: ,&mG & 5 &+" $+D&8 5 . - -& $+&!"%7/3$7::." 1"+7/3$76::j . k. n .k7mG: 7: .=klk. n .k7mG:k=.l 76: +! 5)+ " 7H=: +" $556 B M:, 55!" &5 +&H= ,&&5+-FEG +!+& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &") + 5- . ,&m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`abcadef_dgh . ,+ &") + 5- . ,&m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`abcadef_dgh . ,+ &") + 5- . ,&mG & 5 -& $+&!"% +& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &")$1- ,&+& ,&&, + 5- !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: ,&mG & 5 &+" $+D&8 5 . - -& $+&!"%7/3$7::." 1"+7/3$76::j . k. n .k7mG: 7: .=klk. n .k7mG:k=.l 76: _dgh &")$1- ,&+& ,&&, !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: 5 &+" $+D&8 5 . - !"%7/3$7::." 1"+7/3$76::j . k. n .k7mG: 7: .=klk. n .k7mG:k=.l 76: )$1- ,&+& ,&&, !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: &+" $+D&8 5 . - 7/3$7::." 1"+7/3$76::j . k. n .k7mG: 7: =klk. n .k7mG:k=.l 76: , :j 7: 7: 76: - !&")&+" $+7i-+:j .=klk7mG:n . k=.l 7: 5 &+" $+D&8 5 . - !"%7/3$7::." 1"+7/3$76::j . k. n .k7mG: 7: .=klk. n .k7mG:k=.l 76: :j 7: 7: 76: - !&")&+" $ .=klk7mG:n . k=.l 5 &+" $+D&8 !"%7/3$7::." 1"+ . k. n .k7mG: .=klk. n .k7mG:k=.l %.& !5( $ /0128 * %.& !5( $ /0128 ! "#! $%#$ ! &% ! %$ %' %' $%' ! $$ % (( ! $!&)$%%$ * (+! $ ! ) !')! ! ,#-+,#-&%.& #"$ $/012(3'$% $ %'! %$$"$ %4%' !5( 6 ! &% %$%' ! $ % 7'"'#"% "%! " $ %$ ! &% %'$ ! $%!!! 5)+ " 7H=: +" $556 B M:, 55!" &5 +&H= ,&&5+-FEG +!+& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &") + 5- . ,&mG & 5 -& $+&!"% (+!'% %'%$%%%$;$ >%'%%%$ % ! $ !( +!'% ! &$#%'#%&% #%%%""& ! %%%#$ $" $% $ %'! %$$"$ %4%' !5( ! $ 7'"'#"% "%! " $ $ ! $%!!! %%$ ! %' /0128 %9:;<=9:><? 9:;<=%9:>< 9@< % $% ! %' %'(A % !!%$ ! #08 %9:><=%9:><? 0=09:><( 91< $ % ! 4$ !5(%%%%" ! $ % $%''%$'%(>'#" &%% %% . (+!'% %'%$%%%$;$ % ! $ !( '#%&% #%%%""& ! %%%#$ $" $%(+! )# & ! %&% #$ $%BCB'$ %( % $% $# % $% $ % !%%D'E# & !5( $ /0128 9@< $ ! 91< ! % $ " %( & $ ! &% %'$ ! $ !!! %%$ ! %' /0128 %9:;<=9:><? 9:;<=%9:>< 9@< ! 9:><%$ %$% $% ! %' %'(A % !!%$ ! %' $#"$ #08 %9:><=%9:><? 0=09:><( 91< +!% !"$ % ! 4$ !5(%%%%" ! &% $%' $ % $%''%$'%(>'#" &%% 7'% ! %%% . (+!'% %'%$%%%$;$ >%'%%%$ % ! $ !( +!'% ! &$#%'#%&% #%%%""& ! %%%#$ $" % $ $%$%(+! )# & ! %&% #$ $%BCB'$ %( ;$%$) ! '%'%D% $% $# % $% $ % !%%D'E# & # ! ! 1BC1B$# $ ! %'(F#$% !%%D$$ "$%'%$) % $%' $%"' )$')! 5)+ " 7H=: +" $556 B M:, 55!" &5 +&H= ,&&5+-FEG +!+& NOPQRSTQUVWXYQZRQQY[V \] ]^_`abcadef_dgh . ,+ &") + 5- . ,&m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`abcadef_dgh . ,+ &") + 5- . ,&m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efghijklmnopqorstmurtvwxpyz{\|nzp\|nmn}t~s}tnoxpyzyss{osowzxpyzxopo}sp vysrntsyzozvp\|nsry~nttopnrtvwxpyztqp\|nswrrtptp\|nxzwtopyzppo rtvwxpyztp\|nxzwtopyzrtvwxpyztqp\|nswrrtptp\|nxzwtopyz pportvwxpyztp\|nxzwtopyz m{n~nt owzynvryxpwtnxoznrtnsnzpnv s\|{yzp\|op{\|nzyztq { xopo}spx~ntonzp\|nswrrtp p\|nmurtvwxpyzyzxtnosns {\|ynypwtp\|ntvnxtnosns {\|nzypnxqnsotntp\|ozywtn sry~nttopnsnpppp\|opryzp p\|n wzvot}nzp\|spotpspvnxtnosn\|ntntn yp\|nsry~ntrop\|\|osp\|nvqyzozp tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n~ntomurtvwxpyztopnozvtnswpyzo ~xozs\|ornxwt~np\|nm rtvwxpyztopn~ntswsqnpotoxpyzzp\|nswrrtp xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|norn{wvvnrnzvzp\|noxpwotopns +, /21 !"#$%&&$$&'(&) /01 /21 /41 +KLM?KAM?>?NE+ODPNE?@IAJ+KJDDQ+L@D@,HKIAJ+K+,I@A@ 01CE+GHIAJ+K,E+B AM?DHCC+EA,+EAM?STUI+K,JVHE@AJ+KW HE@AJ+KW/31CE+GHIAJ+K,E+BAM?DHCC+EA,+EAM?UTSI+K,JVH SI+K,JVHE@AJ+KY KDL?I+KDJG?E?GAM??,,?IAD+,B?A@QI+O?E@V?YZJ,,?E @KGDL?E?CQ@I?GHKJ,+EBQN+KAM?DHCC+EA\I@A@QNDA/" ?E@A?D+,> GJ,,HDJ+K+KAM?DHCC+EA/&]H@AJ+K/^11@ z{\|nzp\|nmn}t~s}tnoxpyzyss{osowzxpyzxopo rtvwxpyztqp\|nswrrtptp\|nxzwtopyzp zrtvwxpyztqp\|nswrrtptp\|nxzwtopy wtopyz nxoznrtnsnzpnv s\|{yzp\|op{\|nzyztq p p\|nm rtvwxpyzyzxtnosns {\|ynypwtp\|ntvnxtnos /01 /31 56789:<=>?@AB@CD,+E>FCE+ GJDC?EDJ+K@KGAM?DCJQQ+O?EE A+A@QCE+GHIAJ+K,+EAM?STUI AJ+KW/41A+A@QCE+GHIAJ+K,+EA (KAM?K?!AD?A+,DJBH KHB[?ED+,I@A@QNDA\DHCC+ HE?F@1Y(KC@E@QQ?QL?O@EJ efghijklmnopqorstmurt vysrntsyzozvp\|nsry~nttop rtvwxpyztp\|nxz pportvwxpyztp\|n m{n~nt owzynvry xopo}spx~ntonzp\|nsw /21 /01 /41 @IAJ+K CC+EA BAM?D ,,?IAD+ EBQN+ +KAM? xpyzy tptp nswrr {yz tnosns /31 567 GJD A+A AJ+ KH HE efg vys rt xop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efghijklmnopqorstmurtvwxpyz{\|nzp\|nmn}t~s}tnoxpyzyss{osowzxpyzxopo}sp vysrntsyzozvp\|nsry~nttopnrtvwxpyztqp\|nswrrtptp\|nxzwtopyzppo rtvwxpyztp\|nxzwtopyzrtvwxpyztqp\|nswrrtptp\|nxzwtopyz pportvwxpyztp\|nxzwtopyz m{n~nt owzynvryxpwtnxoznrtnsnzpnv s\|{yzp\|op{\|nzyztq { xopo}spx~ntonzp\|nswrrtp p\|nmurtvwxpyzyzxtnosns {\|ynypwtp\|ntvnxtnosns {\|nzypnxqnsotntp\|ozywtn sry~nttopnsnpppp\|opryzp p\|n wzvot}nzp\|spotpspvnxtnosn\|ntntn yp\|nsry~ntrop\|\|osp\|nvqyzozp tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n~ntomurtvwxpyztopnozvtnswpyzo ~xozs\|ornxwt~np\|nm rtvwxpyztopn~ntswsqnpotoxpyzzp\|nswrrtp xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|norn{wvvnrnzvzp\|noxpwotopns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efghijklmnopqorstmurtvwxpyz{\|nzp\|nmn}t~s}tnoxpyzyss{osowzxpyzxopo}sp vysrntsyzozvp\|nsry~nttopnrtvwxpyztqp\|nswrrtptp\|nxzwtopyzppo rtvwxpyztp\|nxzwtopyzrtvwxpyztqp\|nswrrtptp\|nxzwtopyz pportvwxpyztp\|nxzwtopyz m{n~nt owzynvryxpwtnxoznrtnsnzpnv s\|{yzp\|op{\|nzyztq { xopo}spx~ntonzp\|nswrrtp p\|nmurtvwxpyzyzxtnosns {\|ynypwtp\|ntvnxtnosns {\|nzypnxqnsotntp\|ozywtn sry~nttopnsnpppp\|opryzp p\|n wzvot}nzp\|spotpspvnxtnosn\|ntntn yp\|nsry~ntrop\|\|osp\|nvqyzozp tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n~ntomurtvwxpyztopnozvtnswpyzo ~xozs\|ornxwt~np\|nmurtvwxpyztopn~ntswsqnpotoxpyzzp\|nswrrtp xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|norn{wvvnrnzvzp\|noxpwotopns op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~ntonswrrtp}p\|nxopo}spr\|osn wpswx\| vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nxosnyvnrsypnvztnvwxnvtor\|nzn yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsqnsrnxyxwyvnyznstnotvyzp\|nx\|yxn oswyponxopo}spswrrtpxqyzopyz{yp\|\|y\|m«¬oxpy~yp}\|n}tnopnpp\|nx\|yxn p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}ozvp\|nswrrtp{yp\|{ntmovsoxpy~yp}s p\|opypxozzprtqpnp\|ntqopyzmovswpxozoxxnrpmovstq p\|nxopo}sp p\|nswrrtpsopysnsp\|ysxzvypyz m vywsyzozvtnxqyzopyz{yosnosp sp\|nxzptywpyzp\|nsry~ntrop\|{yn\|y\|zvortrntxopo}sp®swrrtp wzvot}nzp\|spotpspvnxtnosn\|ntn tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n ~xozs\|ornxwt~np\|nmurtvwxpyztopn xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|n op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~nton vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nx yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsq oswyponxopo}spswrrtpxqyzopyz{yp\|\| p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}oz p\|opypxozzprtqpnp\|ntqopyzmovs p\|nswrrtpsopysnsp\|ysxzvypyz movsvyw sp\|nxzptywpyzp\|nsry~ntrop\|{yn xqyzopyz {nozp\|nm«¬~xozyzo ornp\|n~xozxwt~nsp\|opp\|nxopo}sp®sw tnyzm rtvwxpyz p\|ys{yvnxtnosnp\|n~ntom rtvwxpyztopnozvtnswpyzo ~xozs\|ornxwt~np\|nm rtvwxpyztopn~ntswsqnpotoxpyzzp\|nswrrtp xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|norn{wvvnrnzvzp\|noxpwotopns op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~ntonswrrtp}p\|nxopo}spr\|osn wpswx\| vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nxosnyvnrsypnvztnvwxnvtor\|nzn yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsqnsrnxyxwyvnyznstnotvyzp\|nx\|yxn oswyponxopo}spswrrtpxqyzopyz{yp\|\|y\|m«¬oxpy~yp}\|n}tnopnpp\|nx\|yxn p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}ozvp\|nswrrtp{yp\|{ntmovsoxpy~yp}s p\|opypxozzprtqpnp\|ntqopyzmovswpxozoxxnrpmovstq p\|nxopo}sp p\|nswrrtpsopysnsp\|ysxzvypyz movsvywsyzozvtnxqyzopyz{yosnosp sp\|nxzptywpyzp\|nsry~ntrop\|{yn\|y\|zvortrntxopo}sp®swrrtp xqyzopyz {nozp\|nm«¬~xozyzooyznqnvyo¯u°±nxzxnzptopnopp\|n ornp\|n~xozxwt~nsp\|opp\|nxopo}sp®swrrtps}spnq{y\|o~n\|y\|oxpy~yp}n~nz %# 5 6 # 7 ! "#$%#%#&'( #)%#"&#)$) ###+%# +)")""+,-&.-& /-&,$.01 23 4%5 6 7 ( ( $) 8 #)$###+9: ;< =%#### &'(# #)#"&'(# ##+#"+&#)" 7 ##)>##%# YVNWOBlm\ROVObbNbo]N`OV[NXP^NVOPNSUTM X[TTYS[U\U\P^NSYRRUVP@tuYOP[U\@vBBO\X P^NVOPNUTM XNSUVRP[U\TVUQ P^NSYRRUVP@tuYOP[U\@wBBqNP]NN\xO\XxyylMU]N`NVo TUVU\NS[QYbOP[U\o]NSNPP^NVOPNSUTP^NRVUZNSSNS@vBO\X@wBPUqNNuYOb@P^YSoxzxoxyz xyoO\XxyyzxyyBl{OQNb_oX[TTYS[U\qOVV[NVSZObNS][P^MOXSq[\X[\hN\NVh_\|W}~o]^[bN MOXSVNZUQq[\OP[U\][bbObSUqNTOSPNVTUVP^N]NOaNVq[\X[\hSYVTOZNSlsUVNOZ^RO[VUTP^N ZU\S[XNVNXVOPNoSR[bbU`NVVOPNOQUY\P[\hPUylyxoxoO\Xxyy]OSPNSPNXls[hYVNWqzNS^U]S P^NVNSYbPSTUVP^NxyyzxyyRO[VUTVOPNSUTVNOZP[U\@vBO\X@wBl ! "#$%#%#&'( #)%#"&#)$) ###+% +)")""+,-&.-& /-&,$.01 23 4%5 7 ( ( $) 8 #)$###+9: ;< =%### &'(# #)#"&'(# ##+#"+&#)" ##)>##%# YVN@WOBl@m\@ROVObbNbo@]N@`OV[NX@P^N@VOPNS@UT@M @X[TTYS[U\@U\@P^N@SYRRUVP@@tuYOP[U\@@vBB@O\X@ P^N@VOPN@UT@M@XNSUVRP[U\@TVUQ@P^N@SYRRUVP@@tuYOP[U\@@wBB@qNP]NN\@x@O\X@xyyl@MU]N`NVo@ TUV@U\N@S[QYbOP[U\o@]N@SNP@P^N@VOPNS@UT@P^N@RVUZNSSNS@@vB@O\X@@wB@PU@qN@NuYOb@@P^YSo@xzxo@xyz xyo@O\X@xyyzxyyBl@{OQNb_o@X[TTYS[U\@qOVV[NV@SZObNS@][P^@MOXS@q[\X[\h@N\NVh_@\|W}~o@]^[bN@ MOXS@VNZUQq[\OP[U\@][bb@ObSU@qN@TOSPNV@TUV@P^N@]NOaNV@q[\X[\h@SYVTOZNSl@sUV@NOZ^@RO[V@UT@P^N@ ZU\S[XNVNX@VOPNo@SR[bbU`NV@VOPN@OQUY\P[\h@PU@ylyxo@ xo@O\X@xyy@]OS@PNSPNXl@s[hYVN@ WqzN@S^U]S@ P^N@VNSYbPS@TUV@P^N@xyyzxyy@RO[V@UT@VOPNS@UT@VNOZP[U\@@vB@O\X@@wBl@ @ @dB@ @@@ @@@ @ @ @ @ @ @ @ @ ¡¢£¤¥ @§@¨@©ª«¬«®@¯°¬±²³¬´@@µ¶@ª±¯³·¬´@´@¬¸¹@°º°¬¹»@¼¬¸@°®¼@µ¹º±½°¾º@±¹«·¬´@¼¬¸@ ¯¹±¹´¬@·½¹±«¿¹@@·«¬«®º°¬@ª¸«°¹@²±¿¸¬¹±@·®±@°¬«´¯°@±@»±¹@´¬¹´°¹@µ¶@ª±¯³·¬´@¬¸¹@°º°À ¬¹»°@¼¹±¹@¿¹´¹±«¬¹¯@´@«@¬¼À¯»¹´°´«®@®«¬¬·¹@¼¬¸@ÁÂ@ª´¬°@´@@«´¯@º@¯±¹·¬´°@¯°¬«´·¹@ ²¹¬¼¹¹´¬¸¹ª´¬°°¯»¹´°´®¹°°Ãµ¹«¬»«ª°±µ¶ª±¯³·¬´¼¸¹´¬¸¹µ¹º±½°¾º±¹«·¬´° °®¼«°«·«¬«®º°¬·½¹±«¿¹«´¯¬¸¹°ª®®½¹±±«¬¹Äª±¯³·¬´±»¬¸¹°³ªª±¬±¬¸¹ÅÆ©·´À ¿³±«¬´Ç¬¬«®ª±¯³·¬´±¬¸¹ÅÆ©·´¿³±«¬´Çª±¯³·¬´±» ¬¸¹°³ªª±¬±¬¸¹ ©ÆÅ·´¿³±«¬´Ç ¬¬«®ª±¯³·¬´±¬¸¹©ÆÅ·´¿³±«¬´Ã ÎÏÐÑÒÓÔ,Õ.Ö##)&:)+% &+×+# )#"##+#,"#)%&:) +%"#)#Ø)%##;1Ù#)+),) )%. &#%#&:)&+×+# ####+#"#)#Ú,Û.)% Ý"#Þ,ß.#)6ÜÖÝ"#Þ,à.)% ÖÜ6Ý"#Þ,Ó.#)ÖÜ6Ý"# %#%# ) ###+ 1 23 4% ;< =%## ##+#"+&#) %# U\P^NSYRRUVP@tuYOP[U\@@vBB@O @wBBqNP]NN\xO\XxyylMU]N` O\X@wBPUqNNuYOb@P^YSo@xzxo@ ^MOXSq[\X[\hN\NVh_\|W}~o@] X[\hSYVTOZNSlsUVNOZ^RO[V@UT xyy]OSPNSPNXls[hYVNWqzN@S^U \X@wBl ! wzvot}nzp\|spotpspvnxtnosn\|ntn tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n ~xozs\|ornxwt~np\|nmurtvwxpyztopn xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|n op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~nton vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nx yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsq oswyponxopo}spswrrtpxqyzopyz{yp\|\| p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}oz p\|opypxozzprtqpnp\|ntqopyzmovs p\|nswrrtpsopysnsp\|ysxzvypyz movsvyw sp\|nxzptywpyzp\|nsry~ntrop\|{yn xqyzopyz {nozp\|nm«¬~xozyzo ornp\|n~xozxwt~nsp\|opp\|nxopo}sp®sw )#"""( ! "" + ! #&")) ! # " )!"' "#!1 )#."+" & ! #"&2()"( #)! 0(34/5$ 365 375 89:;<=?@A)#& ! ,-)# ( "&# (& ! # " #& !(&1 & " "& "&,"(&&"(#+( ( ! ")) 3")( " B5C365,- )# (&+ ! ")) '375 ,-)# ($ DEFEHIJJKLMNOIPHQOQRSKOTUVWWJXOHJYZLMQOLJM [+ ! (" "'. #("+")#(&(#(("( ! #!(#&"( # " \")) #+( (.( !!(!,]^# (( $%! xyz{\|}~ ¡ ¢£ ¤ ¥¦§£ ¨ §£ ©ª #"$%! &'(& ! ")() !!" ! +( #" ! , )# ( " (# # ( ""+ &# ( ! ")) $/&#"'( (& ! )0.) ! # "& ! ! #&")) ! # " )!"' "#! " & ! #"&2()"( #)! 375 )# ( "&# (& ! # " #& !(& (#+( ( ! ")) 3")(* " B5C365, 5 ,-)# ($ WWJXOHJYZLMQOLJM "'. #("+")#(&(#(("( ! ")) #+( (.( !!(!,]^# (( $%! ¡ !" " (,)# (' !(".( "!)#& ! wzvot}nzp\|spotpspvnxtnosn\|ntn tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n ~xozs\|ornxwt~np\|nmurtvwxpyztopn xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|n op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~nton vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nx yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsq oswyponxopo}spswrrtpxqyzopyz{yp\|\| p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}oz p\|opypxozzprtqpnp\|ntqopyzmovs p\|nswrrtpsopysnsp\|ysxzvypyz movsvyw sp\|nxzptywpyzp\|nsry~ntrop\|{yn xqyzopyz {nozp\|nm«¬~xozyzo ornp\|n~xozxwt~nsp\|opp\|nxopo}sp®sw "#$ "&#)$ "+,-&.-& /-&,$.01 ##+9: ; '(# # # XP^NVOPNSUTM X[TTYS[U\@U UQ P^NSYRRUVP@tuYOP[U\@@w NVOPNSUTP^NRVUZNSSNS@vB@O X[TTYS[U\qOVV[NVSZObNS][P^ qNTOSPNVTUVP^N]NOaNVq[\X PNOQUY\P[\hPUylyxoxoO\X@x O[VUTVOPNSUTVNOZP[U\@vB@O\X @dB @dB @ B @@ @@ @ @ ¡¢£¤¥ §¨©ª«¬«®¯°¬±²³¬´ ¯¹±¹´¬·½¹±«¿¹·«¬«®º°¬ª¸ ¬¹»°¼¹±¹¿¹´¹±«¬¹¯´«¬¼À ²¹¬¼¹¹´¬¸¹ª´¬°°¯»¹´° °®¼«°«·«¬«®º°¬·½¹±«¿¹«´¯ ¿³±«¬´Ç¬¬«®ª±¯³·¬´ ©ÆÅ·´¿³±«¬´Ç ¬¬«®ª± ÎÏÐÑÒÓÔ,Õ.Ö##) )#"##+ +%"#)# )% ####+#"# Ý"#Þ,ß.#) @ @ µ¶ª±¯³·¬´´¬¸¹°º°¬¹»¼¬¸°®¼µ¹º±½°¾º±¹«·¬´¼¬ °¹²±¿¸¬¹±·®±°¬«´¯°±»±¹´¬¹´°¹µ¶ª±¯³·¬´¬¸¹°º° »¹´°´«®®«¬¬·¹¼¬¸ÁÂª´¬°´«´¯º¯±¹·¬´°¯°¬«´· ¹°°Ãµ¹«¬»«ª°±µ¶ª±¯³·¬´¼¸¹´¬¸¹µ¹º±½°¾º±¹«·¬´ ¸¹°ª®®½¹±±«¬¹Äª±¯³·¬´±»¬¸¹°³ªª±¬±¬¸¹ÅÆ©·´ ±¬¸¹ÅÆ©·´¿³±«¬´Çª±¯³·¬´±» ¬¸¹°³ªª±¬±¬¸ ³·¬´±¬¸¹©ÆÅ·´¿³±«¬´Ã &:)+% &+× #,"#)%&:) Ø)%##;1Ù#)+) . &#%#&:)&+× )#Ú,Û.)% 6ÜÖÝ"#Þ,à.)% ¡¢£ ¯ ¬¹» ²¹¬ °®¼ ¿³ ©ÆÅ ÏÐÑÒÓ ) +% Ý" ¼¬¸ »±¹´ ´¬° ·¬´¼ ´±» ª±¯³· ¬´Ã + #) ;1 ) ) Þ,à. @ @ @ ¡¢£¤¥ §¨©ª«¬«®@¯°¬±²³¬´@@µ¶@ª±¯³·¬´@´@¬¸¹@°º°¬¹»@¼¬¸@°®¼@µ¹º±½°¾º@±¹«·¬´@¼¬¸@ ¯¹±¹´¬·½¹±«¿¹@@·«¬«®º°¬@ª¸«°¹@²±¿¸¬¹±@·®±@°¬«´¯°@±@»±¹@´¬¹´°¹@µ¶@ª±¯³·¬´@¬¸¹@°º°À ¬¹»°¼¹±¹¿¹´¹±«¬¹¯´«¬¼À¯»¹´°´«®®«¬¬·¹¼¬¸ÁÂª´¬°´«´¯º¯±¹·¬´°¯°¬«´·¹ ²¹¬¼¹¹´¬¸¹ª´¬°°¯»¹´°´®¹°°Ãµ¹«¬»«ª°±µ¶ª±¯³·¬´¼¸¹´¬¸¹µ¹º±½°¾º±¹«·¬´° °®¼«°«·«¬«®º°¬·½¹±«¿¹«´¯¬¸¹°ª®®½¹±±«¬¹Äª±¯³·¬´±»¬¸¹°³ªª±¬±¬¸¹ÅÆ©·´À ¿³±«¬´Ç¬¬«®ª±¯³·¬´±¬¸¹ÅÆ©·´¿³±«¬´Çª±¯³·¬´±» ¬¸¹°³ªª±¬±¬¸¹ ©ÆÅ·´¿³±«¬´Ç ¬¬«®ª±¯³·¬´±¬¸¹©ÆÅ·´¿³±«¬´Ã ÎÏÐÑÒÓÔ,Õ.Ö##)&:)+% &+×+# )#"##+#,"#)%&:). +%"#)#Ø)%##;1Ù#)+),)# )%. &#%#&:)&+×+# ####+#"#)#Ú,Û.)% 6ÜÖ Ý"#Þ,ß.#)6ÜÖÝ"#Þ,à.)% ÖÜ6Ý"#Þ,Ó.#)ÖÜ6Ý"# ( # "'( & ! "#!1 )! # "'( & ! "#!1 )! !" " #"$%! &'(& ! ")() !!" ! +( (,)# (' !(".(#" ! ,)# ( " (# "!)#& ! ,)# ( ""+ &# ( ! ")) $/&#"'( (" (##( ("' ! )"( (& ! )0.) ! # "& ! wzvot}nzp\|spotpspvnxtnosn\|ntn tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n ~xozs\|ornxwt~np\|nmurtvwxpyztopn xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|n op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~nton vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nx yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsq oswyponxopo}spswrrtpxqyzopyz{yp\|\| p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}oz p\|opypxozzprtqpnp\|ntqopyzmovs p\|nswrrtpsopysnsp\|ysxzvypyz movsvyw sp\|nxzptywpyzp\|nsry~ntrop\|{yn xqyzopyz {nozp\|nm«¬~xozyzo ornp\|n~xozxwt~nsp\|opp\|nxopo}sp®sw ,) (" (##( ("' ! )" )#"""( ! "" + )#."+ 0(34/5$ 365 89:;<=?@A)#& ! , & " "& "&,"(&& )# (&+ ! ")) '3 DEFEHIJJKLMNOIPHQOQRSKOTU [+ ! (" #!(#&"( # " \ xyz{\|}~ ¢£ 365 89: & ) DEF #! xyz{ 375 & ! # ")) ")#(&(# !(!, 365 375 89:;<=?@A)#& ! ,-)# ( "&# (& ! # " #& !(&1 & " "& "&,"(&&"(#+( ( ! ")) 3")(* " B5C365,- )# (&+ ! ")) '375 ,-)# ($ DEFEHIJJKLMNOIPHQOQRSKOTUVWWJXOHJYZLMQOLJM [+ ! (" "'. #("+")#(&(#(("( ! #!(#&"( # " \")) #+( (.( !!(!,]^# (( $%! xyz{\|}~ ¡ ¢£ ¤ ¥¦§£ ¨ §£ ©ª 365 375 89:;<=?@A)#& ! ,-)# ( "&# (& ! # " #& !(&1 & " "& "&,"(&&"(#+( ( ! ")) 3")( " B5C365,- )# (&+ ! ")) '375 ,-)# ($ DEFEHIJJKLMNOIPHQOQRSKOTUVWWJXOHJYZLMQOLJM [+ ! (" "'. #("+")#(&(#(("( ! #!(#&"( # " \")) #+( (.( !!(!,]^# (( $%! wzvot}nzp\|spotpspvnxtnosn\|ntn tnyzmurtvwxpyz p\|ys{yvnxtnosnp\|n ~xozs\|ornxwt~np\|nmurtvwxpyztopn xwtsn yztnoyspyxxzvypyzs p\|nrsypyzp\|n op\|nrtxnssnsyzp\|ns}spnqozvp\|nx~nton vnrnzvnzx}{osotnov}vnqzsptopnvtp\|nx yvnt ¡¢£¤¥¦§§¨©ª ¡ ¡ tqp\|npoyznvtnswps {nxozvnty~nsq oswyponxopo}spswrrtpxqyzopyz{yp\|\| p\|nxopo}sp{yp\|p\|nrpyqomovsozyp}oz p\|opypxozzprtqpnp\|ntqopyzmovs p\|nswrrtpsopysnsp\|ysxzvypyz movsvyw sp\|nxzptywpyzp\|nsry~ntrop\|{yn xqyzopyz {nozp\|nm«¬~xozyzo ornp\|n~xozxwt~nsp\|opp\|nxopo}sp®sw xyz{\|}~ ¡ ¢£ ¤ ¥¦§£ ¨ §£ ©ª §£ ¢ª «¬ ª®¯ ª «¬¬ ¬¢ °±®¯ ²³ ¢© © ³« ®´µ¥ §¢£ ¶´µ ¶§·´µ·¶§£ ª ¸ ¹º»º¼½¾¿ÀÁ¿ÂÀ½ÃÄ¾¿ÅÃÆÁÃ¿Ç¾ÈÁÃ¿ ÉÊ¬ ¸ ²³²³¸ ¨²³ « ²³²³ª © ¥ ¤ Ë Ë³ ³ ¨ ²³ ¢© © ® §¢£ ¶´µ ¶§·´ ¸ ¹º»º¼½¾¿ÀÁ¿ÂÀ½ÃÄ¾¿ÅÃÆÁÃ¿Ç¾ÈÁÃ¿ ÉÊ ²³ ¨ « ²³ © ¥ ¤ Ë ¯ ¸¡ ¡ Ìª¸ "#"&'#& ,30+4 ¨-.'""+ ©ª()&"#&"&"#+,«3¤ &¬ 4 '&& ®¯'"&&',30&4 ¤'""&' '£"¥"°¡± +& ++& !"# +&" ®¯''" ,+"'"4&" ¤+" ®¯&'# "²©"'&& ""'&""+ ®¯&'# ¤&! "' "+'3""! &""# §& "'" ®¯&'#"¥" ¤&() &'#¦+³"+#&&"& +#¦ +"#&""+&&¥ !'" && &¦&+&+&"'# ?C=ABA?T=99T<@;??9P6P7=6UVW6A<P=6?@6O=P<L96K7=A=6UVW6P<i=96BL<T=6<P67DF7698AE<T=6 T?C=A<F=;MU XE?@=6T?@9D>=A96F=@=A<L6PA=@>96D@6UVW6T<P<LMPDT6<TPDCDPD=96D@6P7=6E?AO6?E6 () "&&"#&&'#& §& &() ¦ "3! "#"&'#& ,30+4 ¨-.'""+ ©ª()&"#&"&"#+,«3¤ &¬ 4 '&& ®¯'"&&',30&4 ¤'""&' '£"¥"°¡± +& ++& !"# +&" ®¯''" ,+"'"4&" ¤+" ®¯&'# "²©"'&& ""'&""+ ®¯&'# ¤&! "' "+'3""! &""# §& "'" ®¯&'#"¥" ¤&() &'#¦+³"+#&&"& +#¦ +"#&""+&&¥ !'" && &¦&+&+&"'# "&° ´± §"# #""'"&¦¢ &'&+&£³""#"&&°± ¯&! ²³ ¢© © ® §¢£ ¶´µ ¶§·´ ¸ ¹º»º¼½¾¿ÀÁ¿ÂÀ½ÃÄ¾¿ÅÃÆÁÃ¿Ç¾ÈÁÃ¿ ÉÊ ²³ ¨ « ²³ © ¥ ¤ Ë ¯ ¸¡ ¡ Ìª¸ () ¤&&&¥"&'£ "+"& ¦"#"&'# + () "&&"#&&'#& §& &() ¦ "3! "#"&'#& ,30+4 ¨- '""+ ©ª()&"#&"&"#+,«3¤ &¬ 4 "# $ $ & /. 012 @<>9?ABPD?@ fcXhd=eK= :K7DT7<A=@ P7=9BDLL?C=ABA?T=9 E?AP7=T?@9D>=A=>A {\| s=TPD?@9\X\<@>\ ""'& & &"#"# "& "# jklmnoqrsT7=O<PDTA=BA=9=@P<PD?@?EP7=9BDLL?C=ABA?T=99<@>U\BA?>8TPD?@?@P7=ARSt98BB?AP O=P<LT<P<LM9PKDP7P7==@=AFMBA?EDL=E?AP7=T?@9D>=A=>A=<TPD?@B<P7X uvwvyz{\|}~\|}z{\|\|{\| <9=>?@P7=T?@9D>=A<PD?@9D@s=TPD?@9\X\<@>\XH:DPD9LDi=LMP7<PP7=7M>A?F=@9BDLLt ?C=ABA?T=99T<@;??9PP7=UVWA<P=?@O=P<L9K7=A=UVWP<i=9BL<T=<P7DF798AE<T= T?C=A<F=;MU XE?@=T?@9D>=A9F=@=A<LPA=@>9D@UVWT<P<LMPDT<TPDCDPD=9D@P7=E?AO?E &&""'& ¡&()¢ "&"##£"&& ) ""&"#"#'-. () ¤&&&¥"&'£ "+"& ¦"#"&'# + () "&&"#&&'#& §& &() ¦ "3! "#"&'#& ,30+4 ¨- '""+ ©ª()&"#&"&"#+,«3¤ &¬ 4 6 jklmnoqrsT7=O<PDTA=BA=9=@P<PD?@6?E6P7=69BDLL?C=A6BA?T=996<@>6U\6BA?>8TPD?@6?@6P7=6ARSt98BB?AP6 O=P<LT<P<LM9PKDP7P7=6=@=AFM6BA?EDL=6E?A6P7=6T?@9D>=A=>6A=<TPD?@6B<P7X6 uvwvyz{\|}~\|}z{\|6\|6{\|6 <9=>?@P7=T?@9D>=A<PD?@96D@6s=TPD?@96\X\6<@>6\XH:6DP6D96LDi=LM6P7<P6P7=67M>A?F=@69BDLLt ?C=ABA?T=99T<@;??9PP7=UVWA<P=?@O=P<L9K7=A=UVWP<i=9BL<T=<P7DF798AE<T= T?C=A<F=;MU XE?@=T?@9D>=A9F=@=A<LPA=@>9D@UVWT<P<LMPDT<TPDCDPD=9D@P7=E?AO?E &&""'& ¡&()¢ "&"##£"&& ) ""&"#"#'-. () ¤&&&¥"&'£ "+"& ¦"#"&'# + () "&&"#&&'#& §& &() ¦ "3! ²³ ¢© © ® §¢£ ¶´µ ¶§·´ ¸ ¹º»º¼½¾¿ÀÁ¿ÂÀ½ÃÄ¾¿ÅÃÆÁÃ¿Ç¾ÈÁÃ¿ ÉÊ ²³ ¨ « ²³ © ¥ ¤ Ë ¯ ¸¡ ¡ Ìª¸ © ³« ®´µ¥ §¢£ ¶´µ ¶§·´µ·¶§£ ª ¸ ¹º»º¼½¾¿ÀÁ¿ÂÀ½ÃÄ¾¿ÅÃÆÁÃ¿Ç¾ÈÁÃ¿ ÉÊ¬ ¸ ²³²³¸ ¨²³ « ²³²³ª © ¥ ¤ Ë Ë³ ³ ¨ ¯ ¸¡ ¡ Ìª¸ ©© ¤Í®¯«¬¬ ¬¢ Ê¢®¯Î ¡ ¡ ¥ ¤ Ë Ë Ì²³²³ §¡ Ï§«¬£ª£ " + ?@P7= 8B7DLL D@W=EX <P<LM9P9 ?C=A<LL =<PP7D9 " + ?@P7= 8B7DLL D@W=EX <P<LM9P9 ?C=A<LL =<PP7D9 !"# $ $ %" "&'#() &&+ ,-.++/. 012" "&$ 304 >?=9@?PE<C?AU E?AO<PD?@XU?K=C=A:DPD9B?99D;L=P?7<C=9PA?@F=A;?@>D@F?@P7= 98BB?AP9?P7<PP7=U E?AO<PD?@9P<F=9<L?@FP7=A=<TPD?@B<P7K<M<A=T?OBL=P=LM8B7DLL EA?O P7=P7=AO?>M@<ODTB?D@P?ECD=KXZ?A=[<OBL=:?P7=ATADP=AD<K=A=<BBLD=>D@W=EX G\\J:E?T89D@F?@P7=RD;;9=@=AFME?AP7=7M>A?F=@<>9?ABPD?@BA?T=99]^_U`Xa<P<LM9P9 KDP7bcXId=ef^_U fc<@>98BB?AP9KDP7g^_UgfcXhd=eK=A=T?@9D>=A=>X67=?C=A<LL A<P=>=B=@>9?@P7=iD@=PDT;<AAD=A9?E=<T7BA?T=99:K7DT7<A=@?P=<9MP?=9PDO<P=<PP7D9 B?D@PX !"# $ $ %" "&'#() &&+ ,-.++/. 012" "&$ 304 >?=9@?PE<C?AU E?AO<PD?@XU?K=C=A:DPD9B?99D;L=P?7<C=9PA?@F=A;?@>D@F?@P7= 98BB?AP9?P7<PP7=U E?AO<PD?@9P<F=9<L?@FP7=A=<TPD?@B<P7K<M<A=T?OBL=P=LM8B7DLL EA?O P7=P7=AO?>M@<ODTB?D@P?ECD=KXZ?A=[<OBL=:?P7=ATADP=AD<K=A=<BBLD=>D@W=EX G\\J:E?T89D@F?@P7=RD;;9=@=AFME?AP7=7M>A?F=@<>9?ABPD?@BA?T=99]^_U`Xa<P<LM9P9 KDP7bcXId=ef^_U fc<@>98BB?AP9KDP7g^_UgfcXhd=eK=A=T?@9D>=A=>X67=?C=A<LL A<P=>=B=@>9?@P7=iD@=PDT;<AAD=A9?E=<T7BA?T=99:K7DT7<A=@?P=<9MP?=9PDO<P=<PP7D9 B?D@PX !"# $ $ %" "&'#() &&+ ,-.++/. 012" "&$ 304 >?=9@?PE<C?AU E?AO<PD?@X6U?K=C=A:6DP6D96B?99D;L=6P?67<C=69PA?@F=A6;?@>D@F6?@6P7=6 98BB?AP9?P7<PP7=U6E?AO<PD?@69P<F=96<L?@F6P7=6A=<TPD?@6B<P7K<M6<A=6T?OBL=P=LM68B7DLL6 EA?O P7=P7=AO?>M@<ODT6B?D@P6?E6CD=KX6Z?A6=[<OBL=:6?P7=A6TADP=AD<6K=A=6<BBLD=>6D@6W=EX6 G\\J:E?T89D@F?@P7=RD;;96=@=AFM6E?A6P7=67M>A?F=@6<>9?ABPD?@6BA?T=996]^_U`X6a<P<LM9P96 KDP7bcXId=ef^_U f6c6<@>698BB?AP96KDP76g^_Ug6f6cXhd6=e6K=A=6T?@9D>=A=>X667=6?C=A<LL6 A<P=>=B=@>9?@P7=iD@=PDT6;<AAD=A96?E6=<T76BA?T=99:6K7DT76<A=6@?P6=<9M6P?6=9PDO<P=6<P6P7D96 B?D@PX jklmnoqrsT7=O<PDTA=BA=9=@P<PD?@?EP7=9BDLL?C=ABA?T=99<@>U\BA?>8TPD?@?@P7=ARSt98BB?AP O=P<LT<P<LM9PKDP7P7==@=AFMBA?EDL=E?AP7=T?@9D>=A=>A=<TPD?@B<P7X uvwvyz{\|}~\|}z{\|\|{\| <9=>?@P7=T?@9D>=A<PD?@9D@s=TPD?@9\X\<@>\XH:DPD9LDi=LMP7<PP7=7M>A?F=@9BDLLt ?C=ABA?T=99T<@;??9PP7=UVWA<P=?@O=P<L9K7=A=UVWP<i=9BL<T=<P7DF798AE<T= T?C=A<F=;MU XE?@=T?@9D>=A9F=@=A<LPA=@>9D@UVWT<P<LMPDT<TPDCDPD=9D@P7=E?AO?E &&""'& ¡&()¢ "&"##£"&& ) ""&"#"#'-. ²³ ¢© © ® §¢£ ¶´µ ¶§·´ ¸ ¹º»º¼½¾¿ÀÁ¿ÂÀ½ÃÄ¾¿ÅÃÆÁÃ¿Ç¾ÈÁÃ¿ ÉÊ ²³ ¨ « ²³ © ¥ ¤ Ë ¯ ¸¡ ¡ Ìª¸ ¢#°¡±&""+#""' &&&¥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ghijklmnopqrstuvwxyzw{y\|}~xxv}{{wvy{oxqw\|y{~{x~uux{ oq y{uy{y\|{yy\|}\|ox\|w\|\|x}w{wv{\|vuwx{{uvwxyzw{y\| }~xw{ux\|w{yu{\|{ywvwx\|{ux\|{\|yv~{y\|qoqwvuv{x\|{yx xy}w{wv{o{w\|{w{y\|wxopqqoqv}w\|uv{}\|{x~}{~y\|wv~w\| }wv}~vw{x\|y\|y\|\|xy }~y\|\|{{x\|y\|y\|xw\|} yvy\|{y{ }uv{rstv}w\|}~xsxxxwxy{xw\|wv~xyx\|{uy{y\|{y nhjh ux\|{x~v{}vwxvy\|y}w{{w{y\|{xw}y\|{wv{w{{x\|vy\| x\|y{x\|w\|}{rstxw{wyx{~y{wvy}\|y{\|{x yx\|{uy{y\|{yoy~xqxx y\|}w {{x\|vv{\| uxy}{xwxy\|w}{yy{y\|}xw{w}x{wy\|uy\|{ vv{}xw ~u\|~x{x~uux{}xwy{{}w{wv{~}w{x\|yvvxux\|{y\|{ }wyw\|wvyyv{{\|{xoy~xqxwvv {y\|{xw}y\|ux} }\|{xy~{{rst ~{y{~{\|{{w{{\|xwvwu{v}w\|}~xy uxxw\|y\|w\|{v{xy\|r \|x{y}w{{}w{wv{ ~xw} !"#$%%##%&'%( )+ ./0 .20 I+@IH>HECHJKHC?>ILACM.NO@P"Q0?DJLACMFC@P"IF==@ DJI.+@RD?DJRDO@P"E?C?>SICD>S=?@CI+CMH=>?@AB?CA* ?@HIMLDC=@HGHDCRDJAII>FCAD0U.10V?+H>=>CI+@HDCA@ D?I+@./00U.20G>E?D=>CEDIC@FECHJFIADTW,XG?>FHI?D HI+EFIADTDCMHIC@DTYADJADTY@?DEMLMA>HADIHCTAGHICM Y?@ITAGHCMH@?DTH+W G?>FHI+@JA++H@HDCJH=IACADCAKHI CMH@E?C?>SICIJAI=>?S=IACAGHIMA+CIADW LMHDJH=IACHJ xv}{{wvy{oxqw\|y{~{x~uux \|ox\|w\|\|x}w{wv{\|vuwx{{uvwxyzw{y\| wx\|{ux\|{\|yv~{y\|qoqwvuv{x\|{yx \|wxopqqoqv}w\|uv{}\|{x~}{~y\|wv~w\| }~y\|\|{{x\|y\|y\|xw\|} yvy\|{y{ wxy{xw\| wv~xyx\|{uy{y\|{y ./0 ./0 .10 345678:;./0<%#=>?@AB?CADEF@GH .NQ0AICMHJH=IACADCAKHADIHE* EF@GHI?CJHH=H@DHT?CAGH=CHDCA?> IH@AHI+E?C?>SICI.CMHI?KHDC?CA E?>EF>?CHJMSJ@THDYADJADTHDH@TA EK=>HCH<%#G>E?DEF@GHZ%@@@ "DCMHCMH@M?DJ?>>CMH* ghijklmnopqrstuvwxyzw{y\|}~x oq y{uy{y\|{yy\|} }~xw{ux\|w{yu{\|{ywv xy}w{wv{o{w\|{w{y }wv}~vw{x\|y\|y\|\|xy }uv{rstv}w\|}~xsxxx .10 34567 .NQ0 EF@GH IH@AHI E?>EF EK= ghijk oq }~x xy }wv}~v }u .20 O@P"Q C?>SIC II>FCA CEDIC DJADTY >FHI+@ CAGHIM x wv{\| v~{y }\|{x y\| v~x .10 .20 345678:;./0<%#=>?@AB?CADEF@GHI+@IH>HECHJKHC?>ILACM.NO@P"Q0?DJLACMFC@P"IF==@C .NQ0AICMHJH=IACADCAKHADIHEDJI.+@RD?DJRDO@P"E?C?>SICD>S=?@CI+CMH=>?@AB?CAD EF@GHI?CJHH=H@DHT?CAGH=CHDCA?>I?@HIMLDC=@HGHDCRDJAII>FCAD0U.10V?+H>=>CI+@HDCA@H IH@AHI+E?C?>SICI.CMHI?KHDC?CAD?I+@./00U.20G>E?D=>CEDIC@FECHJFIADTW,XG?>FHI?DJ E?>EF>?CHJMSJ@THDYADJADTHDH@TAHI+EFIADTDCMHIC@DTYADJADTY@?DEMLMA>HADIHCTAGHICMH EK=>HCH<%#G>E?DEF@GHZ%@@@Y?@ITAGHCMH@?DTH+W G?>FHI+@JA++H@HDCJH=IACADCAKHIZ "DCMHCMH@M?DJ?>>CMHCMH@E?C?>SICIJAI=>?S=IACAGHIMA+CIADW LMHDJH=IACHJ ghijklmnopqrstuvwxyzw{y\|}~xxv}{{wvy{oxqw\|y{~{x~uux{ oq y{uy{y\|{yy\|}\|ox\|w\|\|x}w{wv{\|vuwx{{uvwxyzw{y\| }~xw{ux\|w{yu{\|{ywvwx\|{ux\|{\|yv~{y\|qoqwvuv{x\|{yx xy}w{wv{o{w\|{w{y\|wxopqqoqv}w\|uv{}\|{x~}{~y\|wv~w\| }wv}~vw{x\|y\|y\|\|xy }~y\|\|{{x\|y\|y\|xw\|} yvy\|{y{ }uv{rstv}w\|}~xsxxxwxy{xw\|wv~xyx\|{uy{y\|{y nhjh ux\|{x~v{}vwxvy\|y}w{{w{y\|{xw}y\|{wv{w{{x\|vy\| x\|y{x\|w\|}{rstxw{wyx{~y{wvy}\|y{\|{x yx\|{uy{y\|{yoy~xqxx y\|}w {{x\|vv{\| uxy}{xwxy\|w}{yy{y\|}xw{w}x{wy\|uy\|{ vv{}xw ~u\|~x{x~uux{}xwy{{}w{wv{~}w{x\|yvvxux\|{y\|{ }wyw\|wvyyv{{\|{xoy~xqxwvv {y\|{xw}y\|ux} .10 345678:;./0<% .NQ0AICMHJH EF@GHI?CJHH=H IH@AHI+E?C?>SI E?>EF>?CHJMSJ@ EK=>HCH<%#G "DCMHC ghijklmnopqrst oq y{u }~xw{ux xy}w{wv{ }wv}~vw{x }uv{rst nhjh ux x\|y{x yx\|{u uxy}{ ~u\|~x{x }wyw\| }\|{xy~{{ uxxw\| .NQ0AICMHJH=IACADCAKHADIHEDJI.+@RD?DJRDO@P"E?C?>SICD>S=?@CI+CMH=>?@AB?CAD EF@GHI?CJHH=H@DHT?CAGH=CHDCA?>I?@HIMLDC=@HGHDCRDJAII>FCAD0U.10V?+H>=>CI+@HDCA@H IH@AHI+E?C?>SICI.CMHI?KHDC?CAD?I+@./00U.20G>E?D=>CEDIC@FECHJFIADTWG?>FHI?DJ E?>EF>?CHJMSJ@THDYADJADTHDH@TAHI+EFIADTDCMHIC@DTYADJADTY@?DEMLMA>HADIHCTAGHICMH EK=>HCH<%#G>E?DEF@GHZ%@@@Y?@ITAGHCMH@?DTH+WG?>FHI+@JA++H@HDCJH=IACADCAKHIZ "DCMHCMH@M?DJ?>>CMHCMH@E?C?>SICIJAI=>?S=IACAGHIMA+CIADWLMHDJH=IACHJ oq y{uy{y\|{yy\|}\|ox\|w\|\|x}w{wv{\|vuwx{{uvwxyzw{y\| }~xw{ux\|w{yu{\|{ywvwx\|{ux\|{\|yv~{y\|qoqwvuv{x\|{yx xy}w{wv{o{w\|{w{y\|wxopqqoqv}w\|uv{}\|{x~}{~y\|wv~w\| }wv}~vw{x\|y\|y\|\|xy }~y\|\|{{x\|y\|y\|xw\|} yvy\|{y{ }uv{rstv}w\|}~xsxxxwxy{xw\|wv~xyx\|{uy{y\|{y nhjh ux\|{x~v{}vwxvy\|y}w{{w{y\|{xw}y\|{wv{w{{x\|vy\| x\|y{x\|w\|}{rstxw{wyx{~y{wvy}\|y{\|{x yx\|{uy{y\|{yoy~xqxx y\|}w {{x\|vv{\| uxy}{xwxy\|w}{yy{y\|}xw{w}x{wy\|uy\|{ vv{}xw ~u\|~x{x~uux{}xwy{{}w{wv{~}w{x\|yvvxux\|{y\|{ }wyw\|wvyyv{{\|{xoy~xqxwvv {y\|{xw}y\|ux} }\|{xy~{{rst ~{y{~{\|{{w{{\|xwvwu{v}w\|}~xy uxxw\|y\|w\|{v{xy\|rw\|x{y}w{{}w{wv{ ~xw} * ! * )!** !.2* % * ! * )!** !.2* % !"#$ %&' (()+,-./0%123+! 3(44* )! '5617 ! 25! * ! )3(44'5! * .)(..))!2 ! 45%8.! 44) ! (9 42'4+:(-,-./0%2 443( !)!** !'. "#$3'. +.' 4* )(24! !.2* )2 + 45+'5" !* ! 45;.( !% <=>?@ACD&"#$2) 4((!1E! 45** * (.! (4! )F +% '2.453!'. (( !) '2:)+ 445 * )+ 617+.4 3'. .4'!45 .4% 123 !'. (+( ! 4((! .4'45'F (! 4! ((45)) !4,-.G3((! ()03'. * TUVWXYZ[\]^_`ab^cde^feghij]gejkfjgl^c^lmncopkhehiqjejhjhrfkegf^i^kecl^djgs egfegt^u pvbgrjiq e]^kfecgvregfe]^gfe^chk^dhe]vj^cb^l^wd^cgt^fehiiqgjfehj dctgf^fehjgfe]^xmyjgtrihegfj vree]^j^j]rilv^kfjgl^c^lfiqhjzrgl^igf^ju mc^b^c {^fe^e]hee]^kfecgvregfcte]^gfe^chkghi^^kej{rilv^]gz]iqv^f^\|s kghigfe]^khj^\|f^iqlgjd^cj^ldhcegki^j}~gzrc^ ^^kejlgjd^cjgf vree]^j^hc^ lg\|krieevehgfrjgfz^i^kecl^djgegfu~cjt^e]^jrddcekhehiqjejktvgfhegfj dc^j^fe^l]^c^ e]^c^hc^jecfzgflgkhegfje]he]qlcz^fjdgiib^cgjd^chegb^ hjt^fs egf^lv^c^u_{^b^c {^trjefe^e]hee]^c^hc^hijddjgfzbg^{j hei^hje <=>?@ACD&"#$2) 4((!1E! 45* * (.! (4! )F +% '2.453!'. (( !) '2:)+ 445 * )+ 617+.4 3'. .4'!45 .4% 123 !'. (+( ! 4((! .4'45'F (! 4! ((45)) !4,-.G3((! ()03'. * TUVWXYZ[\]^_`ab^cde^feghij]gejkfjgl^c^lmncopkhehiqjejhjhrfkegf^i^kecl^djgs egfegt^u pvbgrjiq e]^kfecgvregfe]^gfe^chk^dhe]vj^cb^l^wd^cgt^fehiiqgjfehj dctgf^fehjgfe]^xmyjgtrihegfj vree]^j^j]rilv^kfjgl^c^lfiqhjzrgl^igf^ju mc^b^c {^fe^e]hee]^kfecgvregfcte]^gfe^chkghi^^kej{rilv^]gz]iqv^f^\|s kghigfe]^khj^\|f^iqlgjd^cj^ldhcegki^j}~gzrc^ ^^kejlgjd^cjgf vree]^j^hc^ lg\|krieevehgfrjgfz^i^kecl^djgegfu~cjt^e]^jrddcekhehiqjejktvgfhegfj dc^j^fe^l]^c^ e]^c^hc^jecfzgflgkhegfje]he]qlcz^fjdgiib^cgjd^chegb^ hjt^fs egf^lv^c^u_{^b^c {^trjefe^e]hee]^c^hc^hijddjgfzbg^{j hei^hje <=>?@ACD&"#$2) 4((!1E! 45* * (.! (4! )F +% '2.453!'. (( !) '2:)+ 445 * )+ 617+.4 3'. .4'!45 .4% 123 !'. (+( ! 4((! .4'45'F (! 4! ((45)) !4,-.G3((! ()03'. .10 345678:;./0<% .NQ0AICMHJH EF@GHI?CJHH=H IH@AHI+E?C?>SI E?>EF>?CHJMSJ@ EK=>HCH<%#G "DCMHC ghijklmnopqrst oq y{u }~xw{ux xy}w{wv{ }wv}~vw{x }uv{rst nhjh ux x\|y{x yx\|{u uxy}{ ~u\|~x{x }wyw\| }\|{xy~{{ uxxw\| TUVWXYZ[\]^_`ab^cde^feghij]gejkfjgl^c^lmncopkhehiqjejhjhrfkegf^i^kecl^djgs egfegt^u pvbgrjiq e]^kfecgvregfe]^gfe^chk^dhe]vj^cb^l^wd^cgt^fehiiqgjfehj dctgf^fehjgfe]^xmyjgtrihegfj vree]^j^j]rilv^kfjgl^c^lfiqhjzrgl^igf^ju mc^b^c {^fe^e]hee]^kfecgvregfcte]^gfe^chkghi^^kej{rilv^]gz]iqv^f^\|s kghigfe]^khj^\|f^iqlgjd^cj^ldhcegki^j}~gzrc^ ^^kejlgjd^cjgf vree]^j^hc^ lg\|krieevehgfrjgfz^i^kecl^djgegfu~cjt^e]^jrddcekhehiqjejktvgfhegfj dc^j^fe^l]^c^ e]^c^hc^jecfzgflgkhegfje]he]qlcz^fjdgiib^cgjd^chegb^ hjt^fs egf^lv^c^u_{^b^c {^trjefe^e]hee]^c^hc^hijddjgfzbg^{j hei^hje ce]^lzchd]^f^jqje^tj u\]^j^e]^c^egkhikhikrihegfjjrzz^je^le]^hvj^fk^ _ jdgiib^cct jthiilkirje^cjl^djge^lfe]^zchd]^f^jrchk^u_{^b^c ge tgz]ehijv^e]hee]^zchd]^f^jrchk^jgfe]^j^jerlg^j{^c^ed^c^keuht^iq dcgjegf^ zchd]^f^vgflj_hljb^cq{^hiq {]gi^e]^vflgfzv^kt^jtrk]jecfz^cfk^l^^kej hflrfkegfhizcrdjhc^gfeclrk^lfe]^jrchk^u\]rj d^c^kezchd]^f^l^jfe jhegjqe]^kcge^cghce]^jrddcezgb^fgf^kegfuufhke {^]hb^dc^bgrjiqj]{f e]hegfe]^khj^gl^djge^lflg^c^fezchd]^f^svhj^ljrchk^j djgegb^^^kejhc^ j^^fccop {]gk]kfehgfjhihcz^frtv^cl^^kejhflhk^cehgfchkegfwqz^f rfkegfhizcrdjgfe]^vhjhidihf^u_{^b^c e]^^^ke{hjhvj^fe{]^fg{hj l^djge^lf{^iiscl^c^lzchd]^f^fhfdihe^i^ej{ge]{^iisl^\|f^lzchd]gegkjecrkerc^ dc^j^cb^ljdvhjhidihf^ hflwqz^frfkegfhizcrdjhee]^^lz^jge^ju\]rj {]gi^{^ ]hb^l^tfjeche^le]^^^kechj^cg^jt^ehijfcop {^v^ig^b^e]hee]^cjrddcej khfhijv^regig^levje_`abghjdgiib^c jrk]hji{skjehiiqj lg^c^fefhfjecrks erc^lkhcvfj t^ehiwgl^j hfle]^cju_{^b^c gegjkcrkghie]hek]j^fjrddcejri\|ii f^k^jjhcqe]^ctlqfhtgkhflgf^egkkflgegfj hjlgjkrjj^lgf^kegfuu \]^ihke^k]fgr^jcbgjrhiggfzd]^ft^fhgjhz^f^chidcvi^tjdgiib^cs krj^ljerlg^j hjfe^lgfu_{^b^c e]^]hctfqv^e{^^fe]^e]^c^egkhihjjrtds egfjhfle]^^wd^cgt^fehic^jriejjehfljgfhbce]^dcdj^lt^k]hfgjtuht^iq gf e]^jdgiib^cdck^jj{hjfelgc^keiqbgjrhig^l vree]^kfcthegf{hjgflgc^ke e]crz]e]^bhir^e]^\h^ijid^u_{^b^c gegjgtdcehfeevj^cb^e]hee]^gl^feg\|^l khehiqjejrddcektvgfhegf e csmj]{^ltrk]i{^c_`afj^ede^feghikts dhc^leegfhfhkglgkt^lgrtvre]gz]^c_`ahkegbgeqheihcz^_`ab^cbiehz^ju\]gj c^jried^c^keiqhzc^^j{ge]rc{c hjgekcc^jdflje]gz]jrchk^kb^chz^vq_ j^^fccop {]gk]kfehgfjhihcz^frtv^cl^^kejhflhk^cehgfchkegfwqz^f rfkegfhizcrdjgfe]^vhjhidihf^u_{^b^c e]^^^ke{hjhvj^fe{]^fg{hj l^djge^lf{^iiscl^c^lzchd]^f^fhfdihe^i^ej{ge]{^iisl^\|f^lzchd]gegkjecrkerc^ dc^j^cb^ljdvhjhidihf^ hflwqz^frfkegfhizcrdjhee]^^lz^jge^ju\]rj {]gi^{^ ]hb^l^tfjeche^le]^^^kechj^cg^jt^ehijfcop {^v^ig^b^e]hee]^cjrddcej khfhijv^regig^levje_`abghjdgiib^c jrk]hji{skjehiiqj lg^c^fefhfjecrks erc^lkhcvfj t^ehiwgl^j hfle]^cju_{^b^c gegjkcrkghie]hek]j^fjrddcejri\|ii f^k^jjhcqe]^ctlqfhtgkhflgf^egkkflgegfj hjlgjkrjj^lgf^kegfuu \]^ihke^k]fgr^jcbgjrhiggfzd]^ft^fhgjhz^f^chidcvi^tjdgiib^cs krj^ljerlg^j hjfe^lgfu_{^b^c e]^]hctfqv^e{^^fe]^e]^c^egkhihjjrtds egfjhfle]^^wd^cgt^fehic^jriejjehfljgfhbce]^dcdj^lt^k]hfgjtuht^iq gf e]^jdgiib^cdck^jj{hjfelgc^keiqbgjrhig^l vree]^kfcthegf{hjgflgc^ke e]crz]e]^bhir^e]^\h^ijid^u_{^b^c gegjgtdcehfeevj^cb^e]hee]^gl^feg\|^l khehiqjejrddcektvgfhegf e csmj]{^ltrk]i{^c_`afj^ede^feghikts dhc^leegfhfhkglgkt^lgrtvre]gz]^c_`ahkegbgeqheihcz^_`ab^cbiehz^ju\]gj c^jried^c^keiqhzc^^j{ge]rc{c hjgekcc^jdflje]gz]jrchk^kb^chz^vq_hlj {]^c^jdgiib^cgj^wd^ke^lev^d^chegb^uehijjrddcejrcjehflgfzdgfee]hee]^ kfkirjgfjl^cgb^lctxmyhc^bhiglcve]hkglgkhflhihigf^jiregfju hc e]^v^f^\|eje]^gcopktvgfhegf]hb^v^^fj]{fce]^copstlg\|^l ghthj{^iihjcgl^djge^lfcop u\c^gfck^e]^^^kejgfe^chkgfz ghflcop {^dc^dhc^lhgncopktdjge^rjgfzlgc^ke^i^keck]^tgkhil^djgegf cte]^opskfehgfgfz^i^kecl^djgegfvhe]hflktdhc^lgejhkegbgeqee]^gl^djge vehgf^lrfl^cgl^fegkhikflgegfjgfe]^hvj^fk^op}~gzrc^huyi^hciq e]^hkegbgeq egfjhfle]^^wd^cgt^fehic^jriejjehfljgfhbce]^dcdj^lt^k]hfgjtuht^iq gf e]^jdgiib^cdck^jj{hjfelgc^keiqbgjrhig^l vree]^kfcthegf{hjgflgc^ke e]crz]e]^bhir^e]^\h^ijid^u_{^b^c gegjgtdcehfeevj^cb^e]hee]^gl^feg\|^l khehiqjejrddcektvgfhegf e csmj]{^ltrk]i{^c_`afj^ede^feghikts dhc^leegfhfhkglgkt^lgrtvre]gz]^c_`ahkegbgeqheihcz^_`ab^cbiehz^ju\]gj c^jried^c^keiqhzc^^j{ge]rc{c hjgekcc^jdflje]gz]jrchk^kb^chz^vq_hlj {]^c^jdgiib^cgj^wd^ke^lev^d^chegb^uehijjrddcejrcjehflgfzdgfee]hee]^ kfkirjgfjl^cgb^lctxmyhc^bhiglcve]hkglgkhflhihigf^jiregfju hc e]^v^f^\|eje]^gcopktvgfhegf]hb^v^^fj]{fce]^copstlg\|^l ghthj{^iihjcgl^djge^lfcop u\c^gfck^e]^^^kejgfe^chkgfz ghflcop {^dc^dhc^lhgncopktdjge^rjgfzlgc^ke^i^keck]^tgkhil^djgegf cte]^opskfehgfgfz^i^kecl^djgegfvhe]hflktdhc^lgejhkegbgeqee]^gl^djge vehgf^lrfl^cgl^fegkhikflgegfjgfe]^hvj^fk^op}~gzrc^huyi^hciq e]^hkegbgeq "' " "!. " %8$ LCEBC "' " "!. " %8$ LCEBC !"#" $"#%&"' $()'"$#$#$!"$+! " $",#-%&$#. "!.$%/$# 0+! 123 $"#2#$ 445#6,6"!. 0+! 13%/$4 #$"!. 4" $ #%7$'."#$"$$%8$ 9"## 2."$$",#- 2. ,"$##$ #4#. #"$# ,$#$$! "%/$#$ $$$##"" ##$#$ #! 4#.'2$!$-#$"#:#$ $!"$"$"$;<<=% [\X]DAHFYAAJKF?@JACFK@BCDFJILNEGF@N[\JDFFCJVcDFLFAJ@BFO@LFMFLfAO?@LCEBC ?@ABCC@B@CFVqDFBQeL[\ G@O?@JACFJELFOEKFXFACDFLSfFHFGCL@KF?@JACA@B@BCDF KL@?UGEJCFK[\ NAHO @LSfG@UKF?@JACA@B@NOFCEHEBKL[\XCDFLFAJEH]EfJHFJJOFCEH KF?@JACFKABCDFGEJF@NL[\UG@BCEABABdKF?@JACJVpNCDFG@BJCEBCGILLFBCG@BKACA@BJELF E??HAFKXEGFLCEABNLEGCA@B@NGDELdFAJIJFKC@LFKIGF[\X]DAGDAJGHFELHfMAJASHFABCDF JAdBANAGEBCKF?@HELArECA@B@NCDFFHFGCL@KFKILABdCDFFHFGCL@KF?@JACA@B?L@GFJJABCDF ?LFJFBGF@N[\^``aV $4 #$ %7$ "## 2. $##$ #4 $! "%/$#$ $ 4#.' $!"$" AAJKF?@JACFK@BCDFJILNEGF?@N?[\?JDF CFVqDFBQeL[\ G@O?@JACFJ?ELF?OEK [\ NAHO @LSfG@UKF?@JACA@B?@N?OFCEH CDFGEJF@NL[\UG@BCEABABd?KF?@JACJV? FLCEABNLEGCA@B@NGDELdFAJ?IJFK?C@?LFKI F?@HELArECA@B@NCDFFHFGCL@KF?KILABd [\^``aV gtj ()"#:#$ ' #$ 4. .10 345678:;./0<% .NQ0AICMHJH EF@GHI?CJHH=H IH@AHI+E?C?>SI E?>EF>?CHJMSJ@ EK=>HCH<%#G "DCMHC ghijklmnopqrst oq y{u }~xw{ux xy}w{wv{ }wv}~vw{x }uv{rst nhjh ux x\|y{x yx\|{u uxy}{ ~u\|~x{x }wyw\| }\|{xy~{{ uxxw\| "$# ""%$"$#$ gsj v 9 gsj gtj ' # " " guj 03 $ $ #. gtj guj vwxyz{\|}0~3()"#:#$ ' #$ . ""#" " $4. '#!#403)9 #!" "03)9 #!"% +#$ "'$$()#'. $!$! 9$# 4 2 "$$ $4$#: ##'#%&$"$##.#' ### 0)&3%& ! #" !2#"$!% #2$#$)&# #$#$#.#6 "#.4$ "$%/#"##$# $#$.4#. $"2 "####. "$# ""%$"$#$ ? gtj? guj? vwxyz{\|}0~3()"#:#$ ' #$ . ""#" " $4. '#!#403)9 #!" "03)9 #!"% +#$ "'$$()#'. $!$! 9$# 4 2 "$$ $4$#: ##'#%&$"$##.#' ### 0)&3%& ! #" !2#"$!% #2$#$)&# #$#$#.#6 "#.4$ "$%/#"##$# $#$.4#. $"2 "####. "$# ""%$"$#$ $$+! 12 $# #"##$22'%(4' . '#.0+! 1# $34##"$4$ !$#$ 4#" 2$! 4##!%7 #.$##)&4#'. ##$# 4# $$"$#2.!"$%/$ $ () #'"$$+! .10 345678:;./0<% .NQ0AICMHJH EF@GHI?CJHH=H IH@AHI+E?C?>SI E?>EF>?CHJMSJ@ EK=>HCH<%#G "DCMHC ghijklmnopqrst oq y{u }~xw{ux xy}w{wv{ }wv}~vw{x }uv{rst nhjh ux x\|y{x yx\|{u uxy}{ ~u\|~x{x }wyw\| }\|{xy~{{ uxxw\| '#!# )9 #!"% #$ "'$$( # 4 2 " '#%&$"$# ! #" "!. 4" '."#$"$$%8 "$$",#- 4#. #"$# $$##"" ##$# '2$!$-#$"#:# $"$;<<=% DFFCJVcDFLFAJ@BFO@LFMFLf?AO?@LCEB EKFXFACDFLSfFHFGCL@KF?@JACA@B?@B?CDF CEHEBKL[\XCDFLFAJEH]EfJ?HFJJ?OFCE JVpNCDFG@BJCEBCGILLFBC?G@BKACA@BJ?ELF KIGF[\X]DAGDAJGHFELHf?MAJASHF?AB?CDF dCDFFHFGCL@KF?@JACA@B?L@GFJJ?AB?CDF guj . ""#" 403)9 #!" ()#'. $!$! $$ $4$#: #.#' # !2#"$! 0+! 13%/$ $ # 9" 2. ," ,$#$$ #! [\X?]DAHF?YA ?@ABC?C@?B@CF KL@?UGEJCFK?[ KF?@JACFK?AB? E??HAFKX?E?GFL JAdBANAGEBC?KF ?LFJFBGF?@N?[ gsj vwxyz{\|}0~3( " $ "03)9 + 9$# 0+! 13%/$4 #$"!. 4" $ #%7$'."#$"$$%8$ 9"## 2."$$",#- 2. ,"$##$ #4#. #"$# ,$#$$! "%/$#$ $$$##"" ##$#$ #! 4#.'2$!$-#$"#:#$ $!"$"$"$;<<=% [\X]DAHFYAAJKF?@JACFK?@B?CDF?JILNEGF?@N?[\?JDFFCJV?cDFLF?AJ?@BF?O@LF?MFLf?AO?@LCEBC? ?@ABCC@B@CFVqDFB?QeL[\?G@O?@JACFJ?ELF?OEKFX?FACDFL?Sf?FHFGCL@KF?@JACA@B?@B?CDF? KL@?UGEJCFK[\?NAHO?@L?Sf?G@UKF?@JACA@B?@N?OFCEH?EBK?L[\X?CDFLF?AJ?EH]EfJ?HFJJ?OFCEH? KF?@JACFKABCDF?GEJF?@N?L[\UG@BCEABABd?KF?@JACJV?pN?CDF?G@BJCEBC?GILLFBC?G@BKACA@BJ?ELF? E??HAFKXEGFLCEAB?NLEGCA@B?@N?GDELdF?AJ?IJFK?C@?LFKIGF?[\X?]DAGD?AJ?GHFELHf?MAJASHF?AB?CDF? JAdBANAGEBCKF?@HELArECA@B?@N?CDF?FHFGCL@KF?KILABd?CDF?FHFGCL@KF?@JACA@B??L@GFJJ?AB?CDF? ?LFJFBGF@N[\^``aV? gsj gtj guj vwxyz{\|}0~3()"#:#$ ' #$ . ""#" " $4. '#!#403)9 #!" "03)9 #!"% +#$ "'$$()#'. $!$! 9$# 4 2 "$$ $4$#: ##'#%&$"$##.#' ### 0)&3%& ! #" !2#"$!% #2$#$)&# #$#$#.#6 "#.4$ "$%/#"##$# $#$.4#. $"2 "####. JKLM 'NOPQ#JKL TQ#$ V)#) Q )$ $ $!'$#Q Q #) X# !) ) $X#S$Y$. @AZAC[\[]HEEHIHI\^H]_[`_abcEF d$$Ke(f$ Q$1 )#'d )'WQ i Wj kRT#Q/.l# $$#$ !h Ri Wj kRT'WQ SiXdTn on p.' V)#) Q 'W! $ Q )$ $ )!V! $!'$#Q #)$%# Q #))X## X# !) )! # $X#S$Y$. T' @AZAC[\[]HEEHIHI\^H]_[`_abcEF d$$Ke(f$$$#$ Q$1QR)gR)LS'JLT )#'d! Q1#)#1# )'WQg) hRi+RShR i Wj kRT#Q/.l#hR)M&1R$V $$#$ ! hfmdi#Q&)Sh Ri Wj kRT'WQ# !) SiXdTniXdonR)gR)Lp.' q0p.'./ 0rX' KKe(f$Q! 1%#!#1 QQ!#)$#.'$'L #Q! Q'(1 ,SRL(f,hQTV$Q1$%## ### ##$'WQ# !L##(f1$VL##'W$ W'#1 K1Ke(f1 Q$' d$$Ke(f$$$#$ Q$1QR)gR)LS'JLT )#'d! .10 345678:;./0<% .NQ0AICMHJH EF@GHI?CJHH=H IH@AHI+E?C?>SI E?>EF>?CHJMSJ@ EK=>HCH<%#G "DCMHC ghijklmnopqrst oq y{u }~xw{ux xy}w{wv{ }wv}~vw{x }uv{rst nhjh ux x\|y{x yx\|{u uxy}{ ~u\|~x{x }wyw\| }\|{xy~{{ uxxw\|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`_abcEF d$$Ke(f$$$#$ Q$1QR)gR)LS'JLT )#'d! Q1#)#1# )'WQg) hRi+RShR i Wj kRT#Q/.l#hR)M&1R$V $$#$ ! hfmdi#Q&)Sh .10 345678:;./0<% .NQ0AICMHJH EF@GHI?CJHH=H IH@AHI+E?C?>SI E?>EF>?CHJMSJ@ EK=>HCH<%#G "DCMHC ghijklmnopqrst oq y{u }~xw{ux xy}w{wv{ }wv}~vw{x }uv{rst nhjh ux x\|y{x yx\|{u uxy}{ ~u\|~x{x }wyw\| }\|{xy~{{ uxxw\| $#)1$# ' 234567895:;5<=476>?=; @A ABCDEFGEHI JKLM 'NOPQ#JKL$#'R$!S TQ#$ )Q! $UV)# V)#) Q 'W! $ Q )$ $ )!V! $!'$#Q #)$%# JKLM 'NOPQ#JKL TQ#$ V)#) Q )$ $ $!'$#Q Q #) X# !) ) $X#S$Y$. @AZAC[\[]HEEHIHI\^H]_[`_abcEF d$$Ke(f$ Q$1 )#'d )'WQ i Wj kRT#Q/.l# $$#$ !h Ri Wj kRT'WQ SiXdTn on p.' Q1#)#1# )'WQg) hRi+RShR i Wj kRT#Q/.l#hR)M&1R$V $$#$ ! hfmdi#Q&)Sh Ri Wj kRT'WQ# !) SiXdTniXdonR)gR)Lp.' q0p.'./ 0rX' KKe(f$Q! 1%#!#1 QQ!#)$#.'$'L #Q! Q'(1 ,SRL(f,hQTV$Q1$%## ### ##$'WQ# !L##(f1$VL##'W$ W'#1 K1Ke(f1 Q$' s5t:7u3d,$K Ke(f'R$ Q'')Q#v#V' w:7x68?=7y?;969?<z56> {7y?;969?<\|?<=969?<; {7y?;969?<s9}7~; JQSR) ,JT Xq .'/R$ +. L O. /. QQ!#)$#.'$ 'L #Q! Q'(1 ,SRL(f,hQTV$Q1$%## ### ##$'WQ# !L##(f1$VL##'W$ W'#1 K1Ke(f1 Q$' s5t:7u3d,$K Ke(f'R$ Q'')Q#v#V' w:7x68?=7y?;969?<z56> {7y?;969?<\|?<=969?<; {7y?;969?<s9}7~; JQSR) ,JT Xq .'/R$ +.L O. /. jf p.'OK jf p.'+K jXhf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`! " #!!! !%2!"!"J aJ U)5SWWU)5)V ^5! '! !""!#"! !%"&2%#!"! &" 3'bVc2!!""! ! $$,$2!!! %!"32&)5S$%$ T1 U)5V$%$ O XU)5SV$%$ OY1 #22!%%)5&%$ Q1(Q2! ! Y Y!3" Y,2!%OPQ15R2!%!"#"!$!%%! "! %"(+V)$-$+S, )"5R2!%!"#"!$!% #.!! %!!! 2!$! #2#%"."!%R%%"O !! %!!%"'$$! '%2!R%5-!2!%!" 2!!! %!"#!! #"2!%#2%!!%#!%"!!%2! 2!!.!! %!!! 2!$! $!"!$! "5 ! 2!$$!"!$! "#!! !%"&ZZ[\ !!]51!! " ! $$! 2!!.!! %!!! 2!$!5-"!%X$!! %! (YX,"! !%"!!! !!! %! %^$ O$ (Q%!#X$3%&! _$!% X$3%&! &%,"!!! %!501 "(Y&$-%2 Y5_" \1 [Y-, $%$+ !!%#2!%!!%#! #""!% !]!$! "5-2!$!"!$! "#!!!$!%$!$!!5`! " #!!! !%2!"!"JaJYXU)5SWWU)5)V ^5! JKLM 'NOPQ#JKL TQ#$ V)#) Q )$ $ $!'$#Q Q #) X# !) ) $X#S$Y$. @AZAC[\[]HEEHIHI\^H]_[`_abcEF d$$Ke(f$ Q$1 )#'d )'WQ i Wj kRT#Q/.l# $$#$ !h Ri Wj kRT'WQ SiXdTn on p.' +.. QQ!#)$#.'$ 'L #Q! Q'(1 ,SRL(f,hQTV$Q1$%## ### ##$'WQ# !L##(f1$VL##'W$ W'#1 K1Ke(f1 Q$' s5t:7u3d,$K Ke(f'R$ Q'')Q#v#V' w:7x68?=7y?;969?<z56> {7y?;969?<\|?<=969?<; {7y?;969?<s9}7~; JQSR) ,JT Xq .'/R$ +.L O. /. jf p.'OK jf p.'+K jXhf X .'/R$ . L +. . .. +.. ,SRL(f,hQTV$Q1$%## ### ##$'WQ# !L##(f1$VL##'W$ W'#1 K1Ke(f1 Q$' s5t:7u3d,$K Ke(f'R$ Q'')Q#v#V' 4765: w:7x68?=7y?;969?<z56> {7y?;969?<\|?<=969?<; {7y?;969?<s9}7~; R) /r.+K R)pK JQSR) ,JT Xq .'/R$ +.L O. /. R# .'./K XR#L Op.'OK jf+p.'OK jf+p.'+K jXhfO X .'/R$ .L +. . .. +.. L / 'O)$+LfOrqXfp)$+jfOp)$+X+f+ '0 'R)gR)LgJL $$# /. q/ .. +.. +..)$+fOrqXfp)$+L rXfp')$+X+f+X v#'/QXfO /R$ $$# /. .. +.. j q)$+jfOrXfp)$+X+f+ '0 'R)gR)LgJL $$# . +. . .. /. .. h ..)$+hL rOXfpjX+,pjXOLX .'+R$ $$# . .. .. +.. h .)$+XNhSfXT Pp/)$+JfX .'q/R$ q/L . /. .. +..)$+fOp..)$+jfOp)$+XfO /R$ $$# +. %!!$!%"&"& ) "5,5$ !)$"+ 5T! $(]$ 2%%!$ "!3d&!3'3 `2!$`fY !! & 6767:ghMG=8> R2!4". "&2!j$Y (QQ-,2!$! !!&'W)'#"! !"&!%2!%!" ())),"!X` !"%&(S^S !!!%W^W^q. &2! dp. !kWl5 (\_, (, #2 &! !"&!%2!%! "!'d!%(,"!iiX$! "(-& - O `% `, 2! ())),"!X`$! "(X%n, %2!i!(),"!(TXX$! ,"&2! !"%&(S^S,!"& ! "!" #2.'!"3"5-\d2"o`dp. ! !!%W^W^q.$!"2#""!%!&! 2!4"!%3!T!kWl5 i2!"%!%"! #!2 ! !"&!%2!%")5SV$'!(\_, !&!5'%&! 3%&! !&' J #"!%"#"r J aJY[iU +JY[i+sSJS (, #2!!JY[iU JY[i %JS"%2! ! !&'2!"!#2%" 2! ! !&'2! !"!%2! ! !&'"!%S$! !5 tuvwxyz{\|}wx\| R2!!! ' '$! "!""%$!%3'2!!"#2 %!! "! !"%!$!%!" 33!!! !'!%3'!&$! '""#2"3'2"! ""5!! #!"!%2!!! "2'%&! " !2!2'%&! ! !d!$!%5["&0! \!X "$"%!! '"o"4&" #!!%23! 3"!%2!"" !2'%&! " ! 3'2!$! !&!3' !"#!!$!"!%"&! $$!' -.&!%$%$+/!"01"&2!3 !.%!"3!%! 67789:;<8==8>8?@=>ABCD>AE;F;GB;H;>8?IJKLM=N= O%OPQ1#!!%!"!%"2!%R%(%$!! $$,$2!!!%!"32&)5S$%$+T OWXU)5SV$%$+OY1W #22!%%)5&%$+Q Y!3" Y,2!%OPQ15R2!%!"# "!%"(+V)$-$+S, )"5R2!%!" #.!!%!!!2!$! #2#%"."!%R !!%!!%"'$$!'%2!R%5-!2!%!" #"2!%#2%!!%#!%"!!%2!2!!.!!%! '! !""!#"! !%"&2%#!"! &" 3'bVc2!!""! ! %! !$!%"&"&!.$!%!$!"!$! + "5YX(]$!' )* "5,5$!"!$! "#!!!$!%"&' $$! '"#!! !)$"+5T!! 2!! "#!! 2!!! %!#"2!%+ "5YX $(]$!')* "5,5R2! 2!!! %!! #""#! 2%%! $!"! '5R2!R%#""!%#%"! ! "!3d&!3'333!"5\2&''""#"!%"&Y\.ef#2 `2! $`fY&! \"!(`2! $ %2 ! R2!O! 2!%",#2 !! &!)d5 6767:ghMG=8> R2!4".!"e! "'iR2!'(eiR,"#!!!$!% "&2!j$Y`YY1d&!kW WWl5R2!Q! !!%Q%! -]$ (QQ-,2!$! 3'`!%!# Td %!2#""!%kWVl5X. ! !&'W)'#"!$'!% #2! 2! 2&!%! "'.#"$!"&!5m! !"&!%2!%! "!'d!%(,"!iiX$! "(-& - O `% `, 2! ())),"!X`$! "(X%n, %2!i!(),"!(TXX$! ,"&2! %!!$!%"&"& ) "5,5$ !)$"+ 5T! $(]$ 2%%!$ "!3d&!3'3 `2!$`fY !! & 6767:ghMG=8> R2!4". "&2!j$Y (QQ-,2!$! !!&'W)'#"! !"&!%2!%!" ())),"!X` !"%&(S^S !!!%W^W^q. !)$" 5T!! 2!! "#!! 2!!! %!#"2!%+* "5YX $(]$!')* "5,5R2! 2!!! %!! #""#! 2%%! $!"! '5R2!R%#""!%#%"! ! "!3d&!3'333!"5\2&''""#"!%"&Y\.ef#2 `2! $`fY&! \"!(`2! $ %2 ! R2!O! 2!%",#2 !! &!)d5 6767:ghMG=8> R2!4".!"e! "'iR2!'(eiR,"#!!!$!% "&2!j$Y`YY1d&!kW WWl5R2!Q! !!%Q%! -]$ (QQ-,2!$! 3'`!%!# Td %!2#""!%kWVl5X. ! !&'W)'#"!$'!% #2! 2! 2&!%! "'.#"$!"&!5m! !"&!%2!%! "!'d!%(,"!iiX$! "(-& - O `% `, 2! ())),"!X`$! "(X%n, %2!i!(),"!(TXX$! ,"&2! !"%&(S^S,!"& ! "!" #2.'!"3"5-\d2"o`dp. ! !!%W^W^q.$!"2#""!%!&! 2!4"!%3!T!kWl5 i2!"%!%"! #!2 ! !"&!%2!%")5SV$'!(\_, !&!5'%&! 3%&! !&' J #"!%"#"r J aJY[iU +JY[i+sSJS (, #2!!JY[iU JY[i %JS"%2! ! !&'2!"!#2%" 2! ! !&'2! !"!%2! ! !&'"!%S$! !5 tuvwxyz{\|}wx\| R2!!! ' '$! "!""%$!%3'2!!"#2 3'`!%!# Td %!2#""!%kWVl5X. '!% #2! 2! 2&!%! "'.#"$!"&!5m! d!%(,"!iiX$! "(-& - O `% `, 2! ! "(X%n, %2!i!(),"!(TXX$! ,"&2! !"& ! "!" #2.'!"3"5-\d2"o`dp. $!"2#""!%!&! 2!4"!%3!T!kWl5 !2 ! !"&!%2!%")5SV$'!(\_, &! !&' J #"!%"#"r J aJY[iU +JY[i+sSJS (, %JS"%2! ! !&'2!"!#2%" 2! !%2! ! !&'"!%S$! !5 '$! "!""%$!%3'2!!"#2 $!%!" 33!!! !'!%3'!&$! 2"! ""5!! #!"!%2!!! "2'%&! . 5m! , 2! JKLM 'NOPQ#JKL TQ#$ V)#) Q )$ $ $!'$#Q Q #) X# !) ) $X#S$Y$. @AZAC[\[]HEEHIHI\^H]_[`_abcEF d$$Ke(f$ Q$1 )#'d )'WQ i Wj kRT#Q/.l# $$#$ !h Ri Wj kRT'WQ SiXdTn on p.' $"+, 5 ! %$ $"+, 5 ! !"#!!$!"!%"&! $$! '()$"+, -.&!%$%$+/!"01"&2!3 !.%!"3!%!4&5 67789:;<8==8>8?@=>ABCD>AE;F;GB;H;>8?IJKLM=N= O%OPQ1#!!%!"!%"2!%R%(%$! !"""! ! $$,$2!!! %!"32&)5S$%$+T1U)5V$%$+ OWXU)5SV$%$+OY1W #22!%%)5&%$+Q1(Q2! ! Y Y!3" Y,2!%OPQ15R2!%!"#"!$!%%! "! %"(+V)$-$+ , )"5R2!%!"#"!$!% #.!! %!!! 2!$! #2#%"."!%R%%"O !!!%W^W^ i2!"%!%" !&!5'% #2!!JY[iU J !!&'2!! tuvwxyz{\|}wx\| R2!!! %!!"!!" '""#2" " !2!2 "$"% 3"!%2!" "2&25Z !&!5'%&! 3%&! !&' J #"!%"#"r J aJY[iU +JY[i+sSJS (, #2!!JY[iU JY[i %JS"%2! ! !&'2!"!#2%" 2! ! !&'2! !"!%2! ! !&'"!%S$! !5 tuvwxyz{\|}wx\| R2!!! ' '$! "!""%$!%3'2!!"#2 %!! "! !"%!$!%!" 33!!! !'!%3'!&$! '""#2"3'2"! ""5!! #!"!%2!!! "2'%&! " !2!2'%&! ! !d!$!%5["&0! \!X "$"%!! '"o"4&" #!!%23! 3"!%2!"" !2'%&! " ! 3'2!$! !&!3' %""2&25Z! !!!%W^W^ i2!"%!%" !&!5'% #2!!JY[iU J !!&'2!! tuvwxyz{\|}wx\|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`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ &"&& >?@ABCDBE@CFG?@FBEHIJ#"(# KL,#&<16M&)$ <16M5&# <16 #&KL,M#$ : 13L ,N, 1K:#&O;PM#5)# :* 13L ,N, 1K:#&O;PM <16#&KL,M #)Q)#&""# KL,#&<16M #)Q5 #&&#) :* 13L ,N, 1K: O;P#&<16M"5# <16M #&#)R# KL,#&<16M15&#&)&"'&5# #" S?ETFEUI #&&'$K#/#&9"'K'269-<K"V 9J14#&#$8&# K# #&#):5"#9"' K'2J#;W.XY.Z[X[[.Z[XX4<16#&'&9/#& K'#1&$K##&1 "V/. X ""#)&$ \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# #)Q5 #&&#) : 13L ,N, 1K: O;P#&<16M"5# <16M #&#)R# KL,#&<16M15&#&)&"'&5# #" S?ETFEUI #&&'$K#/#&9"'K'269-<K"V 9J14#&#$8&# K# #&#):5"#9"' K'2J#;W.XY.Z[X[[.Z[XX4<16#&'&9/#& K'#1&$K##&1 "V/. X ""#)&$ \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## #" S?ETFEUI #&&'$K#/#&9"'K'269-<K"V 9J14#&#$8&# K# #&#):5"#9"' K'2J#;W.XY.Z[X[[.Z[XX4<16#&'&9/#& K'#1&$K##&1 "V/. X ""#)&$ \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## orstuv wx [ =YyJ9z )$n{\|rs{u w X=X[yJ9zy6'&z &#$)##)$}J'##)!"##&$#$ #" S?ETFEUI #&&'$K#/#&9"'K'269-<K"V 9J14#&#$8&# K# #&#):5"#9"' K'2J#;W.XY.Z[X[[.Z[XX4<16#&'&9/#& K'#1&$K##&1 "V/. X ""#)&$ \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## orstuv wx [ =YyJ9z #)$n{\|rs{u w X=X[yJ9zy6'&z &#$)##)$}J'##)!"##&$#$ S?ETFEUI #&&'$K#/#&9"'K'269-<K"V 9J14#&#$8&# K# #&#):5"#9"' K'2J#;W.XY.Z[X[[.Z[XX4<16#&'&9/#& K'#1&$K##&1 "V/. X ""#)&$ \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## orstuv wx [ =YyJ9z )$n{\|rs{u w X=X[yJ9zy6'&z &#$)##)$}J'##)!"##&$#$ [= [WyJ9z ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" 9J14#&#$8&# K# #&#) K'2J#;W.XY.Z[X[[.Z[XX4<16#& K'#1&$K##&1 "V/. X "" \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# 6M,# OMK#05 P9M05 ;:)#"#" orstuv wx [ =YyJ9z #)$n{\|rs{u w X=X[yJ9zy6' &#$)##)$}J'##)!" [= [WyJ9z ,M,#(( 89MK#5# KM1""'$ 1L$&)# lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# A% D D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!" !$("56k:6:2S6;lm< On >?>E@A""BC !" # $%&$'( )($+(!"%,-. ($! $!(,""/""-0""1232345674829:486 ;< = >? >@A""BC@0D-C E %. )D,! DA% FG GA(!.!%%,-. H(!$!$!H!,('(($! (,! "I!%!% %,-. 0-. .,&&!(($"$"J42KLJ4M3K846N936; O P ?E@A""BC Q 0! R224:3S32TKL43UV2W. X" $ EP > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD'H$ ZD#%$.!% \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]` EaD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% DF(, !&'H$ D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" !" # $%&$'( )($+(!"%,-. ($! $!(,""/""-0""1232345674829:486 ;< = >? >@A""BC@0D-C E %. )D,! DA% FG GA(!.!%%,-. H(!$!$!H!,('(($! (,! "I!%!% %,-. 0-. .,&&!(($"$"J42KLJ4M3K846N936; O P ?E@A""BC Q 0! R224:3S32TKL43UV2W. X" $ EP > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD'H$ ZD#%$.!% \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]` EaD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% DF(, !&'H$ * D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# $k:6:2S6 :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE? >Q@A""BC Q D$A, AAo. yDA%!& D! " A&(( fW$%&'.],-. #H(H.B$! -_g,. #H(H.B$!#($! $!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi @A""BC@0D-C Q [- [A% \((-& 0"-&! ("%$-%,-. "+((H-"!. h2K68e;m r PP?P@A""BC "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! $k:6:2S6 :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE? >Q@A""BC Q D$A, AAo. yDA%!& D! " A&(( fW$%&'.],-. #H(H.B$! -_g,. #H(H.B$!#($! $!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi @A""BC@0D-C Q [- [A% \((-& 0"-&! ("%$-%,-. "+((H-"!. h2K68e;m r PP?P@A""BC Qi &0%!% WZ. B !%+!% [B&+0% yH"!.! +((HD$%"& ((-& Y$! - ],-. #g(! -$!(-F+%56k:6:2S6;; = i ?@A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((-& E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# >Q@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! $k:6:2S6 :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE? >Q@A""BC Q D$A, AAo. yDA%!& D! " A&(( fW$%&'.],-. #H(H.B$! -_g,. #H(H.B$!#($! $!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$% :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+ QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$! g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"! $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"y V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# Y0! \ }! y\- DW' [f%$!"$& D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# \]@]>^]F_]GF_F@`a@]@bcbE@I:5'"#R >deEBf_bTUcbE@HI<16 &0#0K50P# g#5$O)&Q/$ 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# lbibCbEdbH K#0 :M6"m" 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## $nopqorstuv wx [ =YyJ9z [ 05 ;<##)$n{\|rs{u w X=X[yJ9zy6'&z Y ,$ LM05 ;&#$)##)$}J'##)!"##&$#$ qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u >deEBf_bTUcbE@HI<16 &0#0K50P# g#5 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# lbibCbEdbH K#0 :M6"m" 6M,# OMK#05 P9M05 ;:)#"#" $nopqorstuv wx [ =YyJ9z [ 05 ;<##)$n{\|rs{u w X=X[yJ9zy6'&z Y ,$ LM05 ;&#$)##)$}J'##)!"# qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5 #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:) ""&$$&)#5##)!"##&#&q YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&) 6$J$ "&# DBEhFd@HBijE@bCbH@I&##k# lbibCbEdbH K#0 :M6"m" 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## $nopqorstuv wx [ =YyJ9z [ 05 ;<##)$n{\|rs{u w X=X[yJ9zy6'&z Y ,$ LM05 ;&#$)##)$}J'##)!"##&$#$ qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 DBEhFd@HBijE@bCbH@I&##k# lbibCbEdbH K#0 :M6"m" 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## $nopqorstuv wx [ =YyJ9z [ 05 ;<##)$n{\|rs{u w X=X[yJ9zy6'&z Y ,$ LM05 ;&#$)##)$}J'##)!"##&$#$ qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ lbibCbEdbH K#0 :M6"m" 6M,# OMK#05 P9M05 ;:)#"#"$&)#5## $nopqorstuv wx [ =YyJ9z [ 05 ;<##)$n{\|rs{u w X=X[yJ9zy6'&z Y ,$ LM05 ;&#$)##)$}J'##)!"##&$#$ qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z [ 05 ;<##)$n{\|rs{u w X=X[yJ9zy6'&z Y ,$ LM05 ;&#$)##)$}J'##)!"##&$#$ qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z qorstqo~spo{{u [= [WyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z :,( LM./ ,M,#(( 89MK#5# KM1""'$ 1L$&)#5###" #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z #0#&'$o{{spt{qorstu w [Y=[yJ9z W N')& LM JMJ# NMO#0 L:M;05 LNM1'&.6&# /ML /:)##)#"5&## ""&$$&)#5##)!"##&#&qorstqo~spo{{u YX[[=YX[ yJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z O#0 L:M7# 9MN')& LMJ0## 6ML /J$(#)7$&)#85#9#27894 $'&#K%&;#{u Y W=Y yJ9z O#0 L:MJ# 3MN(0$ mM/# 1LML /1"$'&#K%&J$8 7$&)#6&#}<#)#-)#J$15${u w = [YyJ9z : 7MN#) 9.M, M /Mm L9#")###""& ##&$ $&)#5#nopru [=[[yJ9z K0## #)5$ #&'5}<<<8$$&)#5##& #{qrspo{r{orsqrspr{u x Y=yJ9z X ;05 LNMO)& Mm)& 1MN# L9MJ# L,M6#&5 KMK#) g#&#8!#)J# 7$&)#85#{qrspr{p{ w L[YyJ9z 4829:486 I!%!% 4829:486 I!%!% ;< = >? lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# A% DF(, !&'H$ D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,- ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(, !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H& :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,( QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""B EQ A%- Y]\|! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. %,-. 0-. .,&&!(($"$"J42KLJ4M3K846N936; O P ?E@A""BC Q 0! R224:3S32TKL43UV2W. X" $ EP > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD'H$ ZD#%$.!% \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]` EaD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% DF(, !&'H$ D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD'H$ ZD#%$.!% \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]`aD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% DF(, !&'H$ D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! $k:6:2S6 E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936t O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,bN 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! $k:6:2S6 :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]`aD!(A ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% D D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^ ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!" !$("56k:6:2S6;lm< On >?>E@A""BC i A )A ($ o +((H]! . "A!(,"":2S6p2M6;llq r= i ?i [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \ %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$% :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y( D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-" QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]`aD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# Y0! \ }! y\- DW' [f%$!"$&-s$! g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" * +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"! $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"y V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,- 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""B Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! - u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$ "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# $k:6:2S6 :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE? >Q@A""BC Q D$A, AAo. yDA%!& D! " A&(( fW$%&'.],-. #H(H.B$! -_g,. #H(H.B$!#($! $!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi @A""BC@0D-C Q [- [A% \((-& 0"-&! ("%$-%,-. "+((H-"!. h2K68e;m r PP?P@A""BC Qi &0%!% WZ. B !%+!% [B&+0% yH"!.! +((HD$%"& ((-& Y$! - ],-. #g(! -$!(-F+%56k:6:2S6;; = i ?@A""BC $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! $k:6:2S6 :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE? >Q@A""BC Q D$A, AAo. yDA%!& D! " A&(( fW$%&'.],-. #H(H.B$! -_g,. #H(H.B$!#($! $!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi @A""BC@0D-C Q [- [A% \((-& 0"-&! ("%$-%,-. "+((H-"!. h2K68e;m r PP?P@A""BC Qi &0%!% WZ. B !%+!% [B&+0% yH"!.! +((HD$%"& ((-& Y$! - ],-. #g(! -$!(-F+%56k:6:2S6;; = i ?@A""BC Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE? >Q@A""BC Q D$A, AAo. yDA%!& D! " A&(( fW$%&'.],-. #H(H.B$! -_g,. #H(H.B$!#($! $!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi @A""BC@0D-C Q [- [A% \((-& 0"-&! ("%$-%,-. "+((H-"!. lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# $%&$((, ,!%"-Z-$-F+% 0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((- 2M6R26l E>PE@A""BC (, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6 >?iP> @A""BC H~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 ""BC +H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'((' f%&$%"&56J29K846:2S6; = v ?E@A""BC "'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. H(!NVK6N936 @A""BC 0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f%&-,&$" !-,b i?E@A""BC ! \ o!& D '-&H ZD!" }! y\y&+H-$!(,"!"%,-. H(! ("(!"!%.%!%($! $%&$(&!$'(/-$-. +%g-! $k:6: ;t r QE?QE Q@A""BC@0D-C H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!%"-Z-$-F+% ! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89: QE? >Q@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D! Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(! "(!X!yf%&$%"&56J29K846:2S6; = v ?E@A""BC Q F(, Z"'H [o. /"$( -"!,$!(!%,/0"-"$.((,"%,-. O P? @A""BC QE .% \[B0" D\G'0" Wy]/ +((H!%.%!%A!(,"! !!X\\0y!f 748l i?E@A""BC QQ F!~ $ Y0! \ o!& D '-&H ZD!" }! y\y&+H-$! $!('("(!"!%.%!%($! $%&$(&!$'(/-$-. +% :2S6k:6;t r QE?QE Q@A""BC@0D-C Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($! $%&$((, ,!% _g-A&+"! A!(,"!"],-. #H(!\('(D-jf%#$!"F+%_g- N936;l u QE? lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$! g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"! $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"y V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,- 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""B Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! - EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" * &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# E %. )D, %,-. 0- Q 0! R !" # $%&$'( )($+(!"%,-. ($! $!(,""/""-0""1232345674829:486 ;< = >? >@A""BC@0D-C E %. )D,! DA% FG GA(!.!%%,-. H(!$!$!H!,('(($! (,! "I!%!% %,-. 0-. .,&&!(($"$"J42KLJ4M3K846N936; O P ?E@A""BC Q 0! R224:3S32TKL43UV2W. X" $ EP > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD'H$ ZD#%$.!% \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]` EaD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% DF(, !&'H$ D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 !" # $%&$'( )($+(!"%,-. ($! $!(,""/""-0""1232345674829:486 ;< = >? >@A""BC@0D-C E %. )D,! DA% FG GA(!.!%%,-. H(!$!$!H!,('(($! (,! "I!%!% %,-. 0-. .,&&!(($"$"J42KLJ4M3K846N936; O P ?E@A""BC Q 0! R224:3S32TKL43UV2W. X" $ EP > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD'H$ ZD#%$.!% \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]` EaD!(A!(,"!"b4L2c6:2S6d46Je6 ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! A% DF(, !&'H$ * D'H$ ZDf-"$!H!,!%I! ($! (,"$!"QeD^Z A D`%,-^g,`g-$!(,"!"76h2K6 ; P?i@A""BC@0D-C [%0 !$(!!$(D. !],-. \!&"(!&j\(&A!(,"!"&!$(!Z.%0. !" # $%&$'( )($+(!"%,-. ($! $!(,""/""- ;< = >? >@A""BC@0D-C E %. )D,! DA% FG GA(!.!%%,-. H(!$!$!H!, %,-. 0-. .,&&!(($"$"J42KLJ4M3K846N936; O P ?E@A" Q 0! R224:3S32TKL43UV2W. X" $ EP > Y(H$ Z 00& B ! &$' YA%. [/A('" \ !&'H$ BD \('(],-. #H(!B$!\$!H!,!%.%!%W$!(!,Z^_]` EaD!(A ; = E> ?E> @A""BC@0D-C 00& Bf+'H$ YA%. [/A ! &$' Y('" \]"! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC [& ) +s$ A[(0! \F0$%! --( #'$ GHW'%H \A!(,"!"++! $!" %,-. "+((H7VK2;t =u ?i@A""BC " B],-. +((H$!"-$!:2S6p2M6; v Ei>?EiQ@A""BC EP ".'" FD-'" F#-&!("!-"-+$+!"!%"+((H&$%"&%,-. "!. :2S68SSV46;; uw i QQ?i >Q@A""BC E ! &$' Yx$%& D). A 00& BY(H$ ZH-(! Y('" \ !&'H$ B D'H$ ZDy&+H.!%%,-. g-!$!! 0,+&!%,-g,(-"+!76:2S6; = QPP?QP@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" * &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC EE f Gz% GA {o Go A% z] Y"$H,],-. +((H/W"-W,#($! $!(,"!" ],-. #H(!X&f%,!#g+&!bN6 v E?E @A""BC EQ A%- Y]\|! Y0! \ }! y\- DW' [f%$!"$&-s$!0,-$-. +% g-!%($! $!(,!$$!H!,$'(!I-"!%%,-. H(!$!k:6:2S6:2S6k:6;q w E>?Ei>@A""BC@0D-C E> F!~ $ HH~ $ \zY0! \ D!+ YB(H~ $ oY}! y\D!" &+(! "!%&/ +H&!%,-. H(!$!H!,Z/D$!(,"!"X&"(/. (-H-+I-$!(,"!"!. +%"++! - $/($! -+"!"d4656eK8L6J42KL; u >?@A""BC E )("$%&-! ]"$%" _ !&&. x],-. B$!"Z"! $!- $"yV22N932492:28K V4eS244e13893)$'I"' \ Z"'H [ #-"%)(, " y$X]0' Z x \ EPP@A""BC E }! y\o!& D '-&H ZF(+$+("-".."++! -$!(,"!"%,-. H(!$! 0"-$$+!([!$D! A(&-(.d4656eK8L6J42KL;m u EE?EQ@A""BC Ei f, ]oW0 \#[!$% B ,'" #A]],-. Y""$!- +((Hy-H-(y"(! -((-& \!&"k:6p2M6R26l E>PE@A""BC E Z(' \xF(, DH''" D\ &+(B(f%&0Y"f"!/D!( $"b4L2c6:2S6d46 Je6m u= iP>?iP> @A""BC E '}~ $ DH~ $ D["!~ $ Z[!$(,""%,-. H(!!Z?D((,($! -"J29K89:3S6b9 u= E?E@A""BC QP ["!~ $ Z+H~ $ DF. WH~ $ D+ Y_!%'!$"!%%,-. H(!$!$'(('( EQ A%- Y]\|! Q> F!~ $ 0H~ $ DD!+ Y}! y\[~ $ D!" #($!$%&$((, ,!%"-Z-$-F+% _g-A&+"!A!(,"!"],-.#H(!\('(D-jf%#$!"F+%_g- ++!d4656J29K89:2S6 N936;l u QE?>Q@A""BC Q D$A, AAo. yDA%!& D!" A&(( fW$%&'.],-.#H(H.B$! -_g,.#H(H.B$!#($!$!(,"!" ()! +(!!.YH$"56bS6:2S6N896;q w >Q>i?>Qi @A""BC@0D-C Q [- [A% \((-& 0"-&!("%$-%,-."+((H-"!.h2K68e;m r PP?P@A""BC Qi &0%!% WZ. B !%+!% [B&+0% yH"!.! +((HD$%"& ((-& Y$!- ],-.#g(!-$!(-F+%56k:6:2S6;; = i ?@A""BC lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# '# ""()!# +,,-# .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0 /00#* !678979:;7<:7== > ???@??ABC#D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 13 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! * 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 1*3 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# ""# # " $ %#& ! '# .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0* /00#* !678979:;7<:7== > E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ' 1*3 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOB* OP Q* R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()* +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#) )J J3#++0 )0#/ /,+)0# J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC# O Q HH 1E E 5, )* $* C-HH3J ##, 200 Q 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 ! "# ! ""# # " $ %#& ! '# ""()!# +,,-# .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0 /00#* !678979:;7<:7== > ???@??ABC#D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 13 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! * 5)#,, $, ) $* C-HH ##, 2 ! "# ! ""# # " $ %#& ! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# '# ""()!# +,,-# .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0 /00#* !678979:;7<:7== > ???@??ABC#D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 13 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! * 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D ! "# ! ""# # " $ %#& ! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D O $ &0 "F1/& F2+#/,+ 0E,, H # 0#!0 897o:p7r=mnL ABC#D stuvwxtyz{\|}~wtuz{uz#00#J# 0 + ))0/ 0 #) ,,+3,/0 #,#,(0#0# )-), 0* )/ 030 ) 0$214 )0##)0$214 )0##)0)/,J#+ 3,0( ( R(0 +#+,#++#0(#,0 !J ()# J#0) 0/0 +)/0##*#)0 0#/ 0# 0 O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! * 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D O $ &0 "F1/& F2+#/,+ 0E,, H # 0#!0 897o:p7r=mnL ABC#D stuvwxtyz{\|}~wtuz{uz#00#J# 0 + ))0/ 0 #) ,,+3,/0 #,#,(0#0# )-), 0* )/ 030 ) 0$214 )0##)0$214 )0##)0)/,J#+ 3,0( ( R(0 +#+,#++#0(#,0 !J ()# J#0) 0/0 +)/0##*#)0 0#/ 0# 0 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D O $ &0 "F1/& F2+#/,+ 0E,, H # 0#!0 897o:p7r=mnL ABC#D stuvwxtyz{\|}~wtuz{uz#00#J# 0 + ))0/ 0 #) ,,+3,/0 #,#,(0#0# )-), 0* )/ 030 ) 0$214 )0##)0$214 )0##)0)/,J#+ 3,0( ( R(0 +#+,#++#0(#,0 !J ()# J#0) 0/0 +)/0##*#)0 0#/ 0# 0 lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# * 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D O $ &0 "F1/& F2+#/,+ 0E,, H # 0#!0 897o:p7r=mnL ABC#D +,,)J/,0# /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! * 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D O $ &0 "F1/& F2+#/,+ 0E,, H # 0#!0 897o:p7r=mnL ABC#D stuvwxtyz{\|}~wtuz{uz#00#J# 0 + ))0/ 0 #) ,,+3,/0 #,#,(0#0# )-), 0* )/ 030 ) 0$214 )0##)0$214 )0##)0)/,J#+ 3,0( ( R(0 +#+,#++#0(#,0 !J ()# J#0) 0/0 +)/0##*#)0 0#/ 0# 0 O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 4#0,%h'5h$ .1C.ic'J),* )+# /#0I#+R#/0j 0JJ,0 J0#,67897 dek:e7l:Xm > AAP?BC#DB13$#)D OO Q HH 1' )# iEJJ# E iE ! lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# h2K68e;m r PP?P@A""BC Qi &0%!% WZ. B !%+!% [B&+0% yH"!.! +((HD$%"& ((-& Y$! - ],-. #g(! -$!(-F+%56k:6:2S6;; = i ?@A""BC .#/0102 E, /C#( $ 1*3 OP Q R2#,CS +,,-#* C#D 0!# J,, ! "# ! ""# # " $ %#& ! '# ""()!# +,,-#* .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0 /00#* !678979:;7<:7== > ???@??ABC#D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 13 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D ! "# ! ""# # " $ %#& ! '# ""()!# +,,-# .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0 /00#* !678979:;7<:7== > ???@??ABC#D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 13 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D O Q HH 1E* E 5, )* $* C-HH3J ##, 200 Q//// $23 ! "# ! ! "# ! ""# # " $ %#& ! '# ""()!# +,,-# .#/0102+#)'/0-0#)3 0)#)3(4 #,0/5#0 /00#* !678979:;7<:7== > ???@??ABC#D E, /C#( $F0 F4',)/ $G+#H $F', F'4+,,-#C#,#- 0"()!# ')+0 )0!# 13 2+#)I01,,)J5 +0/,#K678979:;7=L >M NAN@NOBC#D OP Q R2#,CS '',)/ $F0 F4G+#H $F', F''3# /#+,,-#()!# )3#) J,, +,,)J/,0#* /#)0!+# #-/ /#TUU7VWXY7VZ[7= \\] O ABC#D O 230 '1^0 4'$# 0 _&)J- 5_'!# #,-#I 0##/0-0(0#`(!# /0 ,H#)!+# # 3,+, #8979:;79:;7897=L a AP@ABC#D O? ##0J $&)J- 5_b$3c.`0# )#)* )J J3#++0 )0#/ /,+)0#00#b$3!# #, J#I&& #0/$ 0#,J,0 d;UW7897d;;We7=f ]g @BC#D lbibCbEdbH K#0 $ [ 05 Y ,$ qorstq :,( #0# 5)#,, $, ) $* C-HH ##, 2 /// $#0,')- /#)/+3,0#J0#,J)#,, !I0% 0J.1C.i67897dek:e7l:X=n >] OA PBC#DB13$#)D OA 1#)#I F1E&# b.* H#* $Q# #,H#)Q)# 0'++`J0 $)#J+,#897o:p7<:7=mmL qq A BC#DB13$#)D O $ &0 "F1/& F2+#/,+ 0E,, H # 0#!0 897o:p7r=mnL ABC#D stuvwxtyz{\|}~wtuz{uz#00#J# 0 + ))0/ 0 #) ,,+3,/0 #,#,(0#0# )-), 0* )/ 030 ) 0$214 )0##)0$214 )0##)0)/,J#+ 3,0( ( R(0 +#+,#++#0(#,0 !J ()# J#0) 0/0 +)/0##**#)0 0#/ 0# 0
https://openalex.org/W2968511059	https://www.pure.ed.ac.uk/ws/files/113264305/FordEtalJAMA2019EffectOfPrimaryCare.pdf	English	null	Effect of Primary Care Parent-Targeted Interventions on Parent-Adolescent Communication About Sexual Behavior and Alcohol Use	JAMA network open	2,019	cc-by	8,995	General rights C i h f h General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Edinburgh Research Explorer Effect of primary care parent-targeted interventions on parent- adolescent communication about sexual behavior and alcohol use A randomized clinical trial Citation for published version: Ford, CA, Mirman, JH, García-España, JF, Fisher Thiel, MC, Friedrich, E, Salek, EC & Jaccard, J 2019, 'Effect of primary care parent-targeted interventions on parent-adolescent communication about sexual behavior and alcohol use: A randomized clinical trial', JAMA Netw Open, vol. 2, no. 8, e199535. https://doi.org/10.1001/jamanetworkopen.2019.9535 Citation for published version: Ford, CA, Mirman, JH, García-España, JF, Fisher Thiel, MC, Friedrich, E, Salek, EC & Jaccard, J 2019, 'Effect of primary care parent-targeted interventions on parent-adolescent communication about sexual behavior and alcohol use: A randomized clinical trial', JAMA Netw Open, vol. 2, no. 8, e199535. https://doi.org/10.1001/jamanetworkopen.2019.9535 Digital Object Identifier (DOI): 10.1001/jamanetworkopen.2019.9535 Link: Link to publication record in Edinburgh Research Explorer Document Version: Publisher's PDF, also known as Version of record Published In: JAMA Netw Open g j ( ) 10.1001/jamanetworkopen.2019.9535 Document Version: Publisher's PDF, also known as Version of record Published In: JAMA Netw Open Key Points Question Can parent-targeted interventions delivered within primary care settings affect parent-adolescent communication about sexual health and alcohol use? IMPORTANCE Adolescent well care visits provide opportunities for clinicians to facilitate parent- adolescent communication (PAC) to reduce pregnancy, sexually transmitted infections, and alcohol- related harm among adolescents. OBJECTIVE To test the effect of brief parent-targeted interventions delivered in primary care settings on PAC about sexual and alcohol use behaviors. Findings This randomized clinical trial included 118 parent-adolescent dyads and found that brief parent-targeted interventions in primary care pediatric settings were acceptable, were feasible, and significantly increased adolescent- reported frequency of parent- adolescent communication about sexual health and alcohol use. DESIGN, SETTING, AND PARTICIPANTS Randomized clinical trial conducted at a primary care pediatric practice from January 4, 2016, to April 10, 2017. Adolescents who were scheduled for a well care visit were recruited, along with their parent or guardian. Data analyses continued through April 30, 2018. INTERVENTIONS During well care visits, parents in sexual health intervention and alcohol prevention intervention groups received coaching to discuss written intervention materials encouraging PAC about sex or alcohol, respectively, with their adolescent within 2 weeks, followed by a brief clinician endorsement. After 2 weeks, parents received a follow-up telephone call. Control group parents received usual care. Meaning Delivering parent-targeted interventions in primary care settings may be an important strategy for influencing parent-adolescent communication and adolescent behaviors and improving health outcomes. MAIN OUTCOMES AND MEASURES Participants were surveyed 4 months after the well care visit. Parent-reported and adolescent-reported quality of PAC was measured using the 20-item Parent- Adolescent Communication Scale, in which a higher score indicates better PAC; and frequency of PAC about sex or alcohol was measured using a 4-point Likert-type scale with 1 indicating not at all or never, and 4 indicating a lot or often. + Supplemental content Author affiliations and article information are listed at the end of this article. RESULTS Of 196 parent-adolescent dyads assessed for eligibility, 118 (60.2%) were eligible to participate. These 118 dyads were randomized to 1 of 3 groups: (1) sexual health intervention (n = 38 [32.2%]); (2) alcohol prevention intervention (n = 40 [33.9%]); and control (n = 40 [33.9%]); 104 parents (88.1%) and 99 adolescents (83.9%) completed the study. Parents included 112 women (94.9%) and had a mean (SD) age of 45.8 (6.9) years. Original Investigation \| Pediatrics Effect of Primary Care Parent-Targeted Interventions on Parent-Adolescent Communication About Sexual Behavior and Alcohol Use A Randomized Clinical Trial Carol A. Ford, MD; Jessica H. Mirman, PhD; J. Felipe García-España, PhD; Megan C. Fisher Thiel, MPH; Elizabeth Friedrich, BA; Elyse an, PhD; J. Felipe García-España, PhD; Megan C. Fisher Thiel, MPH; Elizabeth Friedrich, BA; Elyse C. Salek, MSEd; James Jaccard, Ph Original Investigation \| Pediatrics Effect of Primary Care Parent-Targeted Interventions on Parent-Adolescent Communication About Sexual Behavior and Alcohol Use A Randomized Clinical Trial Take down policy Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 24. Oct. 2024 + Supplemental content Author affiliations and article information are listed at the end of this article. August 16, 2019 1/12 Author affiliations and article information are listed at the end of this article. Overview This randomized clinical trial included adolescents aged 14 to 15 years and their parent or caregiver (hereafter called parent) and focused on improving PAC about sexual health and alcohol use. The trial was conducted from January 4, 2016, to April 10, 2017. Data analyses continued until April 30, 2018. Another study arm involved older adolescents and focused on PAC about safe driving; that study is reported elsewhere.16 The study protocol (Supplement 1) was approved by the Children’s Hospital of Philadelphia Institutional Review Board. Parents provided verbal informed consent, and adolescents provided verbal assent. This study is reported following the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. Abstract (continued) in the control group (2.32 [95% CI, 1.97-2.66] vs 1.79 [95% CI, 1.50-2.08]; P = .02); adolescents in the alcohol prevention intervention group reported a higher mean frequency score for PAC about alcohol compared with those in the control group (2.93 [95% CI, 2.60-3.25] vs 2.40 [95% CI, 2.08-2.72]; P = .03). Parent-reported frequency scores for PAC about sex or alcohol did not differ by group. CONCLUSIONS AND RELEVANCE Brief parent-targeted interventions in primary care settings increased adolescent-reported frequency of PAC about sexual health and alcohol use and may be an important strategy for parents to influence adolescent behaviors and health outcomes. CONCLUSIONS AND RELEVANCE Brief parent-targeted interventions in primary care settings increased adolescent-reported frequency of PAC about sexual health and alcohol use and may be an TRIAL REGISTRATION ClinicalTrials.gov identifier: NCT02554682 JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 August 16, 2019 2/12 Introduction Engaging in sexual behaviors and drinking alcohol during the second decade of life is common.1-4 Community-, school-, and home-based interventions involving direct contact between staff and parents or caregivers can favorably affect parent-adolescent communication (PAC) and a wide range of adolescent risk-associated behaviors.4,5 Parents can influence adolescents’ risk of unwanted pregnancy, sexually transmitted infection, and alcohol-related injury that cause substantial morbidity and mortality among adolescents and young adults.5-8 Enthusiasm about these interventions must be balanced with acknowledgment that such programs can be difficult to scale and sustain. Delivering effective PAC interventions in conjunction with annual adolescent well care visits in primary care clinics could provide a recurring mechanism to systematically reach sizable portions of adolescents and their parents or guardians.9 This would also align with goals of patient-centered and family-centered care because adolescents and parents or guardians are interested in receiving information from primary care clinicians to facilitate increased PAC about a variety of issues, including sexual health and alcohol use.10,11 To our knowledge, few primary care clinic–based PAC interventions exist.12,13 We tested the efficacy, feasibility, and acceptability of delivering parent-targeted interventions in a primary care pediatric practice for PAC about sexual health or alcohol use. Our study was not sufficiently large or long enough to test influences of PAC on adolescent behaviors, but we adapted and tested interventions shown to improve both PAC and behavior (ie, initiation of first sexual intercourse, condom use,14 and alcohol-associated injury15) in other settings. Recruitment Parents of all patients aged 14 to 15 years with a scheduled annual well care visit between January and September 2016 at the selected primary care pediatric practice were identified. The practice is August 16, 2019 2/12 Key Points Adolescents included 60 girls (50.9%); 67 adolescents (56.8%) were aged 14 years, and 51 adolescents (43.2%) were aged 15 years. Participant race/ethnicity reflected that of the practice (63 black adolescents [53.4%]; 46 white adolescents [38.9%]; 111 non-Hispanic adolescents [94.1%]). At baseline, 15 adolescents (12.7%) reported a history of sexual behavior and 16 adolescents (13.6%) reported a history of alcohol use. Intention-to- treat analyses found that 4 months after the intervention, adolescents in the sexual health intervention group reported a higher mean frequency score for PAC about sex compared with those ( ti d) Open Access. This is an open access article distributed under the terms of the CC-BY License. August 16, 2019 1/12 JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol Procedures Eligible parent-adolescent dyads were enrolled after providing informed parental consent and adolescent assent, and they each privately completed a telephone survey before the well care visit. The dyads were randomized into 1 of 3 groups: sexual health intervention group, alcohol prevention intervention group, or control group. Randomization was performed using a computer-generated random listing of the arms using a prespecified seed. Intervention group parents were instructed to arrive at the clinic 15 minutes early, taken to a quiet area in the waiting room away from their adolescent, and given intervention materials based on group assignment (ie, sexual health or alcohol prevention). Parents later joined their adolescent in the examination room. Parents in the control group received usual care. Two weeks after the well care visit, intervention group parents were contacted via telephone call. Research assistants (RAs) covered material that was not completed during the in-person health coaching visit, if applicable, and administered a short survey. If parents had not had a conversation with their adolescent about the intervention topic in the preceding 2 weeks, barriers were identified and RAs discussed how the materials might be used to overcome them. Four months after the well care visit, parents and adolescents in all groups were contacted to complete a final survey. Parents and adolescents who completed the baseline survey each received $10. Parents who completed the 2-week telephone survey received $5. Parents and adolescents who completed the final follow-up survey each received $20. Figure. Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 community based and does not include trainees, and clinician salaries are linked to patient volume. Parents were mailed an introductory letter and invited to contact the study team; telephone calls were placed to all parents who did not contact the study team. Families interested in participating were screened for eligibility. The Figure presents a diagram of participant recruitment. To be eligible, adults had to be the parent or legal guardian of the adolescent scheduled for the well care visit, planning to attend the appointment, and fluent in written and spoken English. For adolescents to be eligible, they had to be aged 14 to 15 years at the well care visit, an established practice patient, fluent in written and spoken English, able to complete study procedures, and not pregnant. JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolesce JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol community based and does not include trainees, and clinician salaries are linked to patient volume. Parents were mailed an introductory letter and invited to contact the study team; telephone calls were placed to all parents who did not contact the study team. Families interested in participating were screened for eligibility. The Figure presents a diagram of participant recruitment. To be eligible, adults had to be the parent or legal guardian of the adolescent scheduled for the well care visit, planning to attend the appointment, and fluent in written and spoken English. For adolescents to be eligible, they had to be aged 14 to 15 years at the well care visit, an established practice patient, fluent in written and spoken English, able to complete study procedures, and not pregnant. August 16, 2019 3/12 Procedures CONSORT Flow Diagram of Study Recruitment 196 Dyads assessed for eligibility 76 Excluded 33 Were ineligible 2 Withdrew 1 Was ineligible 1 Did not participate 20 Declined to participate 23 Were unavailable to contact 118 Dyads were randomized 38 Dyads in sexual health intervention group 36 Received clinic intervention 29 Received follow-up telephone call 2 Did not receive clinic intervention 9 Did not receive follow-up telephone call 12 Were lost to follow-up (9 parents, 11 teens) 1 Discontinued intervention (1 parent, 2 teens) 25 Completed the study (28 parents, 25 teens) 4 Were lost to follow-up (3 parents, 4 teens) 36 Completed the study (37 parents, 36 teens) 2 Were lost to follow-up (1 parent, 2 teens) 38 Completed the study (39 parents, 38 teens) 40 Dyads in alcohol prevention intervention group 39 Received clinic intervention 37 Received follow-up telephone call 1 Did not receive clinic intervention 3 Did not receive follow-up telephone call 40 Dyads in control group JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 (Reprinted) Figure. CONSORT Flow Diagram of Study Recruitment Figure. CONSORT Flow Diagram of Study Recruitment 118 Dyads were randomized August 16, 2019 3/12 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol JAMA Network Open \| Pediatrics Sexual Health Intervention Physician or nurse practitioner provides a direct endorsement and written prescription reinforcing the key messages in the clinic room with the parent and adolescent at the end of the visit. Interventions Alcohol Prevention Intervention Clinic Intervention Health coach talks with parent in lobby to discuss: Alcohol prevention brochure and handbookb (provided in written form, also available in electronic form) General communication handbook (provided in written form, also available in electronic form) Key Messages: PAC about alcohol use makes a difference PAC about alcohol can be hard Use materials to help you talk with your adolescent within 2 weeks Anticipated barriers to PAC and how to overcome them (ie, demonstrate how materials can be used) Research suggests it is important to convey these main points: Wait until you’re older to drink alcohol If you do drink alcohol, minimize risk of injury or harm Do not drive or get into a vehicle with someone who has been drinking Physician or nurse practitioner provides a direct endorsement and written prescription reinforcing key messages in clinic room with the parent and adolescent at the end of the visit Follow-up Telephone Call Follow-up telephone call to the parent at 2 weeks about PAC about alcohol, inquire about barriers, and review how materials can be used to overcome barriers. Box. Description of Parent-Targeted Sexual Health and Alcohol Prevention Interventions Delivered in Primary Care Pediatric Clinic Setting in Context of Adolescent Well Care Visits Follow-up Telephone Call Follow-up telephone call to the parent at 2 weeks about PAC about alcohol, inquire about barriers, and review how materials can be used to overcome barriers. Follow-up telephone call to the parent at 2 weeks to ask about PAC about sexual health, inquire about barriers, and review how materials can be used to overcome barriers. Abbreviation: PAC, parent-adolescent communication. a Content developed based on previous research identifying parental attitudes and beliefs that influence PAC about sexual health and adolescent attitudes and beliefs that influence sexual behaviors. a Content developed based on previous research identifying parental attitudes and beliefs that influence PAC about sexual health and adolescent attitudes and beliefs that influence sexual behaviors. b Content developed based on previous research identifying parental attitudes and beliefs that influence PAC about alcohol and adolescent attitudes and beliefs that influence alcohol use. b Content developed based on previous research identifying parental attitudes and beliefs that influence PAC about alcohol and adolescent attitudes and beliefs that influence alcohol use. August 16, 2019 4/12 August 16, 2019 4/12 Interventions Intervention materials were adapted from parent-targeted interventions shown to influence PAC and adolescent sexual behavior or alcohol use. Interventions selected were informed by extensive theoretically grounded research identifying parental attitudes and beliefs affecting communication with adolescents about sexual health or alcohol use and adolescent attitudes and beliefs affecting sexual behaviors or alcohol use.14,15,17,18 Further, selected interventions could be pragmatically adapted for use in busy primary care settings. The intervention process included a health coach discussing written materials with parents in the clinic lobby during a well care visit, conveying key messages, and encouraging PAC within 2 weeks (Box). This was followed by a brief direct verbal and written endorsement of the intervention from the adolescent’s clinician and a 2-week follow-up telephone call from the health coach to the parent. Health coaches were college graduates without specific health care training. Written materials were adapted to parents of our targeted age group, provided local data on sexual behaviors or alcohol use, and showed visual images reflecting the racial/ethnic composition of Box. Description of Parent-Targeted Sexual Health and Alcohol Prevention Interventions Delivered in Primary Care Pediatric Clinic Setting in Context of Adolescent Well Care Visits Sexual Health Intervention Clinic Intervention Health coach talks with parent in lobby to discuss: Sexual health brochure and handbooka (provided in written form, also available in electronic form) General communication handbook (provided in written form, also available in electronic form) Key Messages: PAC about sexual health makes a difference PAC about sexual health can be hard Use materials to help you talk with your adolescent within 2 weeks Anticipated barriers to PAC and how to overcome them (ie, demonstrate how materials can be used) Research suggests it is important to convey these main points: Wait until you’re older to have sex If you do have sex, use condoms If you do have sex, go to a physician or clinic to talk about sexually transmitted infections Physician or nurse practitioner provides a direct endorsement and written prescription reinforcing the key messages in the clinic room with the parent and adolescent at the end of the visit. Follow-up Telephone Call Follow-up telephone call to the parent at 2 weeks to ask about PAC about sexual health, inquire about barriers, and review how materials can be used to overcome barriers. Intervention Feasibility and Acceptability Feasibility was assessed by measuring length of time the RA spent with the parent (RA report) and checklists to document content delivered (ie, RA report, parent report at 2-week follow-up, and clinician report). Acceptability was assessed by final surveys. Parents were asked about general helpfulness and intentions to use materials over the next 12 months (ie, use with the adolescent participant, use with other children, or give to other parents). Adolescents were asked how helpful they thought the materials were to their parent. Clinicians received an anonymous web-based survey asking open-ended questions about implementing the study in their clinic and using these interventions in real life if they were found to be effective. Communication There were 4 measures for reported PAC on the final survey. Quality of PAC was assessed by the 20-item Parent-Adolescent Communication Scale,19 in which a higher score indicates better PAC. Adolescents answered similar questions with minor word changes. Frequency of PAC about sex was measured with an item asking parents and adolescents how much they had talked about sex since the adolescent’s last well care visit using a 4-point Likert-type response categories with 1 indicating not at all or never and 4 indicating a lot or often. Frequency of PAC about alcohol was measured with a similar item asking parents and adolescents how much they had talked about alcohol since the adolescent’s last well care visit. Communication about a specific safety plan was measured by asking parents and adolescents about PAC since the adolescent’s last well care visit about strategies to avoid getting in a car being driven by someone who has been drinking, with responses coded as no communication or some PAC. Sociodemographic Characteristics Adolescents’ age was calculated based on date of birth; sex at birth, race, and ethnicity were measured by self-report. Parental age, marital status, and highest level of education were measured by self-report. Adolescent Behaviors Adolescent behaviors were measured by self-report at baseline. Sexual behavior was assessed by 4 items: “In your lifetime, have you ever engaged in vaginal sexual intercourse? …had anal sexual intercourse? …given oral sex? …received oral sex?” Alcohol use was assessed with the question, “Have you ever had a drink of alcohol, other than a few sips?” Adolescents were coded as having engaged in adolescent risk behavior if they responded yes to any of these items. Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 the local community and of gender-nonconforming adolescents and parents. Materials included colorful spiral-bound handbooks with discussion guides and activity sheets on general PAC and a similarly formatted handbook and brochure focused on PAC about sexual health or alcohol use prevention; written materials were also available electronically. JAMA Network Open \| Pediatrics Parent-Targeted Interventions for P JAMA Network Open \| Pediatrics JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol the local community and of gender-nonconforming adolescents and parents. Materials included colorful spiral-bound handbooks with discussion guides and activity sheets on general PAC and a similarly formatted handbook and brochure focused on PAC about sexual health or alcohol use prevention; written materials were also available electronically. August 16, 2019 5/12 Statistical Analysis Adolescent participants were evenly split by sex (60 [50.9%] girls), and almost evenly split by age (67 adolescents [56.8%] aged 14 years; 51 adolescents [43.2%] aged 15 years). Race and ethnicity reflected practice demographics (63 black Table 1. Demographic Characteristics and Baseline Comparison by Experimental Group Participant Characteristic No. (%) Total (N = 118) Sexual Health Intervention (n = 38) Alcohol Prevention Intervention (n = 40) Control (n = 40) Parent Women 112 (94.9) 35 (92.1) 39 (97.5) 38 (95.0) Family structure Married or living with partner 91 (77.1) 30 (79.0) 29 (72.5) 32 (80.0) Other 27 (22.9) 8 (21.1) 11 (27.5) 8 (20.0) Highest level of education High school, GED, or some college 21 (17.8) 8 (21.1) 9 (22.5) 4 (10.0) Associate or 4-y degree 36 (30.5) 10 (26.3) 12 (30.0) 14 (35.0) Master’s or doctoral degree 61 (51.7) 20 (52.6) 19 (47.5) 22 (55.0) Adolescent Girls 60 (50.9) 18 (47.4) 21 (52.5) 21 (52.5) Age, y 14 67 (56.8) 25 (65.8) 23 (57.5) 19 (47.5) 15 51 (43.2) 13 (34.2) 17 (42.5) 21 (52.5) Race Black 63 (53.4) 19 (50.0) 24 (60.0) 20 (50.0) White 46 (38.9) 16 (42.1) 13 (32.5) 17 (42.5) Other 9 (7.6) 3 (7.9) 3 (7.5) 3 (7.5) Ethnicity Hispanic 6 (5.1) 2 (5.3) 1 (2.5) 3 (7.5) Non-Hispanic 111 (94.1) 36 (94.7) 38 (95.0) 37 (92.5) Unsure 1 (0.9) 0 1 (2.5) 0 History of sexual behaviora Yes 15 (12.7) 4 (10.5) 4 (12.5) 6 (15.0) No 103 (87.3) 34 (89.5) 35 (87.5) 34 (85.0) History of alcohol useb Yes 16 (13.6) 4 (10.5) 4 (10.0) 8 (20.0) No 102 (86.4) 34 (89.5) 36 (90.0) 32 (80.0) Abbreviation: GED, general equivalency diploma. a Assessed by reporting to have engaged in vaginal sexual intercourse, engaged in anal sexual intercourse, given oral sex, or received oral sex in their lifetime. b Assessed with the question “Have you ever had a drink of alcohol, other than a few sips?” Table 1. Demographic Characteristics and Baseline Comparison by Experimental Group Participant Characteristic No. Statistical Analysis We tested for differences in sociodemographic characteristics between participants who completed the follow-up survey and those who were lost to follow-up. We examined frequencies of parent- and adolescent-reported PAC and evaluated bivariate associations with adolescent sex, age, race/ethnicity, and risk behaviors using analysis of variance. Unadjusted and adjusted models were estimated. Covari- ates in the adjusted models were adolescent age, sex, race/ethnicity, and sexual behavior or alcohol use. Generalized linear models were conducted to compare differences between each intervention group vs the control group in quality of PAC and frequency of PAC about sex or alcohol. Depending on the distri- bution of the outcome variable being analyzed, different distributions and link functions were specified for the generalized linear models: linear to estimate outcome means, bivariate and adjusted, with 95% JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol JAMA Network Open \| Pediatrics Parent-Targeted Interventions for P JAMA Network Open \| Pediatrics JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol CIs or log-binomial to estimate unadjusted and adjusted risk ratios (RRs) with 95% CIs. We conducted similar analyses specifically for safety plan PAC content. Analyses were performed separately for adolescents and parents. Data were analyzed using an intent-to-treat principle and multiple imputation. To account for missing outcomes, multiple imputation with 100 imputations was used, analyses were conducted on each of the 100 imputed data set, and the parameter estimates across the data sets were combined to produce a unique point estimate and SE taking into account the uncertainty of the imputation process. Statistical analysis was performed using SAS statistical software version 14.2 (SAS Institute). Our sample size of 40 participants per group had 80% power to detect a difference in 2 population means corresponding to a Cohen d of 0.625, using a 2-sided P value of less than .05 for statistical significance. Further details on the trial protocol and statistical analysis can be found in Supplement 1. A total of 118 parents and 118 adolescents participated. Among parents, 112 were women (94.9%), and the mean (SD) age was 45.8 (6.9) years (Table 1). Statistical Analysis (%) Total (N = 118) Sexual Health Intervention (n = 38) Alcohol Prevention Intervention (n = 40) Control (n = 40) Parent Women 112 (94.9) 35 (92.1) 39 (97.5) 38 (95.0) Family structure Married or living with partner 91 (77.1) 30 (79.0) 29 (72.5) 32 (80.0) Other 27 (22.9) 8 (21.1) 11 (27.5) 8 (20.0) Highest level of education High school, GED, or some college 21 (17.8) 8 (21.1) 9 (22.5) 4 (10.0) Associate or 4-y degree 36 (30.5) 10 (26.3) 12 (30.0) 14 (35.0) Master’s or doctoral degree 61 (51.7) 20 (52.6) 19 (47.5) 22 (55.0) Adolescent Girls 60 (50.9) 18 (47.4) 21 (52.5) 21 (52.5) Age, y 14 67 (56.8) 25 (65.8) 23 (57.5) 19 (47.5) 15 51 (43.2) 13 (34.2) 17 (42.5) 21 (52.5) Race Black 63 (53.4) 19 (50.0) 24 (60.0) 20 (50.0) White 46 (38.9) 16 (42.1) 13 (32.5) 17 (42.5) Other 9 (7.6) 3 (7.9) 3 (7.5) 3 (7.5) Ethnicity Hispanic 6 (5.1) 2 (5.3) 1 (2.5) 3 (7.5) Non-Hispanic 111 (94.1) 36 (94.7) 38 (95.0) 37 (92.5) Unsure 1 (0.9) 0 1 (2.5) 0 History of sexual behaviora Yes 15 (12.7) 4 (10.5) 4 (12.5) 6 (15.0) No 103 (87.3) 34 (89.5) 35 (87.5) 34 (85.0) History of alcohol useb Yes 16 (13.6) 4 (10.5) 4 (10.0) 8 (20.0) No 102 (86.4) 34 (89.5) 36 (90.0) 32 (80.0) August 16, 2019 6/12 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex JAMA Network Open \| Pediatrics JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol adolescents [53.4%], 46 white adolescents [38.9%]; 111 non-Hispanic adolescents [94.1%]). At baseline, 15 adolescents (12.7%) reported a history of sexual behavior and 16 adolescents (13.6%) reported a history of drinking more than a few sips of alcohol. adolescents [53.4%], 46 white adolescents [38.9%]; 111 non-Hispanic adolescents [94.1%]). At baseline, 15 adolescents (12.7%) reported a history of sexual behavior and 16 adolescents (13.6%) reported a history of drinking more than a few sips of alcohol. The study was completed by 104 parents (88.1%) and 99 adolescents (83.9%). Intervention Influence on Reported Qualtiy of PAC and Topic-Specific PAC Neither intervention influenced parent-reported quality of PAC, frequency of PAC about sex, or frequency of PAC about alcohol at the 4-month follow-up survey (Table 2). Adolescents with parents in the sexual health intervention group reported a higher mean score for frequency of PAC about sex compared with the control group (unadjusted: 2.32 [95% CI, 1.97-2.66] vs 1.79 [95% CI, 1.50-2.08]; P = .02; adjusted: 2.22 [95% CI, 1.84-2.60] vs 1.75 [95% CI, 1.45-2.05]; P = .05). Adolescents with parents in the sexual health intervention group did not report higher scores for quality of PAC or frequency of PAC about alcohol. Adolescents with parents in the alcohol prevention intervention group reported a higher mean score for frequency of PAC about alcohol during the 4 months after their last well care visit when compared with the control group (unadjusted: 2.93 [95% CI, 2.60-3.25] vs 2.40 [95% CI, 2.08-2.72]; P = .03; adjusted: 2.94 [95% CI, 2.56-3.29] vs 2.42 [95% CI, 2.09-2.75]; P = .03); they did not report higher frequency of general PAC or PAC about sex (Table 2). Statistical Analysis Fewer parents and adolescents in the sexual health intervention group were available for the 4-month interview (25 of 38 dyads [66%]) than the alcohol prevention intervention group (36 of 40 dyads [90%]) (P = .01) or the control group (38 of 40 dyads [95%]) (P = .01) (overall χ2 = 13.97; P = .001). Sociodemographic and behavior characteristics were similar between dyads who completed and those who did not complete the study. Bivariate Associations Between Demographic Characteristics and Communication Responses to quality of PAC were summed into an index ranging from 41 to 96 for parents (α = .84) and 43 to 96 for adolescents (α = .87). Adolescent-reported quality of PAC varied by adolescent age; younger adolescents reported a significantly higher mean (SD) score for quality of PAC than older adolescents (72.3 [9.7] vs 66.9 [9.3]; P = .009). Parents of younger adolescents reported a higher mean (SD) score for frequency of PAC about sex compared with parents of older adolescents (2.7 [0.9] vs 2.2 [0.9]; P = .008). Parents of black adolescents reported a higher mean (SD) frequency score for PAC about sex compared with parents of white adolescents (2.7 [1.0] vs 2.1 [0.7]; P = .001). Black adolescents also reported a higher mean (SD) score for frequency of PAC about sex compared with white adolescents (2.2 [1.0] vs 1.8 [0.7]; P = .05). We found no significant bivariate associations of adolescent sex with parent-reported or adolescent-reported quality of PAC, frequency of PAC about sex, or frequency of PAC about alcohol. We found no significant bivariate associations of adolescent risk behaviors with parent-reported or adolescent-reported quality of PAC, frequency of PAC about sex, or frequency of PAC about alcohol. August 16, 2019 7/12 Feasibility and Acceptability e Assessed using a Likert scale for single items with response range from 1, indicating not at all or never, to 4, a lot or often. Table 3. Effect of Parent-Targeted Interventions on Parent-Reported and Adolescent-Reported PAC About Specific Safety Strategy Plan Communication RR (95% CI) Sexual Health Intervention (n = 38) Alcohol Prevention Intervention (n = 40) Control (n = 40) Parent Report Reported PAC about avoidance of getting into car with someone who has been drinking alcohol, No. (%) 24 (63) 24 (62) 13 (33) Unadjusted 1.79 (1.25-2.32)b 1.84 (1.25-2.42)c 1 [Reference] Adjusteda 1.75 (1.22-2.29)b 2.02 (1.39-2.65)c 1 [Reference] Adolescent Report Reported PAC about avoidance of getting into car with someone who has been drinking alcohol, No. (%) 27 (71) 27 (68) 17 (42) Unadjusted 1.60 (1.15-2.06)b 1.73 (1.25-2.22)c 1 [Reference] Adjusteda 1.64 (1.20-2.09)b 1.76 (1.22-2.29)c 1 [Reference] Abbreviations: PAC, parent-adolescent communication; RR, risk ratio. a Multivariable analyses are adjusted for adolescent sex, age, race, and baseline risk characteristics; analyses were stratified by reporter. b Statistically significant difference between sexual health intervention group and control group (P = .001). c Statistically significant difference between alcohol prevention intervention group and control group (P = .001). JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 (Reprinted) August 16, 2019 8/12 ded From: https://jamanetwork com/ by a University of Edinburgh Library User on 09/26/2019 Table 2. JAMA Network Open \| Pediatrics JAMA Network Open \| Pediatrics Feasibility and Acceptability Parents spent a median of 10 (range, 7-24) minutes in coaching sessions. Among 78 parents in the 2 intervention groups, 75 (96%) received an entire health coaching session in the clinic and 74 (95%) reported receiving an in-person clinician endorsement, which was consistent with clinician reports. We were able to contact 66 intervention parents (85%) for the 2-week follow-up call. Sixty-two parents in the intervention groups rated the intervention materials as moderately or very helpful (80%), 66 parents (85%) reported they would probably or definitely refer to materials Sixty-two parents in the intervention groups rated the intervention materials as moderately or very helpful (80%), 66 parents (85%) reported they would probably or definitely refer to materials Table 2. Effect of Parent-Targeted Interventions on Parent-Reported and Adolescent-Reported Quality of PAC and Frequency of PAC About Sexual Health and Alcohol Use Reported Communication Sexual Health Intervention (n = 38) Alcohol Prevention Intervention (n = 40) Control (n = 40), Mean Score (95% CI) Mean Score (95% CI) P Valuea Mean Score (95% CI) P Valueb Parent Report Quality of PACc Unadjusted 77.48 (73.89-81.06) .98 77.27 (74.06-80.48) .95 77.41 (74.32-80.50) Adjustedd 76.49 (72.53-80.45) .80 76.29 (72.71-79.87) .72 77.11 (73.93-80.30) Frequency of sex PACe Unadjusted 2.69 (2.37-3.01) .19 2.28 (1.99-2.58) .57 2.40 (2.11-2.69) Adjustedd 2.73 (2.40-3.07) .18 2.29 (1.99-2.60) .40 2.46 (2.18-2.73) Frequency of alcohol PACe Unadjusted 2.50 (2.18-2.82) .96 2.58 (2.30-2.85) .73 2.51 (2.24-2.78) Adjustedd 2.61 (2.25-2.96) .77 2.66 (2.35-2.97) .57 2.54 (2.26-2.82) Adolescent Report Quality of PACc Unadjusted 72.30 (68.51-76.09) .13 69.54 (66.37-72.72) .65 68.50 (65.39-71.61) Adjustedd 70.90 (66.81-74.99) .32 68.71 (65.22-72.21) .87 68.33 (65.19-71.47) Frequency of sex PACe Unadjusted 2.32 (1.97-2.66) .02 2.09 (1.80-2.39) .16 1.79 (1.50-2.08) Adjustedd 2.22 (1.84-2.60) .05 1.98 (1.65-2.31) .29 1.75 (1.45-2.05) Frequency of alcohol PACe Unadjusted 2.78 (2.39-3.18) .15 2.93 (2.60-3.25) .03 2.40 (2.08-2.72) Adjustedd 2.79 (2.34-3.24) .17 2.94 (2.56-3.29) .03 2.42 (2.09-2.75) Abbreviation: PAC, parent-adolescent communication. a P values are of sexual health intervention group vs control group. b P values are of alcohol prevention intervention group vs control group. c Quality of PAC was scored using the Parent-Adolescent Communication Scale in which a higher number indicated better PAC (range: parents, 41-96; adolescents, 43-96). d Multivariable analyses are adjusted for adolescent sex, age, race, and baseline risk characteristics. Analyses were stratified by reporter. Interventions’ Influence on Safety Strategy Communication Among 40 parents in the alcohol prevention intervention group, 25 (62.5%) reported that since their last well care visit they had communicated about a specific safety strategy plan to help adolescents avoid being in a car with a driver who had been drinking alcohol compared with 13 of 40 parents (33%) parents in the control group (RR, 1.84 [95% CI, 1.25-2.42]; P = .001; adjusted RR [aRR], 2.02 [95% CI, 1.39-2.65]; P = .001) (Table 3). Similarly, adolescents with parents in the alcohol prevention interven- tion group were significantly more likely to report these conversations compared with adolescents with parents in the control group (27 of 40 adolescents [68%] vs 17 of 40 adolescents [43%]; RR, 1.73 [95% CI, 1.25-2.22]; P = .001; aRR, 1.76 [95% CI, 1.22-2.29]; P < .001). Parents and adolescents in the sexual health intervention group were more likely to report PAC about a specific safety strategy plan com- pared with those in the control group (parents: 24 of 38 [63%] vs 13 of 40 [33%]; RR, 1.79 [95% CI, 1.25- 2.32]; P = .001; aRR, 1.75 [95% CI, 1.22-2.29]; P = .001; adolescents: 27 of 38 [71%] vs 17 of 40 [43%]; RR, 1.60 [95% CI, 1.15-2.06]; P = .001; aRR, 1.64 [95% CI, 1.20-2.09];P = .001). JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 August 16, 2019 7/12 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol Feasibility and Acceptability (%) 27 (71) 27 (68) 17 (42) Unadjusted 1.60 (1.15-2.06)b 1.73 (1.25-2.22)c 1 [Reference] Adjusteda 1.64 (1.20-2.09)b 1.76 (1.22-2.29)c 1 [Reference] Abbreviations: PAC, parent-adole communication; RR, risk ratio. a Multivariable analyses are adjus sex, age, race, and baseline risk analyses were stratified by repo b Statistically significant differenc health intervention group and c (P = .001). c Statistically significant differenc prevention intervention group (P = .001). JAMA Network Open 2019 2(8) e199535 doi 101001/jamanetworkopen 2019 9535 (Reprinted) Aug August 16, 2019 8/12 Feasibility and Acceptability Effect of Parent-Targeted Interventions on Parent-Reported and Adolescent-Reported Quality of PAC and Frequency of PAC About Sexual Health and Alcohol Use Reported Communication Sexual Health Intervention (n = 38) Alcohol Prevention Intervention (n = 40) Control (n = 40), Mean Score (95% CI) Mean Score (95% CI) P Valuea Mean Score (95% CI) P Valueb Parent Report Quality of PACc Unadjusted 77.48 (73.89-81.06) .98 77.27 (74.06-80.48) .95 77.41 (74.32-80.50) Adjustedd 76.49 (72.53-80.45) .80 76.29 (72.71-79.87) .72 77.11 (73.93-80.30) Frequency of sex PACe Unadjusted 2.69 (2.37-3.01) .19 2.28 (1.99-2.58) .57 2.40 (2.11-2.69) Adjustedd 2.73 (2.40-3.07) .18 2.29 (1.99-2.60) .40 2.46 (2.18-2.73) Frequency of alcohol PACe Unadjusted 2.50 (2.18-2.82) .96 2.58 (2.30-2.85) .73 2.51 (2.24-2.78) Adjustedd 2.61 (2.25-2.96) .77 2.66 (2.35-2.97) .57 2.54 (2.26-2.82) Adolescent Report Quality of PACc Unadjusted 72.30 (68.51-76.09) .13 69.54 (66.37-72.72) .65 68.50 (65.39-71.61) Adjustedd 70.90 (66.81-74.99) .32 68.71 (65.22-72.21) .87 68.33 (65.19-71.47) Frequency of sex PACe Unadjusted 2.32 (1.97-2.66) .02 2.09 (1.80-2.39) .16 1.79 (1.50-2.08) Adjustedd 2.22 (1.84-2.60) .05 1.98 (1.65-2.31) .29 1.75 (1.45-2.05) Frequency of alcohol PACe Unadjusted 2.78 (2.39-3.18) .15 2.93 (2.60-3.25) .03 2.40 (2.08-2.72) Adjustedd 2.79 (2.34-3.24) .17 2.94 (2.56-3.29) .03 2.42 (2.09-2.75) Abbreviation: PAC, parent-adolescent communication. a P values are of sexual health intervention group vs control group. b P values are of alcohol prevention intervention group vs control group. c Quality of PAC was scored using the Parent-Adolescent Communication Scale in which a higher number indicated better PAC (range: parents, 41-96; adolescents, 43-96). d Multivariable analyses are adjusted for adolescent sex, age, race, and baseline risk characteristics. Analyses were stratified by reporter. e Assessed using a Likert scale for single items with response range from 1, indicating not at all or never, to 4, a lot or often. Abbreviation: PAC, parent-adolescent communication. Table 3. Effect of Parent-Targeted Interventions on Parent-Reported and Adolescent-Reported PAC About Specific Safety Strategy Plan Communication RR (95% CI) Sexual Health Intervention (n = 38) Alcohol Prevention Intervention (n = 40) Control (n = 40) Parent Report Reported PAC about avoidance of getting into car with someone who has been drinking alcohol, No. (%) 24 (63) 24 (62) 13 (33) Unadjusted 1.79 (1.25-2.32)b 1.84 (1.25-2.42)c 1 [Reference] Adjusteda 1.75 (1.22-2.29)b 2.02 (1.39-2.65)c 1 [Reference] Adolescent Report Reported PAC about avoidance of getting into car with someone who has been drinking alcohol, No. Discussion These parent-targeted primary care-based interventions increased adolescent-reported frequency of PAC about sexual health and alcohol up to 4 months after a well care visit. Interventions were feasible to deliver and highly acceptable to participants. Mirman et al16 similarly adapted a parent- targeted teenage driver safety intervention focused on older patients in the same clinic, which also demonstrated feasibility, acceptability, and initial evidence for efficacy. All interventions were adapted from interventions that, in other settings, influence PAC and adolescent behaviors associated with unintended pregnancy, sexually transmitted infections, alcohol-related injury, or motor vehicle crashes.14,15,20 Together, this provides strong support for future research testing the influence of pragmatic primary care–based parent-targeted interventions on PAC and adolescent health outcomes. Several additional findings were notable. In stratified analyses, increases in frequency of PAC were reported by adolescents but not parents. This was an unexpected finding. Parents and adolescents may recognize or label PAC differently, especially about sensitive topics. Regardless, research has shown that adolescent perceptions of parental messages affect adolescent behavior, highlighting the importance of our adolescent-reported results.21-23 We observed that the reported frequency of PAC was higher for alcohol use than for sexual behavior, and more dyads in the alcohol prevention intervention and control groups were willing to complete follow-up surveys compared with those in our sexual health intervention group. This finding highlights that there are important nuances to take into consideration for PAC interventions about different topics (particularly for sensitive topics, such as sexual health) and suggests that tailoring choice of PAC interventions based on real or perceived value from the perspective of parents, adolescents, and clinicians should be considered. For example, instead of an age-based approach as used in this study, delivering PAC interventions linked to specific health topics could be informed by assessment of individual adolescent risk or adolescent or parent interest. Finally, compared with the control group, approximately 2-fold as many parents and adolescents in the intervention groups reported increased PAC about strategies for adolescents to avoid getting into a car with someone who has been drinking alcohol, although this specific content was only included in alcohol prevention intervention materials. This finding suggests that there may be broader beneficial effects of topic-specific parent- targeted interventions on PAC, which warrants further study. Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 during the next 12 months, 65 parents (83%) reported they would probably or definitely use materials with other children in their family, and 43 parents (55%) reported they would probably or definitely give materials to other parents during the next 12 months. Of the 78 adolescents in the intervention groups, 66 (85%) were aware that their parents had been given intervention materials; of these, 53 adolescents (80%) reported their parent showed them materials and 43 adolescents (65%) reported that the materials were moderately or very helpful to their parent. JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol during the next 12 months, 65 parents (83%) reported they would probably or definitely use materials with other children in their family, and 43 parents (55%) reported they would probably or definitely give materials to other parents during the next 12 months. Of the 78 adolescents in the JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol during the next 12 months, 65 parents (83%) reported they would probably or definitely use materials with other children in their family, and 43 parents (55%) reported they would probably or definitely give materials to other parents during the next 12 months. Of the 78 adolescents in the intervention groups, 66 (85%) were aware that their parents had been given intervention materials; of these, 53 adolescents (80%) reported their parent showed them materials and 43 adolescents (65%) reported that the materials were moderately or very helpful to their parent. Seven of 9 participating clinicians (78%) completed the feedback survey and reported materials were high quality, important, well received, and easy to disseminate. Clinicians expressed needs for additional resources (eg, staff or nurse time) to support the health coach role if interventions were to be provided in routine clinical care. August 16, 2019 9/12 Limitations This study has several limitations. It was conducted at a single site, which may limit generalizability. Parent-adolescent communication was measured by self-report and could be affected by recall or self-report biases. Clinicians were not blinded in this study, which may have influenced care to adolescents in all groups. A multisite study with additional measures of PAC and a longer follow-up period is needed to evaluate intervention impact on behaviors. Discussion It is important to note that we found these interventions feasible within the context of a research study, which supplied health coach staffing resources to provide the interventions as well as the 2-week follow-up telephone call to parents. Clinicians reported that resources for health coach staffing would be needed to implement similar strategies in real-life conditions. Our health coaches were trained in the use of materials and protocols but were not trained in health care, suggesting that it may be possible to consider a range of staffing models. Our study was intentionally designed to leverage the influence of person-to-person contact within trusted, credible clinic settings, but strategies to minimize staff time to do this could be tested (eg, text messages to replace the 2-week follow-up telephone call). Finally, future research showing that parent-targeted interventions August 16, 2019 9/12 JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 delivered in or through primary care settings effectively change adolescent behavior and improve adolescent health outcomes would justify reimbursement for staff time to deliver interventions. JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Pare delivered in or through primary care settings effectively change adolescent behavior and improve adolescent health outcomes would justify reimbursement for staff time to deliver interventions. JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol delivered in or through primary care settings effectively change adolescent behavior and improve adolescent health outcomes would justify reimbursement for staff time to deliver interventions. delivered in or through primary care settings effectively change adolescent behavior and improve adolescent health outcomes would justify reimbursement for staff time to deliver interventions. Conclusions Results suggest that there are innovative strategic opportunities for clinicians in primary care settings to join with parents to effectively achieve better health outcomes among sizable portions of adolescent patients.9 The paucity of existing primary care parent-targeted interventions needs to be addressed, with a focus on developing a portfolio of interventions that effectively address a range of adolescent health issues. Developing and evaluating such interventions, and developing feasible models for incorporating them into practice, are important avenues for future research. ARTICLE INFORMATION Accepted for Publication: June 25, 2019. Accepted for Publication: June 25, 2019. Published: August 16, 2019. doi:10.1001/jamanetworkopen.2019.9535 Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Ford CA et a JAMA Network Open. Corresponding Author: Carol A. Ford, MD, Perelman School of Medicine, Department of Pediatrics, Division of Adolescent Medicine, University of Pennsylvania, Children’s Hospital of Philadelphia, Buerger Building, 12th Floor, Room 12249, Philadelphia, PA 19104-4399 (fordc@email.chop.edu). Corresponding Author: Carol A. Ford, MD, Perelman School of Medicine, Department of Pediatrics, Division of Adolescent Medicine, University of Pennsylvania, Children’s Hospital of Philadelphia, Buerger Building, 12th Floor, Room 12249, Philadelphia, PA 19104-4399 (fordc@email.chop.edu). Author Affiliations: Perelman School of Medicine, Department of Pediatrics, Division of Adolescent Medicine, University of Pennsylvania, Children’s Hospital of Philadelphia, Philadelphia (Ford, García-España, Fisher Thiel, Friedrich, Salek); Center for Parent and Teen Communication, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania (Ford, Friedrich, Salek); Department of Psychology, University of Alabama at Birmingham, Birmingham (Mirman); Silver School of Social Work, New York University, New York (Jaccard). Author Contributions: Drs Ford and García-España had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: All authors. Acquisition, analysis, or interpretation of data: Ford, Mirman, García-España, Fisher Thiel. Drafting of the manuscript: Ford, García-España, Fisher Thiel, Salek, Jaccard. Drafting of the manuscript: Ford, García-España, Fisher Thiel, Salek, Jaccard. Critical revision of the manuscript for important intellectual content: Ford, Mirman, García-España, Fisher Thiel, Friedrich, Salek. Critical revision of the manuscript for important intellectual content: Ford, Mirman, García-España, Fisher Thiel, Friedrich, Salek. Statistical analysis: Ford, García-España, Jaccard. Obtained funding: Ford. Administrative, technical, or material support: Mirman, Fisher Thiel, Friedrich, Salek. Administrative, technical, or material support: Mirman, Fisher Thiel, Friedrich, Salek. Conflict of Interest Disclosures: Dr Ford reported receiving grants from the Health Resources and Services Administration during the conduct of the study. No other disclosures were reported. Funding/Support: Primary funding for this work was provided by the Children’s Hospital of Philadelphia. This project was supported in part by the Health Resources and Services Administration (HRSA) of the US Department of Health and Human Services (HHS) under the Leadership Education in Adolescent Health program (T71MC30798). JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 August 16, 2019 10/12 August 16, 2019 10/12 REFERENCES 1. Kann L, McManus T, Harris WA, et al. Youth risk behavior surveillance—United States, 2017. MMWR Surveill Summ. 2018;67(8):1-114. doi:10.15585/mmwr.ss6708a1 2. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System: prevalence data & data analysis tools. https://www.cdc.gov/brfss/data_tools.htm. Accessed December 17, 2018, 2018. 3. Copen CE, Chandra A, Febo-Vazquez I. Sexual behavior, sexual attraction, and sexual orientation among adults aged 18-44 in the United States: data from the 2011-2013 National Survey of Family Growth. Natl Health Stat Report. 2016;(88):1-14. 4. Banspach S, Zaza S, Dittus P, Michael S, Brindis CD, Thorpe P. CDC Grand Rounds: adolescence—preparing for lifelong health and wellness. MMWR Morb Mortal Wkly Rep. 2016;65(30):759-762. doi:10.15585/mmwr. mm6530a2 5. Burrus B, Leeks KD, Sipe TA, et al; Community Preventive Services Task Force. Person-to-person interventions targeted to parents and other caregivers to improve adolescent health: a community guide systematic review. Am J Prev Med. 2012;42(3):316-326. doi:10.1016/j.amepre.2011.12.001 6. Jones RK, Jerman J. Population group abortion rates and lifetime incidence of abortion: United States, 2008- 2014. Am J Public Health. 2017;107(12):1904-1909. doi:10.2105/AJPH.2017.304042 7. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance 2017: STDs in adolescents and young adults. https://www.cdc.gov/std/stats17/adolescents.htm. Accessed December 17, 2018. 8. Park MJ, Scott JT, Adams SH, Brindis CD, Irwin CE Jr. Adolescent and young adult health in the United States in the past decade: little improvement and young adults remain worse off than adolescents. J Adolesc Health. 2014; 55(1):3-16. doi:10.1016/j.jadohealth.2014.04.003 9. Harris SK, Aalsma MC, Weitzman ER, et al. Research on clinical preventive services for adolescents and young adults: where are we and where do we need to go? J Adolesc Health. 2017;60(3):249-260. doi:10.1016/j. jadohealth.2016.10.005 10. Ford CA, Cheek C, Culhane J, et al. Parent and adolescent interest in receiving adolescent health communication information from primary care clinicians. J Adolesc Health. 2016;59(2):154-161. doi:10.1016/j. jadohealth.2016.03.001 11. Patient-Centered Outcomes Research Institute. PCORI’s stakeholders. https://www.pcori.org/funding- opportunities/what-we-mean-engagement/pcoris-stakeholders. Accessed July 17, 2018. 12. Santa Maria D, Markham C, Bluethmann S, Mullen PD. Parent-based adolescent sexual health interventions and effect on communication outcomes: a systematic review and meta-analyses. Perspect Sex Reprod Health. 2015;47 (1):37-50. doi:10.1363/47e2415 13. Allen ML, Garcia-Huidobro D, Porta C, et al. Effective parenting interventions to reduce youth substance use: a systematic review. Pediatrics. 2016;138(2):e20154425. doi:10.1542/peds.2015-4425 14. Guilamo-Ramos V, Bouris A, Jaccard J, Gonzalez B, McCoy W, Aranda D. A parent-based intervention to reduce sexual risk behavior in early adolescence: building alliances between physicians, social workers, and parents. Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 JAMA Network Open \| Pediatrics Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol JAMA Network Open \| Pediatrics Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Disclaimer: This information or content and conclusions are those of the authors and should not be construed as the official position or policy of, nor should any endorsements be inferred by HRSA, HHS or the US government. Sexual health and alcohol prevention intervention materials were modified with permission from Dr Jaccard. Data Sharing Statement: See Supplement 2. Data Sharing Statement: See Supplement 2. Additional Contributions: Sharon Hunter, MSN, CRNP, and Wendy Ingersoll, MD (Children’s Hospital of Philadelphia Primary Care Practice at Chestnut Hill), provided feedback on conducting the study. James Massey, RN, MBA (Children’s Hospital of Philadelphia Pediatric Research Consortium), provided assistance in clinic selection and integrating the study with the clinic’s practice. Valerie McGoldrick, RN (Children’s Hospital of Philadelphia Pediatric Research Consortium), served as a research assistant. They were not compensated for their contributions. August 16, 2019 11/12 REFERENCES J Adolesc Health. 2011;48(2):159-163. doi:10.1016/j.jadohealth.2010.06.007 15. Turrisi R, Jaccard J, Taki R, Dunnam H, Grimes J. Examination of the short-term efficacy of a parent intervention to reduce college student drinking tendencies. Psychol Addict Behav. 2001;15(4):366-372. doi:10.1037/0893- 164X.15.4.366 JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 August 16, 2019 11/12 JAMA Network Open \| Pediatrics 16. Mirman JH, Goodman ES, Friedrich E, Ford CA. Talking with teens about traffic safety: Initial feasibility, acceptability, and efficacy of a parent-targeted intervention for primary care settings. J Safety Res. 2018;66: 113-120. doi:10.1016/j.jsr.2018.06.008 17. Fishbein M, Ajzen I. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley; 1975. 18. Fishbein M, Ajzen I. Predicting and Changing Behavior: The Reasoned Action Approach. New York, NY: Psychology Press; 2010. 19. Barnes HL, Olson DH. Parent-adolescent communication and the Circumplex Model. Child Dev. 1985;56: 438-447. doi:10.2307/1129732 20. Mirman JH, Curry AE, Winston FK, et al. Effect of the teen driving plan on the driving performance of teenagers before licensure: a randomized clinical trial. JAMA Pediatr. 2014;168(8):764-771. doi:10.1001/ jamapediatrics.2014.252 21. Jaccard J, Dittus PJ. Adolescent perceptions of maternal approval of birth control and sexual risk behavior. Am J Public Health. 2000;90(9):1426-1430. doi:10.2105/AJPH.90.9.1426 22. Dittus PJ, Jaccard J. Adolescents’ perceptions of maternal disapproval of sex: relationship to sexual outcomes. J Adolesc Health. 2000;26(4):268-278. doi:10.1016/S1054-139X(99)00096-8 23. Jaccard J, Dodge T, Dittus P. Parent-adolescent communication about sex and birth control: a conceptual framework. New Dir Child Adolesc Dev. 2002;(97):9-41. doi:10.1002/cd.48 August 16, 2019 12/12 JAMA Network Open. 2019;2(8):e199535. doi:10.1001/jamanetworkopen.2019.9535 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Parent-Targeted Interventions for Parent-Adolescent Communication About Sex and Alcohol Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019 Downloaded From: https://jamanetwork.com/ by a University of Edinburgh Library User on 09/26/2019
https://openalex.org/W3161335305	https://www.researchsquare.com/article/rs-20026/latest.pdf	English	null	Development of Statistical Literacy and Scientific Reasoning &amp; Argumentation in Physicians	Research Square (Research Square)	2,020	cc-by	6,359	Development of Statistical Literacy and Scientific Reasoning & Argumentation in Physicians Felicitas Maria Schmidt (  Felicitasmariaschmidt@gmail.com ) Ludwig-Maximilians-Universitat Munchen Medizinische Fakultat https://orcid.org/0000-0001-8284- 5549 Jan Zottmann Ludwig-Maximilians-Universitat Munchen Medizinische Fakultat Maximilian Sailer Universitat Passau Martin R Fischer Ludwig-Maximilians-Universitat Munchen Medizinische Fakultat Markus Berndt Ludwig-Maximilians-Universitat Munchen Medizinische Fakultat Research article Keywords: evidence-based practice, scientific reasoning and argumentation, statistical literacy, medical education, postgraduate medical education Posted Date: April 8th, 2020 DOI: https://doi.org/10.21203/rs.3.rs-20026/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Research article Keywords: evidence-based practice, scientific reasoning and argumentation, statistical literacy, medical education, postgraduate medical education License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License License:   This work is licensed under a Creative Commons Attribution 4.0 International License. R d F ll Li Page 1/16 Page 1/16 Abstract Background: Statistical literacy (SL) of physicians, i.e. the ability to understand and apply statistics, is an essential prerequisite for risk estimation and communication. Together with scientific reasoning and argumentation (SRA) skills, SL provides the basis for evidence-based practice. Several studies suggest that in medical students both skills are developed merely on a medium level. Methods: The aim of the present study was to investigate these skills in practicing physicians ( N = 71, M Age = 40.00, SD = 9.59) and when and how these skills were acquired. Biographical data was collected with an online survey tool. SL was assessed with multiple-choice items. SRA skills evidence evaluation (EE) and drawing conclusions (DC) were measured with a decision scenario. Results: Study results indicated that physicians have medium to high levels of SL ( M = 17.58, SD = 6.92, max 30 pts.) and SRA (EE-Score: M = 7.75, SD = 1.85, max 10 pts.; DC-Score: M = 37.20, SD = 5.35, max 60 pts.). Skills development via autodidactic learning activities ( M = 4.78, SD = 1.125, range 1-6) was reported significantly more often than development during medical education ( M = 2.31, SD = 1.456, t (71) = -9.915, p < .001) or in extracurricular activities ( M = 3.34, SD = 1.869, t (71) = 4.673, p = .000). The active involvement in research seemed decisive: The number of publications and time spent in research predicted SL to a large extent (β = .355, p = .002; β = .280, p = .018). SRA skills were associated with the type of doctoral thesis (β = -.380 ± .154, p = .016) and working in research (β = 3.355 ± 1.229, p = .008). Conclusion : The development of SL and SRA skills needs to be systematically fostered during medical education combined with active involvement in research. 1. Background In a patient-physician relationship on equal footing, transparent information about risks is key. Statistical literacy (SL) of the physician is a prerequisite for effective risk communication (1, 2). Statistical literacy is not only the ability to understand statistical information, but also to apply it in decision-making (3). It comprises the aptitude of critical reflection about statistics as evidence in arguments (4) and is intertwined with scientific reasoning and argumentation skills (SRA), to provide the basis for evidence- based decision-making (5, 6). However, a collective statistical illiteracy has been observed among physicians (3, 7) and SRA skills needed for evidence-based practice (8), such as evidence evaluation (EE) or drawing conclusions (DC), are underdeveloped (3, 9). The aim of the present study was thus to gain insights into the status quo of SL and SRA skills in medical doctors (physicians), and to identify influencing factors that should be considered in focused educational training to foster these skills. Statistical literacy and Scientific reasoning & argumentation Statistical literacy and Scientific reasoning & argumentation Page 2/16 Page 2/16 Statistical literacy is based on numeracy, the aptitude of mathematical operations (10), and encompasses the use and interpretation of statistics and its results (11, 12) and(Walker, 1951) the ability to explain and critically evaluate them (13, 14). In a medical context, it ranges from relative risks and conditional probabilities (15, 16), over test constructs like sensitivity and specificity (3), to complex inferential statistical procedures applied in empirical research studies (17-19). Statistical literacy is based on numeracy, the aptitude of mathematical operations (10), and encompasses the use and interpretation of statistics and its results (11, 12) and(Walker, 1951) the ability to explain and critically evaluate them (13, 14). In a medical context, it ranges from relative risks and conditional probabilities (15, 16), over test constructs like sensitivity and specificity (3), to complex inferential statistical procedures applied in empirical research studies (17-19). Several existing instruments for the assessment of SL have been employed in a variety of studies. Domain-general assessment tools, such as the Berlin Numeracy Test (20), or the Three Item Test (7) are focusing mainly on basic numeracy and are considered less reliable in highly educated samples, such as physicians, due to ceiling effects (20, 21). Such effects have not been observed in domain-specific tests, such as the Obstetrician-Gynecologist Statistical Literacy Questionnaire (OGSLQ) (9) which aims at assessing clinically relevant numerical facts, concepts, and relations in physicians, or the Medical Data Interpretation Test assessing the aptitude to weigh treatment alternatives based on evidence (22). In a study by Anderson et al. (2014), 52% of the participating physicians answered two (or fewer) of four questions regarding statistical concepts correctly (9). This is in line with findings by Windish and colleagues (2007) reporting only 40% of resident physicians demonstrating adequate understanding of biostatistical concepts (23). Similar results were found by Gigerenzer and Wegwarth (2008), showing that 33% of gynecologists were not aware of the benefits of mammography screening, with 79% being unable to interpret the positive predictive value (PPV) (24). Gigerenzer and colleagues (2008) summarized various studies on the concept of PPV and its dependence on the prevalence, finding that 50% of participants were under the impression that false positive test results in HIV testing do not exist. They also found that only 2 of 20 urologists have sufficient knowledge about the reliability of a PSA-test. Statistical literacy and Scientific reasoning & argumentation Overall, physicians´ SL is not below-average (9). However, it can be considered comparable to other educated samples (7, 25) and was found superior to that of residents in research training (23) or medical students (26). The few studies that have assessed medical students’ SL are supporting the findings of superiority of physicians (27). The authors (2020) compared medical students to those of social sciences and economics and found that medical students in their first years of study scored higher in comparison to social science students and comparable to economics students (28). This study not only assessed SL, but also the two skills: evidence evaluation (EE) and drawing conclusions (DC), as part of the conceptual framework for scientific reasoning and argumentation (SRA) by Fischer and colleagues (2014). Based on this framework, SRA can be defined as the competence of comprehending and applying scientific working methods and their results when solving problems (29, 30). The intertwining of SRA and SL was assumed by Anderson et al. (2013) stating that the latter is needed to evaluate scientific evidence (31) and by Franklin et al. (2005), who stated that SL itself encompasses also SRA skills, like formulating questions and answering them on an informed basis (32). In a medical context, possible links between SL and SRA have been examined with the Medical Data Interpretation Test (22). Participating physicians scored higher overall than others with postgraduate degrees (89 out of 100 score points). Johnson et al. (2014) assessed numeracy of medical students and residents and found students with poor numeracy being more likely to misjudge risks of different Page 3/16 Page 3/16 treatment alternatives and that the confidence in treatment recommendation increased during medical school (26). In the study by the authors (2020), the skills did not improve over the course of undergraduate studies (28). In summary, SL and SRA seem to be underdeveloped in medical students and better developed in physicians. There are indicators that the development of SL and SRA does not necessarily happen within formal education and the question remains how, where, and when this development occurs. This study should contribute to the discussion when and how to best foster SL and SRA skills in lifelong learning of physicians. It aimed at providing further insights into these skills in physicians and to identify demographic factors and learning opportunities that may be associated with the development of these skills. 2.2. Test instrument / Measures For the assessment of SL and SRA skills, we used a tool previously developed in the context of a study with medical students (28). The instrument combines multiple choice items to assess SL with a decision scenario to assess the participants´ skills in EE and DC (SRA skills) (33). We added items on relevant demographic factors. Statistical literacy and Scientific reasoning & argumentation Our guiding research questions were: 2.1. Design and sample Our study followed a quasi-experimental, causal-comparative design with two dependent variables: SL and SRA. We included N = 71 German and Austrian physicians (31 female, 34 male) from different work settings and locations in our study: hospital (N = 43), outpatient sector (N = 2), research (N = 16). A doctoral thesis was completed by 58 and 9 participants were currently working on it. The mean age of participants was 40 years (SD = 9.59, range = 26–65) (Table 1). 1. SL and SRA 1. To what extent are SL and SRA skills developed in physicians? 1. To what extent are SL and SRA skills developed in physicians? 2. To what extent does SL predict SRA skills of physicians? 2. Education and Development: 1. How, where, and when do physicians develop SL and SRA skills? 1. How, where, and when do physicians develop SL and SRA skills? Which demographic factors are related to the development of SL and SRA? 2. Which demographic factors are related to the development of SL and SRA? 2.2.1. Demography Demographic parameters of the participants were assessed with a special interest in their working history and environment (hospital, out-patient care, research). Questions were adapted from a study by Epstein Page 4/16 Page 4/16 and colleagues (34) and comprised multiple choice items, some with the opportunity to fill in additional free text; Five items on the doctoral thesis (qualifications fostered meanwhile), three items on the professional career, two items on the publication record (type of authorship, number of publications), and three items on the current job description. 2.2.3. SRA skills The assessment of SRA skills focused on the two epistemic activities EE and DC being assessed with a decision-making scenario in a medical context (general medicine, out-patient care) and provided a separate overall score for EE and DC (maximum score 10 and 60, respectively, Cronbach's α for EE items .87; for DC items .74). For EE, participants had to evaluate four pieces of evidence in terms of their scientificness (35–37) including one authentic pharmaceutical brochure. Then, the participants rated the persuasiveness (Likert 1–6) of 20 arguments, which were extracted from the presented evidence and assigned a level of argument strength from 1 (lowest) to 4 (highest). The participants’ ratings for scientificness were compared to the ratings of scientificness by the authors, resulting in a measure of similarity for EE and DC. The range of these measures was from 0–10 (EE-Score) and 0–60 (DC score) with zero indicating no similarity. 2.3. Procedure The study was completed by the participants either online with LamaPoll (https://www.lamapoll.de/; accessed March 22, 2020), a survey tool optimized for mobile applications, or with pen and paper (return rate 16.5% online and 66.7% pen and paper). Average duration was 45 minutes. Participants were invited l l d l 2.4. Statistical Procedures and Analyses Statistical analysis was performed with IBM SPSS 25. Descriptive and frequency data were computed for primary analysis and Cronbach´s alpha for internal reliability. Extensive outlier analyses were conducted and all required prerequisites for statistical analyses, such as normal distribution and homoscedasticity of variables, were tested. One-factorial ANOVAs and linear regression models were calculated to assess the association of demographic factors on SL and SRA. P values less than .05 were considered significant. Data in natural verbal language underwent thematic analysis to extract common themes. 2.2.2. Statistical Literacy Statistical literacy was measured with multiple choice items based on validated instruments (9, 20, 22), to assess a broad spectrum from basic numeracy to conditional probabilities and statistical concepts. Reliability of the SL test was .82 (Cronbach's α) in our sample with a maximum score of 30 points. All items were framed in a medical context; however, no medical content knowledge was necessary to answer them correctly. 3.1. Statistical literacy and SRA The 71 physicians' average score in SL was M = 17.58, SD ± 6.92, with a range of 5 to 30 out of 30 attainable points. On average, physicians evaluated the evidences concordantly with the authors' evaluation, EE-Score M = 7.75, SD = 1.85. The ratings for argument quality were in accordance with the authors' rating, DC-Score M = 37.20, SD = 5.35. Statistical literacy and DC were significantly inversely correlated (DC-Score r(71) = − .272, p = .022). However, no correlation was found between SL and EE, r(71) = .198, p = .098, nor EE and DC, r(71) = .138, p = .256. 3. Results Page 5/16 Page 5/16 Page 5/16 We included 71 complete cases (see Table 1). For 13 participants, we imputed missing values in the evaluation of the 20 arguments from the average of the respective item. The entire data set was checked for univariate outliers, skewness and kurtosis for all variables was within the ± 2 range (38). The prerequisites for ANOVA were fulfilled, all variables were normally distributed and homoscedastic. We included 71 complete cases (see Table 1). For 13 participants, we imputed missing values in the evaluation of the 20 arguments from the average of the respective item. The entire data set was checked for univariate outliers, skewness and kurtosis for all variables was within the ± 2 range (38). The prerequisites for ANOVA were fulfilled, all variables were normally distributed and homoscedastic. 4.1. Research question 1 We found average SL of physicians (59%), a rather high-level EE score (77%) and a medium-level DC score (62%). SL did not predict the SRA skills of physicians´ which could indicate that these are independent concepts. Due to the focus on SL rather than the combination of basic numeracy and SL(20), the test discriminated well and we did not find any ceiling effects as observed in other educated samples (7, 21). A comparison to other studies assessing SL of physicians is not easily done as every test focuses on different aspects of SL. Schmidt et al. (2017) focused on knowledge of 18 different statistical tests among pathologists and observed a rather low level of SL (39). Anderson et al. (2014) did not create an overall score but distinguished between fact, concept, and relation questions and found altering levels of SL (9). A study with Greek residents also concentrated on knowledge questions and reported a rather low SL (40). A similar tool was applied with German medical students (28). Comparing the two studies, the EE and DC score of medical students and physicians were almost on the same level. Riegelman and Hoveland (2012) found that residents struggled when critical reflection upon research was required (41), whereas the participants in our study showed medium to high levels of SRA skills. SRA and SL scores were not correlated, and DC and SL correlated inversely. A possible explanation would be that the evidences provided in the scenario did not contain any statistical information, as suggested by Anderson et al. (2014). Nevertheless, in theoretical frameworks, SL has been regarded as prerequisite for SRA (42). Moreover, in a Dutch community-based study, more numerate participants showed enhanced performance in SRA due to increased evaluation of pros and cons in decision making and evaluation of judgments (43). Future research could incorporate numerical evidence in the assessment of SRA in order to further analyze this connection. 3.2. Education & Development We explored how, where, and when physicians developed scientific skills. Significantly more participants indicated to have acquired scientific skills in an autodidactic manner (M = 4.78, SD = 1.125, Likert 1–6 scale) rather than during their studies (M = 2.31, SD = 1.456, Likert 1–6 scale) (t(71) = -9.915, p < .001) or in extracurricular activities (M = 3.34, SD = 1.869, Likert 1–6 scale) (t(71) = 4.673, p < .001) (Fig. 1). In a free-text box, the participants added various other learning opportunities, such as massive open online courses, higher education, workshops, and learning through peer reviews and feedback (Fig. 2). Regarding the doctoral thesis, a univariate ANOVA showed that having completed or working on a doctoral thesis had an effect on the scoring in EE, F(3,71) = 10.494, p < .001, partial η2 = .320, but not on DC, F(3,71) = 1.133, p = .342, nor SL, F(3,71) = 1.812, p = .153, partial η2 = .075. The fostering of critical scrutinization of study results presented by other researchers during the preparation of the doctoral thesis was positively correlated with the scoring in SL, r(71) = .271, p = .033. Regarding the postgraduate phase, a one-factorial ANOVA showed a significant main effect of having worked in research on SL, F(1,70) = 12.737, p = .001, partial η2 = .156. The type of authorship, F(5,71) = 3.886, p = .004, partial η2 = .230, and the time spent in research, F(23,71) = 2.262, p = .009, partial η2 = .525, were significantly associated with better SL, as well as the number of publications, r(71) = .36, p = .002. Regarding SRA, linear regression models revealed that the corresponding score in EE increased by β = .314 ± .150, p = .041, when the Likert value of the doctoral thesis supervisor’s content-related support was increased by one point. Additionally, the form of the doctoral thesis (experimental, clinical, empirical, statistical, or literature review) was associated with EE, with experimental and clinical design being positively related to EE skills, β = − .380 ± .154, p = .016, R² = .187, F(1,59) = 4.353, p = .041. DC was higher when participants indicated to have already worked in research, β = 3.355 ± 1.229, p = .008, R² = .314 F(1,68) = 7.448, p = .008. Page 6/16 Page 6/16 4.3. Strengths and Limitations This study built upon an innovative approach by the authors (2020) to assess SL and SRA skills in physicians (28). As this participant group is not easily recruited, the sample size of N = 71 was satisfactory. The recruited sample may not have been representative of all physicians. Furthermore, sociocultural factors and specifics of the medical educations system may limit generalizability of results The addition of numerous demographic variables yielded insights on how, where and when scientific skills were acquired and helped to identify potential associated factors. Although first attempts were made to validate the assessment tool with medical students, it could be considered too extensive and time consuming for physicians (as was indicated by 11 participants in their feedback). 4.2. Research question 2 We explored how, where, and when physicians developed SL and SRA skills. They indicated to have acquired scientific skills mostly in an autodidactic manner, in higher education outside of medical studies or in extracurricular activities. Better SL was associated with the fostering of critical thinking during the doctoral thesis, having worked in research, time spent in research, number of publications, and the type of authorship. Our findings are in line with Schmidt et al. (2017), who found that having an advanced degree other than MD or statistic courses were positively associated with SL. A study with physicians, residents, and final year medical students in Thailand showed – not surprisingly – that having recently completed a statistical workshop led to higher SL scores (44). However, additional courses are often hard to integrate in medical training. A study showed that 37% of American Ob-Gyn residents do not receive formal training (31), a study with neurology residents observed a lack of acceptance for interventions on SL (45). Page 7/16 Page 7/16 In the present study, better EE was associated with having worked on a doctoral thesis (with favorably experimental or clinical design) and having content-related support by the supervisor. These findings are in line with the subjective impression of German medical graduates with a doctoral thesis ranking their scientific skills higher compared to those working on it (34). However, the participants in the study by Epstein and colleagues (2018) did not feel confident enough to conduct research on their own (34). This is particularly important because having already worked in research was associated with a higher SL and DC score in the present study and in Schmidt et al. (2017). Moreover, Epstein and colleagues (2018) found, that medical graduates self-estimate their scientific skills after medical school as rather low (34). In the United States, only 68.1% of medical students in their final year participated in research during medical school and only 42% had (co-)authored a paper submitted for publication (46). It seems important that medical students should be involved in the publishing process and learn how to write papers already during the completion of the doctoral thesis, as this might enhance their SL and SRA skills in the long run. 5. Conclusion/ Outlook We assessed SL and SRA skills of German and Austrian physicians together with a thorough analysis of the predictive value of demographic variables. The active involvement in research seems to play an important role to foster these skills and might in consequence enhance adequate risk communication and evidence-based practice. As most participants indicated to have acquired these skills post-graduate and in an autodidactic manner, we argue to formalize and intensify the acquisition of these skills during undergraduate medical studies. Medical education curricula should include more and more effective statistical training and the active involvement of students in research. Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Consent for publication Not applicable Abbreviations Page 8/16 Abbreviations ANOVA Analysis of Variance DC Drawing conclusions DV Dependent variable ANOVA Analysis of Variance DC Drawing conclusions DV Dependent variable EE Evidence evaluation EL Evidence level IV Independent variable LV Likert value MME Master of Medical Education MOOCS Massive Open Online Courses PPV Positive Predictive Value SR Scientific reasoning SRA Scientific reasoning and argumentation Competing Interests The authors declare that they have no competing interests. Ethics approval and consent to participate The study was approved with student subjects by the ethics committee of the Medical Faculty of LMU Munich (approval reference no. 527–16). The additional clearance to conduct the study with physicians was given by the Bavarian State Chamber of Physicians. The consent being obtained from study participants was written. Funding This work was supported by an intramural grant of the Medical Faculty of LMU Munich (WiFoMed). WiFoMed is a sponsorship association of LMU Munich´s medical faculty promoting excellent research of early stage researchers. It provided financial means for participant acquisition, survey assistants, students support staff, and printing costs for the paper-based survey. The German Federal Ministry of Page 9/16 Education and Research (grant no. 01PB14004C) financed project members (one author full time, one author part time) until the 31st of March 2018, as well as costs for conference presentations and project meetings. However, none of these funding agencies provided any kind of content-related directive that could have influenced the research and its outcomes. Education and Research (grant no. 01PB14004C) financed project members (one author full time, one author part time) until the 31st of March 2018, as well as costs for conference presentations and project meetings. However, none of these funding agencies provided any kind of content-related directive that could have influenced the research and its outcomes. Authors´ contributions FS was the principal author in writing the manuscript and analyzed and interpreted the data. MB was a major contributor in writing the manuscript and supervising the data analyses and interpretations. All authors collectively designed the measurement instrument. In particular, JZ, MS and MB designed the statistical literacy test, while FS and MF were principally involved in writing the SRA decision scenario. All authors read and approved the final manuscript. Acknowledgement The authors would like to thank Wolfgang Gaissmaier for his valuable comments and suggestions to the present study. Authors' information Felicitas M. Schmidt MPH, attended medical school at the Ludwig-Maximilian University Munich. This paper constitutes a central part of her doctoral thesis “Statistical literacy as basis for scientific reasoning and argumentation in medicine”. During her undergraduate degree in Philosophy, Politics and Economics (PPE) from the Free University of Bolzano and the HEC Lausanne, she developed a profound understanding of statistics and its value for the scientific community and everyday life. She holds a Master´s degree in Public Health and is a scholar of the Friedrich Naumann-Foundation for Freedom. Jan M. Zottmann holds a master’s degree and a doctoral degree in Educational Sciences (LMU Munich, Germany). He worked at the Knowledge Media Research Center (Tübingen, Germany), the University of Tübingen, and LMU Munich before he became Assistant Director of the Institute for Medical Education at the University Hospital of LMU Munich in 2016. Dr. Zottmann is also a member of the Munich Center of the Learning Sciences at LMU. His areas of academic interest include clinical reasoning, case-based learning, instructional support for simulation-based learning, and computer-supported collaborative learning. Prof. Dr. phil. Maximilian Sailer is a University Professor of Educational Sciences at the University of Passau, Germany. He studied Educational Sciences at the University of Eichstätt and earned his doctoral degree in 2002. After receiving his post-doctoral degree (Habilitation) in 2009, he worked as a senior researcher & Adjunct Professor at the University of Augsburg. He has been involved in teaching and research at LMU Munich since 2011, where served five years as an interim Professor for various positions. He is also a member of the Munich Center of the Learning Sciences. His areas of academic Page 10/16 interest include Technology-Enhanced Learning (TEL), Blended Learning and Empirical Educational Research. Professor Dr. Martin R. Fischer is a board-certified internist and endocrinologist and a medical educator. He graduated from the Albert-Ludwigs-University in Freiburg, Germany and earned his doctoral degree in neuropharmacology in Freiburg. He did his residency at Medizinische Klinik-Innenstadt in internal medicine and endocrinology and earned a Master´s Degree in Medical Education at the University of Berne in Switzerland. From 2008 until 2011 he was a Professor for Health Sciences Education at Witten- Herdecke University. Since 2011 until today he is the Director of the Institute for Medical Education at the University Hospital and Medical Faculty of LM Munich. Since 2012 he also is Dean for Clinical Education at LMU. Dr. Authors' information Markus Berndt earned his PhD in Educational Psychology (Ludwig-Maximilians-Universität of Munich, Germany), a Master´s degree in Educational Sciences with a major in adult education (University of Regensburg, Germany), and a Master´s degree in Human Resource Management (Murray State University, USA). He has been involved in teaching and research at LMU Munich since 2010. Research foci are on clinical reasoning, scientific reasoning and argumentation, peer assessment and peer feedback, assessment and evaluation, and the use of eye tracking technology to investigate cognitive processes and interpretation of radiographic images and electrocardiograms. Currently, Dr. Berndt holds a position as research fellow at the University Hospital of LMU Munich and a position as Senior Contributing Faculty at Walden University. 7. References 1. Reyna VF, Nelson WL, Han PK, Dieckmann NF. How numeracy influences risk comprehension and medical decision making. Psychol Bull. 2009;135(6):943-73. 1. Reyna VF, Nelson WL, Han PK, Dieckmann NF. How numeracy influences risk comprehension and medical decision making. Psychol Bull. 2009;135(6):943-73. 2. Nelson W, Reyna VF, Fagerlin A, Lipkus I, Peters E. Clinical implications of numeracy: theory and practice. Ann Behav Med. 2008;35(3):261-74. 2. Nelson W, Reyna VF, Fagerlin A, Lipkus I, Peters E. Clinical implications of numeracy: theory and practice. Ann Behav Med. 2008;35(3):261-74. 3. Gigerenzer G, Gaissmaier W, Kurz-Milcke E, Schwartz LM, Woloshin S. Helping Doctors and Patients Make Sense of Health Statistics Psychological Science In The Public Interest. 2008;8(2):53-96. 3. Gigerenzer G, Gaissmaier W, Kurz-Milcke E, Schwartz LM, Woloshin S. Helping Doctors and Patients Make Sense of Health Statistics Psychological Science In The Public Interest. 2008;8(2):53-96. 4. Shield M. Statistical literacy: thinking critically about statistics. Of Significance. 1999;1(1):15-20. Shield M. Statistical literacy: thinking critically about statistics. Of Significa 5. Sedlmeier P, Gigerenzer G. Teaching Bayesian Reasoning in Less Than Two Hours. Journal of Experimental Psychology: General. 2001;130(3):380-400. 6. Fischer F, Kollar I, Ufer S, Sodian B, Hussmann H, Pekrun R, et al. Scientific Reasoning and Argumentation: Advancing an Interdisciplinary Research Agenda in Education. Frontline Learning Research,. 2014;2(3):28-45. 7. Lipkus IM, Samsa G, Rimer BK. General performance on a numeracy scale among highly educated samples. Med Decis Making. 2001;21(1):37-44. 7. Lipkus IM, Samsa G, Rimer BK. General performance on a numeracy scale among highly educated samples. Med Decis Making. 2001;21(1):37-44. 8. Monahan J. Statistical literacy: A prerequisite for evidence-based medicine. SAGE Publications Sage CA: Los Angeles, CA; 2007. Page 11/16 Page 11/16 9. Anderson BL, Gigerenzer G, Parker S, Schulkin J. Statistical literacy in obstetricians and gynecologists. J Healthc Qual. 2014;36(1):5-17. 10. Peters E. Beyond Comprehension: The Role of Numeracy in Judgments and Decisions. Current Directions in Psychological Science. 2012;21(1):31-5. 11. Walker HM. Statistical literacy on the social science. The American Statistician. 1951;5(1):6-12. 12. Wallman KK. Enhancing statistical literacy: enriching our society. Journal of the American Statistical Association. 1993;88(421):1-8. 13. Ben-Zvi D, Garfield J. Statistical literacy, reasoning, and thinking: Goals, definitions, and challenges. . In: Ben-Zvi D, Garfield J, editors. The challenge of developing statistical literacy, reasoning and thinking. Dordrecht, The Netherlands: Kluwer; 2004. p. 3-25. 14. Watson JM. Assessing Statistical Thinking Using the Media. In: Gal I, Garfield JB, editors. The Assessment Challenge in Statistics Education. 7. References Amsterdam: IOS Press and The International Statistical Institute 1997. p. 107-21. 15. Covey J. A meta-analysis of the effects of presenting treatment benefits in different formats. Med Decis Making. 2007;27(5):638-54. 16. Naylor CD, Chen E, Strauss B. Measured enthusiasm: Does the method of reporting trial results alter perceptions of therapeutic effectiveness? Annals of Internal Medicine. 1992;117(916-921). 17. Sotos AEC, Vanhoof S, Van den Noortgate W, Onghena P. Students’ misconceptions of statistical inference: A review of the empirical evidence from research on statistics education. Educational research review. 2007;2(2):98-113. 18. Gal I. Expanding conceptions of statistical literacy: An analysis of products from statistics agencies. Statistics Education Research Journal. 2003;2(1):3-21. 19. Wyatt CM, Fallar R. Medical Students Learn to Apply Bayesian Reasoning for Clinical Inference but Not Statistical Inference: A Mixed-Methods Study. 20. Cokely ET, Galesic M, Schulz E, Ghazal S, Garcia-Retamero R. Measuring Risk Literacy: The Berlin Numeracy Test. Judgment and Decision Making. 2012;7(1):25-47. 21. Hanoch Y, Miron-Shatz T, Cole H, Himmelstein M, Federman AD. Choice, numeracy and physician-in- training performance: The case of Medicare part D. Health Psychology. 2010;29:454-9. 22. Schwartz LM, Woloshin S, Welch HG. Can patients interpret health information? An assessment of the medical data interpretation test. Med Decis Making. 2005;25(3):290-300. 23. Windish DM, Huot SJ, Green ML. Medicine residents' understanding of the biostatistics and results in the medical literature. JAMA. 2007;298(9):1010-22. 24. Gigerenzer G, Wegwarth O. Risikoabschätzung in der Medizin am Beispiel der Krebsfrüherkennung. Zeitschrift für Evidenz, Fortbildung und Qualität im Gesundheitswesen. 2008;102(9):513-9. 25. Okamoto M, Kyutoku Y, Sawada M, Clowney L, Watanabe E, Dan I, et al. Health numeracy in Japan: measures of basic numeracy account for framing bias in a highly numerate population. BMC Med Inform Decis Mak. 2012;12:104. Page 12/16 26. Johnson TV, Abbasi A, Schoenberg ED, Kellum R, Speake LD, Spiker C, et al. Numeracy among trainees: are we preparing physicians for evidence-based medicine? Journal of surgical education. 2014;71(2):211-5. 27. Msaouel P, Kappos T, Tasoulis A, Apostolopoulos AP, Lekkas I, Tripodaki E-S, et al. Assessment of cognitive biases and biostatistics knowledge of medical residents: a multicenter, cross-sectional questionnaire study. Medical Education Online. 2014;19:23646-. 27. Msaouel P, Kappos T, Tasoulis A, Apostolopoulos AP, Lekkas I, Tripodaki E-S, et al. Assessment of cognitive biases and biostatistics knowledge of medical residents: a multicenter, cross-sectional questionnaire study. Medical Education Online. 2014;19:23646-. 28. Berndt M, Schmidt FM, Sailer M, Fischer F, Fischer MR, Zottmann JM. 7. References Investigating Statistical Literacy and Scientific Reasoning & Argumentation in Medical-, Social Sciences-, and Economics Students. Submitted 2020. 29. Hetmanek A, Engelmann K, Opitz A, Fischer F. Beyond intelligence and domain knowledge: Scientific reasoning and argumentation as a set of cross-domain skills. In: Fischer F, Chinn CA, Engelmann K, Osborne J, editors. Scientific reasoning and argumentation: The roles of domain-specific and domain-general knowledge. New York: Routledge; 2018. p. 203-26. 30. Rudolph JL, Horibe S. What do we mean by science education for civic engagement? Journal of Research in Science Teaching. 2016;53(6):805-20. 30. Rudolph JL, Horibe S. What do we mean by science education for civic engagement? Journal of Research in Science Teaching. 2016;53(6):805-20. 31. Anderson BL, Williams S, Schulkin J. Statistical literacy of obstetrics-gynecology residents. Journal of graduate medical education. 2013;5(2):272-5. 31. Anderson BL, Williams S, Schulkin J. Statistical literacy of obstetrics-gynecology residents. Journal of graduate medical education. 2013;5(2):272-5. 32. Franklin C, Kader G, Mewborn D, Moreno J, Peck, R. , Perry M, et al. Guidelines for Assessment and Instruction in Statistics Education (GAISE) report: A Pre-K-12 curriculum framework. Alexandria: VA: American Statistical Association; 2005. 32. Franklin C, Kader G, Mewborn D, Moreno J, Peck, R. , Perry M, et al. Guidelines for Assessment and Instruction in Statistics Education (GAISE) report: A Pre-K-12 curriculum framework. Alexandria: VA: American Statistical Association; 2005. 33. Trempler K, Hetmanek A, Wecker C, Kiesewetter J, Wermelt M, Fischer F, Fischer M, Gräsel C. Nutzung von Evidenz im Bildungsbereich – Validierung eines Instruments zur Erfassung von Kompetenzen der Informationsauswahl und der Bewertung von Studien. Z Pädagogik. 2015;61:144-66. 33. Trempler K, Hetmanek A, Wecker C, Kiesewetter J, Wermelt M, Fischer F, Fischer M, Gräsel C. Nutzung von Evidenz im Bildungsbereich – Validierung eines Instruments zur Erfassung von Kompetenzen der Informationsauswahl und der Bewertung von Studien. Z Pädagogik. 2015;61:144-66. 34. Epstein N, Huber J, Gartmeier M, Berberat P, Reimer M, Fischer M. Investigating on the acquisition of scientific competencies during medical studies and the medical doctoral thesis. Journal of Medical Education 2018;35(2). 35. Simon D. Chronische Herzinsuffizienz Apotheken Umschau2016 [Available from: http://www.apotheken.de/gesundheit-heute-news/article/chronische-herzinsuffizien 35. Simon D. Chronische Herzinsuffizienz Apotheken Umschau2016 [Available from: http://www.apotheken.de/gesundheit-heute-news/article/chronische-herzinsuffizienz/ 36. Ärzteblatt. Neue Empfehlungen zum Einsatz von Weißdornpräparaten bei Herzinsuffizienz 2017 [Available from: https://www.aerzteblatt.de/fachgebiete/kardiologie/news?nid=72259]. 36. Ärzteblatt. Neue Empfehlungen zum Einsatz von Weißdornpräparaten bei Herzinsuffizienz 2017 [Available from: https://www.aerzteblatt.de/fachgebiete/kardiologie/news?nid=72259]. 37. Meissner T. Weißdorn: Effekt auf das Endothel im Fokus. ÄrzteZeitung 2017 15. März 2017. 37. Meissner T. Tables Table 1: Description of study group Variable Options Frequency Percentage Gender Male 34 47.9% female 31 43.7% No answer 6 8.4% German mother tongue Yes 68 95.8% No 3 4.2% NA 5 7.0% MD in Germany Yes 68 95.8% No 3 4.2% Thesis Yes 58 81.7% Currently working on 9 12.7% No thesis or skipped 4 5.6% Academic qualification Habilitation 12 16.9% Professorship 7 9.9 Ever worked as researcher Yes 36 50.7% No 35 49.3% Working Environment Hospital 2 2.8% Out-patient care 43 60.6% Research 16 22.5% NA 10 14.1% 7. References Weißdorn: Effekt auf das Endothel im Fokus. ÄrzteZeitung 2017 15. März 2017. 38. Tabachnick BG, Fidell LS. Using multivariate statistics. 4th ed. ed. Needham Heights, MA: Allyn & Bacon; 2001. 38. Tabachnick BG, Fidell LS. Using multivariate statistics. 4th ed. ed. Needham Heights, MA: Allyn & Bacon; 2001. 39. Schmidt RL, Chute DJ, Colbert-Getz JM, Firpo-Betancourt A, James DS, Karp JK, et al. Statistical Literacy Among Academic Pathologists: A Survey Study to Gauge Knowledge of Frequently Used Statistical Tests Among Trainees and Faculty. Arch Pathol Lab Med. 2017;141(2):279-87. 39. Schmidt RL, Chute DJ, Colbert-Getz JM, Firpo-Betancourt A, James DS, Karp JK, et al. Statistical Literacy Among Academic Pathologists: A Survey Study to Gauge Knowledge of Frequently Used Statistical Tests Among Trainees and Faculty. Arch Pathol Lab Med. 2017;141(2):279-87. Page 13/16 40. Msaouel P, Kappos T, Tasoulis A, Apostolopoulos AP, Lekkas I, Tripodaki ES, et al. Assessment of cognitive biases and biostatistics knowledge of medical residents: a multicenter, cross-sectional Page 13/16 40. Msaouel P, Kappos T, Tasoulis A, Apostolopoulos AP, Lekkas I, Tripodaki ES, et al. Assessment of cognitive biases and biostatistics knowledge of medical residents: a multicenter, cross-sectional Page 13/16 questionnaire study. Medical Education Online. 2014;19:10. questionnaire study. Medical Education Online. 2014;19:10. questionnaire study. Medical Education Online. 2014;19:10. 41. Riegelman RK, Hovland K. Scientific Thinking and Integrative Reasoning Skills (STIRS): Essential outcomes for medical education and for liberal education. Peer Review. 2012;14(4):10. 42. Watson JM, Callingham R. Statistical literacy: A complex hierarchical construct. Statistics Education Research Journal. 2003;2(2):3-46. 43. Ghazal S, Cokely ET, Garcia-Retamero R. Predicting biases in very highly educated samples: Numeracy and metacognition. Judgment and decision making. 2014. 44. Laopaiboon M, Lumbiganon P, Walter SD. Doctors' statistical literacy: a survey at Srinagarind Hospital, Khon Kaen University. Journal of the Medical Association of Thailand (Chotmaihet thangphaet). 1997;80(2):130-7. 44. Laopaiboon M, Lumbiganon P, Walter SD. Doctors' statistical literacy: a survey at Srinagarind Hospital, Khon Kaen University. Journal of the Medical Association of Thailand (Chotmaihet thangphaet). 1997;80(2):130-7. 45. Leira E, Granner M, Torner J, Callison R, Adams H. Education research: the challenge of incorporating formal research methodology training in a neurology residency. Neurology. 2008;70(20):e79-e84. 46. Karani R, Ognibene FP, Fallar R, Gliatto P. Medical students’ experiences with authorship in biomedical research: a national survey. Academic Medicine. 2013;88(3):364-8. 46. Karani R, Ognibene FP, Fallar R, Gliatto P. Medical students’ experiences with authorship in biomedical research: a national survey. Academic Medicine. 2013;88(3):364-8. Figures Page 14/16 Figure 2 Where have you acquired scientific skills (free text analysis)? Page 16/16 Figure 2 Where have you acquired scientific skills (free text analysis)? Figure 1 In which context have you acquired scientific skills? In which context have you acquired scientific skills? Page 15/16 Figure 2 Where have you acquired scientific skills (free text analysis)? Where have you acquired scientific skills (free text analysis)? Page 16/16
https://openalex.org/W2133853168	https://jneuroinflammation.biomedcentral.com/counter/pdf/10.1186/1742-2094-8-2	English	null	Plasmacytoid dendritic cells are increased in cerebrospinal fluid of untreated patients during multiple sclerosis relapse	Journal of neuroinflammation	2,011	cc-by	2,606	* Correspondence: leonilda@unicamp.br † Contributed equally 1Neuroimmunology Unit, Dept Genetics, Evolution and Bioagents, Biology Institute University of Campinas - UNICAMP Full list of author information is available at the end of the article Abstract The plasmacytoid dendritic cells (pDCs) express a high level of Toll-like receptor 9 (TLR-9), which recognizes viral DNA. Activated via TLR-9, pDCs also secrete large amounts of type I interferon which are involved either in stimulation or down regulation of immune response in multiple sclerosis (MS). In the present study, we determinate pDCs levels by flow cytometry in Cerebrospinal Fluid (CSF) and Peripheral Blood from MS patients in relapsing and in remitting phases of the disease, comparing with other non-inflammatory diseases (OND). We provide evidence that MS patients in relapse without any treatment have a significantly (p < 0.01) higher percentage of pDCs in CSF than do patients in remission or those with OND. No change in the percentage of pDCs was observed in the peripheral blood of any of these patients. The increase of pDCs in central nervous system during relapse may be explained either by a virus infection or a down regulatory process. Plasmacytoid dendritic cells are increased in cerebrospinal fluid of untreated patients during multiple sclerosis relapse Ana Leda F Longhini1†, Felipe von Glehn1,2†, Carlos Otávio Brandão1, Rosemeire FO de Paula1, Fernando Pradella1, Adriel S Moraes1, Alessandro S Farias1, Elaine C Oliveira1,3, Juan G Quispe-Cabanillas1, Cassiana Horta Abreu1,2, Alfredo Damasceno2, Benito P Damasceno2, Konstantin E Balashov4, Leonilda MB Santos1* SHORT REPORT Plasmacytoid dendritic cells are increased in cerebrospinal fluid of untreated patients during multiple sclerosis relapse © 2011 Longhini et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Introduction this cytokine by the pDCs may suggest an important immunomodulatory function of these cells. The pathogenesis of multiple sclerosis (MS) is mainly driven by central nervous system-invading encephali- togenic CD4 T lymphocytes of both the Th1 and Th17 types. These effector cells can be down- regulated by regulatory T lymphocytes [1]. One subset of dendritic cells, the plasmacytoid dendritic cells (pDCs), has been given particular emphasis due to its importance in stimulating or down regulating effectors T cells in MS [2]. In the present study, the concentration of pDCs in the CSF and peripheral blood of MS patients during relapsing and remitting phases of the disease was deter- mined and compared to what is present in other non- inflammatory neurological diseases (OND). Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 JOURNAL OF NEUROINFLAMMATION Patients and Methods Peripheral venous blood (5 ml) and CSF (5-10 ml) sam- ples were collected from patients with RRMS, as defined by the revised McDonald criteria [5]. The MS patients were divided into two groups: relapsing (six patients) and in remission (eleven patients). Moreover, samples were collected from 8 patients with other non- inflammatory neurological diseases (OND). Relapse was defined as recent onset (within 1-7 days) of clinical neu- rological symptoms, but without any clinical or labora- torial signs of infection at the time of lumbar puncture. All patients included agreed to participate in the study, which was approved by the University of Campinas Committee for Ethical Research, and they signed a term These pDCs are present in the cerebrospinal fluid (CSF), leptomeninges and demyelinating lesions of patients with MS [3]. These cells express a high level of Toll-like receptor 9 (TLR-9), which recognizes viral DNA. Activated via TLR-9, pDCs secrete large amounts of type I interferon [4]. The use of type I interferon as an immunomodulator in the treatment of MS patients has proved beneficial for patients with the relapsing/ remitting form of MS (RRMS), and the production of * Correspondence: leonilda@unicamp.br † Contributed equally 1Neuroimmunology Unit, Dept Genetics, Evolution and Bioagents, Biology Institute University of Campinas - UNICAMP Full list of author information is available at the end of the article Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 Page 2 of 4 Table 1 Demographic and baseline clinical characteristics of patients and controls Patients # Age (years) Gender F/M Time from first relapse (Years) CSF cells/μl * Oligoclonal Bands RRMS - Relapse 6 34 (30-47) 4/2 5 (1-8) 6 (0-17) 6+/0- RRMS- remission 11 34 (26-61) 9/2 3 (1-8) 3 (1-23) 8+/3- OND** 8 46 (30-64) 7/1 - 2 (0-5) Median (range) Other Neurological Diseases Table 1 Demographic and baseline clinical characteristics of patients and controls Table 1 Demographic and baseline clinical characteristics of patients and controls Patients # Age (years) Gender F/M Time from first relapse (Years* of Consent. The clinical characteristics of the patients are presented in Table 1. a model for studying MS. A recent report shows that they promote priming of autoimmune Th17 in EAE, whereas depletion of pDC prior to induction of the disease decreases its severity [7]. Flow Cytometry Analysis The proportion of pDCs (in %) in relation to other mononuclear cells was determined by staining the CSF and peripheral blood mononuclear cells (PBMC) with anti-human BDCA2-mAb conjugated with APC (Milte- nyi Biotec, Germany). Data were acquired for gating mononuclear cells using a BD FACSCanto cytometer (BD Biosciences, USA) and analyzed using BD FACS- Diva software (BD Biosciences, USA). The p value was determined using unpaired T-test. Also of relevance is the impact of treatment with immunomodulators and the length of its use on the presence of pDCs in CSF. There is no significant differ- ence in the percentage of pDCs in the CSF when we analyzed the group of MS patients treated and not trea- ted with interferon beta IFNb (data not shown), sug- gesting that the increase of pDCs is restricted to patients in relapse. Patients and Methods Another recent study has demonstrated that clinical signs of EAE are exacer- bated considerably if the pDCs are depleted during the peak period of the disease [2]. Thus, pDC depletion significantly enhances the activation of CNS cells and the production of cytokines such as IL-17 and IFN-g, but not peripheral CD4 T cells [2]. Recent study devel- oped in the EAE model, demonstrated that the tolero- genic property of pDCs is associated with MHC class II molecule in the presenting of neuroantigen to CD4 T cells. This specific-antigen stimulation induces regu- latory T cells, which results in the reduction of the dis- ease severity [8]. Moreover, pDCs also produce indolamine 2, 3 dioxigenase (IDO), which is an enzyme activated by both type I and type II Interferon, and is involved in tryptophan catabolism. Its immunosuppres- sive effect is linked to the reduction of local trypto- phan concentration and to the activation of regulatory T cells [9,10]. Patients using corticosteroids or other immunosup- pressive and immunomodulatory drugs at the time of investigation were excluded from the study. The group with OND consisted of eight patients with no clinical evidence of any inflammatory process in the central ner- vous system (CNS). Two patients had had an ischemic stroke, two patients had pseudotumor cerebri, one had psychiatric disorders, one had epilepsy, one had normal pressure hydrocephalus and one patient had post trauma headache. Results and Discussion (A) BDCA2 positive gated CSF cells, dot plot representative of MS patients in relapse and in remission. (B) Percentage of pDCs in CSF, with data points representing individuals of MS patients in relapse (6) and in remission (11) and in other neurological diseases (8). (C) Concentration of pDCs in peripheral blood, with data points representing individuals. Data are shown as median and the statistically significant differences are indicated by (p < 0.01). ** * p<0,01 ** p<0,01 Figure 1 Concentration of pDCs in relation to other mononuclear cells in CSF and peripheral blood. (A) BDCA2 positive gated CSF cells, dot plot representative of MS patients in relapse and in remission. (B) Percentage of pDCs in CSF, with data points representing individuals of MS patients in relapse (6) and in remission (11) and in other neurological diseases (8). (C) Concentration of pDCs in peripheral blood, with data points representing individuals. Data are shown as median and the statistically significant differences are indicated by (p < 0.01). ** * p<0,01 ** p<0,01 Figure 1 Concentration of pDCs in relation to other mononuclear cells in CSF and peripheral blood. (A) BDCA2 positive gated CSF cells, dot plot representative of MS patients in relapse and in remission. (B) Percentage of pDCs in CSF, with data points representing individuals of MS patients in relapse (6) and in remission (11) and in other neurological diseases (8). (C) Concentration of pDCs in peripheral blood, with data points representing individuals. Data are shown as median and the statistically significant differences are indicated by (p < 0.01). References 1. Ramagopalan SV, Dobson R, Meier UC, Giovannoni G: Multiple sclerosis: risk factors, prodromes, and potential causal pathways. Lancet Neurol 2010, 9:727-739. 1. Ramagopalan SV, Dobson R, Meier UC, Giovannoni G: Multiple sclerosis: risk factors, prodromes, and potential causal pathways. Lancet Neurol 2010, 9:727-739. 1. Ramagopalan SV, Dobson R, Meier UC, Giovannoni G: Multiple sclerosis: risk factors, prodromes, and potential causal pathways. Lancet Neurol 2010, 9:727-739. Received: 22 November 2010 Accepted: 7 January 2011 Published: 7 January 2011 Received: 22 November 2010 Accepted: 7 January 2011 Published: 7 January 2011 Results and Discussion The number of pDCs is significantly elevated in the CSF of patients in the relapse phase of untreated MS com- pared to patients in remission (Figure 1). Since there are no differences in the number of these cells neither in the PBMC nor in total number of cells in the CSF of the same patients, pDCs must selectively increase in the CNS during the relapse phase of disease. As far as we know, this is the first observation of such an increase in the percentage of pDCs in the CSF of MS patients in a specific phase of the disease. A previous study reported an elevated concentration of dendritic cells, mainly pDCs, in patients with infections and other inflamma- tory neurological diseases, including MS, but no men- tion was made of variations during different phases of the disease [6]. Although the exact function of pDCs in the CNS needs to be elucidated in future studies, the presence of pDCs during the phase of relapse may be explained either by a virus infection or by the regulation of inflam- matory process. As the immune response evolves, the increase in the production of pro-inflammatory cyto- kines such as IFNs stimulates the secretion of IDO by pDCs, which in turn will activate regulatory T lympho- cytes. This immunomodulatory response will probably contribute to the reduction of inflammation in the CNS, thus preparing this microenvironment for the remission phase of MS. The ambivalent function of pDCs has been observed in experimental autoimmune encephalomyelitis (EAE), Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 Page 3 of 4 Relapse Remission BDCA-2 SSC-A ** * p<0,01 ** p<0,01 Figure 1 Concentration of pDCs in relation to other mononuclear cells in CSF and peripheral blood. (A) BDCA2 positive gated CSF cells, dot plot representative of MS patients in relapse and in remission. (B) Percentage of pDCs in CSF, with data points representing individuals of MS patients in relapse (6) and in remission (11) and in other neurological diseases (8). (C) Concentration of pDCs in peripheral blood, with data points representing individuals. Data are shown as median and the statistically significant differences are indicated by (p < 0.01). Relapse Remission BDCA-2 SSC-A Remission BDCA-2 ** * p<0,01 ** p<0,01 Figure 1 Concentration of pDCs in relation to other mononuclear cells in CSF and peripheral blood. Author details 1 1Neuroimmunology Unit, Dept Genetics, Evolution and Bioagents, Biology Institute University of Campinas - UNICAMP. 2Dept of Neurology, University of Campinas - UNICAMP, Campinas, SP, Brazil. 3FATEC-Sorocaba, SP, Brazil. 4Robert Wood Johnson Medical School, New Brunswick, NJ, USA. Acknowledgements p g The authors declare that they have not competing interests. KEB has served as consultants for Biogen, TEVA Neuroscience, Bayer Healthcare, and EMD Serono. g This study received financial support from the Brazilian government agencies FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo), CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) and CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior). Authors’ contributions 2. Bayley-Bucktrout SL, Caulkins SC, Goings G, Fischer JAA, Dzionek A, Miller SD: Cutting edge: central nervous system plasmacytoid dendritic cells regulate the severity of dendritic cells in multiple sclerosis. J Immunol 2008, 180:6457-6461. LMBS designed the study. FVG and COB selected the patients and collected the CSF and peripheral blood. ALFL, FVG, FP, ASM, RFOP, JGQC, ECO and ASF performed the experiments. ALFL made the flow cytometry analysis. ALFL, ASF, FVG, LMBS and COB analyzed the results. ALFL, FVG, COB, ASF, CHA, AD, BPD, KEB and LMBS helped write the paper. All authors have read and approved the final version of the manuscript. 3. Lande R, Gava V, Serafini B, Giacomini E, Visconti A, Remoli ME, Severa M, Parmentier M, Ristori G, Salvetti M, Aloisi F, Coccia EM: Plasmacytoid dendritic cells in multiple sclerosis: intracerebral recruitment and Page 4 of 4 Longhini et al. Journal of Neuroinflammation 2011, 8:2 http://www.jneuroinflammation.com/content/8/1/2 doi:10.1186/1742-2094-8-2 Cite this article as: Longhini et al.: Plasmacytoid dendritic cells are increased in cerebrospinal fluid of untreated patients during multiple sclerosis relapse. Journal of Neuroinflammation 2011 8:2. impaired maturation in response to interferon-beta. J Neuropathol Exp Neurol 2008, 67:388-401. impaired maturation in response to interferon-beta. J Neuropathol Exp Neurol 2008, 67:388-401. impaired maturation in response to interferon-beta. J Neuropathol Exp Neurol 2008, 67:388-401. 4. Moseman EA, Liang X, Dawson AJ, Panoskaltsis-Mortari A, Krieg AM, Liu YJ, Blazar BR, Chen W: Human plasmacytoid dendritic cells activated by CpG oligodeoxynucleotides induce the generation of CD4+CD25+ regulatory T cells. J Immunol 2004, 173:4433-4442. 5. Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP, Kappos L, Lublin FD, Metz LM, McFarland HF, O’Connor PW, Sandberg-Wollheim M, Thompson AJ, Weinshenker BG, Wolinsky JS: Diagnostic criteria for multiple sclerosis: 2005 revisions to the ´McDonald Criteria’. Ann Neurol 2005, 58:840-846. 6. Pashenkow M, Huang YM, Kostulas V, Haglund M, Sodestrom M, Link H: Two subsets of dendritic cells are present in human cerebrospinal fluid. Brain 2001, 124:480-492. 7. Isaksson M, Ardesjo B, Ronnblom L, Kämpe O, Lassmann H, Eloranta ML, Lobell A: Plasmacytoid DC promote priming of autoimmune Th17 cells and EAE. Eur J Immunol 2009, 39:2925-2935. 8. Irla M, Kuepfer N, Suter T, Lissilaa R, Benkhoucha M, Skupsky J, Lalive PH, Fontana A, Reith W, Hugues S: MHC class II-restricted antigen presentation by plasmacytoid dendritic cells inhibits T cell-mediated autoimmunity. J Exp Med 2010, 207:1891-1905. 9. Bayas A, Stasiolek M, Kruse N, Toyka KV, Selmaj K, Gold R: Altered innate immune response of plasmacytoid dendritic cells in multiple sclerosis. Clin Expl Immunol 2009, 157:332-342. 10. Kwidzinski E, Bechmann I: IDO expression in the brain: a double-edge sword. J Mol Med 2007, 85:1351-1359. doi:10.1186/1742-2094-8-2 Cite this article as: Longhini et al.: Plasmacytoid dendritic cells are increased in cerebrospinal fluid of untreated patients during multiple sclerosis relapse. Journal of Neuroinflammation 2011 8:2. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit
https://openalex.org/W2938172788	https://europepmc.org/articles/pmc6520747?pdf=render	English	null	Gastric Cancer in Northern Canadian Populations: A Focus on Cardia and Non-Cardia Subsites	Cancers	2,019	cc-by	8,786	Received: 28 February 2019; Accepted: 4 April 2019; Published: 15 April 2019 Abstract: In northern Canada where there is a high prevalence of Helicobacter pylori infection, there is a paucity of information on gastric cancer by the topographical subsites cardia (CGC) and non-cardia (NCGC). Here we describe the incidence of CGC and NCGC, separately, among northern Canadian populations. We used data from the Cancer Incidence in Five Continents Volumes X (CI5X) and XI (CI5XI) to obtain CGC and NCGC incidence for Canada and for Yukon (YT), a northern Canadian territory. Using these data with those provided by the Government of the Northwest Territories (NT), we estimated standardized incidence ratios comparing northern populations to Canada as a whole. We also estimated age-standardized incidence rates to permit comparisons across populations globally. NT and YT populations were disproportionately impacted by gastric cancer, particularly NCGC. This was especially true for Indigenous populations: NCGC incidence rates among NT Indigenous men were 2.7 times the rates among all men in Canada, while rates among NT Indigenous women were 3.1 times the rates among all women in Canada. Similarly, age-standardized rates of NCGC among Indigenous NT residents were comparable to global regions where there is a high burden of NCGC. This study has, for the ﬁrst time, quantiﬁed the incidence of CGC and NCGC for the NT and YT, providing new insights into the burden of these cancers among northern Canadian populations. Abstract: In northern Canada where there is a high prevalence of Helicobacter pylori infection, there is a paucity of information on gastric cancer by the topographical subsites cardia (CGC) and non-cardia (NCGC). Here we describe the incidence of CGC and NCGC, separately, among northern Canadian populations. We used data from the Cancer Incidence in Five Continents Volumes X (CI5X) and XI (CI5XI) to obtain CGC and NCGC incidence for Canada and for Yukon (YT), a northern Canadian territory. Using these data with those provided by the Government of the Northwest Territories (NT), ti t d t d di d i id ti i th l ti t C d h l Keywords: gastric cancer; cardia; incidence; epidemiology; Arctic; Canada; Indigenous health; Helicobacter pylori cancers cancers cancers cancers Amy Colquhoun 1,* , Heather Hannah 2, André Corriveau 2, Brendan Hanley 3, Yan Yuan 1 , Karen J. Goodman 1,4 and The CANHelp Working Group 5 1 School of Public Health, University of Alberta, Edmonton, AB T6G 1C9, Canada; yan.yuan@ualberta.ca (Y.Y.); karen.goodman@ualberta.ca (K.J.G.) 1 School of Public Health, University of Alberta, Edmonton, AB T6G 1C9, Canada; y yan.yuan@ualberta.ca (Y.Y.); karen.goodman@ualberta.ca (K.J.G.) y y g 2 Department of Health and Social Services, Government of the Northwest Territories, Yellowknife, NT X1A 2L9, Canada; heather_hannah@gov.nt.ca (H.H.); andre_corriveau@gov.nt.ca (A.C.) g ( ) g ( ) 3 Department of Health and Social Services, Government of Yukon, Whitehorse, YT Y1A 1Z4, Canada; brendan.hanley@gov.yk.ca 4 Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB T6G 2E1, Canada 5 4 Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB T6G 2E1, Canada 5 The CANHelp Working Group, University of Alberta, Edmonton, AB T6G 2E1, Canada; canhelp@ualberta.ca * Correspondence: amy.colquhoun@ualberta.ca; Tel.: +1-780-492-2525 4 Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB T6G 2E1, Canada 5 The CANHelp Working Group, University of Alberta, Edmonton, AB T6G 2E1, Canada; canhelp@ualberta.ca * Correspondence: amy.colquhoun@ualberta.ca; Tel.: +1-780-492-2525 ceived: 28 February 2019; Accepted: 4 April 2019; Published: 15 April 2019 Received: 28 February 2019; Accepted: 4 April 2019; Published: 15 April 2019 Received: 28 February 2019; Accepted: 4 April 2019; Published: 15 April 2019 Gastric Cancer in Northern Canadian Populations: A Focus on Cardia and Non-Cardia Subsites Amy Colquhoun 1,* , Heather Hannah 2, André Corriveau 2, Brendan Hanley 3, Yan Yuan 1 , Karen J. Goodman 1,4 and The CANHelp Working Group 5 1. Introduction Gastric cancer is a major contributor to the global burden of cancer: in 2018, there were an estimated 1,033,701 new cases and 782,685 deaths due to gastric cancer, making it the 6th most commonly diagnosed and 3rd most fatal cancer worldwide [1]. Gastric cancer incidence varies by sex, with rates among men on average twice those observed among women. Variation also exists across geographic regions: Gastric cancer incidence rates are high in countries within Asia, Eastern Europe, and parts of Central and South America, while rates are comparatively low in many higher-resourced regions such as North America and Western Europe [1]. Although often reported as a single entity, gastric cancers can be subdivided into two topographical categories: cardia gastric cancer (CGC) arising in the proximal area of the stomach, and non-cardia Cancers 2019, 11, 534; doi:10.3390/cancers11040534 www.mdpi.com/journal/cancers www.mdpi.com/journal/cancers 2 of 13 Cancers 2019, 11, 534 gastric cancers (NCGC) arising in more distal regions. As with all gastric cancers combined, there is considerable geographic variation in incidence across these subsites [2]. This may result, in part, from diﬀerences in risk factors by subsite that vary by geography. These might include genetic risk factors: For example, genetic polymorphisms in certain DNA repair processes can contribute to an increased risk of carcinogenesis [3,4]. Environmental factors also play a role: NCGC risk is increased by Helicobacter pylori infection [5–7] and cancers of this subsite are more common in less-resourced countries [8,9]. In contrast, CGC is positively associated with obesity and gastroesophageal reﬂux but not with H. pylori infection, evidence that suggests a distinct etiology from H. pylori-associated NCGC. Due at least in part to its association with H. pylori infection, NCGC is also associated with indicators of low socioeconomic status such as household crowding, low income, low education levels and increased number of siblings [10]. The mechanisms by which genetic and environmental risk factors interact to modify gastric cancer risk remain unclear [3,4]. Lower and steadily declining incidence of NCGC among higher-resourced countries may be attributable to improved sanitation and other factors that reduced the prevalence of H. pylori infection, as well as other technological advances of the early 20th century that reduced the occurrence of diseases of the pre-modern era [11–13]; however, some populations within more developed areas are still at high risk. 1. Introduction Studies have documented, for example, a higher burden of gastric cancer in Indigenous populations worldwide compared to their non-Indigenous counterparts [14,15]. Similarly, others have shown elevated incidence rates across the circumpolar region, particularly among Indigenous populations [16,17]. Given the disproportionate impact of H. pylori on NCGC incidence relative to CGC incidence, the distribution of gastric cancer by subsite varies by H. pylori prevalence within higher-resourced countries. Ethnic minorities in the US, such as African-Americans or Asian-Americans, have higher rates of NCGC and lower rates of CGC compared to non-Hispanic white Americans [18,19]. This diﬀering pattern in NCGC and CGC incidence rates has also been observed among Indigenous populations compared to their country-speciﬁc non-Indigenous counterparts [14]. There remains a paucity of information, however, on diﬀerences in gastric cancer incidence by subsite among high-risk populations. In northern Canada, where Indigenous groups comprise a high proportion of the population, communities have expressed concern over their perceived high frequency of stomach cancer [20]. To address these concerns, a community-driven research program conducted by the Canadian North Helicobacter pylori (CANHelp) Working Group brings together academic researchers, healthcare providers, health oﬃcials, and members of communities in the Northwest Territories (NT) and Yukon (YT), Canada to conduct community H. pylori projects. These projects have estimated a high prevalence of H. pylori infection in participating communities [20,21]. Furthermore, participants assessed pathologically through biopsies of the stomach lining had a high prevalence of precancerous lesions of the stomach [20]. To characterize the burden of stomach cancer in the northern regions where these communities reside, we aimed to describe the incidence of gastric cancer overall and by subsite and compare these estimates to those observed elsewhere. 2.1. Gastric Cancer Overall During 2003 through 2012, there was an annual average of 2638 cases of gastric cancer in Canada; in each of the two northern territories included here, YT during 2003–2012 and NT during 1997–2015, there was an annual average of 3 cases (Table 1). In all populations deﬁned by geography or Indigenous status, there were proportionally more cases of gastric cancer observed among men than women. Disparities in the distribution of age at diagnosis were apparent by geography and Indigenous status. In Canada, less than 25% of gastric cancer cases were diagnosed in individuals under the age of 60 (Figure 1). Among northern populations, however, over 40% of cases were observed in those younger than 60 years of age: 42% of total cases in YT and the non-Indigenous NT population, and 48% in the 3 of 13 Cancers 2019, 11, 534 Indigenous NT population. The disproportionate impact on younger populations was particularly pronounced among Indigenous people in the NT: 16% of cases among NT Indigenous people were observed in 0–39-year-olds (7/44) compared to 2% in Canada as a whole (646/26,382). Cancers 2019, 11, x FOR PEER REVIEW 3 of 12 Table 1. Number of gastric cancer cases and person-years by gastric cancer subsite, population and sex; northern territories, Canada; various years. ulation, and 48% in the Indigenous NT population. The disproportionate impact on younge ulations was particularly pronounced among Indigenous people in the NT: 16% of cases amon ndigenous people were observed in 0–39-year-olds (7/44) compared to 2% in Canada as a whol Table 1. Number of gastric cancer cases and person-years by gastric cancer subsite, population and sex; northern territories, Canada; various years. Population Gastric Cancer Subsite 1 Calendar Years Men Women Number of Cases Person-Years Number of Cases Person-Years Canada All 2003–2012 16,872 144,495,378 9510 147,171,135 CGC 2003–2007 3151 79,864,738 857 81,320,815 NCGC 2003–2007 6489 79,864,738 4544 81,320,815 Yukon All 2003–2012 21 169,680 12 162,500 CGC 2003–2012 7 169,680 3 162,500 NCGC 2003–2012 14 169,680 9 162,500 NT—non- Indigenous All 1997–2015 16 225,396 3 199,893 CGC 1997–2015 7 225,396 0 199,893 NCGC 1997–2015 9 225,396 3 199,893 NT— Indigenous All 1997–2015 26 215,886 18 214,022 CGC 1997–2015 1 215,886 0 214,022 NCGC 1997–2015 25 215,886 18 214,022 1 CGC cardia gastric cancer; NCGC non-cardia gastric cancer CGC deﬁned as C16 0; non-cardia gastric NCGC population, and 48% in the Indigenous NT population. 2.1. Gastric Cancer Overall Cardia and Non-Cardia Subsites 2.2. Cardia and Non-Cardia Subsites 1 CGC, cardia gastric cancer; NCGC, non-cardia gastric cancer. CGC deﬁned as C16.0; non-cardia gastric NCGC deﬁned as C16.1-9; NT, Northwest Territories. 1 CGC, cardia gastric cancer; NCGC, non-cardia gastric cancer. CGC defined as C16.0; non-cardia gastric NCGC defined as C16.1-9; NT, Northwest Territories. 2.1. Gastric Cancer Overall The disproportionate impact on younger populations was particularly pronounced among Indigenous people in the NT: 16% of cases among NT Indigenous people were observed in 0–39-year-olds (7/44) compared to 2% in Canada as a whole (646/26,382). Table 1. Number of gastric cancer cases and person-years by gastric cancer subsite, population and sex; northern territories, Canada; various years. Population Gastric Cancer Subsite 1 Calendar Years Men Women Number of Cases Person-Years Number of Cases Person-Years Canada All 2003–2012 16,872 144,495,378 9510 147,171,135 CGC 2003–2007 3151 79,864,738 857 81,320,815 NCGC 2003–2007 6489 79,864,738 4544 81,320,815 Yukon All 2003–2012 21 169,680 12 162,500 CGC 2003–2012 7 169,680 3 162,500 NCGC 2003–2012 14 169,680 9 162,500 NT—non- Indigenous All 1997–2015 16 225,396 3 199,893 CGC 1997–2015 7 225,396 0 199,893 NCGC 1997–2015 9 225,396 3 199,893 NT— Indigenous All 1997–2015 26 215,886 18 214,022 CGC 1997–2015 1 215,886 0 214,022 NCGC 1997–2015 25 215,886 18 214,022 1 CGC, cardia gastric cancer; NCGC, non-cardia gastric cancer. CGC defined as C16.0; non-cardia Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years Canada and Yukon include cases diagnosed 2003 2012; Northwest Territories (NT) Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Figure 1. Age distribution of total gastric cancer cases by population; northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. 2.2. 2.3. Comparisons Across Populations 2.3. Comparisons Across Populations NCGC incidence rates were higher among Indigenous NT men compared to all Canadian men (SIR 2.7, 95% CI 1.7–3.9) (Table 2). Similarly, Indigenous NT women had a higher incidence of NCGC compared to all women in Canada (SIR 3.1, 95% CI 1.8–4.8). For non-Indigenous NT men and women, incidence rates were similar to Canada. However, in YT, NCGC rates among both men and women were elevated compared to Canada (30% and 50% higher, respectively). It should be noted, however, that for YT and non-Indigenous NT groups, these differences were statistically imprecise. For CGC, with the exception of YT men and women and non-Indigenous NT men in whom there was an apparent elevation, rates were lower compared to Canadian rates. However, for most groups there were no or very few cases of CGC, resulting in imprecise estimates vulnerable to small fluctuations NCGC incidence rates were higher among Indigenous NT men compared to all Canadian men (SIR 2.7, 95% CI 1.7–3.9) (Table 2). Similarly, Indigenous NT women had a higher incidence of NCGC compared to all women in Canada (SIR 3.1, 95% CI 1.8–4.8). For non-Indigenous NT men and women, incidence rates were similar to Canada. However, in YT, NCGC rates among both men and women were elevated compared to Canada (30% and 50% higher, respectively). It should be noted, however, that for YT and non-Indigenous NT groups, these diﬀerences were statistically imprecise. For CGC, with the exception of YT men and women and non-Indigenous NT men in whom there was an apparent elevation, rates were lower compared to Canadian rates. However, for most groups there were no or very few cases of CGC, resulting in imprecise estimates vulnerable to small ﬂuctuations in incidence. were no or very few cases of CGC, resulting in imprecise estimates vulnerable to small fluctuations in incidence. Table 2. Number of cases and estimated standardized incidence ratios (SIR) comparing northern populations to Canada as a whole on gastric cancer incidence rates by gastric cancer subsite, population, and sex; various years. 2.2. Cardia and Non-Cardia Subsites 2.2. Cardia and Non-Cardia Subsites The proportion of total gastric cancer cases classified as cardia varied by geography and Indigenous status (Figure 2). In Canada, YT, and the non-Indigenous NT population, the proportional CGC incidence ranged from 29–37% of total gastric cancer cases. In contrast, only 2% of gastric cancer cases in the Indigenous NT population were classified as CGC (1/44). This population also had the The proportion of total gastric cancer cases classiﬁed as cardia varied by geography and Indigenous status (Figure 2). In Canada, YT, and the non-Indigenous NT population, the proportional CGC incidence ranged from 29–37% of total gastric cancer cases. In contrast, only 2% of gastric cancer cases in the Indigenous NT population were classiﬁed as CGC (1/44). This population also had the lowest 4 of 13 Cancers 2019, 11, 534 proportion of cases classiﬁed as ‘other and not otherwise speciﬁed’. The distribution of gastric cancer by subsite also varied by sex. Among men, CGC cancers comprised an average across subgroups of 25% (range 4–44%) of total gastric cancer cases; among women, this proportion was 9% (range 0–25%) on average across subgroups. Cancers 2019, 11, x FOR PEER REVIEW 4 of 12 Figure 2. Proportional NCGC incidence (NCGC cases divided by total gastric cancer cases); northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Non-cardia gastric (NCGC) defined as C16.1-9. 95% confidence intervals are based on the binomial exact distribution. Figure 2. Proportional NCGC incidence (NCGC cases divided by total gastric cancer cases); northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Non-cardia gastric (NCGC) deﬁned as C16.1-9. 95% conﬁdence intervals are based on the binomial exact distribution. Figure 2. Proportional NCGC incidence (NCGC cases divided by total gastric cancer cases); northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Non-cardia gastric (NCGC) defined as C16.1-9. 95% confidence intervals are based on the binomial exact distribution. Figure 2. Proportional NCGC incidence (NCGC cases divided by total gastric cancer cases); northern territories, Canada; various years. Canada and Yukon include cases diagnosed 2003–2012; Northwest Territories (NT) includes cases diagnosed 1997–2015. Non-cardia gastric (NCGC) deﬁned as C16.1-9. 95% conﬁdence intervals are based on the binomial exact distribution. 2.3. Comparisons Across Populations 2.3. Comparisons Across Populations Population Gastric Cancer Subsite 1 Men Women Cases 2 SIR 3 95% CI 4 Cases 2 SIR 3 95% CI 4 Yukon All 21 1.3 0.78 1.9 12 1.8 0.89 3.0 CGC 7 1.2 0.48 2.5 3 2.6 0.53 7.5 NCGC 14 1.3 0.71 2.2 9 1.5 0.71 2.9 NT— non-Indigenous All 16 1.1 0.61 1.7 3 0.52 0.11 1.5 CGC 7 1.3 0.51 2.6 0 0.00 0.00 3.8 NCGC 9 0.94 0.43 1.8 3 0.62 0.13 1.8 NT—Indigenous All 26 1.9 1.2 2.7 18 2.6 1.5 4.1 CGC 1 0.21 0.01 1.2 0 0.00 0.00 3.3 Comparisons with other regions across the globe revealed similar patterns of variation in the distribution of gastric cancer subsites by geography and Indigenous status. For CGC among men, ASRs for Canada, YT, and the non-Indigenous NT population (Table 3) were comparable to regions such as Oceania, Southern Europe, Northern Africa & Western Asia (ASRs ranging from 2.5 to 3.3 per 100,000). These estimates were low relative to high-incidence regions such as Eastern/Southeastern Asia and Eastern Europe. Among Indigenous and non-Indigenous NT women, there were no cases of CGC. For Indigenous NT men, CGC incidence rates were based on a very small number of cases and, as a result, were very low. Conversely, NCGC incidence rates were markedly high in the Indigenous NT population (ASRs 13.3 and 7.7 per 100,000 among men and women, respectively), similar to high-incidence regions like Eastern Europe. For all of Canada, and in YT men and women and non-Indigenous NT men, NCGC rates were comparable to regions such as Northern and Western Europe (ASRs in Canada, YT and non-Indigenous NT groups ranged from 5.0 to 6.1 per 100,000 men and 2.8 to 4.5 per 100,000 women). The NCGC incidence rate among non-Indigenous NT women was even lower (ASR 2.0 per 100,000) (Figure 3a,b). 5 of 13 Cancers 2019, 11, 534 Table 2. Number of cases and estimated standardized incidence ratios (SIR) comparing northern populations to Canada as a whole on gastric cancer incidence rates by gastric cancer subsite, population, and sex; various years. 2.3. Comparisons Across Populations 2.3. Comparisons Across Populations Population Gastric Cancer Subsite 1 Men Women Cases 2 SIR 3 95% CI 4 Cases 2 SIR 3 95% CI 4 Yukon All 21 1.3 0.78 1.9 12 1.8 0.89 3.0 CGC 7 1.2 0.48 2.5 3 2.6 0.53 7.5 NCGC 14 1.3 0.71 2.2 9 1.5 0.71 2.9 NT—non- Indigenous All 16 1.1 0.61 1.7 3 0.52 0.11 1.5 CGC 7 1.3 0.51 2.6 0 0.00 0.00 3.8 NCGC 9 0.94 0.43 1.8 3 0.62 0.13 1.8 NT— Indigenous All 26 1.9 1.2 2.7 18 2.6 1.5 4.1 CGC 1 0.21 0.01 1.2 0 0.00 0.00 3.3 NCGC 25 2.7 1.7 3.9 18 3.1 1.8 4.8 SIR, standardized incidence ratio; CI, conﬁdence interval; CGC, cardia gastric cancer; NCGC, non-cardia gastric cancer; NT, Northwest Territories. 1 CGC deﬁned as C16.0; NCGC deﬁned as C16.1-9; 2 Yukon includes cases diagnosed during 2003–2012; NT includes cases diagnosed during 1997–2015; 3 SIRs compare study population to Canada during 2003–2007; 4 95% CIs are based on the Poisson distribution. SIR, standardized incidence ratio; CI, conﬁdence interval; CGC, cardia gastric cancer; NCGC, non-cardia gastric cancer; NT, Northwest Territories. 1 CGC deﬁned as C16.0; NCGC deﬁned as C16.1-9; 2 Yukon includes cases diagnosed during 2003–2012; NT includes cases diagnosed during 1997–2015; 3 SIRs compare study population to Canada during 2003–2007; 4 95% CIs are based on the Poisson distribution. Table 3. Estimated gastric cancer age-standardized incidence rates (ASRs per 100,000) by subsite, population, and sex; northern territories and Canada; various years. Table 3. Estimated gastric cancer age-standardized incidence rates (ASRs per 100,000) by subsite, population, and sex; northern territories and Canada; various years. ASR, age-standardized incidence rate; CI, conﬁdence interval; CGC, cardia gastric cancer; NCGC, non-cardia gastric cancer; NT, Northwest Territories; 1 CGC deﬁned as C16.0; NCGC deﬁned as C16.1-9; 2 For Canada and Yukon, all gastric cancer cases combined include cases diagnosed during 2003–2012; gastric cancer subsites include cases diagnosed during 2003–2007. NT includes cases diagnosed during 1997–2015; 3 ASRs standardized to the World Population; 4 95% CIs are based on the Poisson distribution. 2.3. Comparisons Across Populations 2.3. Comparisons Across Populations World an regional cardia and non-cardia gastric cancer ASR estimates from Colquhoun, et al. (2014) [2]. Figure 3. (a) Estimated cardia gastric cancer age-standardized incidence rates (ASRs per 100,000) by population and sex; (b) Estimated non-cardia gastric cancer ASRs (per 100,000) by population and sex; northern territories, Canada, and world regions, various years. Age-standardized incidence rates (ASRs) standardized to the World Population. Canada and Yukon include cases diagnosed during 2003–2007; Northwest Territories (NT) includes cases diagnosed during 1997–2015. World and regional cardia and non-cardia gastric cancer ASR estimates from Colquhoun, et al. (2014) [2]. 2.3. Comparisons Across Populations 2.3. Comparisons Across Populations Population Gastric Cancer Subsite 1 Men Women Cases 2 ASR 3 95% CI 4 Cases 2 ASR 3 95% CI 4 Canada All 16,872 7.0 6.9 7.1 9510 3.2 3.1 3.3 CGC 3151 2.5 2.4 2.6 857 0.6 0.6 0.6 NCGC 6489 5.0 4.9 5.1 4544 2.8 2.7 2.9 Yukon All 21 9.2 5.2 13.2 12 5.8 2.4 9.2 CGC 3 3.3 0.0 7.2 1 0.9 0.0 2.7 NCGC 6 6.1 1.1 11.1 4 4.5 0.0 9.1 NT—non- Indigenous All 16 8.8 3.9 13.7 3 2.0 0.0 4.3 CGC 7 3.0 0.6 5.4 0 0.0 0.0 0.0 NCGC 9 5.8 1.5 10.1 3 2.0 0.0 4.3 NT— Indigenous All 26 13.8 8.4 19.2 18 7.7 4.1 11.2 CGC 1 0.5 0.0 1.4 0 0.0 0.0 0.0 NCGC 25 13.3 8.0 18.6 18 7.7 4.1 11.3 6 of 13 on Cancers 2019, 11, 534 during 1997–2 d b (a) ncers 2019, 11, x FOR PEER REVIEW 6 of (b) Figure 3. (a) Estimated cardia gastric cancer age-standardized incidence rates (ASRs per 100,000) by population and sex; (b) Estimated non-cardia gastric cancer ASRs (per 100,000) by population and sex; northern territories, Canada, and world regions, various years. Age-standardized incidence rates (ASRs) standardized to the World Population. Canada and Yukon include cases diagnosed during 2003–2007; Northwest Territories (NT) includes cases diagnosed during 1997–2015. World and regional cardia and non-cardia gastric cancer ASR estimates from Colquhoun, et al. (2014) [2]. Figure 3. (a) Estimated cardia gastric cancer age-standardized incidence rates (ASRs per 100,000) by population and sex; (b) Estimated non-cardia gastric cancer ASRs (per 100,000) by population and sex; northern territories, Canada, and world regions, various years. Age-standardized incidence rates (ASRs) standardized to the World Population. Canada and Yukon include cases diagnosed during 2003–2007; Northwest Territories (NT) includes cases diagnosed during 1997–2015. World and regional cardia and non-cardia gastric cancer ASR estimates from Colquhoun, et al. (2014) [2]. (a) Cancers 2019, 11, x FOR PEER REVIEW 6 (a) ( ) (b) (b) Figure 3. (a) Estimated cardia gastric cancer age-standardized incidence rates (ASRs per 100,000) b population and sex; (b) Estimated non-cardia gastric cancer ASRs (per 100,000) by population an sex; northern territories, Canada, and world regions, various years. Age-standardized incidence rate (ASRs) standardized to the World Population. Canada and Yukon include cases diagnosed durin 2003–2007; Northwest Territories (NT) includes cases diagnosed during 1997–2015. 2.4. Sensitivity Analyse 2.4. Sensitivity Analyses As part of a sensitivity analysis to assess the impact of the attribution of all C16.8-9 (gastri cancers classified as overlapping and not-otherwise specified) cases to NCGC, we estimated SIRs fo NCGC defined as C16.1-4 (Table S1). SIR estimates based on this revised NCGC case definition were similar to those estimated using the original definition for NCGC (C16.1-9). For Indigenous people in the NT, SIR estimates increased with the more restricted case definition: In men, the SIR for NCGC increased 33.3% to 3.6 (95% CI 1.9–6.0); in women, the SIR increased 58% to 4.9 (95% CI 2.5–8.6). C d t ASR ti t d i th b d d fi iti ASR ti t d i NCGC As part of a sensitivity analysis to assess the impact of the attribution of all C16.8-9 (gastric cancers classiﬁed as overlapping and not-otherwise speciﬁed) cases to NCGC, we estimated SIRs for NCGC deﬁned as C16.1-4 (Table S1). SIR estimates based on this revised NCGC case deﬁnition were similar to those estimated using the original deﬁnition for NCGC (C16.1-9). For Indigenous people in the NT, SIR estimates increased with the more restricted case deﬁnition: In men, the SIR for NCGC increased 33.3% to 3.6 (95% CI 1.9–6.0); in women, the SIR increased 58% to 4.9 (95% CI 2.5–8.6). As part of a sensitivity analysis to assess the impact of the attribution of all C16.8-9 (gastri cancers classified as overlapping and not-otherwise specified) cases to NCGC, we estimated SIRs fo NCGC defined as C16.1-4 (Table S1). SIR estimates based on this revised NCGC case definition wer similar to those estimated using the original definition for NCGC (C16.1-9). For Indigenous peopl in the NT, SIR estimates increased with the more restricted case definition: In men, the SIR for NCGC increased 33.3% to 3.6 (95% CI 1.9–6.0); in women, the SIR increased 58% to 4.9 (95% CI 2.5–8.6). C d ASR i d i h b d d fi i i ASR i d i NCGC As part of a sensitivity analysis to assess the impact of the attribution of all C16.8-9 (gastric cancers classiﬁed as overlapping and not-otherwise speciﬁed) cases to NCGC, we estimated SIRs for NCGC deﬁned as C16.1-4 (Table S1). SIR estimates based on this revised NCGC case deﬁnition were similar to those estimated using the original deﬁnition for NCGC (C16.1-9). 3. Discussion To characterize stomach cancer burden in NT and YT, we estimated the incidence of gastric cancer overall and by subsite and compared these estimates to rates estimated for Canada nationwide. We observed that NT and YT populations are disproportionately impacted by gastric cancer compared to Canada, particularly at younger ages when a larger proportion of NT and YT cases are diagnosed. This was especially true for the Indigenous NT population, in whom 16% of cases were observed in those under the age of 40, in contrast to 2% of cases in Canada as a whole diagnosed in this age group. Further, NCGC incidence rates among NT Indigenous men and women were 2.7 and 3.1 times, respectively, rates of Canada as a whole. To provide a global context for these observations, we also compared the estimated NT and YT rates to those observed elsewhere globally and found that NCGC incidence rates in the Indigenous NT population were similar to regions such as Eastern Europe where there is a high burden of gastric cancer. Our analysis is consistent with previous work that demonstrated elevations in gastric cancer rates across northern populations. Young, et al. (2016), for example, reported that gastric cancer rates among Inuit populations in the circumpolar region of Canada are higher than global rates [16]. Similarly, other work has shown that gastric cancer rates are higher in northern regions compared to rates observed in more southern parts of the same country [22,23]. In addition to geographic variation, our ﬁndings are consistent with reports of diﬀerences in the stomach cancer burden among Indigenous populations compared to their non-Indigenous counterparts [14]. In some regions, gastric cancer rates among Indigenous peoples were over two times the rates of non-Indigenous people living in the same area [14]: in Alberta for example, a Canadian province south of NT, during 2009-2016 gastric cancer rates among First Nations men and women were 1.5 and 2.3 times, respectively, rates of their non-First Nations counterparts [24]. As with all gastric cancers combined, there is also geographic variation in the incidence of gastric cancer subsites [2]. Because H. pylori is a risk factor for gastric cancer arising in distal portions of the stomach [6], it is not surprising that regions with higher prevalence of this infection are impacted more greatly by NCGC. Further, elevations in NCGC incidence have been observed to coincide with an increased burden of H. Cancers 2019, 11, 534 Cancers 2019, 11, 534 Cancers 2019, 11, 534 Cancers 2019, 11, 534 7 of 13 Compared to ASRs estimated using the broader case deﬁnition, ASRs estimated using NCGC deﬁned as C16.1-4 were lower for Canada and for the study populations in the NT and YT. This, in turn, resulted in lower estimates among the study groups and for Canada relative to the global populations that maintained a broader deﬁnition of NCGC (Table S2). 2.4. Sensitivity Analyse 2.4. Sensitivity Analyses For Indigenous people in the NT, SIR estimates increased with the more restricted case deﬁnition: In men, the SIR for NCGC increased 33.3% to 3.6 (95% CI 1.9–6.0); in women, the SIR increased 58% to 4.9 (95% CI 2.5–8.6). 3. Discussion The relatively small number of gastric cancer cases observed in northern Canadian populations created imprecise ASR estimates that are vulnerable to large ﬂuctuations if the numerator is oﬀ by as little as one case. To account for this, we used SIRs to compare the observed number of cases in the study populations with the number of cases expected based on Canadian rates. This method of age-adjustment is beneﬁcial when age-speciﬁc rates are unstable due to small numbers [37]. Unfortunately, because the age distribution of the study population is used as the standard for these methods, age adjustment through the use of SIRs limits comparisons over time or across multiple populations [38]. To support comparisons with global rates, we also estimated ASRs, using CIs to acknowledge imprecision in these age-standardized rates due to small numbers of cases. It should be noted that we used data on gastric cancer in Canada during 2003–2007 as a comparison to rates that included cases diagnosed in more recent years in NT and YT. Given that gastric cancer rates are decreasing in Canada [39], in line with observations that NCGC rates are declining in developed regions of the world [40–42], our SIRs comparing NT and YT rates to Canadian rates are likely to be underestimated. While gastric cancer rates are decreasing overall in developed regions of the world, Anderson, et al. (2018) demonstrated through stratiﬁcation by age group that NCGC rates may be increasing in younger non-Hispanic white populations in the United States, particularly among women [33]. Although the number of cases in NT and YT populations were too small to examine age-speciﬁc incidence rates in detail, increasing incidence in younger age groups over time could explain why the proportion of gastric cancer cases corresponding to younger age groups was larger in NT and YT than in Canada as a whole, given that NT and YT data included cases diagnosed in more recent years. This study has, for the ﬁrst time, estimated the incidence of CGC and NCGC among NT and YT populations. Consistent with other work describing gastric cancer among northern and Indigenous populations, we found that NT and YT populations were disproportionately impacted by gastric cancer, particularly of the non-cardia gastric region. These estimates provide new insights into the burden of gastric cancer in northern Canadian populations. As a result, this work supports the on-going community-driven CANHelp Working Group research program. 3. Discussion pylori infection among Indigenous populations [25–27]. Comparing Indigenous ethnic groups (designated as American Indians and Alaska Natives) to non-Hispanic whites in the United States, for example, estimated NCGC incidence rates during 1999–2004 were 5.8- and 14.9-fold higher among men and women, respectively [28]. Other reports similarly estimated NCGC incidence rates to be higher among Native Americans compared to other ethnic groups in the United States [29,30] and among Maori New Zealanders compared to non-Maori New Zealanders [31]. Despite the higher burden of H. pylori infection and gastric cancer among Indigenous populations, there is relatively little information on CGC and NCGC in these groups across the globe. While our analysis of gastric cancer incidence by subsite in northern Canada contributed to reducing this information gap, there were data restrictions that limited this assessment. When disaggregating rates by age category, the CI5X and CI5XI datasets combined gastric cancers of the lesser and greater curvature (C16.5 and C16.6) with those classiﬁed as ‘overlapping lesion of stomach’ and ‘stomach, not otherwise speciﬁed’ (C16.8 and C16.9). Because C16.5-6 are typically classiﬁed as cancers of the non-cardia gastric region [19,32,33], and because other analysts have reclassiﬁed all or some C16.8-9 cases as NCGC when these classiﬁcations are relatively common [2,34–36], we chose to attribute all cases classiﬁed as C16.5-9 to NCGC. It is possible that some of the cases classiﬁed as overlapping or unknown topography (C16.8-9) are, in fact, cancers of the cardia and were misclassiﬁed Cancers 2019, 11, 534 8 of 13 here; however, given the low proportion of cases classiﬁed as CGC in global populations, the degree of any potential misclassiﬁcation is likely small. To assess the extent to which misclassiﬁcation of cases could have impacted comparisons of NCGC incidence estimates across populations, we also performed a sensitivity analysis where NCGC cases were redeﬁned as C16.1–4: We found that SIR estimates with the more restricted deﬁnition were similar to those estimated using the original deﬁnition for NCGC (C16.1-9). In contrast, ASRs based on the restricted deﬁnition of NCGC (C16.1-4) resulted in lower estimated rates for the study groups and Canada relative to the global populations, which maintained a broader deﬁnition of NCGC. Despite this, ASRs of NCGC among Indigenous populations in the NT remained high relative to global estimates. 3. Discussion In particular, it will be used to inform gastric cancer control strategies aimed at achieving the goal identiﬁed by impacted Indigenous Arctic communities of implementing eﬀective interventions to reduce their disproportionate gastric cancer disease burden. 4.3. Statistical Analyses We aimed to describe the incidence of gastric cancer overall and by subsite across study groups in the NT and YT and to compare these estimates to those observed in other populations. Our analyses used an estimation approach: We estimated relevant parameters to describe their magnitude and used conﬁdence intervals (CIs) as measures of precision, to report the range of values compatible with the expected degree of random error in the data [47,48]. Accordingly, our comparisons of rates across populations were not based on statistical signiﬁcance testing; instead, we focused on the magnitude, direction and precision of estimated associations. We estimated the proportional incidence of NCGC as the number of NCGC cases divided by the total number of gastric cancer cases; to report the precision of estimated proportional incidence, we present 95% conﬁdence intervals (CIs) based on the binomial exact distribution. To permit comparisons across groups with varying population age distributions, we estimated a weighted average of age-speciﬁc incidence rates using weights from a standard population, a process known as standardization. Here, we used two methods to accomplish this: standardized incidence ratios (SIRs) and age-standardized cancer incidence rates (ASRs). SIRs are the functional equivalent of age- and sex-standardized incidence rate ratios using the study population distribution as the standard. We estimated SIRs by sex for all gastric cancer combined, CGC, and NCGC in YT (2003–2012) and NT Indigenous and non-Indigenous populations (1997–2015); to report the precision of estimated SIRs, we present 95% CIs based on the Poisson distribution. As the referent, which is the denominator, we used incidence rates in Canada; these were restricted to the years 2003–2007 due to the limited availability of data by gastric cancer subsite, age group, and sex. For SIRs, the numerator of the ratio is the total number of cases in the study group, which does not require stratiﬁcation of cases by age group; thus, for SIRs we were able to include cases from a longer time period for YT despite the shorter time period for which the required data were available for Canada. SIRs provide a comparison of rates in the numerator and denominator populations unconfounded by the variables used for standardization; however, one SIR cannot be compared to another if the numerator populations diﬀer substantially on the distribution of these variables. 4.2. Data Sources We used data from the Cancer Incidence in Five Continents (CI5), volumes X (CI5X) and XI (CI5XI) published online by the International Agency for Research on Cancer [44,45]. These volumes are the most recent CI5 publications: They report data from select registries around the world for 2003–2007 and 2008–2012, respectively. We used CI5 data on numbers of cases and person-years from the YT and nationwide Canadian registries, available by cancer site, sex, and 5-year age group. CI5X includes disaggregated data by gastric cancer subsite; however, at the time of analysis, CI5XI did not include stratiﬁcation within subsite groupings, precluding the identiﬁcation of subsite-speciﬁc cases by age group and sex. Furthermore, these data sources do not contain histological information that would permit estimates based on the Lauren histological classiﬁcation of gastric cancers [46]. For NT-speciﬁc analyses, we used data provided directly by the NT government. These data included gastric cancer cases by subsite, sex, 5-year age group, Indigenous status, and year of diagnosis during 1997 and 2015. The NT dataset included dichotomized Indigenous status linked from vital statistics data housed by the NT government and mid-year populations for estimating person-years. 4.1. Comparison Populations The northern populations available for comparison in this analysis are the NT population dichotomized as Indigenous or non-Indigenous, and the YT population without classiﬁcation by Indigenous status. As a reference population, we use the total population of Canada, including the NT and YT populations, without classiﬁcation by Indigenous status. In 2016, Indigenous groups comprised approximately 51% of the NT population of 41,135, 23% of the YT population of 35,110, and 4.9% of the total Canadian population of 34,460,065 [43]. Sex-speciﬁc frequencies were available for 9 of 13 Cancers 2019, 11, 534 all comparison populations. Classiﬁcation by Indigenous status was available for the NT population, but not for the YT or total Canadian populations. NT residents were dichotomized as Indigenous, including individuals who identify as Inuit, First Nations, or Métis, or non-Indigenous, including everyone else. 4.3. Statistical Analyses For global comparisons, we estimated age-standardized cancer incidence rates (ASRs) by gastric cancer subsite, geography, and sex; to report the precision of estimated ASRs, we present 95% CIs based on the Poisson distribution. Due to limitations in the years of data available by subsite that were 10 of 13 Cancers 2019, 11, 534 also disaggregated by age group and sex, ASRs by gastric cancer subsite for Canada and YT were limited to 2003–2007; ASRs for NT included data from 1997–2015. ASRs were age-standardized using the World Standard Population as deﬁned by Segi [49]; we used world and regional CGC and NCGC ASR estimates reported by Colquhoun, et al. (2014) [2]. All data analyses, including SIR and ASR estimates and their CIs, were performed using SAS 9.4 software as outlined in the SAS/STAT® 14.3 User’s Guide [50]. 4.4. Case Deﬁnitions and Sensitivity Analyses In the CI5 and NT datasets, gastric cancer cases were classiﬁed according to the International Classiﬁcation of Diseases, 10th Revision (ICD-10): Gastric cancer (C16), CGC (C16.0), and NCGC (C16.1-9). NCGC typically includes the following anatomical regions of the stomach: fundus, body, antrum, pylorus, lesser curvature, and greater curvature (C16.1-6) [19,32,33]. However, the CI5X datasets group cancers of the greater and lesser curvature (C16.5 and C16.6) with cancers classiﬁed as overlapping and not-otherwise speciﬁed (C16.8 and C16.9). Knowing that a portion of this group includes NCGC cases (C16.5-6), and that other works have allocated some or all C16.8-9 sites to NCGC when data were limited [2,34–36], here we chose to classify C16.5-9 as NCGC. Because some of these cases might have been misclassiﬁed CGC cases, we also estimated SIRs and ASRs for NCGC deﬁned as C16.1-4. Furthermore, in addition to age-standardization using the world population, we also repeated analyses using the Canadian 2011 standard population to permit comparisons across populations within Canada (Table S3). Conﬂicts of Interest: The authors declare no conﬂict of interest. 5. Conclusions This study has, for the ﬁrst time, estimated the incidence of CGC and NCGC in NT and YT populations. Consistent with other work describing gastric cancer among northern and Indigenous populations, we found that NT and YT populations were disproportionately impacted by gastric cancer, particularly of the non-cardia gastric subsite. This was especially true for Indigenous populations: NCGC incidence rates among NT Indigenous men were 2.7 times the rates among all men in Canada, while rates among NT Indigenous women were 3.1 times the rates among all women in Canada. Similarly, age-standardized rates of NCGC among Indigenous NT residents were comparable to global regions where there is a high burden of NCGC. These estimates provide new insights into the burden of gastric cancer in northern Canadian populations; as a result, this work has value for informing gastric cancer control strategies aimed at reducing the disproportionate gastric cancer disease burden of these populations. Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6694/11/4/534/s1, Table S1: Comparison of the number of cases and estimated standardized incidence ratios (SIRs) using two diﬀerent deﬁnitions of non-cardia gastric cancer by population and sex, Table S2: Comparison of the number of cases and estimated age-standardized incidence rates (ASRs per 100,000) using two diﬀerent deﬁnitions of non-cardia gastric cancer by population and sex, Table S3: Estimated gastric cancer age-standardized incidence rates (ASRs per 100,000) by subsite, population, and sex using two diﬀerent standard populations. Author Contributions: Conceptualization, A.C. (Amy Colquhoun) and K.J.G.; data curation, A.C. (Amy Colquhoun), H.H. and A.C. (André Corriveau); formal analysis, A.C. (Amy Colquhoun); funding acquisition, CANHelp Working Group; project administration, CANHelp Working Group; resources, H.H., A.C. (André Corriveau) and CANHelp Working Group; software, A.C. (Amy Colquhoun); supervision, K.J.G.; validation, Y.Y. and K.J.G.; visualization, A.C. (Amy Colquhoun) and K.J.G.; writing—original draft, A.C. (Amy Colquhoun); Writing—review and editing, H.H., A.C. (André Corriveau), B.H., Y.Y. and K.J.G. Funding: This work was supported by grants from the Canadian Institutes for Health Research (MOP115031) and Alberta Innovates Health Solutions (201201159). Amy Colquhoun was supported by a Vanier Canada Graduate Scholarship. Karen J Goodman was supported by a Senior Health Scholar award from the Alberta Heritage Foundation for Medical Research. Acknowledgments: The authors would like to thank all members of the Canadian North Helicobacter pylori (CANHelp) Working Group. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Acknowledgments: The authors would like to thank all members of the Canadian North Helicobacter pylori (CANHelp) Working Group. 5. Conclusions Conﬂicts of Interest: The authors declare no conﬂict of interest. 11 of 13 11 of 13 Cancers 2019, 11, 534 References 1. Bray, F.; Ferlay, J.; Soerjomataram, I.; Siegel, R.L.; Torre, L.A.; Jemal, A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 2018, 68, 394–424. [CrossRef] [PubMed] 2. Colquhoun, A.; Arnold, M.; Ferlay, J.; Goodman, K.J.; Forman, D.; Soerjomataram, I. Global patterns of cardia and non-cardia gastric cancer incidence in 2012. Gut 2015, 64, 1881–1888. [CrossRef] [PubMed] 3. González, C.A.; Sala, N.; Rokkas, T. Gastric Cancer: Epidemiologic Aspects. Helicobacter 2013, 18, 34–38. [CrossRef] [PubMed] 4. Karimi, P.; Islami, F.; Anandasabapathy, S.; Freedman, N.D.; Kamangar, F. Gastric cancer: Descriptive epidemiology, risk factors, screening, and prevention. Cancer Epidemiol. Biomark. Prev. 2014, 23, 700–713. [CrossRef] 5. The EUROGAST Study Group. An international association between Helicobacter pylori infection and gastric cancer. Lancet 1993, 341, 1359–1362. [CrossRef] 6. Plummer, M.; Franceschi, S.; Vignat, J.; Forman, D.; de Martel, C. Global burden of gastric cancer attributable to Helicobacter pylori. Int. J. Cancer 2015, 136, 487–490. [CrossRef] [PubMed] 7. Crew, K.D.; Neugut, A.I. Epidemiology of gastric cancer. World J. Gastroenterol. WJG 2006, 12, 354–362. [CrossRef] [PubMed] 8. Parkin, D.M. International variation. Oncogene 2004, 23, 6329–6340. [CrossRef] 9. De Martel, C.; Ferlay, J.; Franceschi, S.; Vignat, J.; Bray, F.; Forman, D.; Plummer, M. Global burden of cancers attributable to infections in 2008: A review and synthetic analysis. Lancet Oncol. 2012, 13, 607–615. [CrossRef] 0. Barker, D.J.; Coggon, D.; Osmond, C.; Wickham, C. Poor housing in childhood and high rates of stom cancer in England and Wales. Br. J. Cancer 1990, 61, 575–578. [CrossRef] 11. Howson, C.P.; Hiyama, T.; Wynder, E.L. The decline in gastric cancer: Epidemiology of an unplanned triumph. Epidemiol. Rev. 1986, 8, 1–27. [CrossRef] [PubMed] 12. Paik, D.C.; Saborio, D.V.; Oropeza, R.; Freeman, H.P. The epidemiological enigma of gastric cancer rates in the US: Was grandmother’s sausage the cause? Int. J. Epidemiol. 2001, 30, 181–182. [CrossRef] 13. Ushijima, T.; Sasako, M. Focus on gastric cancer. Cancer Cell 2004, 5, 121–125. [CrossRef] 14. Arnold, M.; Moore, S.P.; Hassler, S.; Ellison-Loschmann, L.; Forman, D.; Bray, F. The burden of stomach cancer in indigenous populations: A systematic review and global assessment. Gut 2014, 63, 64–71. [CrossRef] 15. Moore, S.P.; Antoni, S.; Colquhoun, A.; Healy, B.; Ellison-Loschmann, L.; Potter, J.D.; Garvey, G.; Bray, F. Cancer incidence in indigenous people in Australia, New Zealand, Canada, and the USA: A comparative population-based study. Lancet Oncol. 2015, 16, 1483–1492. References [CrossRef] p p y 16. Young, T.K.; Kelly, J.J.; Friborg, J.; Soininen, L.; Wong, K.O. Cancer among circumpolar populations: An emerging public health concern. Int. J. Circumpolar Health 2016, 75. [CrossRef] [PubMed] 17. Department of Health and Social Services, Government of the Northwest Territories. Cancer in the Northwest Territories, 1990–2000. A Descriptive Report; Department of Health and Social Services, Government of the Northwest Territories: Yellow Knife, NT, Canada, 2003. 18. Wu, X.; Chen, V.W.; Ruiz, B.; Andrews, P.; Su, L.J.; Correa, P. Incidence of esophageal and gastric carcinomas among American Asians/Paciﬁc Islanders, whites, and blacks: Subsite and histology diﬀerences. Cancer 2006, 106, 683–692. [CrossRef] [PubMed] 19. Wu, H.; Rusiecki, J.A.; Zhu, K.; Potter, J.; Devesa, S.S. Stomach carcinoma incidence patterns in the United States by histologic type and anatomic site. Cancer Epidemiol. Biomark. Prev. 2009, 18, 1945–1952. [CrossRef] 20. Cheung, J.; Goodman, K.J.; Girgis, S.; Bailey, R.; Morse, J.; Fedorak, R.N.; Geary, J.; Fagan-Garcia, K.; van Zanten, S.V.; Munday, R.; et al. Disease manifestations of Helicobacter pylori infection in Arctic Canada: Using epidemiology to address community concerns. BMJ Open 2014, 4, e003689. [CrossRef] 21. Hastings, E.V.; Yasui, Y.; Hanington, P.; Goodman, K.J.; CANHelp Working Group. Community-driven research on environmental sources of H. pylori infection in arctic Canada. Gut Microbes 2014, 5, 606–617. [CrossRef] 22. Friborg, J.; Koch, A.; Wohlfarht, J.; Storm, H.-H.; Melbye, M. Cancer in Greenlandic Inuit 1973–1997: A cohort study. Int. J. Cancer 2003, 107, 1017–1022. [CrossRef] 23. Lanier, A.P.; Kelly, J.J.; Maxwell, J.; McEvoy, T.; Homan, C. Cancer in Alaska Native people, 1969–2003. Alaska Med. 2006, 48, 30–59. 12 of 13 Cancers 2019, 11, 534 24. Health Standards, Quality and Performance Division, Analytics and Performance Reporting Branch. Stomach Cancer Incidence among First Nations in Alberta; First Nations—Health Trends Alberta; Government of Alberta: Alberta, Canada, 2018. 25. Windsor, H.M.; Abioye-Kuteyi, E.A.; Leber, J.M.; Morrow, S.D.; Bulsara, M.K.; Marshall, B.J. Prevalence of Helicobacter pylori in Indigenous Western Australians: Comparison between urban and remote rural populations. Med. J. Aust. 2005, 182, 210–213. 26. McDonald, A.M.; Sarfati, D.; Baker, M.G.; Blakely, T. Trends in Helicobacter pylori infection among M¯aori, Paciﬁc, and European Birth cohorts in New Zealand. Helicobacter 2015, 20, 139–145. [CrossRef] 27. McMahon, B.J.; Bruce, M.G.; Koch, A.; Goodman, K.J.; Tsukanov, V.; Mulvad, G.; Borresen, M.L.; Sacco, F.; Barrett, D.; Westby, S.; et al. References The diagnosis and treatment of Helicobacter pylori infection in Arctic regions with a high prevalence of infection: Expert Commentary. Epidemiol. Infect. 2016, 144, 225–233. [CrossRef] 28. Wiggins, C.L.; Perdue, D.G.; Henderson, J.A.; Bruce, M.G.; Lanier, A.P.; Kelley, J.J.; Seals, B.F.; Espey, D.K. Gastric cancer among American Indians and Alaska Natives in the United States, 1999–2004. Cancer 2008, 113, 1225–1233. [CrossRef] 9. Kubo, A.; Corley, D.A. Marked Multi-Ethnic Variation of Esophageal and Gastric Cardia Carcinomas wi the United States. Am. J. Gastroenterol. 2004, 99, 582–588. [CrossRef] 0. Brown, L.M.; Devesa, S.S. Epidemiologic trends in esophageal and gastric cancer in the United St Surg. Oncol. Clin. N. Am. 2002, 11, 235–256. [CrossRef] 31. Armstrong, R.W.; Borman, B. Trends in incidence rates of adenocarcinoma of the oesophagus and gastric cardia in New Zealand, 1978–1992. Int. J. Epidemiol. 1996, 25, 941–947. [CrossRef] 32. Mahadevan, V. Anatomy of the stomach. Surg. Oxf. Int. Ed. 2014, 32, 571–574. [CrossRef] 33. Anderson, W.F.; Rabkin, C.S.; Turner, N.; Fraumeni, J.F.; Rosenberg, P.S.; Camargo, M.C. The Changing Face of Noncardia Gastric Cancer Incidence Among US Non-Hispanic Whites. J. Natl. Cancer Inst. 2018, 110, 608–615. [CrossRef] [PubMed] 34. Freedman, N.D.; Abnet, C.C.; Leitzmann, M.F.; Mouw, T.; Subar, A.F.; Hollenbeck, A.R.; Schatzkin, A. A prospective study of tobacco, alcohol, and the risk of esophageal and gastric cancer subtypes. Am. J. Epidemiol. 2007, 165, 1424–1433. [CrossRef] [PubMed] 35. Abnet, C.C.; Freedman, N.D.; Kamangar, F.; Leitzmann, M.F.; Hollenbeck, A.R.; Schatzkin, A. Non-steroidal anti-inﬂammatory drugs and risk of gastric and oesophageal adenocarcinomas: Results from a cohort study and a meta-analysis. Br. J. Cancer 2009, 100, 551–557. [CrossRef] 36. Keszei, A.P.; Schouten, L.J.; Goldbohm, R.A.; Brandt, V.D. Red and processed meat consumption and the risk of esophageal and gastric cancer subtypes in The Netherlands Cohort Study. Ann. Oncol. 2012, 23, 2319–2326. [CrossRef] 37. Szklo, M.; Nieto, J. Epidemiology: Beyond the Basics, 3rd ed.; Jones & Bartlett Learning: Burlington, MA, USA, 2014; ISBN 978-1-4496-0469-1. 38. Julious, S.A.; Nicholl, J.; George, S. Why do we continue to use standardized mortality ratios for small area comparisons? J. Public Health Med. 2001, 23, 40–46. [CrossRef] 39. Xie, L.; Semenciw, R.; Mery, L. Cancer incidence in Canada: Trends and projections (1983–2032). Health Promot. Chronic Dis. Prev. Can. 2015, 35 (Suppl. 1), 2–186. [CrossRef] pp 40. Devesa, S.S.; Blot, W.J.; Fraumeni, J.F. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998, 83, 2049–2053. References [CrossRef] 41. Steevens, J.; Botterweck, A.A.M.; Dirx, M.J.M.; van den Brandt, P.A.; Schouten, L.J. Trends in incidence of oesophageal and stomach cancer subtypes in Europe. Eur. J. Gastroenterol. Hepatol. 2010, 22, 669–678. [CrossRef] 42. Holster, I.L.; Aarts, M.J.; Tjwa, E.T.T.L.; Lemmens, V.E.P.P.; Kuipers, E.J. Trend breaks in incidence of non-cardia gastric cancer in the Netherlands. Cancer Epidemiol. 2014, 38, 9–15. [CrossRef] 43. Government of Canada, S.C. Aboriginal Peoples Highlight Tables, 2016 Census. Available online: http://www12.statcan.gc.ca/census-recensement/2016/dp-pd/hlt-fst/abo-aut/Table.cfm?Lang=Eng& T=101&S=99&O=A (accessed on 30 May 2018). 44. Forman, D.; Bray, F.; Brewster, D.; Gombe Mbalawa, C.; Kohler, B.; Pineros, M.; Steliarova-Foucher, E.; Swaminathan, R.; Ferlay, J. Cancer Incidence in Five Continents, Vol. X (Electronic Version). Available online: http://ci5.iarc.fr (accessed on 6 July 2014). 13 of 13 Cancers 2019, 11, 534 13 of 13 45. Bray, F.; Colombet, M.; Mery, L.; Pineros, M.; Znaor, A.; Zanetti, R.; Ferlay, J. Cancer Incidence in Five Continents, Vol. XI; International Agency for Research on Cancer: Lyon, France; Available online: http://ci5.iarc.fr (accessed on 30 January 2018). 46. Lauren, P. The two histological main types of gastric carcinoma: Diﬀuse and so-called intestinal-type carcinoma. An attempt at a histo-clinical classiﬁcation. Acta Pathol. Microbiol. Scand. 1965, 64, 31–49. [CrossRef] 47. Rothman, K.J.; Greenland, S.; Lash, T.L. Precision and Statistics in Epidemiologic Studies. In Modern Epidemiology; Wolters Kluwer: Philadelphia, PA, USA, 2008; pp. 151–152. 48. Rothman, K.J. Random Error and the Role of Statistics. In Epidemiology: An Introduction; Oxford University Press: New York, NY, USA, 2012; pp. 151–152. 49. Segi, M. Cancer Mortality for Selected Sites in 24 Countries (1950–1957); Department of Public Health, Tohoku University School of Medicine: Sendai, Japan, 1960. 50. SAS Institute Inc. The STDRATE Procedure. In SAS/STAT(R) 14.3 User’s Guide; SAS Institute Inc.: Cary, NC, USA, 2017; p. 78. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W4230011558	https://www.qeios.com/read/5OJRZM/pdf	English	null	Baraitser-Winter cerebrofrontofacial syndrome	Definitions	2,020	cc-by	119	Qeios · Definition, February 10, 2020 Open Peer Review on Qeios Open Peer Review on Qeios INSERM INSERM Qeios ID: 5OJRZM · https://doi.org/10.32388/5OJRZM Source INSERM. (1999). Orphanet: an online rare disease and orphan drug data base. Baraitser- Winter cerebrofrontofacial syndrome. ORPHA:2995 Baraitser-Winter syndrome (BWS) is a malformation syndrome, characterized by facial dysmorphism (hypertelorism with ptosis, broad bulbous nose, ridged metopic suture, arched eyebrows, progressive coarsening of the face), ocular coloboma, pachygyria and/or band heterotopias with antero-posterior gradient, progressive joint stiffening, and intellectual deficit of variable severity, often with severe epilepsy. Pachygyria - epilepsy - intellectual disability - dysmorphism (Fryns-Aftimos syndrome (FA); see this term) corresponds to the appearance of BWS in elderly patients. Qeios ID: 5OJRZM · https://doi.org/10.32388/5OJRZM 1/1
W3004548250.txt	https://riorpub.com/en/storage/download/48865	en		SAFETY OF COMPLEX OF HYDRAULIC STRUCTURES OF CHERNORECHENSKIY RESERVOIR IN SEVASTOPOL	Stroitelʹstvo i arhitektura	2,019	cc-by	2,831	При цитировании этой статьи ссылка на DOI обязательна Construction and Architecture (2019) Vol. 7. Issue 3 (24): 13–18 Construction and (2019) Vol. 7. Issue 3 (24) DOIArchitecture 10.29039/2308-0191-2019-7-3-13-18 Безопасность комплекса гидротехнических сооружений Чернореченского водохранилища в г. Севастополе УДК 556.166 Волосухин Я.В. Генеральный директор ИКЦ «Безопасность гидротехнических сооружений» (г. Новочеркасск), e-mail: safety@ibgts.ru Статья получена: 01.10.2019. Рассмотрена: 11.11.2019. Одобрена: 15.11.2019. Опубликована онлайн: 26.11.2019. ©РИОР Аннотация. Рассмотрены вопросы безопасности комплекса гидротехнических сооружений Чернореченского водохранилища — основного источника водоснабжения города Севастополя. Исследован вопрос пропуска паводка Р = 0,1% обеспеченности с учетом трансформации части паводкового объема водохранилищем. Дана оценка пропуска максимального паводка с Q0,01% (с учетом гарантийной поправки) всеми водопропускными сооружениями гидроузла при наполненном водохранилище до отметки НПУ как наиболее ответственного мероприятия. Ключевые слова: безопасность, комплекс гидротехнических сооружений, водохранилище, паводок, пропускная способность. Введение. Крымский полуостров относится к территориям с низкой водообеспеченностью для среднего года Р = 50 %, водообеспеченность составляет 0,4 тыс. м3/год на одного человека, в целом по РФ — 29,1 тыс. м3/год на одного человека. На территории Крымского полуострова расположено 15 водохранилищ естественного стока с суммарным объемом при НПУ W = 253,12 млн м3 [1]. Суммарное водопотребление Крыма на 1990 г. составляло 3909,09 млн м3, в том числе 3202 млн м3 (81,91%) — вода СевероКрымского канала; 296,73 млн м3 (7,59%) — речные воды (местный сток, аккумулированный в водохранилищах и прудах); 257,85 млн м 3 (6,60%) — подземные воды; 152,51 млн м 3 (3,90%) — морские воды [3; 6]. С 2014 г. Украиной подача воды по СевероКрымскому каналу прекращена и водопотребление изменено. В Крыму ныне эксплуатируются восемь водохранилищ Северо-Крымского канала с суммарным объемом при НПУ W = 185 млн м3, на сегодняшний день они наполнены менее чем на 50%. На территории Крымского полуострова имеется 1657 рек, ручьев и балок общей длиной около 6 тыс. км. Гидрологическая сеть Крыма включает 34 гидропоста (один гидропост в Крыму приходится на 794 км2 площади водосбора). Для сопоставления. Площадь водосбора, в км2, приходящейся на один гидропост, составляет: во Франции — 203 км2, в Великобритании — 175 км2, в Германии — 119 км2, в Японии — 67 км2. Малая река Черная имеет длину 35 км и площадь водосбора 427 км2. Первая рукописная работа о р. Черная была подготовлена 80 лет назад (1939 г.) сотрудником Крымского гидрометеобюро П.М. Шликарь [6]. Гидропост в низовье малой р. Черная у с. Хмельницкое был установлен в 1946 г. (площадь водосбора F = 342 км2, средняя высота водосбора Нср = 520 м). На период проектирования и строительства Чернореченского водохранилища на SAFETY OF COMPLEX OF HYDRAULIC STRUCTURES OF CHERNORECHENSKIY RESERVOIR IN SEVASTOPOL Yakov Volosukhin General Director, Engineering and Consulting Centre (ECC) «Hydraulic structures Safety», Novocherkassk; e-mail: safety@ibgts.ru Manuscript received: 01.10.2019. Revised: 11.11.2019. Accepted: 15.11.2019. Published online: 26.11.2019. ©RIOR Abstract. The issues of safety of the complex of hydraulic structures of the Chernorechenskiy reservoir - the main source of water supply of the city of Sevastopol are considered. The issue of flood passage P = 0,1% of the provision taking into account the transformation of the flood volume by the reservoir has been investigated. The maximum flood pass with Q 0,01% (taking into account the warranty amendment) of all water passage structures of the hydro-node at the filled reservoir up to the NPP elevation as the most responsible measure has been estimated. Keywords: reservoir, flood, capacity, safety, complex of hydraulic structures of reservoir. Строительство и архитектура (2019). Том 7. Выпуск 3 (24) RIOR 13 Construction and Architecture (2019) Vol. 7. Issue 3 (24) р. Черная гидрологическая информация была достаточно скудной. Грунтовая плотина была построена в 1956 г., имела высоту Н = 28 м, длину L = 1082 м и создавала водохранилище объемом W = 33,2 млн м3. В период с 1979 по 1984 г. была проведена реконструкция грунтовой плотины Чернореченского водохранилища, в результате которой увеличена высота плотины на 8 м (Н = 36 м), доведя полный объем при НПУ до W = 64,2 млн м3, а полезный — до 61,2 млн м3. Чернореченское водохранилище является основным источником водоснабжения в городе Федерального значения Российской Федерации Севастополе и самым крупным водохранилищем естественного стока (площадь зеркала составляет 604 га, объем воды при НПУ — 64,2 млн м3) Крымского полуострова, предназначено для водообеспечения населения города (443 212 человек на 01.01.2019). Комплекс ГТС Чернореченского водохранилища относится к ГТС I класса — чрезвычайно высоко опасности и находится под постоянным государственным надзором Ростехнадзора. Комплекс ГТС Чернореченского водохранилища эксплуатируется 63 года и на нем начали проявляться проблемы старения сооружений. Цель исследования — обосновать безопасность комплекса гидротехнических сооружений (ГТС) Чернореченского гидроузла при пропуске паводка Р = 0,1% обеспеченности (основной расчетный случай) с учетом трансформации части объема водохранилищем и пропуске паводка Р = 0,01% (поверочный расчетный случай с учетом гарантированной поправки) всеми водопропускными сооружениями гидроузла [2–12; 14; 15]. Основные характеристики дождевого паводка за весь жизненный цикл Чернореченского гидроузла, используемые при расчетах трансформации максимального расхода воды водохранилищем: максимальный расход Q0,01% = 660 м3/с; объем паводка W0,01% = 89,4 млн м3; продолжительность паводка T = 5,5 сут; продолжительность подъема t1 = 1,25 сут; продолжительность спада t2 = 4,25 сут. В данных расчетах водохранилище характеризуется «батиграфическими кривыми» (рис. 1), из которых устанавливается емкость форсиров- 14 RIOR ки (трансформации) расходов воды и глубина слоя форсировки: Vф = WФПУ – WНПУ = 81,0 – – 64,2 = 16,8 млн м 3; hф = 263,30 – 261,00 = = 2,30 м. Кривая объёмов Чернореченского водохранилища Рис. 1. Фрагменты батиграфических кривых водохранилища на р. Черная В состав водосбросных и водопропускных сооружений Чернореченского ГУ входят: автоматический водосброс; донный водовыпуск из четырех ниток трубопровода; водозаборное сооружение из двух ниток трубопровода. Водохозяйственный расчет водохранилища на пропуск максимальных расходов выполняется на основе уравнения баланса воды в водохранилище. В общем виде баланс воды в водохранилище за время dt может быть выражен следующим дифференциальным уравнением: Qdt = Qсб dt ± ΩdZ , где Q — расчетный расход во входном створе водохранилища, м 3 /с; Qсб — расход в створе водосбросного сооружения (сбросной расход); Ω — площадь водной поверхности водохранилища; уровень воды в водохранилище, м. При детальных балансовых расчетах применяют различные приемы приближенного интегрирования. При этом весь период половодья делят на конечные малые интервалы времени, в течение которых расходы притока и сброса можно считать изменяющимися линейно. Гидрограф модели паводка приведен на рис. 2. Уравнение с конечными интервалами времени примет вид: Q + Qсб.к Qн + Qк ∆t − сб.н ∆t = ∆W , 2 2 где ΔW — изменение объема водохранилища; индексы н и к означают начало и конец расчетного интервала времени. Строительство и архитектура (2019). Том 7. Выпуск 3 (24) Construction and Architecture (2019) Vol. 7. Issue 3 (24) В уравнении неизвестны сбросной расход Qсб.к и объем водохранилища Wк на конец интервала. Расчет выполнен численными табличными способами. В начале первого расчетного интервала расходы притока и сброса равны нулю. В конце интервала приток определяют, как ординату гидрографа половодья (паводка) и вычисляют объем стока за интервал. Сбросной расход на конец интервала неизвестен; задаются отметкой уровня воды в водохранилище на конец интервала Z к′ и по графику пропускной способности водосброса находят в первом приближении сбросной расход. Решая уравнение баланса воды в водохранилище, находим его объем на конец интервала, и по батиграфической кривой Z = f(W) устанавливаем соответствующую отметку уровня воды ZУВ. Если ZУВ и Z к′ не совпадают, то изменяется отметка Z к′ , и расчет повторяется. Конечный расход стока Qк, сбросного расхода Qсб.к и емкости Wк одного интервала являются начальными для последующего. Результат расчета показывает, что даже при столь глубокой сработке уровня (до отметки 256,40 м БС) возможно переполнение водохранилища от отметки ФПУ на 0,1 м (до отметки 263,4 м БС). Период стояния максимальных уровней составляет 0,5 суток. При таком режиме возможно еще избежать катастрофических последствий. Основные характеристики дождевого паводка, используемые при расчете трансформации максимального расхода воды водохранилищем: максимальный расход Q0,1% = 370 м3/с; объем паводка W0,01% = 50,1 млн м3; коэффициент несимметричности гидрографа ks = 0,33; продолжительность паводка T = 5,5 сут; продолжительность подъема t1 = 1,25 сут; продолжительность спада t2 = 4,25 сут. Гидрограф модели паводка 0,1% приведен на рис. 3. р. Черная — створ гидроузла Q0,1% = 370 м3/с р. Черная — створ гидроузла Q0,01 = 660 м3/с (с гарантийной поправкой) Рис. 2. Расчетный гидрограф паводка 0,01% обеспеченности В результате расчета находят ординаты гидрографа сбросных расходов (расходы и уровни). Сбросные расходы вычисляют в зависимости от уровня воды в водохранилище. При заданных условиях в указанных границах возможно переполнение водохранилища; уровень может подняться до отметки 264,30 м, т.е. на 1,0 м выше отметки ФПУ. По оценке баланса притока и сброса, необходима предпаводковая сработка водохранилища ~ 20,0 млн м3, до отметки 256,40 м БС. В этом варианте предусмотрено включение водовыпуска на пропуск расходов с началом паводка; величина этих расходов будет зависеть от устанавливающегося уровня воды в водохранилище. Строительство и архитектура (2019). Том 7. Выпуск 3 (24) Рис. 3. Расчетный гидрограф паводка 0,1% обеспеченности Расчет трансформации паводка представлен в табл. 1. Расчетные графики показаны на рис. 4 и 5. р. Черная - створ гидроузла Q0,1% = 370 м3/с Рис. 4. Гидрографы притока и сброса паводка в створе Чернореченского гидроузла существующими водосбросными сооружениями RIOR 15 Construction and Architecture (2019) Vol. 7. Issue 3 (24) Таблица 1 Wакк (сраб.) за интервал (по формуле), млн м3 Сброс Водохранилище Δt, сут Приток № интервала ∑Δt нарастающим итогом Расчет пропуска паводка 0,1% ВП в створе Чернореченского гидроузла существующими сооружениями ZНПУ = 261,00 м; ZФПУ = 263,30 м; hф = 2,30 м; WНПУ = 64,2 млн м3; WФПУ = 81,0 млн м3; Vф = 16,8 млн м3; Q0,1% = 370 м3/с; Δt = 0,25 сут. (21,6 · 103 с) 1 0,25 0,25 2 0,50 0 44,4 0,48 0 261,10 30,0 0,19 0,29 64,2 64,49 261,10 3 0,75 44,4 199,8 2,64 30,0 261,15 64,0 1,02 1,62 64,49 66,11 261,15 Q н, м3/с Q к, м3/с Qн + Qк ∆t , 2 млн м 0 0 Z к′ Qсб.н, м3/с , м БС 3 Qсб.к, м3/с Qсб.н + Qсб.к ∆t , 2 млн м 3 0 W н, млн м3 W к, млн м3 ZУВ, м БС 64,20 64,20 261,00 4 1,00 199,8 329,3 5,71 64,0 261,8 125,0 2,04 3,67 66,11 69,78 261,80 5 1,25 329,3 370,0 7,55 125,0 262,45 220,0 3,73 3,82 69,78 73,60 262,45 6 1,50 370,0 340,4 7,67 220,0 262,90 265,0 5,24 2,43 73,60 76,03 262,95 7 1,75 340,4 284,9 6,75 265,0 263,00 246,5 5,52 1,23 76,03 77,26 263,00 8 2,00 284,9 218,3 5,43 246,5 263,00 246,5 5,32 0,11 77,26 77,37 263,05 9 2,25 218,3 162,8 4,12 246,5 262,90 228,0 5,12 –1,00 77,37 76,37 262,90 10 2,50 162,8 114,7 3,00 228,0 262,80 210,0 4,73 –1,73 76,37 74,64 262,70 11 2,75 114,7 81,4 2,12 210,0 262,20 114,6 3,51 –1,39 74,64 73,25 262,40 12 3,00 81,4 55,5 1,48 114,6 262,10 100,6 2,32 –0,84 73,25 72,41 262,20 13 3,25 55,5 37,0 1,00 100,6 262,00 87,2 2,03 –1,03 72,41 71,38 262,10 14 3,50 37,0 25,2 0,67 87,2 261,90 74,4 1,75 –1,08 71,38 70,30 261,90 15 3,75 25,2 16,7 0,45 74,4 261,80 62,4 1,48 –1,03 70,30 69,27 261,80 16 4,00 16,7 11,5 0,30 62,4 261,60 40,5 1,11 –0,81 69,27 68,46 261,70 5,00 11,5 1,9 0,58 40,5 261,30 14,0 2,35 –1,77 68,46 66,69 261,30 18 6,00 1,9 0 0,10 14,0 7,8 0,94 –0,84 66,69 65,85 19 7,00 17 1,0 р. Черная — створ гидроузла Пропуск паводка 0,1% обеспеченности Рис. 5. Динамика уровня воды в водохранилище при пропуске паводка существующими сооружениями Выводы 1. Обоснованы расходы редкой обеспеченности (Р = 0,1% и Р = 0,01%) р. Черная в створе Чернореченского водохранилища с учетом всего периода наблюдений на гидропосте в с. 16 RIOR 261,10 261,00 Хмельницкое (бывшее Чернореченское, Т = 63 года) и данных службы эксплуатации гидроузла за весь его жизненный цикл. 2. Комплекс гидротехнических сооружений (ГТС) Чернореченского водохранилища и его противопаводковая емкость позволяет трансформировать паводок Р = 0,1% обеспеченностью без превышения отметки НПУ. 3. На спаде основной волны паводка Р = 0,1% возможно отключение донного водовыпуска, так как водосбросное сооружения самостоятельно справится с пропуском оставшейся части гидрографа притока в Чернореченское водохранилище. 4. В целях контроля за состоянием водохранилища и его водосбора и своевременного принятия мер по пропуску паводков редкой обеспеченности рекомендуется создание автомаСтроительство и архитектура (2019). Том 7. Выпуск 3 (24) Construction and Architecture (2019) Vol. 7. Issue 3 (24) тизированной системы мониторинга и прогнозирования состояния водных объектов, влияющих на безопасность гидротехнических сооружений гидроузла Чернореченского водохранилища. В состав такой системы входят три автоматизированных гидрологических поста (АГК) — уровнемеры и осадкомеры, а также подсистема гидрологического моделирования на основе математических моделей, позволяющая в ре- жиме реального времени давать прогноз притока с пошаговой коррекцией и оценивать ситуацию для принятия управляющих решений. Автор выражает благодарность профессорам, д-рам техн. наук М.М. Мордвинцеву, В.Л. Бондаренко и В.А. Волосухину за ценные замечания и пожелания по статье, улучшившие ее содержание. Литература 1. Волосухин Я.В. Обеспечение безопасности водопользования в Республике Крым [Текст] / Я.В. Волосухин, Д.Ю. Наволокин // Водоснабжение и санитарная техника. — 2017. — № 6. — С. 4–9. 2. Асарин А.Н. Расчетные паводки и безопасность плотин (по материалам СИГБ) [Текст] / А.Е. Асарин, В.М. Семенов // Гидротехническое строительство. — 1992. — № 8. — С. 55 3. О состоянии сооружений и возможности их эксплуатации объекта «Чернореченское водохранилище»: технический отчет [Текст] / Я.В. Волосухин; ИКЦ «Безопасность ГТС». — Новочеркасск, 2016. — 121 с. 4. Калустян Э.С. Оценка и роль рисков в плотиностроении (с использованием материалов СИГБ) [Текст] / Э.С. Калустян // Гидротехническое строительство. — 1999. — № 12. — С. 27–31. 5. Безопасность России. Энциклопедический словарьсправочник / науч. рук. Н.А. Махутов. — М.: Знание, 2008. — 528 с. 6. Тимченко З.В. Гидрография и гидрология рек Крыма [Текст]: монография / З.В. Тимченко. — Симферополь: АРИАЛ, 2012. — 290 с. 7. Паводок, половодье, наводнение — есть ли разница? — URL: https://fireman.club/statyi-polzovateley/pavodok-polovode-navodnenie-est-li-raznitsa 8. Паводки на реках Севастополя. — URL: http://xn-b1amnebsh.xn--80adi0aoagldk8i.xn--p1ai/glavnoe/ item/15950-2018-10-08-10-59-28.html 9. Государственный водный кадастр. Многолетние данные о режиме и ресурсах поверхностных вод суши [Текст]. — Т. 2. — Вып. 3. Бассейны Северского Донца, рек Крыма и Приазовья. — Л.: Гидрометеоиздат, 1985. — 362 с. 10. Чернореченское водохранилище. Правила эксплуатации [Текст] / Укргипроводхоз. — Киев, 1980. — Т. 1. — 271 с. 11. СП 33-101-2003. Определение основных расчетных гидрологических характеристик: введен взамен СНиП 2.01.14-83: дата введения 01.01.2004. — URL: http://docs. cntd.ru/document/1200035578 12. СП 58.13330.2012 Гидротехнические сооружения. Основные положения: веден взамен СНиП 33-01-2003: дата введения 2013-01-01. — URL: http://docs.cntd.ru/ document/1200094156 13. Энциклопедия безопасности гидротехнических сооружений (около 800 терминов) [Текст] / под общ. ред. В.А. Волосухина. –5-е изд., испр. и доп. — Новочеркасск: Изд-во ЮРГПУ (НПИ), 2017. — 138 с. 14. Schnitter N.J. A history of dams: the useful Pyramids. — Aa Balkema, 1994. — 282 p. 15. Proceedings of the Twenty-Sixth International Congress on Large Dams (4th–6th july 2018, Vienna — Austria). URL: https://www.taylorfrancis.com/books/e/9780429465086 References 1. Volosukhin Ya.V. Obespechenie bezopasnosti vodopol’zovaniya v Respublike Krym [Ensuring the safety of water use in the Republic of Crimea]. Vodosnabzhenie i sanitarnaya tekhnika [Water supply and sanitary equipment]. 2017, I. 6, pp. 4–9. 2. Asarin A.N. Raschetnye pavodki i bezopasnost’ plotin (po materialam SIGB) [Design floods and dam safety (based on SIGB materials)]. Gidrotekhnicheskoe stroitel’stvo [Hydrotechnical construction]. 1992, I. 8, p. 55. 3. O sostoyanii sooruzhenij i vozmozhnosti ih ekspluatacii ob”ekta «Chernorechenskoe vodohranilishche»: tekhnicheskij otchet [On the state of structures and the possibility of their operation of the facility “Chernorechenskoe reservoir”: technical report]. IKC «Bezopasnost’ GTS» [ICC “Safety of GTS”]. Novocherkassk, 2016. 121 p. 4. Kalustyan E.S. Ocenka i rol’ riskov v plotinostroenii (s ispol’zovaniem materialov SIGB) [Assessment and the role of risks in dam construction (using SIGB materials)]. Gidrotekhnicheskoe stroitel’stvo [Hydrotechnical construction]. 1999, I. 12, pp. 27–31. Строительство и архитектура (2019). Том 7. Выпуск 3 (24) 5. Bezopasnost’ Rossii. Enciklopedicheskij slovar’-spravochnik [Security of Russia. Encyclopedic Dictionary-Reference]. Moscow: Znanie Publ., 2008. 528 p. 6. Timchenko Z.V. Gidrografiya i gidrologiya rek Kryma [Hydrography and hydrology of the rivers of Crimea]. Smferopol’: IT «ARIAL» Publ., 2012. 290 p. 7. Pavodok, polovod’e, navodnenie — est’ li raznica? [Flood, flood, flood - is there any difference?]. Available at: https:// fireman.club/statyi-polzovateley/pavodok-polovode-navodnenie-est-li-raznitsa 8. Pavodki na rekah Sevastopolya [Floods on the rivers of Sevastopol]. Available at: http://xn--b1amnebsh.xn--80adi0aoagldk8i.xn--p1ai/glavnoe/item/15950-2018-10-08-10-59-28. html 9. Gosudarstvennyj vodnyj kadastr. Mnogoletnie dannye o rezhime i resursah poverhnostnyh vod sushi [The State Water Cadastre. Long-term data on the regime and resources of land surface water]. Bassejny Severskogo Donca, rek Kryma i Priazov’ya [The basins of the Seversky Donets, the rivers RIOR 17 Construction and Architecture (2019) Vol. 7. Issue 3 (24) of Crimea and the Sea of Azov]. L.: Gidrometeoizdat, 1985, V. 2, I. 3, 362 p. 10. Chernorechenskoe vodohranilishche. Pravila ekspluatacii [Chernorechenskoe reservoir. Rules of operation]. Ukrgiprovodhoz [Ukrgiprovodkhoz]. Kiev, 1980, V. 1, 271 p. 11. SP 33-101-2003. Opredelenie osnovnyh raschyotnyh gidrologicheskih harakteristik: vveden vzamen SNiP 2.01.14-83: data vvedeniya 01.01.2004 [SP 33-101-2003. Determination of the main calculated hydrological characteristics: introduced instead of SNiP 2.01.14-83: date of introduction 01.01.2004]. Available at: http://docs.cntd.ru/document/1200035578 12. SP 58.13330.2012 Gidrotekhnicheskie sooruzheniya. Osnovnye polozheniya: veden vzamen SNiP 33-01-2003: data vvedeniya 2013-01-01 [SP 58.13330.2012 Waterworks. Key points: replaced instead of SNiP 33-01-2003: date of introduction 2013-01-01]. Available at: http://docs.cntd.ru/document/1200094156 13. Enciklopediya bezopasnosti gidrotekhnicheskih sooruzhenij (okolo 800 terminov) [Encyclopedia of safety of hydraulic structures (about 800 terms)]. Novocherkassk: YuRGPU(NPI) Publ., 2017. 138 p. 14. Schnitter N.J. A history of dams: the useful Pyramids. Aa Balkema, 1994. 282 p. 15. Proceedings of the Twenty-Sixth International Congress on Large Dams (4th–6th july 2018, Vienna – Austria). Available at: https://www.taylorfrancis.com/books/e/9780429465086 18 Строительство и архитектура (2019). Том 7. Выпуск 3 (24) RIOR
https://openalex.org/W3120917944	https://www.iberpapers.org/jour/article/download/406/314	Russian	null	Yuri V. Dubinin: a biography	Iberoamerikanskie tetradi	2,021	cc-by	1,132	No. 2 No. 2 Cuadernos Iberoamericanos Vol. VIII 2020: 81-83 Золотой фонд • Fondo de oro Золотой фонд • Fondo de oro И.В. Попов https://doi.org/10.46272/2409- 3416-2020-8-2-81-83 7 октября 2020 года исполнилось 90 лет со дня рождения видного советского и российского дипломата Юрия Владимировича Дубинина. Редакционный материал Ю.В. Дубинин оставил яркий след в советской дипломатии периодов «холодной войны», разрядки и пе рестройки. Не менее значителен его вклад в реализацию внешнеполитического курса постсоветской России. Иван Владимирович Попов, к.ист.н., доцент кафедры дипломатии МГИМО, заместитель главного редактора журнала «Ибероамериканские тетради», Москва (Россия) Говоря о Юрии Владимировиче, необходимо, пре жде всего, отметить, что он принадлежал к плеяде замеча тельных советских дипломатов, которые сформировались как настоящие профессионалы в системе Московского государственного института международных отношений. Восьмой выпуск 1955 года, как и предыдущие семь, дал Ми нистерству перспективную и профессионально подготов ленную молодежь, из которой очень быстро сформировался новый костяк ведомства. Этой молодежи предстояло решать принципиально иные, чем ранее, задачи в сфере внешней политики Советского государства, которое по итогам войны превратилось в реальный центр силы мирового уровня. E-mail: ivanvladpopov@gmail.com Для корреспонденции: 119454, Россия, Москва, просп. Вернадского, 76 Для цитирования: Попов И.В. Юрий Владимирович Дубинин: биографический очерк. Ибероамериканские тетради. 2020, 8(2): 81-83. https://doi. org/10.46272/2409-3416-2020- 8-2-81-83. Материал поступил в редакцию: 26.10.2020 Принят к публикации: 14.11.2020 Принят к публикации: 14.11.2020 В стенах МИД Ю.В. Дубинин сделал блестящую дипломатическую карьеру, которую можно изучать как об разец для последующих поколений, несмотря на все отличия современных условий от периода послевоенного становления нашей дипломатической службы. Важно отметить, что эта карьера стала возможной благодаря объективным факторам, помноженным на выдающиеся личные качества, которые будущий посол неустанно развивал. Принят к публикации: 14.11.2020 В стенах МИД Ю.В. Дубинин сделал блестящую дипломатическую карьеру, которую можно изучать как об разец для последующих поколений, несмотря на все отличия современных условий от периода послевоенного становления нашей дипломатической службы. Важно отметить, что эта карьера стала возможной благодаря объективным факторам, помноженным на выдающиеся личные качества, которые будущий посол неустанно развивал. Принят к публикации: 14.11.2020 Стартовая площадка Можно сказать, что Юрий Владимирович личным примером наглядно проде монстрировал работу социальных лифтов. Провинциальный молодой человек поступил в набиравший престиж вуз самостоятельно, за счет личных качеств. Движение к верши нам дипломатической карьеры он начал с самых начальных ступеней служебной лест ницы, пройдя путь от стажера Посольства СССР во Франции до полномочного предста вителя страны в нескольких важных для советской (российской) дипломатии столицах мира: Мадрид, Нью-Йорк, Вашингтон, Париж, Киев. 81 Ибероамериканские тетради Том VIII 2020: 81-83 № 2 Личные качества Ю.В. Дубинин имел отличную память, выдающиеся способности к иностран ным языкам и, главное, – способность обретать новое знание на протяжении всей жизни. К этому необходимо добавить безупречный стиль делового письма, столь нужный в ди пломатической профессии, умение расположить к себе самых разных собеседников: от убежденных западных антисоветчиков до высоких советских и российских (а позже – и украинских) бюрократов. Тонкий психолог, Ю.В. Дубинин умел использовать субъективные моменты для решения практических задач любого уровня, которые Центр ставил перед своими зару бежными представительствами. Красивый и элегантный тандем, который он образовал со своей супругой Лианой Завеновной (также выпускницей МГИМО), был в центре обще ственного внимания, где и когда это было необходимо для дела. Многогранность личности Ю.В. Дубинина подтверждается его широким круго зором. Он профессионально занимался практической политикой, страноведением, по литической наукой и теорией переговоров; профессионально переводил художествен ные произведения на русский язык с его любимого французского. В то же время, все это органично было «заточено» под профессиональный успех. No. 2 советских времен, Ю.В. Дубинин вел жесткую и последовательную линию на защиту жиз ненных интересов России, добился эффективных договоренностей с украинскими пар тнерами по важнейшим вопросам обеспечения национальной безопасности РФ. Будучи человеком долга, посол Ю.В. Дубинин жестко требовал того же от своих подчиненных. Работавшие с ним дипломаты отмечают его умение нацелить сотрудников на достижение конкретных результатов. Неукоснительная исполнительская дисциплина возглавлявшихся им коллективов в центральном аппарате или за рубежом достигалась «железной рукой в бархатной перчатке». Спокойствие и безукоризненная выдержка ру ководителя не делали менее убедительными его требования к качеству и объему по ставленных задач. Корректно, но твердо посол мог поправить дипломата любого ранга, если на то была «производственная необходимость». Наконец, нельзя не отметить исключительно плодотворный период творчества Ю.В. Дубинина после ухода в отставку из МИД. Вернувшись в альма-матер, Юрий Влади мирович не только успел передать сотням будущих международников свой неоценимый опыт, но и оставил богатое мемуарное наследие, описав основные этапы своей работы на профессиональной стезе дипломата. Чувство долга Вероятно, это – важнейшее качество для честного государственного служаще го, тем более на стезе дипломатии. Ю.В. Дубинин образцово исполнял самые сложные и непростые поручения, которые ему выпадали на протяжении дипломатической карьеры. Блестящий эксперт по Франции при необходимости переквалифицировался в специалиста по многосторонним переговорам самого высокого уровня в рамках процес са общеевропейской безопасности. После успеха в рамках СБСЕ – новая задача: разви тие фактически с нуля советско-испанских отношений. Ю.В. Дубинин не только мастерски овладел испанским языком: он сумел расположить к себе представителей самых разных политических и общественных кругов страны, выходившей из многолетнего провинциа лизма франкистской полуизоляции – от коммунистов и левых республиканцев до полити ческой, социальной и экономической элиты «глубинного» испанского государства. По существу вся служебная карьера Ю.В. Дубинина – это смена, подчас ра дикальная, политического досье, которым ему надо было заниматься на уровне руко водителя. В этом смысле особо следует выделить завершающий, «украинский» аккорд дипломатической службы Ю.В. Дубинина. С 1992 года он непосредственно участвовал в двусторонних переговорах с представителями Украины по судьбе Черноморского фло та, а с 1996 до выхода в отставку в 1999 году – непосредственно в Киеве в качестве посла России закладывал основы межгосударственных отношений между двумя странами. В свое время Министр иностранных дел С.В. Лавров охарактеризовал эту мис сию Дубинина как: «…наверное, наиболее драматическое назначение … в период, когда никто не понимал, что делать и что произошло. Это, конечно, был серьезный вызов для классического дипломата, который великолепно знал западное политическое направле ние, болел душой за события в своей стране и которому выпала доля решать серьезней шие проблемы, начиная с вопроса о судьбе ядерного оружия на Украине, договоренно стей по Черноморскому флоту». Сумев подняться над иллюзиями и шаблонами восприятия 82 Cuadernos Iberoamericanos No. 2 Vol. VIII 2020: 81-83 Vol. VIII 2020: 81-83 I.V. Popov https://doi.org/10.46272/2409-3416-2020-8-2-81-83 Editorial Ivan V. Popov, PhD (History), Associate Professor, Department of Diplomacy, MGIMO University; Deputy Editor-In- Chief, Cuadernos Iberoamericanos, Moscow (Russia) E-mail: ivanvladpopov@gmail.com For correspondence: 119454, Russia, Moscow, av. Vernadskogo, 76 For citation: Popov I.V. Yuri V. Dubinin: a biography. Cuadernos Iberoamericanos 8, no. 2 (2020): 81-83. https://doi.org/10.46272/2409-3416-2020-8-2-81-83. Received: 26.10.2020 Accepted: 14.11.2020 https://doi.org/10.46272/2409-3416-2020-8-2-81-83 https://doi.org/10.46272/2409-3416-2020-8-2-81-83 Editorial Ivan V. Popov, PhD (History), Associate Professor, Department of Diplomacy, MGIMO University; Deputy Editor-In- Chief, Cuadernos Iberoamericanos, Moscow (Russia) E-mail: For correspondence: 119454, Russia, Moscow, av. Vernadskogo, 76 For correspondence: 119454, Russia, Moscow, av. Vernadskogo, 76 For citation: Popov I.V. Yuri V. Dubinin: a biography. Cuadernos Iberoamericanos 8, no. 2 (2020): 81-83. https://doi.org/10.46272/2409-3416-2020-8-2-81-83. For citation: Popov I.V. Yuri V. Dubinin: a biography. Cuadernos Iberoamericanos 8, no. 2 (2020): 81-83. https://doi.org/10.46272/2409-3416-2020-8-2-81-83. Received: 26.10.2020 Received: 26.10.2020 Accepted: 14.11.2020 Accepted: 14.11.2020 83
https://openalex.org/W2207092416	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0133104&type=printable	English	null	Correction: A Continent-Wide Migratory Divide in North American Breeding Barn Swallows (Hirundo rustica)	PloS one	2,015	cc-by	466	CORRECTION OPEN ACCESS Citation: Hobson KA, Kardynal KJ, Van Wilgenburg SL, Albrecht G, Salvadori A, Cadman MD, et al. (2015) Correction: A Continent-Wide Migratory Divide in North American Breeding Barn Swallows (Hirundo rustica). PLoS ONE 10(7): e0133104. doi:10.1371/ journal.pone.0133104 Published: July 14, 2015 Correction: A Continent-Wide Migratory Divide in North American Breeding Barn Swallows (Hirundo rustica) Keith A. Hobson, Kevin J. Kardynal, Steven L. Van Wilgenburg, Gretchen Albrecht, Antonio Salvadori, Michael D. Cadman, Felix Liechti, James W. Fox Fig 1 is incorrect. Please see the corrected Fig 1 here. a1111 a1111 Fig 1 is incorrect. Please see the corrected Fig 1 here. PLOS ONE \| DOI:10.1371/journal.pone.0133104 July 14, 2015 Published: July 14, 2015 Copyright: © 2015 Hobson et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. 1 / 2 PLOS ONE \| DOI:10.1371/journal.pone.0133104 July 14, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0133104 July 14, 2015 Fig 1. Estimated fall and spring migration routes and wintering sites for barn swallows from five populations where archival light-level geolocators were deployed. Predicted fall (black lines) and spring (red lines) migration routes and mean wintering locations (black dots) from Kalman filter state-space models. Dashed lines are estimated routes during periods when locations could not be accurately assessed (e.g., equinox, unrealistic points). Ellipses represent directional standard deviations of points used to calculate mean wintering locations. Breeding locations are denoted (stars). Data are for 13 females, 13 males and 1 unknown sex. The last panel indicates mean ± SD and SE (hash marks) for estimated mean wintering locations for five populations (blue = Washington State, green = Saskatchewan, red = Ontario, black = New Brunswick, orange = Alabama). Fig 1. Estimated fall and spring migration routes and wintering sites for barn swallows from five populations where archival light-level geolocators were deployed. Predicted fall (black lines) and spring (red lines) migration routes and mean wintering locations (black dots) from Kalman filter state-space models. Dashed lines are estimated routes during periods when locations could not be accurately assessed (e.g., equinox, unrealistic points). Ellipses represent directional standard deviations of points used to calculate mean wintering locations. Breeding locations are denoted (stars). Data are for 13 females, 13 males and 1 unknown sex. The last panel indicates mean ± SD and SE (hash marks) for estimated mean wintering locations for five populations (blue = Washington State, green = Saskatchewan, red = Ontario, black = New Brunswick, orange = Alabama). doi:10.1371/journal.pone.0133104.g001 1. Hobson KA, Kardynal KJ, Van Wilgenburg SL, Albrecht G, Salvadori A, Cadman MD, et al. (2015) A Continent-Wide Migratory Divide in North American Breeding Barn Swallows (Hirundo rustica). PLoS ONE 10(6): e0129340. doi: 10.1371/journal.pone.0129340 PMID: 26065914 Reference 2 / 2
https://openalex.org/W2046330376	https://europepmc.org/articles/pmc3795348?pdf=render	English	null	Diminished acquired equivalence yet good discrimination performance in older participants	Frontiers in psychology	2,013	cc-by	9,660	INTRODUCTION Dc−). Both groups’ performances were reduced by the rever- sal from the original stages and both recovered; however, group Whole’s performance recovered more quickly than group Part’s did. It is this feature of the data that challenges alternative conﬁg- ural learning theories (e.g., Rescorla, 1976; Pearce, 2002). Notice that in the pre-reversed discrimination these Skinner boxes indi- cate the equivalent reinforcement arrangements for the tone and click. Informally expressed, it is as though rats’ representations of Skinner boxes A and C (and B and D) had “acquired equiva- lence” during pre-reversal training. Thus, new learning during the reversal may transfer between A and C (and between B and D). For group Part, the acquired equivalence between A and C (and between B and D) will lead to conﬂicting information because A and C no longer indicated equivalent tone/click reinforcement relationships. But for group Whole, although the tone/click rein- forcement relationships have all reversed, A and C (and B and D) remain equivalent. It is notable that non-conﬁgural forms of acquired equivalence are possible (e.g., Honey and Hall, 1989) but they are interpretable in simpler terms than those considered here (e.g., Ward-Robinson and Hall, 1999). Dc−). Both groups’ performances were reduced by the rever- sal from the original stages and both recovered; however, group Whole’s performance recovered more quickly than group Part’s did. It is this feature of the data that challenges alternative conﬁg- ural learning theories (e.g., Rescorla, 1976; Pearce, 2002). Notice that in the pre-reversed discrimination these Skinner boxes indi- cate the equivalent reinforcement arrangements for the tone and click. Informally expressed, it is as though rats’ representations of Skinner boxes A and C (and B and D) had “acquired equiva- lence” during pre-reversal training. Thus, new learning during the reversal may transfer between A and C (and between B and D). For group Part, the acquired equivalence between A and C (and between B and D) will lead to conﬂicting information because A and C no longer indicated equivalent tone/click reinforcement relationships. But for group Whole, although the tone/click rein- forcement relationships have all reversed, A and C (and B and D) remain equivalent. It is notable that non-conﬁgural forms of acquired equivalence are possible (e.g., Honey and Hall, 1989) but they are interpretable in simpler terms than those considered here (e.g., Ward-Robinson and Hall, 1999). ORIGINAL RESEARCH ARTICLE published: 11 October 2013 doi: 10.3389/fpsyg.2013.00726 Reviewed by: Correspondence: Jasper Robinson, School of Psychology, The University of Nottingham, Room B33, University Park, NG7 2RD Nottingham, UK e-mail: jasper.robinson@ nottingham.ac.uk Keywords: acquired equivalence, attentional set, ageing, discrimination learning, connectionism, associative learning, healthy aging, conﬁgural processing Jasper Robinson and Emma Owens School of Psychology, The University of Nottingham, Nottingham, UK School of Psychology, The University of Nottingham, Nottingham, UK We asked younger and older human participants to perform computer-based conﬁgural discriminations that were designed to detect acquired equivalence. Both groups solved the discriminations but only the younger participants demonstrated acquired equivalence. The discriminations involved learning the preferences [“like” (+) or “dislike” (−)] for sports [e.g., tennis (t) and hockey (h)] of four ﬁctitious people [e.g., Alice (A), Beth (B), Charlotte (C), and Dorothy (D)]. In one experiment, the discrimination had the form: At+, Bt−, Ct+, Dt−, Ah−, Bh+, Ch−, Dh+. Notice that, e.g., Alice and Charlotte are “equivalent” in liking tennis but disliking hockey. Acquired equivalence was assessed in ancillary components of the discrimination (e.g., by looking at the subsequent rate of “whole” versus “partial” reversal learning). Acquired equivalence is anticipated by a network whose hidden units are shared when inputs (e.g., A and C) signal the same outcome (e.g., +) when accompanied by the same input (t). One interpretation of these results is that there are age-related differences in the mechanisms of conﬁgural acquired equivalence. Keywords: acquired equivalence, attentional set, ageing, discrimination learning, connectionism, associative learning, healthy aging, conﬁgural processing Edited by: Edited by: Robin A. Murphy, University of Oxford, UK Edited by: Robin A. Murphy, University of Oxford, UK Reviewed by: Irina Baetu, University of Adelaide, Australia Sam Hannah, University of Saskatchewan, Canada *Correspondence: Jasper Robinson, School of Psychology, The University of Nottingham, Room B33, University Park, NG7 2RD Nottingham, UK e-mail: jasper.robinson@ nottingham.ac.uk INTRODUCTION Experiments on “acquired equivalance” have revealed important information about the way in which animals encode stimulus representations. For example, Honey and Ward-Robinson (2001) gave rats acquired equivalence training in which a tone would signal food delivery (t+) and a clicker would not (c−) in two distinctly decorated Skinner boxes (A and C). But in two other Skinner boxes (B and D), the tone and click signalled the alterna- tive outcome (i.e., t−and c+). The complete discrimination can be represented as: At+, Ac−, Bt−, Bc+, Ct+, Cc−, Dt−, Dc+. It was evident that rats had learned the discrimination because they anticipated the delivery of food on reinforced (+) trials by approaching the site of delivery and refrained from this on the non-reinforced (−) trials. Notice that no single stimulus uniquely predicts either outcome: all stimuli are equally often reinforced and non-reinforced and it is necessary for rats to learn about spe- ciﬁc conﬁgurations of stimuli. Inﬂuential theoretical accounts of such learning (e.g., Rescorla, 1976; Pearce, 2002) provide accounts of solution of the At+, Ac−, Bt−, Bc+, Ct+, Cc−, Dt−, Dc+ discrimination in which the eight trial types are represented by eight “conﬁgural” stimuli, each being associated with the appro- priate outcome. However, Honey and Ward-Robinson gave an additional stage of training that produced results not anticipated by these models. In the subsequent stage, rats were split into two groups to receive different types of “reversal training,” in which at least some of the trial outcomes were switched. For group Whole, all trial types were reversed (i.e., At−, Ac+, Bt+, Bc−, Ct−, Cc+, Dt+, Dc−) but for group Part only half of the trial types were reversed (At+, Ac−, Bt−, Bc+, Ct+, Cc−, Dt+, Honey et al. (2010) describe this ﬁnding, and others like them (e.g., Ward-Robinson and Honey, 2000; Hodder et al., 2003), in terms of a three-layer connectionist network, which will be described in detail in the Discussion. Those authors also note that their model will adequately explain the ﬁnding that dis- criminations involving “intra-dimensional” shifts are mastered more quickly than those involving “extra-dimensional” shifts (e.g., Owen et al., 1991; Barense et al., 2002). This sugges- tion is theoretically signiﬁcant because, if substantiated, it would October 2013 \| Volume 4 \| Article 726 \| 1 www.frontiersin.org Ageing and acquired equivalence Robinson and Owens FIGURE 1 \| Example of treatments given to the younger and older participants in Experiment 1. EXPERIMENT 1 d d Honey and Ward-Robinson (2001) demonstrated acquired equiv- alence in rats using an appetitive conditioning procedure. We adapted their procedure for use with older and younger partici- pants in Experiment 1 whose design is summarized in Figure 1. Older and younger participants were required to learn about four ﬁctitious characters’ like or dislike of two sports. Two of the char- acters liked the same two sports and disliked the two alternative sports (Stage 1). Acquired equivalence could be demonstrated over a series of “reversals” (Stage 2 and Stage 3) in which some or all of the previously liked sports became disliked and vice versa. If the two pairs of characters had acquired equivalence, participants’ performance should recover more rapidly from the whole reversal than from the part reversal. The new question we asked here was: would this acquired equivalence effect be different in a group of older participants? FIGURE 1 \| Example of treatments given to the younger and older FIGURE 1 \| Example of treatments given to the younger and older participants in Experiment 1. Participants are required to learn whether four ﬁctitious characters, Alice, Beth, Charlotte, and Dorothy, like or dislike the sports tennis and hockey. In each of three stages, two characters like and two dislike each of the two sports; these patterns of liking and disliking are complemented by the remaining two characters. In the example of a “whole reversal” treatment in the top panel, the stage-2 treatment consists of a full reversal of the pattern of the characters’ liked and disliked sports. But in the example of a “part reversal” only two of the characters’ liked and disliked two sports reverses (viz., Beth, and Charlotte), whereas the other two characters’ (viz., Alice, and Dorothy) liked and disliked sports remain unchanged from stage 1. Stage-3 training was identical to Stage-1 training, and was intended to offer an additional attempt to examine the effect of whole or part reversal. FIGURE 1 \| Example of treatments given to the younger and older participants in Experiment 1. Participants are required to learn whether four ﬁctitious characters, Alice, Beth, Charlotte, and Dorothy, like or dislike the sports tennis and hockey. In each of three stages, two characters like and two dislike each of the two sports; these patterns of liking and disliking are complemented by the remaining two characters. PARTICIPANTS Group Y comprised two men and fourteen women with a mean age of 20.8 years (range: 20–24 years); Group O comprised seven men and eight women with a mean age of 64.4 years (range: 55– 77 years). Participants were a self-selected sample of respondents to recruitment posters in public places (cafés, Post Ofﬁces, etc., Group O) or were University of Nottingham students who gained course credit for participation (Group Y). All participants were naive with respect to the stimuli used in the experiment. y The following cartoon depictions were used as stimuli: (a) a pair of crossed tennis rackets and ball with “Tennis” written below them; (b) a hockey stick and ball with “Hockey” written below them; (c) four characters’ faces with “neutral” facial expres- sions, “happy” facial expressions and “sad” facial expressions (i.e., twelve images of the characters). Each character’s neutral image had her name (Alice, Beth, Charlotte or Dorothy) written below it. Each character’s happy image had “X Likes this Sport” written above it where X is the character’s name; sad images were simi- larly accompanied by text that read “X Dislikes this Sport.” These stimuli occupied a screen area of around 400 mm2. Auditory data ﬁles had been created on a computer (iMac, Apple Computers) using a synthetic voice (“Victoria”). These ﬁles read aloud the EXPERIMENT 1 d d In the example of a “whole reversal” treatment in the top panel, the stage-2 treatment consists of a full reversal of the pattern of the characters’ liked and disliked sports. But in the example of a “part reversal” only two of the characters’ liked and disliked two sports reverses (viz., Beth, and Charlotte), whereas the other two characters’ (viz., Alice, and Dorothy) liked and disliked sports remain unchanged from stage 1. Stage-3 training was identical to Stage-1 training, and was intended to offer an additional attempt to examine the effect of whole or part reversal. INTRODUCTION Participants are required to learn whether four ﬁctitious characters, Alice, Beth, Charlotte, and Dorothy, like or dislike the sports tennis and hockey. In each of three stages, two characters like and two dislike each of the two sports; these patterns of liking and disliking are complemented by the remaining two characters. In the example of a “whole reversal” treatment in the top panel, the stage-2 treatment consists of a full reversal of the pattern of the characters’ liked and disliked sports. But in the example of a “part reversal” only two of the characters’ liked and disliked two sports reverses (viz., Beth, and Charlotte), whereas the other two characters’ (viz., Alice, and Dorothy) liked and disliked sports remain unchanged from stage 1. Stage-3 training was identical to Stage-1 training, and was intended to offer an additional attempt to examine the effect of whole or part reversal. undermine claims that non-human animals’ demonstrations of intra-dimensional transfer are not actually demonstrations of a genuine attentional process (cf., Mackintosh, 1974). It is also clinically signiﬁcant because intra-/extra-dimensional shift exper- iments in rats demonstrate the role of the prefrontal cortex in “attentional set” in frontal lobe disorders (Owen et al., 1991; Dias et al., 1997; Birrel and Brown, 2000; Hampshire and Owen, 2010). Deﬁcits in performance in intra-/extra-dimensional shift have been reported in apparently healthy, older human volunteers (Owen et al., 1991; see also, Barense et al., 2002). That observation and Honey et al.’s assertion that the same psychological pro- cesses outlined in their model, govern not only acquired equiv- alence and intra-/extra-dimensional set shifting, make several predictions. In particular, manipulations that affect intra-/extra- dimensional set shifting, should also affect acquired equivalence. We report here results of two experiments that support that pre- diction by demonstrating acquired equivalence performance to be diminished in (healthy) older participants relative to younger participants. MATERIALS AND METHODS PARTICIPANTS side. With the exception of the keys numbered 1, 2, 3, 4, and 5 that ran along the top of the QWERTY part of the keyboard and the space key, black stickers covered the letter/number of each key. Our intention here was to direct participants’ responses to the keys 1, 2, 3, 4, and 5 during the experiment. A pair of headphones (Panasonic RP-HT225) was plugged into the laptop and was used to present the auditory stimuli described below. PROCEDURE Participants from Groups Y and O were randomly assigned to whole-reversal (W) or part-reversal (P) groups, to create Group YW, Group YP, Group OW, and Group OP, see Figure 1. Participants’ mean ages in groups YW, YP, OW, and OP were, respectively 20.1, 21.5, 64.6, and 64.3 years. There were seven woman and one man in both Group YW and YO; there were ﬁve women and two men in Group OW; and there were three women and ﬁve men in Group OP. During stage-1 training, each of the eight trial types was given 24 times in an irregular sequence (i.e., 96 liked and 96 disliked trials). During stage-2 training, participants received an identical treatment but on a reversed version of the task: for participants in Groups YW and OW, all four of the characters now liked the sports they had previously disliked and now disliked the sports they had previously liked. For participants in Groups YP and OP, only two of the characters’ opinions of the sports reversed. Stage 3 was the ﬁnal stage of training and was identical to the ﬁrst except that only sixteen trials of each of the eight trial types was given. Group YP and OPs’ partial reversals were systematically varied and were arranged so that each particular discrimination was matched by a subgroup of YW and OW. Thus, performance differences between whole/part reversals could not be attributed to differences in difﬁculty of the speciﬁc trials types in their discrimination. All participants were given training in which they were asked to learn which sports (Tennis and Hockey) four ﬁctitious char- acters (Alice, Beth, Charlotte, and Dorothy) liked. Participants keyed “5” for liked sports and “1” for disliked sports. Keys in- between could be used for less conﬁdent responses. For the purposes of feedback (see below), keying 4 or 5 were “cor- rect” on like trials and “incorrect” on dislike trials; and keying 1 or 2 were “correct” on dislike trials and “incorrect” on like trials. Keying 3 was neither correct nor incorrect. For all par- ticipants, each sport was liked by two of the characters and disliked by the other two characters; each character liked one sport and disliked the other. Thus, each character agreed in her opinion of the two sports with one other character and disagreed with the two remaining characters. APPARATUS AND STIMULI Experiments were run in a small quiet room in the School of Psychology, University of Nottingham. Stimuli were presented and responses were recorded on a laptop (Toshiba Portégé A200). From opposite corners its screen measured 31 cm. Participants used a separate keyboard that was connected to the laptop and positioned such that the participant could use the keyboard while looking at the laptop screen. The keyboard consisted of a stan- dard QWERTY keyboard with a number pad to the right hand October 2013 \| Volume 4 \| Article 726 \| 2 Frontiers in Psychology \| Personality Science and Individual Differences Robinson and Owens Ageing and acquired equivalence text “Correct,” “Incorrect” and “Have a guess next time” and had durations of between 0.5 and 1.5 s. During the experiment, the character and sport stimuli were presented side by side and ver- tically central. The character appeared only on the left-hand side; the sport appeared only on the right-hand side. A numbered scale comprising the numbers 1, 2, 3, 4, and 5 could be presented below the neutral images of the characters’ faces with 1 on the left and 5 on the right. The word “Dislike” appeared to the left of “1” and the word “Like” appeared to the right of “5.” text “Correct,” “Incorrect” and “Have a guess next time” and had durations of between 0.5 and 1.5 s. During the experiment, the character and sport stimuli were presented side by side and ver- tically central. The character appeared only on the left-hand side; the sport appeared only on the right-hand side. A numbered scale comprising the numbers 1, 2, 3, 4, and 5 could be presented below the neutral images of the characters’ faces with 1 on the left and 5 on the right. The word “Dislike” appeared to the left of “1” and the word “Like” appeared to the right of “5.” The sequence of one type of trial is exempliﬁed in Figure 2. Each trial consisted of: (1) The 1.5-s, centrally located presen- tation of the keyboard character “+,” (2) The presentation of a character and a sport, during which the participant had unlim- ited time to select a response from 1 (dislike) to 5 (like) of the scale that was presented below them, (3) Information about the characters like/dislike of the sport was given for 2.9 s. APPARATUS AND STIMULI This com- prised presentation of text (e.g., Alice likes this Sport) with the accompanying “happy” or “sad” version of the character and the Tennis or Hockey picture, (4) Auditory feedback was given (“Correct,” “Incorrect” or, where 3 was keyed, “Have a guess next time”). www.frontiersin.org PROCEDURE We counterbal- anced stimulus arrangements such that for some participants Alice and Charlotte (and therefore, Beth and Dorothy) had equivalent sports opinions and for others Alice and Beth (and therefore, Charlotte and Dorothy) has equivalent sports opin- ions. Neither Alice and Dorothy nor Beth and Charlotte shared sports opinions for any participants. For some participants the shared opinions were based on liking Tennis (and, therefore, dis- liking Hockey); for others participants the shared opinions were based on liking Hockey (and therefore, disliking Tennis). The orthogonal arrangement of this counterbalancing created four different discriminations, which were given to similar numbers of participants. FIGURE 2 \| Illustration of the the sequence of events on each trial of training in Experiments 1 and 2. (1) The leftmost panel represents the presentation on the computer’s screen of the ﬁxation cross (“+”), which occurred at the beginning of each trial for 1.5 s. There was no requirement of the participant; (2) The central panel represents the presentation on the computer screen of the character, the sport and the rating scale. This example represents a trial in which a participant was asked to rate Alice’s like/dislike of tennis, however, other trial types occurred (see, e.g., Figures 1, 3). This slide remained until the participant had made their rating; (3) The rightmost panel represents the feedback given to the participant following their previous rating. In this example, the participant correctly rated Alice as liking tennis (i.e., the participant gave Alice a rating of ≥4 for tennis), which was accompanied by the spoken word “Correct” and by the statement “Alice likes this sport.” Full details of the feedback given on incorrect trials and on the other types of trial are given above. FIGURE 2 \| Illustration of the the sequence of events on each trial of training in Experiments 1 and 2. (1) The leftmost panel represents the i h ’ f h ﬁ i (“ ”) hi h FIGURE 2 \| Illustration of the the sequence of events on each trial of training in Experiments 1 and 2. (1) The leftmost panel represents the presentation on the computer’s screen of the ﬁxation cross (“+”), which occurred at the beginning of each trial for 1.5 s. There was no requirement of the participant; (2) The central panel represents the presentation on the computer screen of the character, the sport and the rating scale. PROCEDURE This example represents a trial in which a participant was asked to rate Alice’s like/dislike of tennis, however, other trial types occurred (see, e.g., Figures 1, 3). This slide remained until the participant had made their rating; (3) The rightmost panel represents the feedback given to the participant following their previous rating. In this example, the participant correctly rated Alice as liking tennis (i.e., the participant gave Alice a rating of ≥4 for tennis), which was accompanied by the spoken word “Correct” and by the statement “Alice likes this sport.” Full details of the feedback given on incorrect trials and on the other types of trial are given above. FIGURE 2 \| Illustration of the the sequence of events on each trial of training in Experiments 1 and 2. (1) The leftmost panel represents the presentation on the computer’s screen of the ﬁxation cross (“+”), which occurred at the beginning of each trial for 1.5 s. There was no requirement of the participant; (2) The central panel represents the presentation on the computer screen of the character, the sport and the rating scale. This example represents a trial in which a participant was asked to rate Alice’s like/dislike of tennis, however, other trial types occurred (see, e.g., Figures 1, 3). This slide remained until the participant had made their rating; (3) The rightmost panel represents the feedback given to the participant following their previous rating. In this example, the participant correctly rated Alice as liking tennis (i.e., the participant gave Alice a rating of ≥4 for tennis), which was accompanied by the spoken word “Correct” and by the statement “Alice likes this sport.” Full details of the feedback given on incorrect trials and on the other types of trial are given above. FIGURE 2 \| Illustration of the the sequence of events on each trial of training in Experiments 1 and 2. (1) The leftmost panel represents the presentation on the computer’s screen of the ﬁxation cross (“+”), which occurred at the beginning of each trial for 1.5 s. There was no requirement of the participant; (2) The central panel represents the presentation on the computer screen of the character, the sport and the rating scale. This example represents a trial in which a participant was asked to rate Alice’s like/dislike of tennis, however, other trial types occurred (see, e.g., Figures 1, 3). RESULTS interaction, F(5, 135) = 2.3, p < 0.046, η2 p = 0.079 were signiﬁ- cant. No other main effect was signiﬁcant. The source of the interaction involving the age variable was examined using simi- lar analyses separated for young and older participants. Analysis of older participants’ data yielded a main effect of reversal stage only, F(1, 13) = 15.0, p < 0.003, η2 p = 0.537. No other statistic was signiﬁcant and, of most importance, none was signiﬁcant that involved the reversal-group variable, smallest p > 0.121. However, the corresponding analysis of the younger participants’ data yielded reliable main effects of cycle, reversal, and trial, and reliable Reversal × Trial, and Reversal × Trial × Reversal Group interactions, largest p < 0.024, F(5, 70) = 2.7, η2 p = 0.167. The source of the younger participants’ Reversal × Trial × Reversal Group interaction was examined using a pair of ANOVAs with data split across the reversal stage variable (i.e., on established dis- crimination data and reversed data) with only cycle and reversal group as variable. No signiﬁcant statistics were obtained for the established discrimination data, smallest p > 0.134. The corre- sponding ANOVA for the reversed data yielded a signiﬁcant main effects of cycle and trial and a signiﬁcant Trial × Reversal Group interaction, largest p < 0.012, F(1, 29) = 7.3, η2 p = 0.203. Simple main-effects analysis on this interaction using separate error terms for each trial, revealed younger participants’ whole rever- sal performance to be superior to partial reversal performance on the ﬁfth trial, F(1, 14) = 13.7, p < 0.003, η2 p = 0.495. Initial examination of raw trial-by-trial data revealed that all participants mastered the task rapidly, for example reaching asymptotic discrimination after about six blocks of the eight trial types. Our analysis focuses, therefore, on the terminal six trials of established training and the initial six trials of the reversal. Here, group differences were not masked by rapid discrimina- tion learning. Data on “dislike” trials were transformed to match the scale of the “like” trials. That is, 1 (the correct response) was recoded as 5, 2 as 4, 3 remained as 3, 4 as 2 and 5 as 1. This obtained a like/dislike-independent response measure in which 5 s indicate the correct response and 1 s indicate the incorrect response. These data are summarized in Figure 3. RESULTS We see that all four groups’ performance before both reversals was good (around the asymptote of 5) and that it declined on both of the reversals, recovering quickly. Inspection of the two Y groups, indicates that group YW recovered from the disruption of the reversal more quickly than group YP. However, no such pattern can be seen in the O groups: groups OW and OP show no obvious differ- ence in recovery. This description of the data was supported by an analysis of variance (ANOVA) with within-subject variables of: (1) cycle (i.e., the ﬁrst and second twelve-trial cycles of estab- lished discrimination and subsequent reversal), (2) established training (end of stages 1 and 2) versus reversal stage (begin- nings of stages 2 and 3), and (3) trial; and between-subject variables of: (1) age (i.e., Y versus O), and (2) reversal group (i.e., W versus P). The analysis revealed main effects of trial, F(5, 135) = 4.9, p < 0.001, η2 p = 0.156, reversal stage, F(1, 27) = 40.7, p < 0.001, η2 p = 0.602 and cycle, F(1, 27) = 5.0, p < 0.034, η2 p = 0.158. The Cycle × Trial × Age interaction, F(5, 135) = 2.9, p < 0.017. η2 p = 0.097, and the Cycle × Trial × Reversal Stage EXPERIMENT 2 The results of Experiment 1 join those of Honey and Ward- Robinson (2001) and Hodder et al. (2003) in showing an acquired equivalence effect by an improved rate of “whole” reversal learn- ing relative to “part” reversal learning in younger participants. FIGURE 3 \| Means, and one standard error of each mean, of data from Experiment 1. The leftmost four sets of six trials of data are from the two younger participant groups (groups YW and YP); the rightmost data are the corresponding data from two older participant groups (groups OW and OP). For all four groups, and running from left to right, the four sets of data represent: (1) The ﬁnal six liked and ﬁnal six disliked trials of stage 1 (i.e., trials 91 through to 96 of stage 1); (2) the ﬁrst six liked and ﬁrst six disliked trials of stage 2 (i.e., trials 1 through to 6 of stage 2); (3) the ﬁnal six liked and ﬁnal six disliked trials of stage 2 (i.e., trials 91 through to 96 of stage 2); (4) the ﬁrst six liked and ﬁrst six disliked trials of stage 3 (i.e., trials 1 through to 6 of stage 3). The ratings are expressed in a like/dislike-independent form such that data from dislike trials were transformed to match the scale of the like trials. Thus, here data are pooled over the like and dislike trial types and scores of 5 is (maximally) correct and a score of 1 is (maximally) incorrect. trials of stage 2 (i.e., trials 1 through to 6 of stage 2); (3) the ﬁnal six liked and ﬁnal six disliked trials of stage 2 (i.e., trials 91 through to 96 of stage 2); (4) the ﬁrst six liked and ﬁrst six disliked trials of stage 3 (i.e., trials 1 through to 6 of stage 3). The ratings are expressed in a like/dislike-independent form such that data from dislike trials were transformed to match the scale of the like trials. Thus, here data are pooled over the like and dislike trial types and scores of 5 is (maximally) correct and a score of 1 is (maximally) incorrect. PROCEDURE This slide remained until the participant had made their rating; (3) The rightmost panel represents the feedback given to the participant following their previous rating. In this example, the participant correctly rated Alice as liking tennis (i.e., the participant gave Alice a rating of ≥4 for tennis), which was accompanied by the spoken word “Correct” and by the statement “Alice likes this sport.” Full details of the feedback given on incorrect trials and on the other types of trial are given above. Participants read a standard instruction sheet that gave an indication of the rationale of the experiment and emphasized par- ticipants’ entitlement to leave the experiment. Instructions were then presented on the laptop. A scenario was described involving the participant learning which of two sports the four characters liked. Instructions described making 1–5 key responses to indicate each character’s like/dislike of the sports. At the end of the instruc- tion phase the experimenter went through an example of how to use the keyboard to register responses. The experimenter checked that the participant understood and was comfortable with the task, and then left the room. The instruction phase repeated then the participants pressed the spacebar to initiate the trials. October 2013 \| Volume 4 \| Article 726 \| 3 Ageing and acquired equivalence Robinson and Owens PROCEDURE Participants from Groups Y and O were randomly assigned to congruent (C) or incongruent (I) groups, to create Group YC, Group YI, Group OC, and Group OI. The mean ages and num- bers of women and men in these groups was, respectively: 21.0, 22.5, 64.5, and 65.1 years; and 6:2, 4:4, 5:3, and 6:2. All partici- pants were given training in which they were asked to learn which of the four sports (Tennis, Hockey, Bowling, and Netball) the four ﬁctitious characters (Alice, Beth, Charlotte, and Dorothy) liked. For all participants, each of the four sports was liked by two of the characters and disliked by the other two characters; each char- acter liked two sports and disliked the other two sports. For the congruent groups, each character shared her pattern of sport lik- ing and disliking with one other character and the two remaining characters had the complementary pattern of liking and dislik- ing of sports. For all participants in the congruent groups Alice was equivalent to Charlotte and Beth was equivalent to Dorothy. For approximately half of the participants in the two congruent groups this was based upon Alice and Charlotte’s shared liking of Bowling and Tennis (and shared disliking of Netball and Hockey); for the remainder of the participants in the two congruent groups, equivalence was based upon Alice and Charlotte’s shared dislik- ing of Bowling and Tennis (and their shared liking of Netball and Hockey). The arrangements of Alice and Charlotte’s liking and disliking of Bowling and Netball was the same for the two incon- gruent groups as for the two congruent groups. The incongruent groups’ treatment differed from the congruent groups’ treatment in the four characters’ liking and disliking of Tennis and Hockey: for approximately half of the participants in the two incongruent groups, Alice and Beth liked Tennis (and disliked Hockey); but for the remainder Alice and Beth disliked Tennis (and liked Hockey). Notice that for the incongruent groups no two characters were exactly alike in their pattern of sports liking. FIGURE 4 \| Example of treatments given to the younger and older participants in Experiment 2. Participants are required to learn whether four ﬁctitious characters, Alice, Beth, Charlotte and Dorothy, like or dislike the sports tennis, hockey bowling, and netball. Unlike Experiment 1, training consisted of the single stage represented here. EXPERIMENT 2 trials of stage 2 (i.e., trials 1 through to 6 of stage 2); (3) the ﬁnal six liked and ﬁnal six disliked trials of stage 2 (i.e., trials 91 through to 96 of stage 2); (4) the ﬁrst six liked and ﬁrst six disliked trials of stage 3 (i.e., trials 1 through to 6 of stage 3). The ratings are expressed in a like/dislike-independent form such that data from dislike trials were transformed to match the scale of the like trials. Thus, here data are pooled over the like and dislike trial types and scores of 5 is (maximally) correct and a score of 1 is (maximally) incorrect. FIGURE 3 \| Means, and one standard error of each mean, of data from FIGURE 3 \| Means, and one standard error of each mean, of data from Experiment 1. The leftmost four sets of six trials of data are from the two younger participant groups (groups YW and YP); the rightmost data are the corresponding data from two older participant groups (groups OW and OP). For all four groups, and running from left to right, the four sets of data represent: (1) The ﬁnal six liked and ﬁnal six disliked trials of stage 1 (i.e., trials 91 through to 96 of stage 1); (2) the ﬁrst six liked and ﬁrst six disliked October 2013 \| Volume 4 \| Article 726 \| 4 Frontiers in Psychology \| Personality Science and Individual Differences Ageing and acquired equivalence Robinson and Owens Our new ﬁnding is that this difference in whole/part learning rate was absent in older participants. Before considering fully the implications of this ﬁnding, we sought to replicate it using similar logic to that of Experiment 1. For Experiment 1 to reveal acquired equivalence, it is necessary for the beneﬁt of acquired equivalence to more than offset the cost of relearning new character-sport relationships. In Experiment 2, which is summarized in Figure 4, we followed Honey and Ward-Robinson (2001) in the use of a design that avoids this compromise. Older and younger partic- ipants were required to learn the four characters like/dislike of four sports. For the Congruent treatment, each of the characters’ like/dislike of the the four sports was matched with one other character. For the Incongruent treatment, no one character’s sport like/dislike was matched with any other character. EXPERIMENT 2 Acquired equivalence could be demonstrated by the ﬁnding that the discrimination was mastered more rapidly in the congruent than the incongruent condition. Again, we asked whether the extent of acquired equivalence would be different in the two age groups. MATERIALS AND METHODS PARTICIPANTS, APPARATUS, AND STIMULI Group Y comprised six men and ten women with a mean age of 21.2 years (range: 18–24 years); Group O comprised ﬁve men and eleven women with a mean age of 64.8 years (range: 53–77 years). The apparatus and stimuli were those used in Experiment 1. Experiment 2 used an additional two sport stimuli, bowling and netball to make a total of four sports for the four characters. All unspeciﬁed details of participants, apparatus and stimuli were identical to those of Experiment 1. PROCEDURE Two characters like two of the four sports and dislike the other two sports; these patterns of liking and disliking are complemented by the remaining two characters. In the example of a “congruent” treatment in the top panel, two pairs of characters like the same two sports (Alice and Charlotte both like tennis and bowling, and Beth and Dorothy both like hockey and netball) and dislike the same sports (Alice and Charlotte both dislike hockey and netball, and Beth and Dorothy both dislike tennis and bowling).But in the example of an “incongruent” treatment in the bottom panel, no two pairs of characters share patterns of liking and disliking of the sports. For example, although Alice and Beth both like hockey and dislike tennis, Alice likes bowling, whereas Beth dislikes it. FIGURE 4 \| Example of treatments given to the younger and older FIGURE 4 \| Example of treatments given to the younger and older participants in Experiment 2. Participants are required to learn whether four ﬁctitious characters, Alice, Beth, Charlotte and Dorothy, like or dislike the sports tennis, hockey bowling, and netball. Unlike Experiment 1, training consisted of the single stage represented here. Two characters like two of the four sports and dislike the other two sports; these patterns of liking and disliking are complemented by the remaining two characters. In the example of a “congruent” treatment in the top panel, two pairs of characters like the same two sports (Alice and Charlotte both like tennis and bowling, and Beth and Dorothy both like hockey and netball) and dislike the same sports (Alice and Charlotte both dislike hockey and netball, and Beth and Dorothy both dislike tennis and bowling).But in the example of an “incongruent” treatment in the bottom panel, no two pairs of characters share patterns of liking and disliking of the sports. For example, although Alice and Beth both like hockey and dislike tennis, Alice likes bowling, whereas Beth dislikes it. All participants received 256 trials in random sequence with the constraint that each of the sixteen trial types created by the combinations of the four characters and four sports occurred once in each block of sixteen trials. Unspeciﬁed procedural details were identical to those of Experiment 1. DISCUSSION We sought to test Honey et al.’s (2010) claim that acquired equivalence of conﬁgural learning and intra-dimensional/extra- dimensional set-shifting experiments may employ a common mechanism. We reasoned that because performance at atten- tional set-shifting is reduced in healthy, relatively aged subjects (Owen et al., 1991; Barense et al., 2002), if Honey et al.’s asser- tion is correct, performance at acquired equivalence should also be reduced. Our new ﬁndings supported that suggestion. They do not unambiguously conﬁrm that there is a relationship between conﬁgural learning and intra-dimensional/extra-dimensional set- shifting (e.g., one brought about by their reliance on a common psychological process). For example, conﬁgural learning and set shifting could be governed by independent psychological pro- cesses, each being affected by some aspect of ageing. Nonetheless, our new results represent a ﬁrst and necessary step in the conclu- sion that conﬁgural learning and set shifting are governed by a common process. of the like trials and data were collapsed over like and dislike trials. Initial inspection and analysis revealed that the older participants’ discrimination performance was robust, though the response but- tons were often not the most extreme (i.e., responses of 1 and 5). This feature of the data indicates that older participants may have differed from younger participants in their response bias (i.e., tending to make more accurate, but more modest, responses). For example, on the 16th block of training, only three of the six- teen younger participants gave mean responses that were not 5 s or 1 s, however, at that point, ﬁfteen of the sixteen older partici- pants gave scores that were not 5 s or 1 s (χ2 = 15.4, p < 0.001). To correct for this bias each datum was normalized by multiplying it by a normalization ratio (cf., Ringo, 1988; Baxter and Murray, 2001). The normalization ratio was computed for each block by dividing the arithmetic mean of all data for that block (i.e., ignor- ing age and congruency designation) by the mean for the age group (i.e. ignoring only congruency designation) on that block. This process acted to moderate younger participants’ responses and boost older participants’ responses, which was irrespective of congruency designation. The force of that argument relies on the speciﬁcity of the reduction in performance. That is, older participants’ reduced performance at an acquired equivalence task is not theoretically decisive if it is part of a more general pattern of reduction. RESULTS The leftmost 16 8-trial blocks of data are from the two younger participant groups (groups YC and YI); the rightmost data are the corresponding data from two older participant groups (groups OC and OI). The ratings are expressed in a like/dislike-independent form such that data from dislike trials were transformed to match the scale of the like trials. Thus, here data are pooled over the like and dislike trial types and scores of 5 is (maximally) correct and a score of 1 is (maximally) incorrect. ( , ) p The corresponding 2 × 16 ANOVA on older participants’ data yielded only a main effect of block, F(15, 210) = 5.7, p < 0.001, η2 p = 0.288. Neither the congruency main effect nor its interac- tion with block was signiﬁcant, Fs < 1. Frontiers in Psychology \| Personality Science and Individual Differences RESULTS The results of Experiment 2 are summarized in Figure 5. As in Experiment 1, the scale for dislike trials was reversed to match that October 2013 \| Volume 4 \| Article 726 \| 5 www.frontiersin.org Robinson and Owens Ageing and acquired equivalence FIGURE 5 \| Means, and one standard error of each mean, of data from Experiment 2. The leftmost 16 8-trial blocks of data are from the two younger participant groups (groups YC and YI); the rightmost data are the corresponding data from two older participant groups (groups OC and OI). The ratings are expressed in a like/dislike-independent form such that data from dislike trials were transformed to match the scale of the like trials. Thus, here data are pooled over the like and dislike trial types and scores of 5 is (maximally) correct and a score of 1 is (maximally) incorrect. This description of the data was supported by an ANOVA with block as a within-subject variable and age and congruency as between-subject variables, which revealed a main effect of block, F(15, 420) = 16.6, p < 0.001, η2 p = 0.373 and a Block × Age × Congruency interaction, F(15, 420) = 1.8, p < 0.035, η2 p = 0.060. No other statistics were signiﬁcant, smallest p > 0.216, F(1, 28) = 1.6, η2 p = 0.054. p The source of the Block × Age × Congruency interaction was located by performing a pair of separate, 2 × 16 ANOVAs on younger and older participants’ data. The ANOVA on the younger participants’ data yielded a main effect of block and a Block x Congruency interaction, smaller p < 0.014, F(15, 210) = 2.1, η2 p = 0.128. The congruency main effect was not signiﬁcant, F(1, 14) = 3.9, p < 0.066, η2 p = 0.222. The source of the Block × Congruency interaction in younger participants’ data was located using simple main-effects analysis with separate error- terms for each level of block. This showed responding of Group YC to be superior to that of Group YI on blocks, 5, 6, 7, and 8, largest p < 0.049, F(1, 14) = 4.6, η2 p = 0.248. FIGURE 5 \| Means, and one standard error of each mean, of data from Experiment 2. DISCUSSION The presence of tennis in the input layer will tend to provoke activity in the hidden unit “w,” which codes for liked character-sport combina- tions; w’s activity now provokes activity in the like output unit. On this occasion the participant is likely to correctly indicate that “Charlotte likes tennis,” which will again provoke like −→“w” feedback and will improve connection strength between Charlotte and “w” and between tennis and “w.” Of course, some interven- ing trials will involve tennis also being disliked by some characters (viz., Beth and Dorothy). Thus, at intermediate points of training there is no reason to suppose that the presence of tennis on an “Alice likes tennis” trial will correctly activate hidden unit “w”: it could equally well activate hidden unit “y” (which codes for Beth and Dorothy’s dislike of tennis). On such trials in which “y” is incorrectly selected, the dislike output-unit will be activated by “y” but the actual trial outcome will activate the like output-unit. This means that the output-unit −→hidden-unit feedback signal will not sustain activity in the hidden unit “y,” and the capac- ity of Alice to activate it will diminish. Over multiple trials these processes will tend to encourage sharing of hidden units, thus generating acquired equivalence participants (which was evident before data were normalized). Of course, older people may present a general change in per- formance that is especially pronounced in acquired equivalence tasks. Such an interaction between tasks and the effects of age on performance could generate the results obtained. We cannot eliminate such an account but we noted above that the general performance deﬁciency exhibited by older participants appeared to be a response bias and rather than a discrimination deﬁciency. Further evidence against the suggestion that the age-associated change in acquired equivalence is merely part of a general decline comes from Experiment 1. Here, the acquired equivalence deﬁcit was not accompanied by a general change in performance (e.g., the initial ANOVA did not generate any signiﬁcant statistics that involved the variable age). We have no ready explanation for the inconsistency across experiments of the age-related response bias; but because it is uncorrelated with effects on acquired equiva- lence it is, without some additional elaboration, an inadequate explanation of our ﬁndings. DISCUSSION Leaving to one side for a moment the age-related effects of per- formance, the ﬁndings of the acquired equivalence of conﬁgural learning may be accommodated by a connectionist model (Honey et al., 2010), whose main features are summarized in Figure 6. Individual elements of the discrimination, here the characters and the sports, are represented at the input layer of the network. Presentation of two items (e.g., Alice and tennis) will tend to gen- erate activity in the network’s hidden layer. Hidden-unit activity is subject to a “winner-take-all” process in which the single most active unit will suppress activity in less active units. At ﬁrst, hid- den unit selection will be stochastic: one lucky unit (e.g., “w”) FIGURE 6 \| Depiction of theoretical analysis of acquired equivalence. The three-layered network is composed of: input units, that represent the individual components of each stimulus of each trials (e.g., both Alice and tennis); hidden units, that represent the combination of stimuli on that trial (e.g., Alice with tennis); and the output units, that represent the trial’s correct outcome (e.g., that Alice likes tennis). Input unit −→hidden unit and hidden unit −→output unit connections begin with weights of random strength that approximate zero. Weight changes occur as learning progresses. An output unit −→hidden unit connection gives feedback to the hidden unit about the trial’s outcome. It follows from the analysis above that the disruption of the conjoint hidden- and input-unit activity on correct trials will lead to a reduction in the sharing of hidden units. Understanding such a process may be key to understanding age-related changes seen in acquired equivalence here, and, by extension, those seen in attentional set experiments (e.g., Owen et al., 1991; Barense et al., 2002). One way that this could occur has already been proposed to explain similar effects of neural manipulations on conﬁgu- ral acquired equivalence (e.g., Coutureau et al., 2002; Iordanova et al., 2007). Here, older people’s networks function as described above but with the single exception that the feedback signal from the output-layer to the hidden-layer is weakened or is absent. As outlined above, the feedback signal is a necessary step in the sole means by which input units come to share hidden units; the absence of this signal will, therefore, prevent sharing of hidden units and, therefore, prevent acquired equivalence. DISCUSSION This could be obtained by some general disadvantage, perhaps a reduction in working memory performance, inhibition or sim- ply less familiarity with computer-based tasks than the younger participants. Participants may have differed in their motivation to participate (younger participants gained course credit, older participants did not) or in the level or style of their educations (younger participants were current university students, older par- ticipants were not). Examination of performance that is not part of the acquired equivalence task is key to evaluating these pos- sibilities. Older participants’ performance in Experiment 2 did indicate some general deﬁcit in discrimination relative to younger We see that participants in all groups learned the relationships between the characters and the sports and, it seems, less rapidly than the discrimination in Experiment 1, which could be the result of the additional number of trial types. Group YC appeared to master the discrimination more rapidly than Group YI. The question of key interest is whether the older participants would also show acquired equivalence. In particular, would Group OC’s performance show superiority over Group OI’s? As in Experiment 1, the older participants appear to have satisfactorily learned the discrimination but do not demonstrate acquired equivalence. October 2013 \| Volume 4 \| Article 726 \| 6 Frontiers in Psychology \| Personality Science and Individual Differences Ageing and acquired equivalence Robinson and Owens will be active when activity is generated by the outcome (here, the information being that the character likes or dislikes the sport). The development of hidden-unit −→output unit connection strength will be supported by the co-occurrence of activity sus- tained by “feedback” from the output unit back to the hidden unit. It is this feedback process that give this model its capacity to accommodate acquired equivalence. On a correctly answered trial (e.g., one corresponding to “Alice likes tennis”), after some training, “Alice” and “tennis” will generate activity in the hid- den unit, “w,” which will generate activity in the “like” hidden unit. Here “like” is also the outcome of the trial (i.e., the partici- pant is informed that “Alice likes tennis”). This “correct” outcome will tend to stimulate further activity, via a feedback connection, to hidden unit “w.” If we ignore any intervening trials and con- sider next what will happen when the participant receives a trial in which they are asked if Charlotte likes tennis. REFERENCES REFERENCES hierarchical forms of learning. J. Exp. Psychol. Anim. Behav. Process. 26, 358–363. doi: 10.1037/0097- 7403.26.3.358 hierarchical forms of learning. J. Exp. Psychol. Anim. Behav. Process. 26, 358–363. doi: 10.1037/0097- 7403.26.3.358 Mackintosh, N. J. (1974). The Psychology of Animal Learning. London: Academic Press. and Associative Learning From Brain to Behaviour, eds C. J. Mitchel and M. Le Pelley (Oxford: Oxford University Press), 385–406. and Associative Learning From Brain to Behaviour, eds C. J. Mitchel and M. Le Pelley (Oxford: Oxford University Press), 385–406. Barense, M., Fox, M., and Baxter, M. (2002). Aged rats are impaired on an attentional set-shifting task sensitive to medial frontal cortex damage in young rats. Learn. Mem. 9, 191–201. doi: 10.1101/lm.48602 Owen, A. M., Roberts, A. C., Polkey, C. E., Sahakian, B. J., and Robbins, T. W. (1991). Extra-dimensional versus intra-dimensional set shifting performance following frontal lobe excisions, temporal lobe excisions or amygdalo- hippocampectomy in man. Neuropsychologia 29, 993–1003. doi: 10.1016/0028-3932(91)90063-E Hodder, K. I., George, D. N., Killcross, A. S., and Honey, R. C. (2003). Representational blending in human conditional learning: impli- cations for associative theory. Q. J. Exp. Psychol. B 56, 223–238. doi: 10.1080/02724990244000269 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Baxter, M. G., and Murray, E. A. (2001). Opposite relationship of hippocam- pal and rhinal cortex damage to delayed nonmatching-to-sample deﬁcits in monkeys. Hippocampus 11, 61–71. doi:10.1002/1098-1063 (2001)11:1<61::AID-HIPO1021>3. 0.CO;2-Z Honey, R. C., Close, J., and Lin, T. C. E. (2010). “Acquired distinctiveness and equivalence: a synthesis,” in Attention and Associative Learning From Brain to Behaviour, eds C. Mitchel and M. Le Pelley (Oxford: Oxford University Press), 159–186. Pearce, J. M. (2002). Evaluation and development of a connectionist theory of conﬁgural learning. Anim. Learn. Behav. 30, 73–95. doi: 10.3758/BF03192911 (2001)11:1<61::AID-HIPO1021>3. 0.CO;2-Z Received: 07 May 2013; paper pend- ing published: 12 June 2013; accepted: 20 September 2013; published online: 11 October 2013. Birrel, J. M., and Brown, V. J. (2000). Medial frontal cortex mediates per- ceptual attentional set shifting in the rat. J. Neurosci. 20, 4320–4324. Honey, R. C., and Hall, G. (1989). The acquired equivalence and distinctiveness of cues. J. Exp. Psychol. Anim. Behav. Process. 15, 338–346. doi: 10.1037/0097-7403. 15.4.338 Ringo, J. L. (1988). Seemingly discrepant data from hippocampec- tomized macaques are reconciled by detectability analysis. Behav. Neurosci. 102, 173–177. DISCUSSION The absence of shared hidden units will not prevent the engagement of (trial- unique) hidden units in learning. The hidden unit that is most active on a particular trial will still tend to become associated with the output unit and it will not be activated by any other trial type. The mechanism of learning in older people, then, becomes FIGURE 6 \| Depiction of theoretical analysis of acquired equivalence. FIGURE 6 \| Depiction of theoretical analysis of acquired equivalence. The three-layered network is composed of: input units, that represent the individual components of each stimulus of each trials (e.g., both Alice and tennis); hidden units, that represent the combination of stimuli on that trial (e.g., Alice with tennis); and the output units, that represent the trial’s correct outcome (e.g., that Alice likes tennis). Input unit −→hidden unit and hidden unit −→output unit connections begin with weights of random strength that approximate zero. Weight changes occur as learning progresses. An output unit −→hidden unit connection gives feedback to the hidden unit about the trial’s outcome. The three-layered network is composed of: input units, that represent the individual components of each stimulus of each trials (e.g., both Alice and tennis); hidden units, that represent the combination of stimuli on that trial (e.g., Alice with tennis); and the output units, that represent the trial’s correct outcome (e.g., that Alice likes tennis). Input unit −→hidden unit and hidden unit −→output unit connections begin with weights of random strength that approximate zero. Weight changes occur as learning progresses. An output unit −→hidden unit connection gives feedback to the hidden unit about the trial’s outcome. October 2013 \| Volume 4 \| Article 726 \| 7 www.frontiersin.org Ageing and acquired equivalence Robinson and Owens like that described by models such as those of Rescorla (1976) and Pearce (2002): each unique combination of character and sport will require its own, unique hidden-unit. The translation from this model to ageing people is unclear but it seems possible that they reﬂect developmental changes in cortical regions in rhi- nal (e.g., Coutureau et al., 2002 or prefrontal brain-regions e.g., Iordanova et al., 2007) between these facts and age-related deﬁcits in performance on intra-dimensional/extra-dimensional set tasks (e.g., Owen et al., 1991; Barense et al., 2002) could be the result of their being under- pinned by a shared mechanism and that this does not require an attentional component (cf., Honey et al., 2010). Frontiers in Psychology \| Personality Science and Individual Differences hierarchical forms of learning. J. Exp. Psychol. Anim. Behav. Process. 26, 358–363. doi: 10.1037/0097- 7403.26.3.358 ACKNOWLEDGMENTS Whatever the precise detail of the deﬁcit in performance, our current results demonstrate the generality of demonstrations of acquired equivalence reported by others (e.g., Ward-Robinson and Honey, 2000; Honey and Ward-Robinson, 2001; Coutureau et al., 2002; Hodder et al., 2003; Iordanova et al., 2007) and its absence in older participants. We noted also that the parallel This research was supported by an Experimental Psychology Society, University Bursary Scheme grant that supported Emma Owens’ collection experimental work. We gratefully acknowledge Allie Broome, Lone Hørlyck, and Claire Petelczyc for their help. Some results were reported at A Festschrift for Geoffrey Hall, Meeting of the Experimental Psychology Society, Hull, UK, 2012. October 2013 \| Volume 4 \| Article 726 \| 8 REFERENCES doi: 10.1037/0735-7044.102.1.173 Citation: Robinson J and Owens E (2013) Diminished acquired equivalence yet good discrimination performance in older participants. Front. Psychol. 4:726. doi: 10.3389/fpsyg.2013.00726 Coutureau, E., Killcross, A. S., Good, M., Marshall, V. J., Ward-Robinson, J., and Honey, R. C. (2002). Acquired equivalence and dis- tinctiveness of cues: II. Neural manipulations and their impli- cations. J. Exp. Psychol. Anim. Behav. Process. 28, 388–396. doi: 10.1037/0097-7403.28.4.388 This article was submitted to Personality Science and Individual Differences, a section of the journal Frontiers in Psychology. Rescorla, R. A. (1976). Stimulus gen- eralization: some predictions from a model of Pavlovian condition- ing. J. Exp. Psychol. Anim. Behav. Process. 2, 88–96. doi: 10.1037/0097- 7403.2.1.88 Honey, R. C., and Ward-Robinson, J. (2001). Transfer between con- textual conditional discriminations: an examination of how stimu- lus conjunctions are represented. J. Exp. Psychol. Anim. Behav. Process. 27, 196–205. doi: 10.1037/0097- 7403.27.3.196 Copyright © 2013 Robinson and Owens. This is an open-access article dis- tributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Dias, R., Robbins, T. W., and Roberts, A. C. (1997). Dissociable forms of inhibitory control within pre- frontal cortex with an analog of the Wisconsin card sort test: restric- tion to novel situations and inde- pendence from “on-line” process- ing. J. Neurosci. 17, 9285–9297. Ward-Robinson, J., and Hall, G. (1999). The role of mediated condition- ing in acquired equivalence. Q. J. Exp. Psychol. 52, 335–350. doi: 10.1080/027249999393031 Iordanova, M. D., Killcross, A. S., and Honey, R. C. (2007). Role of the medial prefrontal cortex in acquired distinctiveness and equiv- alence of cues. Behav. Neurosci. 121, 1431–1436. doi:10.1037/0735- 7044.121.6.1431 Ward-Robinson, J., and Honey, R. C. (2000). A novel contextual dimension for use with an oper- ant chamber: from simple to Hampshire, A., and Owen, A. M. (2010). “Clinical studies of atten- tion and learning,” in Attention October 2013 \| Volume 4 \| Article 726 \| 8 Frontiers in Psychology \| Personality Science and Individual Differences
https://openalex.org/W4386325525	https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1214928/pdf	English	null	Polish version of the Aesthetic Experience Questionnaire: validation and psychometric characteristics	Frontiers in psychology	2,023	cc-by	12,857	OPEN ACCESS EDITED BY Dominique Makowski, Nanyang Technological University, Singapore REVIEWED BY Nils Myszkowski, Pace University, United States Susana Silva, University of Porto, Portugal Agata Hiacynta Świątek , Małgorzata Szcześniak , Karolina Wojtkowiak , Michał Stempień and Marianna Chmiel Faculty of Social Sciences, Institute of Psychology, University of Szczecin, Szczecin, Poland Introduction: The purpose of the article is to present the results of works on the Polish version of the Aesthetic Experience Questionnaire (AEQ). The AEQ is a 22-item tool for assessing aesthetic experience in the following dimensions: emotional, cultural, perceptual, understanding, and two dimensions about flow (proximal conditions and flow experience). Świątek AH, Szcześniak M, Wojtkowiak K, Stempień M and Chmiel M (2023) Polish version of the Aesthetic Experience Questionnaire: validation and psychometric characteristics. Methods: In the course of works on the Polish version of the AEQ, 3 independent studies with the participation of more than 800 people were carried out. In addition to the AEQ measurement, the tools included: the Emotion Regulation Strategies for Artistic Creative Activities Scale, the Brief Music in Mood Regulation, the Aesthetic Competence Scale, the Aesthetic Processing Preference Scale, the Need for Cognition Scale, the Center for Epidemiologic Studies - Depression Scale, the Material Values Scale and the Multidimensional Existential Meaning Scale. CO G © 2023 Świątek, Szcześniak, Wojtkowiak, Stempień and Chmiel. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Results: The results obtained in the three studies through Confirmatory Factor Analysis indicated the compliance of the factor structure of the Polish version of the AEQ with the original and its good psychometric characteristics. It was also shown that the overall result and individual components of the aesthetic experience correlate positively with emotion regulation through artistic creative activities and mood regulation through music, aesthetic competences (music, literature, plastic arts, film), cognitive curiosity and some dimensions of aesthetic processing preferences. The studies also proved a very weak positive relationship between aesthetic experience and meaning of life. The assumption about a negative correlation between aesthetic experience and depression or materialism was not confirmed. TYPE Original Research PUBLISHED 31 August 2023 DOI 10.3389/fpsyg.2023.1214928 TYPE Original Research PUBLISHED 31 August 2023 DOI 10.3389/fpsyg.2023.1214928 TYPE Original Research PUBLISHED 31 August 2023 DOI 10.3389/fpsyg.2023.1214928 OPEN ACCESS Discussion: The Polish version of the AEQ is a credible psychometric measurement and encourages scientists to design research on the psychology of art and aesthetics in the Polish cultural conditions. KEYWORDS aesthetic experience, measurement, engagement with art, aesthetic competence, artistic creative activities frontiersin.org © 2023 Świątek, Szcześniak, Wojtkowiak, Stempień and Chmiel. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Polish version of the Aesthetic Experience Questionnaire: validation and psychometric characteristics OPEN ACCESS EDITED BY Dominique Makowski, Nanyang Technological University, Singapore REVIEWED BY Nils Myszkowski, Pace University, United States Susana Silva, University of Porto, Portugal CORRESPONDENCE Małgorzata Szcześniak malgorzata.szczesniak@usz.edu.pl RECEIVED 30 April 2023 ACCEPTED 03 August 2023 PUBLISHED 31 August 2023 CITATION Świątek AH, Szcześniak M, Wojtkowiak K, Stempień M and Chmiel M (2023) Polish version of the Aesthetic Experience Questionnaire: validation and psychometric characteristics. Front. Psychol. 14:1214928. doi: 10.3389/fpsyg.2023.1214928 Introduction In Poland, the psychology of art and aesthetics is not a popular research topic, which is reflected in the lack of psychometric tools to conduct research in this area. Therefore, the main issue discussed in the article is the validation of a new tool – the Aesthetic Experience Questionnaire (AEQ; Wanzer et al., 2020) for the Polish cultural conditions. The work presents the issues and the current knowledge of aesthetic experience (AE), and describes the process of Frontiers in Psychology 01 Świątek et al. 10.3389/fpsyg.2023.1214928 10.3389/fpsyg.2023.1214928 meaning of the artist’s intentions in creating the aesthetic experience, i.e., communication of the work function (interpretation of the artist’s message by a viewer). The authors also proposed two dimensions connected with the flow experience – the occurrence of proximal conditions as well as the flow experience itself as a result of coming into contact and engagement in contemplation of the work of art. developing and verifying the psychometric characteristics of the Polish version of the AEQ. It draws attention to the AEQ’s structure, analyzing it with the use of the Confirmatory Factor Analysis (CFA) and answering in this way to the authors’ suggestions which indicated the need for such a verification in future research. The CFA was carried out separately on data from three studies in order to confirm whether the results were coincidental and whether the tool structure is actually confirmed in the Polish reality. Then, the assumptions about the relationships between aesthetic experience and its potential correlations were verified. The results of these three separate studies, their limitations and indications concerning further work, are summarized in the discussion. g g The Aesthetic Experience Questionnaire (AEQ; Wanzer et al., 2020) is a psychological measurement which evaluates the intensity of individual components of aesthetic experience during the contemplation of a work of art. Although it was constructed mainly in relation to visual arts (such as paintings and sculptures), the authors of the original version do not exclude its application to other fields of art, for example music. In such a case, they suggest adjusting the selected items to the specific nature of the given field of art. Introduction 22 statements were grouped into 6 dimensions: emotional (e.g., My emotions change as I continue to view the work of art), cultural (e.g., I compare the past culture of the art with present-day culture), perceptive (e.g., I focus on the subtle aspects of the work of art), understanding (e.g., I gain new insights about the work of art itself), proximal flow conditions (e.g., I have a clear idea of what to look for when viewing works of art) and the flow experience during contact with works of art (e.g., I lose track of time when I view works of art). The study subjects assess how much they agree (or disagree) with each of the statements, using a 7-point Likert scale (from 1 = “I definitely disagree” to 7 = “I definitely agree”). In the original study (Wanzer et al., 2020), all AEQ dimensions were associated with each other and the overall AEQ score, ranging between the values r = 0.240 and r = 0.719. Most of the correlations were at the moderate level (less than r = 0.500). The results for each of the six dimensions are obtained by summing the points gained by the study subject in each scale: emotional (items 1–4), cultural (items 5–8), perceptive (items 9–11), understanding (items 12–15), proximal flow conditions (items 16–18) and the flow experience during contact with works of art (items 19–22). Moreover, the overall result is calculated by adding up the points gained by the study subject in subscales. The questionnaire does not contain reversed questions or buffer questions. According to Marković (2012), aesthetic experience is one of the least clarified concepts in the psychology of art and experimental aesthetics. Similarly, Brattico et al. (2013) note that despite the presence in the literature of research on experiencing aesthetic stimuli, the “aesthetic experience” concept is relatively poorly defined in psychology. These words remain relevant after more than 10 years. Meng and Liang (2022) write that the subjective and objective attribute of AE remains unclear, and they search for stable, reliable grounds for the nature of AE in aesthetic anthropology and neuroaesthetics. Marković (2012, p. 1) writes: “Generally, aesthetic experience can be defined as a special state of mind that is qualitatively different from the everyday experience.” He gives readers insight into three components of AE: aesthetic fascination, aesthetic appraisal, and aesthetic emotion. Introduction He notices that AE does not occur automatically, but it results from the individual’s attitude to the external environment and social context. Everyday objects or works of art will not elicit AE in all people. As Vessel et al. (2012) write, aesthetic experience is connected with the integration of reactions resulting from sensory perception, emotional reactions and personal meaning. On the one hand, it is recognized that reacting to aesthetic stimuli is rooted in human biology. There is a well-known study in which a team of three Italian researchers used the fMRI technique to analyze the biological reactions of viewers who were not experts in response to masterpieces of classical and renaissance sculpture (Di Dio et al., 2007). From a medical point of view, scientists have been able to explain the physiological grounds for aesthetic experience based on neurological studies for more than 20 years (Ramachandran and Hirstein, 1999; Vessel et al., 2012). On the other hand, apart from biological mechanisms, full aesthetic experience seems to require more complex forms of thinking. Bullot and Reber (2013) in their article proposed a psycho-historical framework for the science of art appreciation and tried to bridge the gap between researchers of the perception of art from different fields of science. f Works on the Polish validation of the AEQ scale constituted part of a larger study conducted based on the consent of the Committee for the Ethics of Research of the Institute of Psychology at the University of Szczecin (KB 6/2022 of 27/04/2022). In the first stage of developing the Polish version, three independent translations from English into Polish were commissioned. Based on these translations, a Polish version was prepared jointly by the researchers. This translation was consulted with an art expert. Next, the scale was subjected to reverse translation. After verifying the consistency of the reverse translation with the original, 25 people were initially asked to complete it. These people assessed the clearness and correctness of the individual statements. They also had the possibility to submit their own comments. As no reservations were raised, this Polish version was placed in the questionnaire batteries. In order to exclude the impact of the order of completing the tools by the respondents on the assessment of the tool’s psychometric properties, in individual studies, the AEQ was placed at the beginning, in the middle and at the end of the questionnaire batteries. Introduction While developing their method of measuring aesthetic experience (AEQ), Wanzer et al. (2020) used theoretical frameworks proposed by Csikszentmihalyi and Robinson (1990) based on interviews with employees of museums. Referring to the results of their research, the authors indicate four dimensions of AE connected with works of art. These are the perceptual, emotional, cognitive, and communicative dimensions. The first of them concerns the composition of the work and the aesthetic aspects of its composition (color, texture, form of the painting or sculpture). The second dimension covers different (pleasant and unpleasant) emotions which occur in response to contact with the work of art. The third dimension is the understanding of the work of art through the impact of knowledge of its cultural significance and historical knowledge (general knowledge of art, the artist, and the specific work). The fourth dimension concerns the Frontiers in Psychology frontiersin.org Assumptions adopted While developing their method of measuring aesthetic experience (AEQ), Wanzer et al. (2020) analyzed the relationships between the 02 frontiersin.org Świątek et al. 10.3389/fpsyg.2023.1214928 H2: The aesthetic experience and its components positively correlate with the regulation of emotions through artistic creative activities and their individual types. results obtained by their study subjects in the AEQ and the level of the intensity of openness to experience, inspiration, curiosity and exploration. There is no other measure of the same construct in Poland, nor anything similar enough to check convergent validity. Moreover, if we wanted to replicate the variables from the original study, it would be difficult, because we do not have their Polish versions. Therefore, in our study, we explore the topic, proposing correlates which were not considered by Wanzer and her team. Our studies were aimed at bringing new knowledge of aesthetic experience in addition to the knowledge that already exists. That is why the emphasis in the first study was put on measurements connected with art and the function of aesthetic experience in human life. The following paragraphs present hypotheses and their justification in order to explain the purposefulness of the selection of particular variables. The sense of the meaning of life was the next variable that we wanted to analyze in the AE context. Art frequently touches on universal, timeless motives and themes connected with the physical or mental reality experienced by the author. The viewer interprets the message of the work of art, referring not only to cultural norms but also to personal systems of meanings or specific situations from their own life. Contact with an aesthetic object may stimulate reflection going beyond the aesthetic situation (e.g., in the case where viewing works of art created several centuries ago raises the question of what we and our society leave behind when we pass away). It was exploratorily assumed that a high level of engagement in aesthetic experience (particularly in the aspect of understanding and culture) may help to understand and discover the purpose and meaning of one’s own existence. Therefore, it was assumed that: Using music to regulate the mood was taken into consideration as the first variable. According to Granot et al. (2021), there are many purposes of listening to music, e.g., experiencing aesthetic pleasure, determining one’s own identity, bringing back memories or changing one’s own mood. Assumptions adopted Osowiecka and Gacka (2016) write that while creating music presenting hardships (e.g., hip hop brings up the topics of addictions and poverty), performers cope in this way with problems and give listeners hope for a better future. Furthermore, they refer to the work of Van den Tol and Edwards (2015), who believe that the identification with emotions from a song allows listeners to re-experience something, the experience of connecting with the message of the song may motivate them to be active (undertake actions), the aesthetic value may divert their attention from problems, and memories evoked by the piece of music allow them to experience them again and see them in a new light. Of course, the processing of visual and musical stimuli differs from each other; the very specificity of these stimuli is different (the sculpture is static, the musical work is changing in time). However, referring to the concept of the aesthetic quotient (Dan et al., 2021), which consists of competences in separate fields of art, it can be assumed that if someone is “aesthetically intelligent” in one field of art, he or she is probably better able to engage in the reception of other artistic domains and use them in an adaptive way (emotion regulation). That is why it was assumed that people who have the ability to become emotionally and cognitively involved to a great extent in aesthetic experience are very likely to listen to music in order to change their mood: H3: Aesthetic experience, together with its dimensions, positively correlates with the sense of meaning of life. H3: Aesthetic experience, together with its dimensions, positively correlates with the sense of meaning of life. The frequency of contact with artistic objects or participation in artistic events may sensitize people to artistic experiences. For people who rarely participate in cultural events, the situational context and circumstances may play an important role (the sense of uniqueness of the whole situation, attitude, and conviction of seeing “something unusual” may prevail over conscious focus on the reception of the work). People who are familiar with the arts often visit places intended for art exhibitions and are likely to focus more easily on the art object itself. Assumptions adopted They “immerse” in the work which makes their experience in contact with a specific artwork more reflective, intense, richer, than in people who, for example, rarely visit a museum, go there because of prestige or as tourists. Moreover, the intensity and depth of perceiving artworks probably also varies depending on the role that art plays in a person’s life. People whose paid work is related to the arts (whether they are artists or sellers of art) naturally need to update their art-related competencies and are more sensitive to various aspects of art. By concentrating on the work, they have a chance for stronger experiences, at least in terms of perceiving and understanding art. Similarly, a higher level of aesthetic experience may be characteristic of people who give art an important place in their lives in the category of passion/hobby. It is a way of spending time and personal development dictated by one’s choice, curiosity, the desire to aesthetically experience the world or the feeling of discovering something. Based on these insights, we suppose that this is why the frequency of participation in artistic events and the meaning given to art may differentiate art recipients in terms of the intensity of aesthetic experiences related to art. H1: The aesthetic experience and its individual dimensions positively correlate with using music in order to regulate mood. Frontiers in Psychology frontiersin.org H1: The aesthetic experience and its individual dimensions positively correlate with using music in order to regulate mood. The next variable was the regulation of emotions through artistic creative activities. Also in this case, it was assumed that people who have strong aesthetic experience are more willing to apply strategies of coping with difficulties through their own artistic creativity. Fancourt et al. (2020) showed in their studies that training in doing an artistic activity, regular engagement in the production of art, and enjoyment while engaging are all associated with a greater ability to use artistic activities to regulate our emotions. Osowiecka and Gacka (2016, p. 69) write that “most artists are convinced (and certainly they have such a feeling) that they create to experience emotional well- being” and they draw attention to the human need to create art even in extremely unfavorable conditions, e.g., by concentration camp prisoners or political prisoners. Based on the above intuitions, it was assumed that: H4: Groups with different levels of participation in artistic activities (frequency groups) and giving different meanings to artistic activities (meaning groups) differ in aeasthetic experience. In the second study, we asked about the relationships between AE and aesthetic competences, aesthetic processing preferences, and the need for cognition. 03 frontiersin.org Świątek et al. Świątek et al. 10.3389/fpsyg.2023.1214928 H8: Aesthetic experience and its dimensions negatively correlate with following materialistic values in life. H8: Aesthetic experience and its dimensions negatively correlate with following materialistic values in life. Most likely, people who declare stronger aesthetic experience may also be more competent recipients of a given field of art or several fields of art. The exposure to art and aesthetic experience based on it are conducive to building aesthetic competences. At the same time, a high level of aesthetic quotient could make it easier initially to get engaged in aesthetic experience. It is known, for example, that experts are more differentiated and flexible in the assessment of works of art than people who have little knowledge of art. Moreover, it was stated that experts in general gave higher assessments in almost all scales connected with the appreciation of art (Leder et al., 2012). The studies of Fayn et al. (2018) indicate that deeper engagement of people with wider knowledge of art in the reception of art is connected with more detailed aesthetic experiences. Hence, according to the fifth hypothesis: Depression is an illness that can lead to immense suffering and results in the reduction of activity in daily life (Blomdahl et al., 2013). Frontiers in Psychology Participants The participants were 402 Polish adults (71% women). The age range of the sample was 18–72 years (M = 25.33; SD = 9.69). They were asked about the frequency of artistic activity on a 5-point Likert scale, where 1 = less or not at all, 2 = several times a month, 3 = 1–2 times a week, 4 = several times a week, and 5 = daily or almost daily. Most of them (32%) declared daily or almost daily artistic activity, followed by several times a month (22%), less or not at all (17%), several times a week (16%), and 1–2 times a week (13%). A question was also added about what artistic activity means to the respondent. More than half of the people answered that artistic activity is a passion for them, which they do in their free time from work or study (57%). For the second group of participants, it is something they are learning or developing with a view to earning money in this area (25%). The third group represents people for whom artistic activity is not important (13%). Only 5% of the respondents admitted that they make a living from artistic activity. All the participants were recruited through non-probability convenience sampling and expressed informed and written consent to participate in Studies 1–3. H6: Aesthetic experience and its dimensions positively correlate with appreciation of complexity in art and propensity to contextualize artworks, while they negatively correlate with intolerance for ambiguity in art. The need for cognition leads to “controlled intellect,” and people with a high level of this feature are prone to carefully consider available information (Madrid and Patterson, 2016). The inclination for analytical observation of reality may encourage greater engagement in contemplation of art. It may be assumed that in the case of people with a high level of the need for cognition, contact with art will be connected with stronger aesthetic experience, therefore: H7: Aesthetic experience and its dimensions positively correlate with the need for cognition. H1: The aesthetic experience and its individual dimensions positively correlate with using music in order to regulate mood. Diessner et al. (2008) write that depression is negatively correlated with the overall factor of becoming engaged in beauty and with two subscales. There are also data indicating that engagement in art and culture may be a protective factor, as it is connected with a reduced frequency of the occurrence of depression in the population (Elsden and Roe, 2021). Therefore, it is assumed that: H9: Aesthetic experience and its dimensions negatively correlate with depression indicators. H5: Aesthetic experience and its dimensions positively correlate with aesthetic competences in four fields of art. After presenting the above-described hypotheses, we will present three independent studies in which we analyze the structure of the Polish version of the AEQ and the nomological network of its correlations. Kopatich et al. (2021) write that motivation to become engaged in controlled art processing increases interest in and knowledge of art. Based on reports of other researchers, the same authors indicate that people more willing to deal with aesthetic objects better appreciate complexity and feel greater pleasure while viewing works of art. In their own studies, two dimensions of the aesthetic processing, i.e., propensity to contextualize and appreciation of complexity, co-occurred with interest in art and knowledge of art, while intolerance for ambiguity negatively correlated with knowledge of art. On this basis, it was assumed that: frontiersin.org Procedure, data and statistical analysis While planning the third study, independent of the others, we wondered whether there are negative correlations for aesthetic experience. What features or conditions may be unfavorable for intensive engagement in the reception of art? For the analysis, we selected materialistic values and depression indicators. Before performing a CFA on the data, the skewness and kurtosis values of all 22 AEQ items were considered, to ensure the assumption of a normal distribution required by the structural equation. Although there is no explicit cut-off to denote a symmetric distribution (Bowen and Guo, 2011), the accepted ±2 cut-off was considered as not a cause for concern (Field, 2000). According to the results obtained by Diessner et al. (2008), the higher the person’s engagement in beauty, the less probable it is that they are materialistic. On this basis, it may be assumed that a materialistic attitude to life and finding happiness in possession are not conducive to engagement in the reception of art. If works of art are not treated as investments or material security, the “food for thought” (aesthetic experience) will not be “nourishing” for people who hold materialistic values. Therefore, it is assumed that: In order to examine the adequacy of the hypothesized model, a variety fit indices (Swanson and Holton, 2005; Flynn et al. 2019) were measured: insignificant χ2 test (p > 0.05), the ratio χ2/df (≤ 5), goodness-of-fit index (GFI), Tucker-Lewis index (TLI), and the comparative fit index (CFI) ≥ 0.90, standardized mean square 04 10.3389/fpsyg.2023.1214928 Świątek et al. (α = 0.86), intellectual work (α = 0.88), solace (α = 0.92), and overall mood regulation (α = 0.95). The study subjects take a position on each statement using a 5-point scale, where the “always/almost always” answer is equivalent to the study subject receiving 5 points, while one point is assigned for the “almost never/never” answer. To obtain the results for each individual mood regulation strategy, it is necessary to sum the points received by the study subject from the statements included in that factor. The overall result is obtained by summing all points (α = 0.95). The Polish translation of the tool was used in the study.h residual (SRMS) ≤ 0.06, root mean square error of approximation (RMSEA), LO, and HI ≤ 0.08 (Hooper et al., 2008). Procedure, data and statistical analysis The internal consistency of the six factors and overall AEQ was assessed, adopting the criterion of at least α > 0.75 as a general standard for an acceptable value (Weiner and Greene, 2017). Moreover, we used the corrected item–total correlation to check the consistency between responses to an item and the sum of the other items (Furr and Bacharach, 2008). For loading estimates, the following rule of thumb was assumed: 0.71–excellent, 0.63–very good, 0.55–good, 0.45–fair, and 0.32–poor (Harrington, 2009).h The Emotion Regulation Strategies for Artistic Creative Activities Scale (ERS-ACA; Fancourt et al., 2019) is an 18-item tool to measure engagement in artistic activities as a measure regulating emotions. It provides a calculation of the result for overall engagement in artistic activities as a strategy of dealing with emotions and a reflection of the intensity of using the following distinguished strategies: avoidance strategies (α = 0.92), approach strategies (α = 0.90), self-development strategies (α = 0.90), and overall emotion regulation (α = 0.95). The scale does not concern any specific art domain – the study subject is asked to think about their favorite artistic activity. Then the respondent reads each statement and refers to it, ticking the answer on a 5-point Likert scale (1 = “I definitely disagree,” 5 = “I definitely agree”). The Polish translation of the scale was used in the study. The construct validity of the AEQ and its potential correlates was assessed in Study 1 through the use of the Brief Music in Mood Regulation Scale, the Emotion Regulation Strategies for Artistic Creative Activities Scale, and the Multidimensional Existential Meaning Scale. The basis for the selection of these variables was the assumption that they would correlate positively with aesthetic experience, as some of the constructs are conceptually related or inversely associated. In order to be clear when describing and interpreting the correlation coefficients and their strength, we adopted the following framework: weak between ±0.1 and ±0.3, moderate between ±0.4 and ±0.6, and strong between ±0.7 and ±0.9 (Akoglu, 2018). To determine the suitable sample size, an a priori power analysis was performed, using GPower 3.1.9.4 with a bivariate normal model correlation (Faul et al., 2007). A small effect size of 0.20, an alpha of 0.05, and a power of 0.95 were assumed. The result of the analysis showed that a total sample size would require around 266 participants. Results Table 1 presents the means, standard deviations, values of skewness and kurtosis, CFA loadings, and corrected item-total correlations for the aesthetic experience items. The skewness and kurtosis values did not exceed ±2. The research project (Studies 1–3) was conducted according to the recommendations of the Declaration of Helsinki. Statistical analyses were performed using SPSS Statistics for Windows, version 20, and IBM SPSS AMOS 21. The AEQ subjected to the CFA confirmed the factor structure obtained by Wanzer et al. (2020) through Exploratory Factor Analysis in their original article about the development of the AEQ. All standardized loadings in the CFA model were above 0.70 (between 0.71 and 0.93), indicating excellent values for the AEQ items. The test for goodness-of-fit showed that the specified model had a six-factor structure consisting of emotional (items 1–4), cultural (items 5–8), perceptual (items 9–11), understanding (items 12–15), flow conditions (items 16–18), and flow experience (items 19–22), and represented an acceptable fit with the data: χ2 = 804.91, df = 194, p = 0.000, χ2/df = 4.149, GFI = 0.84, TLI = 0.91, CFI = 0.92, SRMS = 0.04, RMSEA = 0.08, LO = 0.08, HI = 0.09. Although χ2 was significant, this statistic is sensitive to sample size, which for Study 1 Procedure, data and statistical analysis The justification for using the value of Pearson’s r = 0.20 is based on general recommendations that, in individual differences research, such a benchmark is justified (Brydges, 2019). The Multidimensional Existential Meaning Scale (MEMS) is a self-report instrument developed by George and Park (2017). The Polish adaptation was created by Gerymski and Krok (2020). The scale measures meaning in life in three different dimensions: comprehension (e.g., I know what my life is about), purpose (e.g., I have aims in my life that are worth striving for), and mattering (e.g., I am certain that my life is of importance). The respondents assess each of the 9 statements by using answers on a 7-point Likert scale that ranges from 1 – “very strongly disagree” to 7 – “very strongly agree.” The higher the overall score obtained, the higher the level of the comprehension, purpose and mattering in life. The reliability in the individual subscales was sufficient: comprehension (α = 0.75), purpose (α = 0.65), mattering (α = 0.83), as well as in overall multidimensional existential meaning (α = 0.91). A one-way analysis of variance (ANOVA) was carried out to check whether belonging to groups with different levels of participation in artistic activities (frequency groups) and giving different meanings to artistic activities (meaning groups) differentiates aeasthetic experience (H4). A Levene’s test was used to verify the assumptions of data homoscedasticity with a significance level of p < 0.05. A Tukey’s post hoc test or a Games-Howell post hoc test were then computed, depending on the homogeneity of variance value. The Tukey test output was examined when the assumption of homogeneity of variance was not violated (Levene’s test not significant). Instead, the Games-Howell test was considered when this assumption was violated (Leven’s test significant; Allen, 2017). Frontiers in Psychology frontiersin.org Measures The Brief Music in Mood Regulation Scale (B-MMR; Saarikallio, 2012) assesses the degree to which the study subject uses music to regulate their own emotions and estimates the degree to which the study subject uses individual strategies. It consists of 21 statements grouped into 7 strategies of modifying the mood, which in the current study proved to be very reliable, both for the individual subscales and for the overall result: entertainment (α = 0.89), revival (α = 0.88), strong feelings (α = 0.89), distraction (α = 0.86), relieving oneself 05 frontiersin.org Świątek et al. 10.3389/fpsyg.2023.1214928 TABLE 1 Descriptive statistics for AEQ items, CFA loadings, and corrected item-total correlations (N = 402). TABLE 1 Descriptive statistics for AEQ items, CFA loadings, and corrected item-total correlations (N = 402). Item M SD Skewness Kurtosis Loadings Corrected item- total correlations AEQ1 4.60 1.99 −0.39 −1.01 0.71 0.74 AEQ2 4.62 1.86 −0.41 −0.86 0.93 0.78 AEQ3 4.54 1.94 −0.38 −0.97 0.89 0.78 AEQ4 3.76 1.96 0.19 −1.15 0.87 0.61 AEQ5 4.06 2.08 −0.07 −1.32 0.85 0.64 AEQ6 3.94 2.02 −0.07 −1.30 0.79 0.72 AEQ7 3.64 2.11 0.15 −1.37 0.85 0.61 AEQ8 4.20 2.02 −0.24 −1.20 0.77 0.77 AEQ9 4.51 1.97 −0.37 −1.10 0.89 0.76 AEQ10 4.81 2.01 −0.55 −0.97 0.89 0.75 AEQ11 4.59 1.97 −0.49 −0.97 0.86 0.81 AEQ12 4.40 1.93 −0.26 −1.04 0.83 0.79 AEQ13 4.50 1.94 −0.36 −1.02 0.91 0.79 AEQ14 4.48 1.94 −0.42 −0.96 0.86 0.86 AEQ15 4.47 2.04 −0.37 −1.13 0.84 0.78 AEQ16 3.77 1.89 0.13 −1.11 0.87 0.67 AEQ17 3.93 1.83 −0.02 −1.01 0.88 0.71 AEQ18 4.10 1.88 −0.08 −1.07 0.77 0.76 AEQ19 3.52 2.02 0.27 −1.17 0.85 0.70 AEQ20 3.89 1.00 0.02 −1.21 0.92 0.78 AEQ21 3.97 1.99 −0.08 −1.19 0.92 0.80 AEQ22 4.65 1.97 −0.46 0.98 0.83 0.83 AEQ, aesthetic experience questionnaire. AEQ, aesthetic experience questionnaire. The results on one-way ANOVA suggest that there were statistically significant differences for scores on the AEQ between frequency groups [F(401,4) = 5.441, p < 0.001] and between meaning groups [F(401,3) = 9.447, p < 0.001]. Since the Levene statistic for frequency groups (2.295, p = 0.059) showed no violation of the assumption of equality means (p > 0.05), the Tukey’s outcomes were considered. Measures The mean differences in aesthetic experience were noticed between participants who declared participating less or not at all in artistic activities and two other frequency groups: participants practicing artistic activities several times a month (MD = −20.40; p < 0.001) and participants practicing aesthetic activities daily or almost daily (MD = −21.24; p < 0.001). More precisely, people for whom artistic activities have little or no importance presented levels of aesthetic experience significantly lower (M = 77.72) than those who practice artistic activities several times a month (M = 98.12) and daily or almost daily (M = 98.96). There were not other significant differences in aesthetic experience between abovementioned frequency groups and those people who practice artistic activities 1–2 times a week (M = 88.86) and several times a week (M = 93.96). In turn, the Levene statistic for meaning groups (2.778, p = 0.041) showed violation of the assumption of equality means (p < 0.05). Hence, the Games-Howell’s outcomes were examined. The mean differences in AE were observed between respondents who acknowledged that artistic activity was not important for them and two other meaning groups: participants learning or developing with a view to earning money in art (MD = −28.12; p < 0.001) and participants who considered artistic activity as a passion for them was 402. Also, the GFI was slightly below 0.90. However, these statistics tend to present a downward bias with a large number of degrees of freedom (Hooper et al., 2008). Considering all the other results, the model was accepted in its present form. The internal reliability presented via Cronbach’s alpha and composite reliability (CR) for the six factors was as follows: emotional (α = 0.90; CR = 0.91), cultural (α = 0.89; CR = 0.89), perceptual (α = 0.91; CR = 0.91), understanding (α = 0.92; CR = 0.92), flow conditions (α = 0.87; CR = 0.88), flow experience (α = 0.93; CR = 0.93), and overall aesthetic experience (α = 0.96; CR = 0.98). With respect to the correlations among the six dimensions of the AEQ, they ranged from 0.61 to 0.79 (p < 0.001), presenting moderate and strong associations. The strongest link was that between the perceptual dimension of the AEQ and understanding. We obtained positive and varied correlations in terms of strength (Table 2) between the aesthetic experience overall / its six dimensions and mood regulation overall / its seven dimensions. frontiersin.org Frontiers in Psychology Measures They were mostly significant weak, except for a few that were closed to moderate, and two insignificant values (between: AEQ cultural and B-MMR discharge; AEQ perceptual and B-MMR discharge). Similar results were obtained between aesthetic experience overall / its six dimensions and multidimensional existential meaning / its three dimensions. Moderate associations were observed between aesthetic experience overall / its six dimensions and emotion regulation strategies for artistic creative activities / its three dimensions. 06 Świątek et al. 10.3389/fpsyg.2023.1214928 TABLE 2 Correlations between dimensions/overall score of AEQ, B-MMR, ERS-ACA, and MEMS (N = 402). AEQ-E, aesthetic experience emotional; AEQ-C, cultural; AEQ-P, perceptual; AEQ-U, understanding; AEQ-FC, flow proximal conditions; AEQ-FE, flow experience; AEQ-O, overall; B-MMR-E, escapism; B-MMR-R, revival; B-MMR-SF, strong feelings; B-MMR-EN, mood regulation entertainment; B-MMR-D, discharge; B-MMR-I, introspection; B-MMR-S, solace; B-MMR-O, overall; ERS-AS, emotion regulation avoidance strategies; ERS-AP, approach; ERS-SD, self-development; ERS-O, overall; MEMS-C, existential meaning comprehension; MEMS-P, purpose; MEMS-M, mattering; MEMS-O, overall. AEQ-E, aesthetic experience emotional; AEQ-C, cultural; AEQ-P, perceptual; AEQ-U, understanding; AEQ-FC, flow proximal conditions; AEQ-FE, flow experience; AEQ-O, overall; B-MMR-E, escapism; B-MMR-R, revival; B-MMR-SF, strong feelings; B-MMR-EN, mood regulation entertainment; B-MMR-D, discharge; B-MMR-I, introspection; B-MMR-S, solace; B-MMR-O, overall; ERS-AS, emotion regulation avoidance strategies; ERS-AP, approach; ERS-SD, self-development; ERS-O, overall; MEMS-C, existential meaning comprehension; MEMS-P, purpose; MEMS-M, mattering; MEMS-O, overall. (MD = −19.60; p < 0.001). Thus, it can be assumed that people who are not interested in art (M = 71.05) tend to score significantly lower in aesthetic experience than people who work within field of art (M = 99.17) and who are passionate about artistic activity (M = 95.38). These meaning groups did not differ in aesthetic experience from the respondents who admitted that they make a living from artistic activity (M = 90.66). aesthetic quotient. The internal reliability for the four factors were as follows: music (α = 0.90), visual art (α = 0.89), literature (α = 0.91), film (α = 0.92), and overall aesthetic competence (α = 0.96).h The Aesthetic Processing Preference Scale (APPS; Kopatich et al., 2021) is a scale that measures individual differences in the scope of the controlled processing in relation to aesthetics. It consists of 13 statements belonging to 3 factors. They are: appreciation of complexity, intolerance for ambiguity, and propensity to contextualize. Measures The Aesthetic Competence Scale (ACS) is a tool created by a research team led by Dan et al. (2021). In the current study, we used its Polish translation. The instrument measures aesthetic quotient and consists of 20 items, which are divided into four subscales: music (e.g., While listening to music, I can make a certain judgment about the music, performers, and composers), visual art (e.g., I can identify the style and genre of a painting), literature (e.g., I may indulge in the plot of the story while reading a book), and film (e.g., I enjoy the beauty brought forth by films). The participants respond to all statements using a 5-point Likert scale (from 1 – “strongly disagree” to 5 – “strongly agree”). The higher the overall score obtained, the higher the Participants The participants were 201 Polish adults (65% women). The age range of the sample was 18–76 years (M = 26.40; SD = 11.89). The respondents were asked about their level of education. Most of them were still studying (48%), followed by those who had completed higher education (33%), secondary education (9%), diploma of vocational technician (6%), primary education (3%), and basic vocational education (1%). The Need for Cognition Scale (NCS) is a self-report tool used to measure the tendency to engage in and enjoy effortful cognitive tasks. The scale created by Cacioppo and Petty (1982) was adapted into Polish by Matusz et al. (2011). The one-dimensional scale consists of 36 items (e.g., I find it especially satisfying to complete an important task that required a lot of thinking and mental effort). The participants assess each statement by using answers on a 5-point Likert scale (from 1 – “strongly disagree” to 5 – “strongly agree”). The higher the overall score obtained, the higher the level of need for cognition. The internal reliability for the whole scale was α = 0.86. Measures The set of answers for the study subject consists of 6 possible answers, from “I definitely disagree” to “I definitely agree.” One point is assigned to the answer “I definitely disagree,” while 6 points are assigned to the answer “I definitely agree.” The scale does not contain reversed questions. The results are added up and interpreted separately within three subscales, by adding up the points gained by the study subject within each factor. The internal reliability for the three factors and overall result was as follows: appreciation of complexity (α = 0.91), intolerance for ambiguity (α = 0.79), and propensity to contextualize (α = 0.88), overall aesthetic processing preference (α = 0.89). Frontiers in Psychology frontiersin.org Results Table 3 shows selected descriptive statistics for the aesthetic experience items. All standardized loadings in the CFA model were 07 Świątek et al. 10.3389/fpsyg.2023.1214928 TABLE 3 Descriptive statistics for AEQ items, CFA loadings, and corrected item-total correlations (N = 201). TABLE 3 Descriptive statistics for AEQ items, CFA loadings, and corrected item-total correlations (N = 201). Item M SD Skewness Kurtosis Loadings Corrected item- total correlations AEQ1 4.35 1.96 −0.16 −1.18 0.89 0.79 AEQ2 4.32 1.91 −0.34 −0.97 0.90 0.79 AEQ3 4.20 1.89 −0.13 −1.13 0.89 0.77 AEQ4 3.20 1.82 0.55 −0.68 0.73 0.62 AEQ5 3.76 2.01 0.11 −1.24 0.72 0.57 AEQ6 3.92 1.98 0.08 −1.25 0.77 0.68 AEQ7 3.88 2.04 0.06 −1.31 0.81 0.65 AEQ8 4.19 1.89 −0.12 −1.02 0.82 0.69 AEQ9 4.35 1.88 −0.24 −0.94 0.86 0.79 AEQ10 4.85 1.89 −0.64 −0.74 0.79 0.73 AEQ11 4.36 1.93 −0.42 −0.95 0.89 0.84 AEQ12 4.45 1.90 −0.34 −1.02 0.85 0.79 AEQ13 4.76 1.89 −0.55 −0.77 0.85 0.75 AEQ14 4.17 1.84 −0.26 −0.93 0.86 0.80 AEQ15 4.13 1.95 −0.13 −1.13 0.61 0.62 AEQ16 3.55 1.80 0.24 −0.93 0.70 0.65 AEQ17 3.99 1.76 −0.01 −0.93 0.88 0.73 AEQ18 4.00 1.80 −0.03 −0.99 0.91 0.78 AEQ19 3.55 2.02 0.26 −1.13 0.81 0.68 AEQ20 3.84 2.05 0.12 −1.22 0.86 0.74 AEQ21 3.97 1.93 0.03 −1.15 0.87 0.78 AEQ22 4.46 1.93 −0.25 −0.11 0.89 0.83 AEQ, aesthetic experience questionnaire. AEQ, aesthetic experience questionnaire. correlation of a new validated measure with tools that measure other variables (Heale and Twycross, 2015), depression and materialism were selected as variables negatively correlated with engagement with beauty, a construct somewhat similar to aesthetic experiences. Both tools are quite well established in research in Polish cultural conditions. very good (between 0.61 and 0.91). Like in Study 1, the test for goodness-of-fit in Study 2 confirmed a six-factorial structure of the AEQ: χ2 = 447.79, df = 194, p = 0.000, χ2/df = 2.308, GFI = 0.82, TLI = 0.92, CFI = 0.93, SRMS = 0.04, RMSEA = 0.08, LO = 0.07, and HI = 0.09. The internal reliability for the six factors was as follows: emotional (α = 0.91; CR = 0.91), cultural (α = 0.86; CR = 0.86), perceptual (α = 0.88; CR = 0.89), understanding (α = 0.86; CR = 0.87), flow conditions (α = 0.88; CR = 0.87), flow experience (α = 0.91; CR = 0.92), and overall aesthetic experience (α = 0.96; CR = 0.98). Measures The Center for Epidemiologic Studies Depression Scale (CES-D) is a self-report tool used to measure depressive symptoms. The scale was created by Radloff (1977) and adapted by Jankowski (2016). The instrument consists of 20 items, which are divided into four components of depressive symptomatology: somatic symptoms (e.g., Results The correlations among the six dimensions of the AEQ, which ranged from 0.59 to 0.77 (p < 0.001), showed moderate and strong associations. Like in Study 1, the strongest link was between the perceptual dimension of the AEQ and understanding.i Participants The sample consisted of a total of 210 Polish adults (71% women). The age range of all the participants was 18–74 years (M = 26.87; SD = 11.97). Similarly to Study 2, the respondents were asked about their level of education. A relatively large proportion of the participants were still studying (52%), followed by those who had completed higher education (29%), secondary education (8%), diploma of vocational technician (6%), primary education (3%), and basic vocational education (2%). Aesthetic experience showed significant positive and moderate correlations with visual art, overall aesthetic competence, appreciation of complexity, propensity to contextualize, and overall aesthetic processing preference. Moreover, significant positive associations were observed between aesthetic experience and aesthetic competence in music, literature, and film, intolerance for ambiguity, and need for cognition (Table 4). Frontiers in Psychology Results I was bothered by things that usually do not bother me), depressed affect (e.g., I felt that I could not shake off the blues even with help from my family or friends), positive affect (e.g., I felt I was just as good as other people), and interpersonal relations (e.g., I felt that people dislike me). The participants answer how often over the past week they experienced symptoms associated with depression using a 3-point Likert scale from 1 – “some or little of the time” to 3 – “most or almost all the time.” Scores range from 0 (no depressive symptoms) to 60 (maximum depressiveness). The higher the overall score obtained, the greater the depressive symptoms. The reliability of the whole scale was α = 0.84 and for its four subscales was: somatic symptoms (α = 0.83), depressed affect (α = 0.89), positive affect (α = 0.81), and interpersonal relations (α = 0.65). Table 5 presents the descriptive statistics for the aesthetic experience items. All standardized loadings in the CFA model were very good (between 0.67 and 0.93). The test for the goodness-of-fit test in Study 3 indicated a poor fit to the data: χ2 = 542.00, df = 194, p = 0.000, χ2/df = 2.794, GFI = 0.81, TLI = 0.90, CFI = 0.91, SRMS = 0.05, RMSEA = 0.09, LO = 0.08, and HI = 0.10. The internal reliability for the six factors was as follows: emotional (α = 0.91; CR = 0.91), cultural (α = 0.84; CR = 0.84), perceptual (α = 0.91; CR = 0.91), understanding (α = 0.90; CR = 0.90), flow conditions (α = 0.86; CR = 0.91), flow experience (α = 0.92; CR = 0.88), and overall aesthetic experience (α = 0.96; CR = 0.98). The correlations among the six dimensions of the AEQ, which ranged from 0.59 to 0.76 (p < 0.001), presented moderate and strong associations. The strongest link was between flow experience and understanding. The Material Values Scale (MVS) is a self-report tool developed by Richins and Dawson (1992). The Polish adaptation was created by Górnik-Durose (2016). The instrument measures valuing the possession and accumulation of material goods by three factors: success (e.g., I admire people who own expensive homes, cars, and clothes), centrality (e.g., Buying things gives me a lot of pleasure), and happiness (e.g., My life would be better if I owned certain things I do not have). Results The respondents relate to all statements using a 5-point Likert scale (from 1 – “strongly disagree” to 5 – “strongly agree”). The higher the overall score obtained, the greater the appreciation of material values. The reliability of the whole scale was α = 0.82 and for its three subscales was: success (α = 0.65), centrality (α = 0.57), and happiness (α = 0.67). Some dimensions of aesthetic experience showed a few weak negative correlations with happiness related to the possession and accumulation of material goods (Table 6). The other variables displayed no significant associations with aesthetic experience. Study 3 The aim of the third study was to test divergent validity of the AEQ. Since divergent validity is the demonstration of a weak 08 Świątek et al. 10.3389/fpsyg.2023.1214928 TABLE 4 Correlations between dimensions/overall score of AEQ, ACS, APPS, and NCS (N = 201). TABLE 4 Correlations between dimensions/overall score of AEQ, ACS, APPS, and NCS (N = 201). AEQ-E, aesthetic experience emotional; AEQ-C, cultural; AEQ-P, perceptual; AEQ-U, understanding; AEQ-FC, flow conditions; AEQ-FE, flow experience; AEQ-O, overall; ACS-M, aesthetic competence music; ACS-VA, visual art; ACS-L, literature; ACS-F, Film; ACS-O, overall; APPS-AC, aesthetic processing appreciation of complexity; APPS-IA, intolerance for ambiguity; APPS- PC, propensity to contextualize; APPS-O, overall; NCS, need for cognition. AEQ-E, aesthetic experience emotional; AEQ-C, cultural; AEQ-P, perceptual; AEQ-U, understanding; AEQ-FC, flow conditions; AEQ-FE, flow experience; AEQ-O, overall; ACS-M, aesthetic competence music; ACS-VA, visual art; ACS-L, literature; ACS-F, Film; ACS-O, overall; APPS-AC, aesthetic processing appreciation of complexity; APPS-IA, intolerance for ambiguity; APPS- PC, propensity to contextualize; APPS-O, overall; NCS, need for cognition. Frontiers in Psychology frontiersin.org Discussion After finding articles citing the publication of Wanzer et al. (2020), none of them concerned the validation, adaptation or creation of a version in a foreign language. In order to adjust the AEQ to the Polish conditions, 3 studies were conducted with 9 main hypotheses. It is interesting that, in the original study, the authors did not perform a CFA. Therefore, in order to confirm the non-randomness of the factor division for the AEQ, in this study, a CFA was carried out three times, in each of the three independent attempts. foreign language. After finding articles citing the publication of Wanzer et al. (2020), none of them concerned the validation, adaptation or creation of a version in a foreign language. In order to adjust the AEQ to the Polish conditions, 3 studies were conducted with 9 main hypotheses. It is interesting that, in the original study, the authors did not perform a CFA. Therefore, in order to confirm the non-randomness of the factor division for the AEQ, in this study, a CFA was carried out three times, in each of the three independent attempts. With respect to the structure of the AEQ, the results obtained in the three studies support the six-factor model of the original AEQ, obtained through the Exploratory Factor Analyses. The good fit indices of the CFA suggest the psychometric solidity of the Polish version of the AEQ. The reliability values, like for the original ones, confirm the internal consistency of the measure. Thus, the results denote that the Polish version of the scale is a reliable tool and manifests similar psychometric characteristics to Wanzer’s version. Moreover, the correlations between all six subscales presented comparable results, although being slightly stronger. Therefore, the AEQ, with its emotional, cultural, perceptual, understanding, flow conditions, and flow experience dimensions, can be used to assess aesthetic experience. In the first study, it was proven that the overall result and individual components of aesthetic experience positively correlate with mood regulation through music (H1). It can be said that people who have strong, deep aesthetic experience are in general more prone to use music to modify their emotions. Frontiers in Psychology Discussion The article describes (to the best of our knowledge) the first attempt to validate the Aesthetic Experience Questionnaire in a 09 Świątek et al. 10.3389/fpsyg.2023.1214928 TABLE 5 Descriptive statistics for AEQ items, CFA loadings, and corrected item-total correlations (N = 210). Item M SD Skewness Kurtosis Loadings Corrected item- total correlations AEQ1 4.55 1.89 −0.37 −0.88 0.90 0.80 AEQ2 4.57 1.79 −0.44 −0.79 0.89 0.78 AEQ3 4.46 1.86 −0.29 −1.00 0.93 0.81 AEQ4 3.33 1.80 0.37 −0.93 0.67 0.58 AEQ5 4.01 1.93 0.09 −1.20 0.75 0.59 AEQ6 4.13 2.03 −0.11 −1.35 0.80 0.59 AEQ7 3.70 1.87 0.13 −1.15 0.70 0.52 AEQ8 4.16 1.83 −0.18 −1.00 0.77 0.75 AEQ9 4.51 1.91 −0.38 −0.98 0.88 0.77 AEQ10 4.97 1.89 −0.80 −0.46 0.88 0.74 AEQ11 4.49 1.95 −0.31 −1.08 0.87 0.80 AEQ12 4.51 1.88 −0.29 −0.99 0.86 0.82 AEQ13 4.63 1.88 −0.53 −0.77 0.83 0.74 AEQ14 4.41 1.96 −0.28 −1.11 0.90 0.84 AEQ15 4.38 1.90 −0.32 −1.07 0.74 0.70 AEQ16 3.60 1.78 0.21 −0.95 0.73 0.66 AEQ17 3.77 1.71 0.13 −0.80 0.86 0.69 AEQ18 3.91 1.78 −0.01 −1.02 0.90 0.74 AEQ19 3.57 2.02 0.28 −1.18 0.71 0.69 AEQ20 3.94 1.95 −0.03 −1.19 0.82 0.74 AEQ21 4.13 1.81 −0.12 −0.96 0.89 0.82 AEQ22 4.68 1.97 −0.48 −0.92 0.92 0.84 AEQ, aesthetic experience questionnaire. AEQ, aesthetic experience questionnaire. concurrent, convergent and discriminant validity. Due to the difficulty of checking the convergent validity of the AEQ (Polish researchers do not have many fully verified measurements from the scope of the psychology of aesthetics and art), we selected only variables potentially positively and negatively correlated with it. When selecting potential correlations, it was decided to regulate emotions using art [specific for music (H1), and generally for artistic creative activities in any field (H2)], as well as aesthetic competences in four selected fields of art (H5). The relationship between the cognitive aspects of the perception of art (controlled processing) and the intensity of aesthetic experience was considered (H6), and the need for cognition was taken into account (H7). With knowledge of the therapeutic significance of art, the hypothesis on a positive relationship between aesthetic experience and the meaning of life was formulated (H3). The concept of beauty plays an essential role in aesthetics; therefore, referring to knowledge of engagement in beauty (Diessner et al., 2008), it was assumed that aesthetic experience will be negatively correlated with materialism (H8) and depression (H9) to check discriminant validity. foreign language. frontiersin.org Discussion A further issue is that music (especially popular music) has become westernized (Drygas, 2015) and the ability to analyze the work of art in the cultural context is not necessary to use music to cope with difficult emotions.i the opinion that the results fit into the generally available knowledge of the therapeutic role of art. Shaping artistic sensitivity and building resources of aesthetic experience may be conducive to coping with difficulties in a creative and constructive manner. Even in the case of the dominance of avoidance strategies, performing creative activities seems to be a more psycho-physically hygienic strategy of coping with difficulties than running away from problems into, for example, a virtual world or drugs. Furthermore, the assumption about the co-occurrence of strong results in the AEQ and a stronger overall result and in subscales for meaning of life (H3) was also confirmed. Nevertheless, the correlation observed is very weak. We are willing to assume that although engagement in art may be a source of pleasant emotions and may give the recipient of the work grounds for reflections going beyond the aesthetic situation, the sense of understanding, purpose and meaning of one’s own life depend to a greater extent on other psychological factors. Art may be a tool that supports searching for meaning, but most likely, if it is not in the center of the individual’s interests (profession, passion), the role of aesthetic experience for the sense of the meaning of life is marginal. Moreover, it was confirmed that aesthetic experience is positively correlated with emotion regulation through artistic creative activities (H2). The relationships between individual dimensions of both tools range from moderate to weak, while the dimension of practicing creative activities for self-development is most closely correlated with all dimensions of the AEQ. This means that people who declare stronger aesthetic experience in contact with art are more likely to undertake creative activities to cope with everyday challenges. The relationships observed are statistically significant, both in the case of strategies based on self-development, obtaining new insight into the situation, and in the case of avoiding difficulties (distracting attention from them). A question arises whether it would be possible to distinguish profiles of people applying the given strategies more frequently and the manner in which they are connected with aesthetic experience, to which currently we do not have an answer. Frontiers in Psychology frontiersin.org Discussion The weakest correlations for all dimensions of the AEQ are two strategies of using music, i.e., discharge (venting negative emotions) and distraction from worries The literature indicates various types of tool validity verification, including content validity, construct validity, face validity and criterion validity (Bahariniya et al., 2021). The latter includes predictive, 10 Świątek et al. 10.3389/fpsyg.2023.1214928 TABLE 6 Correlations between dimensions/overall score of AEQ, CES, and MVS (N = 210). TABLE 6 Correlations between dimensions/overall score of AEQ, CES, and MVS (N = 210). AEQ-E, aesthetic experience emotional; AEQ-C, cultural; AEQ-P, perceptual; AEQ-U, understanding; AEQ-FC, flow proximal conditions; AEQ-FE, flow experience; AEQ-O, overall; CES-SS, center for epidemiologic studies depression somatic symptoms; CES-DA, depressed affect; CES-PA, positive affect; CES-IR, interpersonal relations; CES-O, overall; MVS-S, material values success; MVS-C, centrality; MVS-H, happiness; MVS-O, overall. AEQ-E, aesthetic experience emotional; AEQ-C, cultural; AEQ-P, perceptual; AEQ-U, understanding; AEQ-FC, flow proximal conditions; AEQ-FE, flow experience; AEQ-O, overall; CES-SS, center for epidemiologic studies depression somatic symptoms; CES-DA, depressed affect; CES-PA, positive affect; CES-IR, interpersonal relations; CES-O, overall; MVS-S, material values success; MVS-C, centrality; MVS-H, happiness; MVS-O, overall. and stress. In the case of discharge and distraction, no relationship with the perceptual dimension of the AEQ was observed. Also, the strategy of using music based on discharge does not correlate with the cultural dimension of the AEQ. The results obtained are not particularly surprising. People who use music to distract themselves from difficulties or to give vent to anger do not need the ability to focus on details of the work or to reflect on the cultural context of the work for this purpose. A further issue is that music (especially popular music) has become westernized (Drygas, 2015) and the ability to analyze the work of art in the cultural context is not necessary to use music to cope with difficult emotions.i and stress. In the case of discharge and distraction, no relationship with the perceptual dimension of the AEQ was observed. Also, the strategy of using music based on discharge does not correlate with the cultural dimension of the AEQ. The results obtained are not particularly surprising. People who use music to distract themselves from difficulties or to give vent to anger do not need the ability to focus on details of the work or to reflect on the cultural context of the work for this purpose. Discussion We are of With respect to H4, the results proved to be consistent with the adopted assumptions. First, it can be said that the frequency of participation in artistic events tends to be associated with stronger and deeper aesthetic experiences. In fact, the frequency of participation turned out to differentiate some frequency groups in relation to the AEQ. Although our justification for the hypothesis seems logical in 11 Świątek et al. 10.3389/fpsyg.2023.1214928 10.3389/fpsyg.2023.1214928 correlate with the appreciation of complexity. Similarly, all dimensions of the AEQ significantly positively correlate with the propensity to contextualize. It may be assumed that people who declare that they have strong aesthetic experience also appreciate the high level of complexity of works of art, have the ability to look at the work of art from many perspectives and like ambiguity. It is also not surprising that in contact with works of art, they are willing to take into account a broader context (the person of the artist, the time during which the work was created and the circumstances). This is also indicated by the positive verification of hypothesis 7 (H7; concerns the relationship between AE and the need for cognition). For people who like to engage in deep reflective thoughts, contemplating complex works of art can provide a stronger aesthetic experience and be cognitively satisfying. The assumption about the negative correlation of the AE dimensions with intolerance for ambiguity was not confirmed. It was observed that there are very weak positive correlations among the emotional dimension, understanding, proximal conditions for flow, flow experience, and intolerance for ambiguity. The result also indicates that people who prefer explicitness in the interpretation and simpler aesthetic objects may also have stronger aesthetic experience. It may be said that less sophisticated aesthetic tastes do not make it impossible to derive deeper experience from contact with art, although in the case of the cultural and perceptual dimension of AE, the relationship with intolerance for ambiguity does not occur at all. In the case of preferring structurally and symbolically uncomplicated works of art, the ability to notice details and to refer the work to works of other artists or times of creation probably is not significant or necessary. the light of contact with art as a goal itself, we are not sure why this is the case. Discussion When it comes to the measurement of the depression indicator, no significant relationships between the CES-D overall result and the AEQ overall result and individual dimensions were observed. Also, the analysis of the relationships between the results obtained from the study subjects in the individual factors of both tools proved that there are no significant relationships. Practical implications from the results obtained refer rather to artistic education at the general level – it is difficult to acquire knowledge of art without the possibility of having close contact with art. In Poland, the “cultural knowledge” subject (combining anthropological issues with history of art) was removed from compulsory subjects and teaching it at schools depends on the school management’s decision. Currently, a student of a secondary school or technical secondary school may choose from music, art and philosophy (taught at the first grade, 1 h per week, within the 4-year cycle of the secondary-school education). In the situation where artistic subjects are treated as a “necessary evil,” it is difficult to build aesthetic competences and knowledge of cultural heritage. Teachers do not have time to go with students to museums, art exhibitions, the theater or concerts. The aesthetic competences and sensitivity of young recipients are shaped mainly through mass culture. It is difficult to acquire aesthetic competences if a person does not experience art in real life, if the contact with art is very rare and reserved only for elites (and due to this, understood to a very limited extent) or marginalized (considered as an insignificant “addition” in the general education). i Similarly, no significant relationship between the MVS overall result and the AEQ overall result and subscales was found. Some statistically significant correlations between the dimension of materialistic happiness and the emotional dimension, cultural dimension, flow experience and the AEQ overall result were observed, although they were very weak negative correlations. People whose sense of happiness is rooted in possession and material well-being declare to a lesser extent that they have aesthetic experience. Aesthetic experience is closer to the spiritual sphere; it does not translate into the increase in material resources and it is therefore probably not perceived as valuable by people with a materialistic value orientation. Hence, people appreciating material values probably do not attach much importance to it and do not get engaged in experiencing art. Discussion The frequency of participation in cultural events may be driven by many motives – from purely cognitive to social (being invited or spending time with friends) or these related to prestige. The type of artistic event can also be important for the depth and type of experience – at a concert of popular music, the aesthetic experiences of the musician and the listener are related to the context of community, as well. The reception of the painting during the vernissage can reflect an immersion into the finished work. Second, it can be suggested that giving different meanings to artistic activities may be related to aesthetic experiences. In fact, the strongest aesthetic experiences are shared by people who learn or develop with a view to earning money in art and those who consider art as their passion. However, it is difficult to say unequivocally whether it is the more intense way of receiving art that predestines to deal with art professionally or out of passion, or rather it is the other way around – that making art the “axis” of life determines the depth of artistic experience. In the second study, 3 hypotheses were verified. According to the assumptions adopted, all dimensions of aesthetic experience correlated significantly and positively with the overall result in the ACS (the level of the Aesthetic Quotient in relation to four fields of art) and the result for each of the dimensions separately (H5). The AEQ measurement was created in relation to visual arts, which is why it is not surprising that all AEQ factors were most closely connected with aesthetic competences in the scope of visual arts (the strength of these relationships is moderate), and then with the overall result of the ACS (also moderate). The results suggest that people with high aesthetic competences also declare stronger cognitive and emotional aesthetic experience. We assume the existence of a two-way relationship, which means that: acquiring greater competences in the scope of art makes people more sensitive to details of works, allows them to better understand those works and to experience more intensive contact with art, but also vice versa, being receptive to aesthetic experience will deepen the aesthetic competences. The third study verified two assumptions concerning potentially negative correlations of aesthetic experience, i.e., materialistic attitude (H8) and depression (H9). Frontiers in Psychology frontiersin.org Author contributions AŚ and MSz conceived and designed the study, analyzed the data, interpreted the results, drafted the manuscript, wrote the manuscript, read, and approved the final version. KW, MSt, and MC performed the study, analyzed the data, drafted the manuscript, read, and approved the final version. All authors contributed to the article and approved the submitted version. Carried out online, the three studies are burdened with all the reservations characteristic for this type of studies (from the randomness of the participant group to difficulty in providing the study subjects with identical conditions when completing the questionnaire). Another point is that the selection of correlation variables did not reflect, in terms of criteria, the AEQ factors (for example, no tool to measure dispositional flow was applied). It seems to us that the novelty was also an advantage and constitutes an added value to the validation process. Some of the measured constructs seem to be similar to AE (e.g., the ACS measuring the level of the Aesthetic Quotient for music, visual arts, literature and film contained questions about knowledge, understanding of art and emotional engagement, which are distinguished as dimensions in the AEQ), but the strength of the correlations does not indicate that they overlap. Limitations In terms of the methodology, the researchers in some cases using Polish translations of certain tools from the scope of art and aesthetics instead of their validations may be considered a fault of this study. We would like to mention that studies were conducted which confirm the good psychometric characteristics of the Polish versions of the ACS, B-MMR and ERS-ACA, and articles concerning these measurements are currently being prepared by Świątek and her research team. Discussion It does not mean that materialists do not appreciate art at all; it probably interests them if it can be treated as an investment or if possessing it is connected with prestige. Also, no relationships between the central character of material values and the dimensions of aesthetic experience were observed, while a very weak positive relationship between material success and proximal conditions for flow is surprising. High results on the scale of material success indicate that people appreciating material values base their sense of success on possessing things that are expensive and make an i The next hypothesis subject to the verification (H6) concerned aesthetic experience with three elements of aesthetic processing preferences. All dimensions of the AEQ significantly positively 12 frontiersin.org Świątek et al. 10.3389/fpsyg.2023.1214928 cognition, some dimensions of aesthetic processing preference, and also to a much lesser extent existential meaning. Also, negative – although very weak – relationships between one of the three dimensions of materialism (happiness based on material values) and the AEQ overall result and three dimensions were observed. impression on other people. The criterion of conditions making it easier to experience flow is connected, e.g., with possessing a clear vision of details of the work of art to which attention should be paid and the conviction of the correctness of one’s own reflections on the work of art. For people who are interested in art for materialistic reasons (collection or investments), certain knowledge of the art market, current trends and fashion and the value of other works of a given artist are necessary when they purchase works of art. Therefore, such a person may have specific expectations toward works of art and, from this point of view, an apparently unexpected correlation seems to be justified. Conclusion All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. The article probably describes the first validation of the Aesthetic Experience Questionnaire for the Polish cultural conditions. Moreover, it is probably the first publication in which a CFA was carried out for this tool. The results of 3 independent studies indicate that the Polish version of the AEQ has very good psychometric characteristics and is highly recommendable for research on the psychology of aesthetics and art as well as individual differences (the Polish version of the questionnaire is available in the Appendix). Additionally, the studies demonstrated the existence of statistically significant, positive correlations between AE and mood regulation through art (listening to music, artistic creative activities), aesthetic competences, need for Data availability statement The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://osf.io/rsby2/?view_only=24860fa 200e44fd09976b78f5b92eab0. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Ethics statement The studies involving humans were approved by the Committee for the Ethics of Research of the Institute of Psychology at the University of Szczecin (KB 6/2022 of 27/04/2022). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. The description of the results lacks statistics comparing groups of professional artists, amateur artists, and people who do not deal with art. People with different levels of knowledge and practice in the scope of art participated in the studies, but uneven representation would make it difficult to obtain reliable results. In the future, this could be taken into account. Frontiers in Psychology References Gerymski, R., and Krok, D. (2020). A Polish adaptation of the multidimensional existential meaning scale: internal structure, reliability, and validity. Roczniki Psychol. 23, 173–190. doi: 10.18290/rpsych20232-4 Akoglu, H. (2018). User's guide to correlation coefficients. Turk. J. Emerg. Med. 18, 91–93. doi: 10.1016/j.tjem.2018.08.001 Akoglu, H. (2018). User's guide to correlation coefficients. Turk. J. Emerg. Med. 18, 91–93. doi: 10.1016/j.tjem.2018.08.001 Allen, R. (2017). Statistics and experimental design for psychologists: A model comparison approach. London: World Scientific Publishing Company. Allen, R. (2017). Statistics and experimental design for psychologists: A model comparison approach. London: World Scientific Publishing Company. Górnik-Durose, M. (2016). Polska adaptacja skali wartości materialnych (MVS)- właściwości psychometryczne wersji pełnej i wersji skróconych. Psychol. Ekonomiczna 9, 5–21. doi: 10.15678/PJOEP.2016.09.01 Bahariniya, S., Ezatiasar, M., and Madadizadeh, F. (2021). A brief review of the types of validity and reliability of scales in medical research. J. Commun. Health Res. 10, 100–102. doi: 10.18502/jchr.v10i2.6582 Granot, R., Spitz, D. H., Cherki, B. R., Loui, P., Timmers, R., Schaefer, R. S., et al. (2021). “Help! I need somebody”: music as a global resource for obtaining wellbeing goals in times of crisis. Front. Psychol. 12:648013. doi: 10.3389/fpsyg.2021.648013 Blomdahl, C., Gunnarsson, A. B., Guregård, S., and Björklund, A. (2013). A realist review of art therapy for clients with depression. Arts Psychother. 40, 322–330. doi: 10.1016/j.aip.2013.05.009 Harrington, D. (2009). Confirmatory factor analysis. New York: Oxford University Press. Bowen, N. K., and Guo, S. (2011). Structural equation modeling. New York: Oxford University Press. Heale, R., and Twycross, A. (2015). Validity and reliability in quantitative studies. Evid. Based Nurs. 18, 66–67. doi: 10.1136/eb-2015-102129 Brattico, E., Bogert, B., and Jacobsen, T. (2013). Toward a neural chronometry for the aesthetic experience of music. Front. Psychol. 4:206. doi: 10.3389/fpsyg.2013.00206 Hooper, D., Coughlan, J., and Mullen, M. R. (2008). Structural equation modeling: guidelines for determining model fit. The electronic journal of business research. Methods 6, 53–60. doi: 10.21427/D7CF7R Brydges, C. C. (2019). Effect size guidelines, sample size calculations, and statistical power in gerontology. Innov. Aging 3:igz036. doi: 10.1093/geroni/igz036 Jankowski, K. S. (2016). Morningness-eveningness and depressive symptoms: test on the components level with CES-D in Polish students. J. Affect. Disord. 196, 47–53. doi: 10.1016/j.jad.2016.02.015 Bullot, N. J., and Reber, R. (2013). The artful mind meets art history: toward a psycho- historical framework for the science of art appreciation. Behav. Brain Sci. 36, 123–137. doi: 10.1017/S0140525X12000489 Kopatich, R. D., Steciuch, C. C., Feller, D. P., Millis, K., and Siegesmund, R. Supplementary material The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1214928/ full#supplementary-material 13 frontiersin.org 10.3389/fpsyg.2023.1214928 10.3389/fpsyg.2023.1214928 Świątek et al. References (2021). Development and validation of the Aesthetic Processing Preference Scale (APPS). Psychol. Aesth. Creat. Arts. doi: 10.1037/aca0000449 [E-pub ahead of print]. Cacioppo, J. T., and Petty, R. E. (1982). The need for cognition. J. Pers. Soc. Psychol. 42, 116–131. doi: 10.1037/0022-3514.42.1.116 Csikszentmihalyi, M., and Robinson, R. E. (1990). The art of seeing: An interpretation of the aesthetic encounter. Malibu, CA: J.P. Getty Museum. Leder, H., Gerger, G., Dressler, S. G., and Schabmann, A. (2012). How art is appreciated. Psychol. Aesthet. Creat. Arts 6, 2–10. doi: 10.1037/a0026396 Dan, Y., Wu, C., and Yang, M. (2021). Development and Validation of the Aesthetic Competence Scale. Advance. Preprint. Madrid, H. P., and Patterson, M. G. (2016). Creativity at work as a joint function between openness to experience, need for cognition and organizational fairness. Learn. Individ. Differ. 51, 409–416. doi: 10.1016/j.lindif.2015.07.010 Di Dio, C., Macaluso, E., and Rizzolatti, G. (2007). The golden beauty: brain response to classical and renaissance sculptures. PLoS One 2:e1201. doi: 10.1371/ journal.pone.0001201 Marković, S. (2012). Components of aesthetic experience: aesthetic fascination, aesthetic appraisal, and aesthetic emotion. i-Perception 3, 1–17. doi: 10.1068/i0450aap Diessner, R., Solom, R. D., Frost, N. K., Parsons, L., and Davidson, J. (2008). Engagement with beauty: appreciating natural, artistic, and moral beauty. J. Psychol. 142, 303–332. doi: 10.3200/JRLP.142.3.303-332 Matusz, P. J., Traczyk, J., and Gąsiorowska, A. (2011). Kwestionariusz Potrzeby Poznania–konstrukcja i weryfikacja empiryczna narzędzia mierzącego motywację poznawczą. Psychol. Społeczna 62, 113–128. Drygas, P. (2015). Zjawisko westernizacji muzyki popularnej. Przyczyny, mechanizmy, skutki. Opuscula Sociologica 4, 21–30. doi: 10.18276/os.2015.4-02 Meng, F., and Liang, Y. (2022). A new interpretation of the essence of aesthetic experience: from the perspective of cognitive neuroaesthetics and aesthetic anthropology. J. Aesthetic Educ. 56, 97–120. doi: 10.5406/15437809.56.4.07 Elsden, E., and Roe, B. (2021). Does arts engagement and cultural participation impact depression outcomes in adults: a narrative descriptive systematic review of observational studies. J. Public Ment. Health 20, 159–171. doi: 10.1108/JPMH-06-2020-0060 Osowiecka, M., and Gacka, I. (2016). Artysta bez skrzydeł? Regulacja emocji poprzez doświadczenia estetyczne. Ars Educandi 13, 67–74. doi: 10.26881/ae.2016.13.06 Fancourt, D., Garnett, C., and Müllensiefen, D. (2020). The relationship between demographics, behavioral and experiential engagement factors, and the use of artistic creative activities to regulate emotions. Psychol. Aesth. Creat. Arts. Advance online publication. Radloff, L. S. (1977). The CES-D scale. Appl. Psychol. Meas. 1, 385–401. doi: 10.1177/014662167700100306 Ramachandran, V. S., and Hirstein, W. (1999). The science of art: a neurological theory of aesthetic experience. J. Conscious. Stud. References 6, 15–51. Fancourt, D., Garnett, C., Spiro, N., West, R., and Müllensiefen, D. (2019). How do artistic creative activities regulate our emotions? Validation of the Emotion regulation Strategies for Artistic Creative Activities Scale (ERS-ACA). PloS One 14:e0211362. doi: 10.1371/journal.pone.0211362 Richins, M. L., and Dawson, S. (1992). A consumer values orientation for materialism and its measurement: scale development and validation. J. Consum. Res. 19, 303–316. doi: 10.1086/209304 Saarikallio, S. (2012). Development and validation of the Brief Music in Mood Regulation Scale (B-MMR). Music Percep. Interdis. J. 30, 97–105. doi: 10.1525/ mp.2012.30.1.97 Faul, F., Erdfelder, E., Lang, A. G., and Buchner, A. (2007). GPower 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav. Res. Methods 39, 175–191. doi: 10.3758/bf03193146 Fayn, K., Silvia, P. J., Erbas, Y., Tiliopoulos, N., and Kuppens, P. (2018). Nuanced aesthetic emotions: emotion differentiation is related to knowledge of the arts and curiosity. Cognit. Emot. 32, 593–599. doi: 10.1080/02699931.2017.1322554 Swanson, R. A., and Holton, E. F. (2005). Research in organizations: Foundations and methods in inquiry. San Francisco: Berrett-Koehler Publishers. Van den Tol, A. J., and Edwards, J. (2015). Listening to sad music in adverse situations: how music selection strategies relate to self-regulatory goals, listening effects, and mood enhancement. Psychol. Music 43, 473–494. doi: 10.1177/0305735613517410 Field, A. (2000). Discovering statistics using SPSS for windows sage publications. London: Sage Publications. Vessel, E. A., Starr, G. G., and Rubin, N. (2012). The brain on art: intense aesthetic experience activates the default mode network. Front. Hum. Neurosci. 6:66. doi: 10.3389/ fnhum.2012.00066 Flynn, P. M., John, M. T., and Sistani, M. M. N. (2019). Confirmation of the unidimensional structure of the Oral Health Literacy Adults Questionnaire. Int. Dent. J. 69, 207–213. doi: 10.1111/idj.12451 Furr, M., and Bacharach, V.R. (2008). Psychometrics: An introduction. California: Sage Publications. Wanzer, D. L., Finley, K. P., Zarian, S., and Cortez, N. (2020). Experiencing flow while viewing art: development of the aesthetic experience questionnaire. Psychol. Aesthet. Creat. Arts 14, 113–124. doi: 10.1037/aca0000203 George, L. S., and Park, C. L. (2017). The multidimensional existential meaning scale: a tripartite approach to measuring meaning in life. J. Posit. Psychol. 12, 613–627. doi: 10.1080/17439760.2016.1209546 Weiner, I. B., and Greene, R. L. (2017). Handbook of personality assessment. New Jersey: John Wiley & Sons. 14 Frontiers in Psychology frontiersin.org
https://openalex.org/W2955385764	https://jmla.pitt.edu/ojs/jmla/article/download/563/877	English	null	Interprofessional collaboration between health sciences librarians and health professions faculty to implement a book club discussion for incoming students	Journal of the Medical Library Association	2,019	cc-by	5,198	403 403 CASE REPORT DOI: dx.doi.org/10.5195/jmla.2019.563 Interprofessional collaboration between health sciences librarians and health professions faculty to implement a book club discussion for incoming t d t Jen Haley, MSN, RN, CNL; Rebecca Carlson McCall, MLS, AHIP; Meg Zomorodi, PhD, RN, CNL; Lisa de Saxe Zerdan, PhD, MSW; Beth Moreton, MLS; Lee Richardson See end of article for authors’ affiliations See end of article for authors’ affiliations. Background: The following case example provides an overview of one innovative way to engage health professions faculty with health sciences librarians in the development of an interprofessional book discussion and identifies strategies to address implementation challenges. Academic health sciences librarians worked with the Interprofessional Education (IPE) Steering Committee to organize interprofessional book discussion groups for incoming health professions students. This inaugural book discussion brought together students and faculty of different disciplines to engage students in “learning from, with, and about” other professions. Case Presentation: When Breath Becomes Air, by Paul Kalanithi, allowed involved discussions on important health sciences issues. The project included outreach, designing pre- and post-surveys, scheduling participants, and communicating with all participants before, during, and after the event. A total of seventy- nine students and thirty-six faculty, representing all health professions schools, participated in the small group IPE book discussions over two weeks. Conclusions: Small group book discussions have been shown to be an effective tool to engage students and faculty in IPE. The results of the participant surveys were positive, and the IPE Steering Committee found value in including health sciences librarians throughout the process. Lessons learned from the pilot project include needing an efficient scheduling system, strongly communicating at all stages of the project, and starting the planning process months ahead of time. The IPE Steering Committee plans to conduct similar book discussions every fall semester moving forward and explore options for other IPE events. See end of article for supplemental content. Rationale for interprofessional education (IPE) activity The IPE Steering Committee comprises faculty and staff representing all health professions schools (allied health, dentistry, medicine, nursing, pharmacy, public health, and social work) at our institution as well as health sciences librarians. Established in 2016, the IPE Steering Committee meets monthly to discuss existing IPE activities at the university and to strategize the implementation of new activities. While the IPE Steering Committee has had success with robust didactic and clinical activities, the committee needed to engage students who had not yet matriculated into their respective programs, as evidence demonstrates the importance of initiating IPE activities early in the curriculum. The IPE Steering Committee comprises faculty and staff representing all health professions schools (allied health, dentistry, medicine, nursing, pharmacy, public health, and social work) at our institution as well as health sciences librarians. Books about patient or health care provider experiences allow readers to reflect on the often sensitive or shocking parts of the story in a safe environment when they are emotionally ready [14, 24]. Book discussions present issues in “an environment that encourages the students to share their insights and solutions to patient and professional dilemmas” [26]. This makes an IPE book discussion an ideal activity to allow students to engage with others from different disciplines, while also achieving the definition of IPE as outlined by the World Health Organization [1]. Creating an IPE opportunity that coincides with the socialization of students to their professional schools encourages students to engage with other professions as well as members of their own profession. This opportunity was intended to help break down stereotypes or myths early in student learning as well give learners an opportunity to ask questions about other professions. The timing of the activity was purposefully planned for the beginning of the fall semester to best establish IPE as an institutional norm. Additionally, given the diverse role of various faculty and librarians from across the health affairs school, the IPE Steering Committee modeled shared leadership and enthusiasm for team work to the students who participated in this event. In the past, librarians have been a part of book discussions, but many times their participation has followed the traditional role of recommending books or teaching participants how to find resources [19, 31]. However, librarians can perform a wide variety of roles in planning educational activities. BACKGROUND cultures, histories, and perspectives are necessary to create robust IPE learning experiences that are meaningful to and appropriate for all learners [4]. In health professions that are frequently siloed, it can be a challenge to plan an event that carries meaning for all participants. Interprofessional education (IPE) occurs when “students from two or more professions learn about, from and with” one another with the purpose of enabling effective collaboration to improve health outcomes [1], whether didactically in course work, in clinical learning environments [2], or both [3]. However, IPE activities require more than just bringing students from diverse disciplines together [4]. Careful considerations about professional The literature has identified several barriers to implementation of IPE activities that can lead to didactic activities being “one and done” and conducted without evaluation metrics. In a Journal of the Medical Library Association 107 (3) July 2019 jmla.mlanet.org jmla.mlanet.org Haley et al. 404 DOI: dx.doi.org/10.5195/jmla.2019.563 IPE activities such as lack of leadership, lack of enthusiasm, and the challenge of time and resources. systematic review of interprofessional education, ten implementation challenges were identified related to: (1) curriculum, (2) leadership, (3) resources, (4) stereotypes, (5) students’ diversity, (6) the concept of IPE, (7) teaching, (8) enthusiasm, (9) professional jargons, and (10) accreditation [5]. In particular, the difficulty of integrating IPE into an already packed curriculum, reconciliation of the different schedules of health professions schools, and the amount of time and resources required to conduct IPE activities are common problems listed in the literature [5]. STUDY PURPOSE This paper outlines a specific IPE book discussion activity related to Paul Kalanithi’s When Breath Becomes Air [36] that included a total of eighty students and thirty-nine faculty representing all health affairs schools. The reasons for selection of this book, the creation of a discussion guide, and the process and implementation of twenty-nine small IPE discussions are described. Additionally, pre− and post−book discussion survey results and information on scheduling and communication strategies used with all participants before, during, and after the activity to assist with replication over time and across institutions are presented. While there has been an influx of literature over the last decade on ways to implement IPE learning activities in academic environments [6–9], an extracurricular activity involving the utilization of a nonfiction book across various professions presents a new IPE learning opportunity. While book discussions have been used many times in health professions education to encourage professionalism [10], cultural competence [11–17], leadership [14–15, 18–21], empathy and compassion [13, 22–26], ethics [12, 27–29], spiritual care [30], and other outcomes [13, 19, 31–34], Kilham and Griffiths found no studies in the literature using a standalone book discussion activity for IPE purposes [35]. CASE PRESENTATION Rationale for interprofessional education (IPE) activity Rationale for interprofessional education (IPE) activity IPE book discussions provide an exciting opportunity for librarians to integrate into IPE initiatives and bring nontraditional skills relating to development of pedagogy, instructional design, and information organization and management. The inclusion of librarians as members of the IPE planning team can help solve traditional barriers to jmla.mlanet.org Journal of the Medical Library Association Journal of the Medical Library Association 107 (3) July 2019 107 (3) July 2019 jmla.mlanet.org Interprofessional book club discussion DOI: dx.doi.org/10.5195/jmla.2019.563 Interprofessional book club discussion DOI: dx.doi.org/10.5195/jmla.2019.563 405 Scheduling the activity was recognized as a potential challenge, as each respective school had a variety of material that needed to be covered during their respective orientation programs. To address this, the IPE Steering Committee needed to identify an activity that could be easily implemented and still enable rich discussion. Because these were incoming students, not all participants would have specific clinical skills, so the activity needed to allow participation from a professional lens without requiring in-depth health care knowledge [10, 29]. A book discussion was ideal because it enabled students who were just beginning their health programs to share their reactions to the book from their professional viewpoints as part of an interprofessional group. book prompted students to contemplate his journey from diagnosis to his passing, while reflecting from their own personal and professional perspectives. From an educational perspective, the book addresses team-based care, coordination, and personal, familial, and system perspectives of a major health diagnosis, making it a good fit for health sciences students. Implementation process The IPE Steering Committee met twice a month from May to July to organize the book discussion. The committee also consulted with coordinators from an existing campus first-year reading program to build on their successful strategies for implementation. Next, each health profession faculty member contacted their individual school’s orientation director and distributed the book information and a brief overview of the activity as part of the orientation packet mailed to all incoming students in mid-May. Additionally, an email about the event was distributed to current students and faculty. Faculty from all health professions schools were notified of the upcoming book discussions and invited to participate as facilitators of small groups. This activity also served as an opportunity to highlight the skills and contributions of the health sciences librarians as part of an interprofessional health sciences education team: three librarians who volunteered on the IPE Steering Committee participated in all aspects of planning, conducting, and assessing the activity. The involvement of the librarians throughout this project has led to additional opportunities for them to share their expertise in health sciences information and education and to participate in IPE initiatives on campus. The health sciences librarians were involved with the project in several ways through developing the discussion framework and materials, assessing students’ learning, and coordinating administrative tasks, all of which expanded on the role of librarians in previous studies [19, 31, 34]. Two of the librarians developed the pre- and post-discussion surveys for students and group facilitators in Qualtrics, using their prior experience with online educational assessment. A librarian was also involved in administrative coordination of the activity, including reserving rooms, designing the initial facilitator and student availability surveys, and scheduling facilitators. The team sorted students into interprofessional groups based on their availability, working to maximize the diversity of disciplines represented in each session and broadly enable IPE to take place in each discussion group. Activity planning The IPE Steering Committee planned to implement the book discussion in August, so organization of the event began the previous March. Topics were brainstormed that could be applicable to all professions, and a focus on end-of-life care was identified as a subject matter that needed more coverage in the university’s curricula. The health sciences librarians and faculty generated a list of books on the topic with a summary, cost of the books, and availability. Using a survey developed by health sciences librarians, committee members then ranked their preferences, and the highest ranked book was selected. The IPE Steering Committee chose to hold as many book discussions as possible in the health sciences library for two reasons. Practically, the committee wanted to utilize the centrally located meeting spaces that were available in the health sciences library as a convenient, familiar place for students and faculty facilitators to hold their jmla.mlanet.org DOI: dx.doi.org/10.5195/jmla.2019.563 DOI: dx.doi.org/10.5195/jmla.2019.563 Table 1 Breakdown of student participants by health professions school Health professions school Percentage of participants (%) Total (n=80) Nursing 16% n=13 Public health 33% n=26 Medicine 16% n=13 Social work 8% n=6 Dentistry 21% n=17 Allied health 3% n=2 Pharmacy 4% n=3 Table 1 Breakdown of student participants by health professions school discussions. Philosophically, the committee wanted to convene these discussion groups in a central learning environment that was accessible to every health program on campus and was “program neutral” [34, 38]. For some discussion group sessions, when a library space was not available, the discussions were scheduled in various health schools on campus, rotating which buildings were utilized. The IPE Steering Committee adapted a discussion guide from the School of Pharmacy and the book’s online reader’s guide [39]. This document was sent to the faculty facilitators shortly before the discussion sessions. The guide included the session’s learning objectives, primary discussion questions, and tips for facilitators to generate dialogue among the students. The supplemental appendix includes the discussion questions, which were adapted from Penguin Random House reader’s guide [39] to be relevant for students across different disciplines and to generate discussion around the key themes of the book. After each discussion group, faculty and student participants were sent thank you emails and a document containing IPE resources, thoughts about resilience, and a link to campus wellness resources and advice on providing end-of-life care, in case the discussions proved to be distressing. Table 2 Breakdown of faculty participants by health professions school Table 2 Breakdown of faculty participants by health professions school Health professions school Percentage of participants (%) Total (n=39) Nursing 13% n=5 Public health 33% n=13 Medicine 31% n=12 Social work 8% n=3 Dentistry 5% n=2 Allied health 3% n=1 Pharmacy 8% n=3 Participants A total of eighty students and thirty-nine faculty, representing seven different health affairs schools, participated in the twenty-nine book discussions held in August and September of 2017. This activity took place over a one-hour period and occurred once. The activity covered the entire book, rather than split it into multiple sessions. One to two health affairs faculty or staff facilitated each discussion session with four to nine students. Students indicated their preferences for timing of the sessions, but they were assigned to groups based on their disciplines to create more diverse groups because previous studies indicate that self-selected groups have less rich conversations [35]. Pre- and post-survey results Students were sent an email prior to the activity that included their scheduled book discussion time and room information, along with a scale to assess comfort levels regarding interprofessional roles and communication. Following the activity, students were sent another survey link with the same questions assessing their comfort with interprofessional roles and communication and evaluation questions that focused on their satisfaction with the activity. This evaluation was sent via Qualtrics approximately one week after the book discussion, and results from student post- surveys are presented in Table 3. The student learners ranged from undergraduate nursing (BSN) students to master’s and doctoral students from all health affairs schools. Table 1 lists the percentage of students participating from each of the health professions schools. Table 2 lists the percentage of faculty members participating from each of the health professions schools. Book selection and purpose When Breath Becomes Air explores the end-of-life experience of Kalanithi, a young neurosurgical resident who was unexpectedly diagnosed with terminal cancer [36, 37]. Students find IPE events most meaningful when they can be translated into current and future practice [38], and Kalanithi’s Journal of the Medical Library Association Journal of the Medical Library Association 107 (3) July 2019 jmla.mlanet.org Haley et al. 406 Interprofessional book club discussion 407 DOI: dx.doi.org/10.5195/jmla.2019.563 407 Table 3 Pre- and post-survey results for student interprofessional education (IPE) knowledge and attitudes (n=33) Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree Currently, I am comfortable with the idea of interprofessional team-based care. Pre 45% 46% 80% 1% — Post 73% 21% 3% 3% — I am comfortable describing my professional role to another team member. Pre 25% 54% 16% 4% 1% Post 50% 38% 9% 5% — I am comfortable describing another professionals’ role. Pre 10% 33% 43% 14% 1% Post 15% 55% 24% 6% — Interprofessional collaboration is essential when providing care for patients and their families. Pre 75% 24% 1% — — Post 88% 12% — — — The aim of this book discussion was to provide an opportunity for students from multiple health professions to learn from, with, and about each other. To what extent do you agree or disagree that we achieved this goal? Post 29% 62% — 9% — Table 3 Pre- and post-survey results for student interprofessional education (IPE) knowledge ey results for student interprofessional education (IPE) knowledge and attitudes (n=33) Faculty and students were also asked questions in the post-survey to evaluate the format and implementation of the book club. Qualitative themes taken from open response post-survey questions found that students appreciated the unique perspectives, dialogues, and interprofessional connections that they made through this book discussion. Faculty valued the thoughtful contributions, interprofessional connections, and the quality of the book selected for this event. Faculty and students both felt the book discussion program could benefit from more interprofessional diversity and a higher number of participants. For both faculty and students, the greatest value of the book discussion was the ability to interact with other disciplines and to engage in collaborative discussion. the beginning of the fall semester, committee members approached problems or needs as they arose in the planning process with little time for anticipation. The IPE Steering Committee could have benefited from additional time for organizing the event, planning the logistics, thinking through various issues, and delving more deeply into the desired student learning objectives of the event [40]. The biggest challenge was lack of time, both due to the steering committee being made up of faculty and staff volunteers and the tight deadline. Interprofessional book club discussion 407 DOI: dx.doi.org/10.5195/jmla.2019.563 Creating subcommittees or smaller teams to address each phase of planning, implementation, and wrap-up might facilitate a smoother organization process and save the steering committee time when planning future events. A predetermined group of questions were sent to facilitators as an optional tool to guide discussions. Feedback shows some facilitators used the questions to an extent, whereas others did not use them at all. A missing piece of our project was a way to gather information about the nature of discussions, which would be helpful in informing project design and creating discussion questions and guides in the future. More information about the discussions might have also helped us understand survey responses from participants. It might also have been helpful to assign an IPE Steering Journal of the Medical Library Association Lessons learned and next steps Overall, the IPE Steering Committee found the book discussion pilot program to be successful based on anecdotal feedback and post-survey results, despite some last-minute challenges. It is important to plan enough time to think through the entire event from learning objectives to dissemination [39]. Because this project needed to be implemented in time for 107 (3) July 2019 jmla.mlanet.org Journal of the Medical Library Association Journal of the Medical Library Association jmla.mlanet.org 107 (3) July 2019 jmla.mlanet.org Haley et al. 408 DOI: dx.doi.org/10.5195/jmla.2019.563 DOI: dx.doi.org/10.5195/jmla.2019.563 Committee member the role of note-taker so that these observations could have been more systematically recorded and analyzed. Additionally, the low post-survey response rate might have been due to timing of the activity with the start-up of the semester. A better response rate could also have been achieved by administering the survey at the end of each session rather than via email later on. Clear and effective communication is an essential part of any project, especially one with many people and schedules to coordinate [38]. It is important to determine in advance what information will be needed, how often it will be needed, and how and with whom it will be shared. Facilitator feedback indicated that they want to better understand what the project goals were, why this project was chosen, whether it was mandatory for students, what participating students’ level of education was, and which health affairs schools were participating. Publicity and dissemination after the activity should also be considered during planning [38]. Photographs, a last-minute thought, were taken at a few groups and shared via Twitter and on the IPE Steering Committee website. Tweets with photos provided a way to keep the university’s health affairs community informed about the project and remind them about the IPE Steering Committee’s work. A report summarizing a broad overview of all IPE projects with results, including outcomes from this learning activity, should be shared widely within the university, even if not formally published. This communication would be particularly impactful since the IPE Steering Committee plans to organize more book discussions and will continue to need support from health affairs faculty, staff, and students. Organized by engaged faculty and staff on the IPE Steering Committee, this event was not an “orientation” activity mandated by any of the health professions schools or disciplines. Given that the home institution is the flagship university in the state, the participation numbers were low. However, in future years, efforts to more systematically coordinate this event to coincide with orientation activities and schedules could boost the participation rate. Future groups planning IPE book discussions may also want to consider alternate formats, such as book discussions through a learning management system or other web tools [41, 42]. Interprofessional book discussion groups as an introductory IPE activity have been well received by health affairs faculty and students at the university. The pilot program was informative in showing how much planning is needed for this type of learning activity, which will ensure a smoother process for future IPE book discussions. The health sciences librarians played an important role in the interprofessional planning process. The librarians’ in-depth, successful involvement with other members of the IPE Steering Committee throughout this project and the hosting of discussions in the library strengthened relationships between the health sciences library and the health affairs schools on campus. The IPE Steering Committee now formally includes three librarians as permanent members. The eighty participants and thirty-nine facilitators each received personal schedules and reminders via email, which was difficult for one person to manage alone and confusing for multiple people to manage together. Requests for faculty or student schedule changes were addressed as quickly as possible, and updated meeting schedules were emailed to groups. Because of the many schedule changes and the large variation in number of students from each discipline, some discussion groups, particularly toward the end of the project, were smaller than or not as diverse as originally intended. Utilizing an IPE group email address that several people can manage or having close communication between the event communicators could solve this problem in the future. Scheduling software might also assist with the many logistics involved in organizing the event [38]. REFERENCES 1. World Health Organization. Framework for action on interprofessional education & collaborative practice [Internet]. Geneva, Switzerland: The Organization [2010; cited 28 Jun 2018]. <http://apps.who.int/iris/bitstream/handle/10665/70185 /WHO_HRH_HPN_10.3_eng.pdf;jsessionid=DD4A4664599 946C6C2242382F4DD6512?sequence=1>. 1. World Health Organization. Framework for action on interprofessional education & collaborative practice [Internet]. Geneva, Switzerland: The Organization [2010; cited 28 Jun 2018]. <http://apps.who.int/iris/bitstream/handle/10665/70185 /WHO_HRH_HPN_10.3_eng.pdf;jsessionid=DD4A4664599 946C6C2242382F4DD6512?sequence=1>. Interprofessional book club discussion DOI: dx.doi.org/10.5195/jmla.2019.563 409 DOI: dx.doi.org/10.5195/jmla.2019.563 2. Laksov KB, Boman LE, Liliedahl M, Bjorck E. Identifying keys to success in clinical learning: a study of two interprofessional learning environments. J Interprof Care. 2015 Mar;29(2):156–8. DOI: http://dx.doi.org/10.3109/13561820.2014.942777. 15. Scott HIR. Toward a greater understanding: utilizing book discussions to effectively engage students in the exploration of women and leadership issues. J Leadersh Educ. 2016;15(2):31–7. DOI: http://dx.doi.org/10.12806/V15/I2/A1. 16. Gillespie GL, Pritchard T, Bankston K, Burno J, Glazer G. An evaluation of forums for discussions on inclusion in a college of nursing. Nurs Outlook. 2017 Jan–Feb;65(1):103– 15. DOI: http://dx.doi.org/10.1016/j.outlook.2016.08.002. 3. Lutfiyya M, Brandt B, Cerra F. Reflections from the intersection of health professions education and clinical practice: the state of the science of interprofessional education and collaborative practice. Acad Med. 2016 Jun;91(6):766–71. DOI: http://dx.doi.org/10.1097/ACM.0000000000001139. 17. McCaughey J. Book clubs as a tool for community building and language enhancement. English Teach Forum. 2017;55(1):22–9. 4. El-Awaisi A, Anderson E, Barr H, Wilby KJ, Wilbur K, Bainbridge L. Important steps for introducing interprofessional education into health professional education. J Taibah Univ Med Sci. 2016 Dec;11(6):546–51. DOI: http://dx.doi.org/10.1016/j.jtumed.2016.09.004. 18. Runyan A, Ellington K, Wershof Schwartz A. A compelling practice: empowering future leaders in the medical humanities. J Med Humanit. 2013 Dec;34(4):493–5. DOI: http://dx.doi.org/10.1007/s10912-013-9240-x. 5. Sunguya BF, Woranich H, Masamine J, Yasuoka J. Interprofessional education for whom? — challenges and lessons learned from its implementation in developed countries and their application to developing countries: a systematic review. PLoS One. 2014 May;9(5):e96724. DOI: http://dx.doi.org/10.1371/journal.pone.0096724. 19. Stephenson PL, Clever S, Coady TR, Ender D, Heyd M, Peth S. Book clubs—outreach opportunities for hospital libraries. Med Ref Serv Q. 2014;33(4):448–59. DOI: http://dx.doi.org/10.1080/02763869.2014.957096. 20. Chappell A, Dervay K. Leadership book club: an innovative strategy to incorporate leadership development into pharmacy residency programs. Hosp Pharm. 2016 Sep;51(8):635–8. DOI: http://dx.doi.org/10.1310/hpj5108- 635. 6. Djukic M, Fulmer T, Adams JG, Lee S, Triola MM. NYU3T: teaching, technology, teamwork: a model for interprofessional education scalability and sustainability. Nurs Clin North Am. 2012 Sep;47(3):333–46. DOI: http://dx.doi.org/10.1016/j.cnur.2012.05.003. 21. White JV, Guthrie KL. Creating a meaningful learning environment: reflection in leadership education. J Leadersh Educ. 2016;15(1):60–75. DOI: http://dx.doi.org/10.12806/V15/I1/R5. 7. Copley JA, Allison HD, Hill AE, Moran MC, Tait JA, Day T. Making interprofessional education real: a university clinic model. Aust Health Rev. 2007 Aug; 31(3):351–7. DOI: http://dx.doi.org/10.1071/AH070351. 22. Lohman H, Griffiths Y, Coppard BM. Intergenerational connections through book discussion groups. J Intergener Relatsh. 2005;3(1):27–34. DOI: http://dx.doi.org/10.1300/J194v03n01_03. 8. Journal of the Medical Library Association 107 (3) July 2019 Interprofessional book club discussion DOI: dx.doi.org/10.5195/jmla.2019.563 MacDonnell CP, Rege SV, Misto K, Dollase R, George P. An introductory interprofessional exercise for healthcare students. Am J Pharm Educ. 2012 Oct;76(8):154. DOI: http://dx.doi.org/10.5688/ajpe768154. 23. Crawley JM. “Once upon a time”: a discussion of children’s picture books as a narrative educational tool for nursing students. J Nurs Educ. 2009;48(1):36–9. DOI: http://dx.doi.org/10.3928/01484834-20090101-02. 9. Zomorodi M, Odom T, Askew NC, Leonard CR, Sanders KA, Thompson D. Hotspotting: development of an interprofessional education and service learning program for care management in home care patients. Nurse Educ. 2018 Sep/Oct;43(5):247−50. 24. Plake KS. Book club elective to facilitate student learning of the patient experience with chronic disease. Am J Pharm Educ. 2010 Apr 12;74(3):37. 10. Bumgarner GW, Spies AR, Asbill CS, Prince VT. Using the humanities to strengthen the concept of professionalism among first-professional year pharmacy students. Am J Pharm Educ. 2007 Apr;71(2):28. 25. Larocque N, Schotsman C, Kaasalainen S, Crawshaw D, McAiney C, Brazil E. Using a book chat to improve attitudes and perceptions of long-term care staff about dementia. J Gerontol Nurs. 2014 May;40(5):46–52. DOI: http://dx.doi.org/10.3928/00989134-20140110-02. 11. Clark L, Zuk J, Baramee J. A literary approach to teaching cultural competence. J Transcult Nurs. 2000 Jun;11(3):199– 203. DOI: http://dx.doi.org/10.1177/104365960001100306. 26. Timm DF, Woodson D, Jones D. Book discussion course: timely topics for medical students. Med Ref Serv Q. 2014;33(2):147–56. DOI: http://dx.doi.org/10.1080/02763869.2014.897513. 12. Gupta A, Cory J, Goldberg MJ. Book clubs in residents’ education. Acad Med. 2000 Jan;75(1):2–3. 13. Furniss KK, Motts E. Literature and medicine: improving health care one book at a time. Nurs Womens Health. 2007 Jun–Jul;11(3):247–9. DOI: http://dx.doi.org/10.1111/j.1751- 486X.2007.00164.x. 27. Cohen R. Building a bridge to liberal learning: using a book club model to foster ethical awareness. J Legal Studies Educ. 2006 Winter/Spring;23(1):87–103. DOI: http://dx.doi.org/10.1111/j.0896-5811.2005.00022.x. 14. Greenawald DA, Adams TM. School nurse book clubs: an innovative strategy for lifelong learning. J Sch Nurs. 2008 Apr;24(2):61–5. DOI: http://dx.doi.org/10.1177/10598405080240020201. 28. Walton KL. Using a popular science nonfiction book to introduce biomedical research ethics in a biology majors course. J Microbiol Biol Educ. 2014 Dec;15(2):240–2. DOI: http://dx.doi.org/10.1128/jmbe.v15i2.767. DOI: dx.doi.org/10.5195/jmla.2019.563 29. Poirier TI, Stamper-Carr C, Newman K. A course for developing interprofessional skills in pre-professional honor students using humanities and media. Curr Pharm Teach Learn. 2017 Sep;9(5):874–80. DOI: http://dx.doi.org/10.1016/j.cptl.2017.05.004. 41. Segrist DJ, Meinz EJ. Looking for a good read? running a psychology book club. Psychology Learn Teach. 2018 Jul;17(2):219–28. DOI: http://dx.doi.org/10.1177/1475725718764486. 42. Francis M. Turning a one campus one book event into a web 2.0 experience: the first two years. Coll Undergrad Libr. 2011;18(4):399–408. DOI: http://dx.doi.org/10.1080/10691316.2011.624948. 30. Cone PH, Giske T. Integrating spiritual care into nursing education and practice: strategies utilizing open journey theory. Nurse Educ Today. 2018 Dec;71:22–5. DOI: http://dx.doi.org/10.1016/j.nedt.2018.08.015. 31. Butell SS, O’Donovan P, Taylor JD. Instilling the value of reading literature through student-led book discussion groups. J Nurs Educ. 2004 Jan;43(1):40–4. 32. Zagier AS. Book clubs for doctors show human side of medicine [Internet]. Columbia, MO: San Diego Union- Tribune [11 Mar 2010; cited 9 Jun 2018]. <http://www.sandiegouniontribune.com/sdut-book- clubs-for-doctors-show-human-side-of-medicine- 2010mar11-story.html>. 32. Zagier AS. Book clubs for doctors show human side of medicine [Internet]. Columbia, MO: San Diego Union- Tribune [11 Mar 2010; cited 9 Jun 2018]. <http://www.sandiegouniontribune.com/sdut-book- clubs-for-doctors-show-human-side-of-medicine- 2010mar11-story.html>. Journal of the Medical Library Association jmla.mlanet.org 107 (3) July 2019 107 (3) July 2019 Haley et al. 410 39. Penguin Random House. When Breath Becomes Air reader’s guide [Internet]. Penguin Random House [cited 8 May 2019]. <https://www.penguinrandomhouse.com/books/258507/ when-breath-becomes-air-by-paul- kalanithi/9780812988406/readers-guide/>. 107 (3) July 2019 jmla.mlanet.org Articles in this journal are licensed under a Creative Commons Attribution 4.0 International License. This journal is published by the University Library System of the University of Pittsburgh as part of its D-Scribe Digital Publishing Program and is cosponsored by the University of Pittsburgh Press. ISSN 1558-9439 (Online) 40. Blakeney E, Zierler B. Assessing interprofessional education. In: Edwards ME, ed. Interprofessional education and medical libraries: partnering for success. Lanham, MD: Rowman & Littlefield; 2016. p. 165–81. ISSN 1558-9439 (Online) ISSN 1558-9439 (Online) AUTHORS’ AFFILIATIONS Jen Haley, MSN, RN, CNL, jenhaley@email.unc.edu, https://orcid.org/0000-0001-6003-821X, Graduate Student, School of Nursing, University of North Carolina at Chapel Hill, Chapel Hill, NC 33. Hanson J. From me to we: transforming values and building professional community through narratives. Nurse Educ Pract. 2013 Mar;13(2):142–6. DOI: http://dx.doi.org/10.1016/j.nepr.2012.08.007. Rebecca Carlson McCall, MLS, AHIP, rcarlson@unc.edu, Clinical Librarian, Health Sciences Library, University of North Carolina at Chapel Hill, Chapel Hill, NC Meg Zomorodi, PhD, RN, CNL, meg_zomorodi@unc.edu, Assistant Provost and Director, Office of Interprofessional Education and Practice; Clinical Associate Professor, School of Nursing; and Josiah Macy Faculty Scholar and Well Care Home Health Faculty Scholar; University of North Carolina at Chapel Hill, Chapel Hill, NC 34. Young LM, Clark SB, Machado CK, Hinton EG, Norris, MR. Medical libraries supporting interprofessional education. In: Edwards ME, ed. Interprofessional education and medical libraries: partnering for success. Lanham, MD: Rowman & Littlefield; 2016. p. 117–40. Lisa de Saxe Zerdan, PhD, MSW, lzerden@email.unc.edu, Senior Associate Dean for Master’s in Social Work Education and Clinical Associate Professor, School of Social Work, University of North Carolina at Chapel Hill, Chapel Hill, NC 35. Kilham JP, Griffiths SP. It takes an academic village: the library’s role in supporting interprofessional communication through a book club. Med Ref Serv Q. 2017 Jan–Mar;36(1):42–8. DOI: http://dx.doi.org/10.1080/02763869.2017.1259903. Beth Moreton, MLS, emoreton@email.unc.edu, Clinical Librarian, Health Sciences Library, University of North Carolina at Chapel Hill, Chapel Hill, NC 36. Kalanithi P, Verghese A. When breath becomes air. New York, NY: Random House; 2016. Lee Richardson, richarlm@email.unc.edu, Information Discovery and Metadata Librarian, Health Sciences Library, University of North Carolina at Chapel Hill, Chapel Hill, NC 37. Maslin J. In ‘When Breath Becomes Air,’ Dr. Paul Kalanithi confronts an early death [review] [Internet]. New York, NY: NY Times [2016; cited 28 Jun 2018] <https://www.nytimes.com/2016/01/07/books/review- in-when-breath-becomes-air-dr-paul-kalanithi-confronts- an-early-death.html>. Correction: 14 Jan 2016. Received July 2018; accepted December 2018 38. Shipman JP, Chase-Cantarini S, Wilson RD, Weber AI. Designing an interprofessional education program from planning to implementation. In: Edwards ME, ed. Interprofessional education and medical libraries: partnering for success. Lanham, MD: Rowman & Littlefield; 2016. p. 69–81. 39. Penguin Random House. When Breath Becomes Air reader’s guide [Internet]. Penguin Random House [cited 8 May 2019]. <https://www.penguinrandomhouse.com/books/258507/ when-breath-becomes-air-by-paul- kalanithi/9780812988406/readers-guide/>. Journal of the Medical Library Association jmla.mlanet.org 107 (3) July 2019
https://openalex.org/W4387017243	https://asp-eurasipjournals.springeropen.com/counter/pdf/10.1186/s13634-023-01053-8	English	null	Deep adaptive temporal network (DAT-Net): an effective deep learning model for parameter estimation of radar multipath interference signals	EURASIP Journal on Advances in Signal Processing	2,023	cc-by	8,204	© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the mate- rial. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by/4.0/. Deep adaptive temporal network (DAT‑Net): an effective deep learning model for parameter estimation of radar multipath interference signals Kang Yan1* , Weidong Jin1, Yingkun Huang2, Pucha Song3 and Zhenhua Li4 Correspondence: kangyan@my.swjtu.edu.cn 1 School of Electrical Engineering, Southwest Jiaotong University, No. 999, Xi’an Road, Chengdu 611756, China 2 National Supercomputing Center in Shenzhen, Shenzhen, China 3 School of Electronic Information and Electrical Engineering, Chengdu University, Chengdu, China 4 Department of Electronic Engineering, College of Electronic & Information Engineering, Guangdong Ocean University, Zhanjiang, China Abstract Accurate parameter estimation in radar systems is critically hindered by multipath interference, a challenge that is amplified in complex and dynamic environments. Traditional methods for parameter estimation, which concentrate on single parameters and rely on statistical assumptions, often struggle in such scenarios. To address this, the deep adaptive temporal network (DAT-Net), an innovative deep learning model designed to handle the inherent complexities and non-stationarity of time series data, is proposed. In more detail, DAT-Net integrates both the pruned exact linear time method for effective time series segmentation and the exponential scaling-based importance evaluation algorithm for dynamic learning of importance weights. These methods enable the model to adapt to shifts in data distribution and provide a robust solution for parameter estimation. In addition, DAT-Net demonstrates the capability to comprehend inherent nonlinearities in radar multipath interference signals, thereby facilitating the modeling of intricate patterns within the data. Extensive valida- tion experiments conducted across parameter estimation tasks and demonstrates the robust applicability and efficiency of the proposed DAT-Net model. The architec- ture yield root mean squared error scores as low as 0.0051 for single-parameter estima- tion and 0.0152 for multiple-parameter estimation. 1 School of Electrical Engineering, Southwest Jiaotong University, No. 999, Xi’an Road, Chengdu 611756, China 2 National Supercomputing Center in Shenzhen, Shenzhen, China Keywords: Deep adaptive temporal network, Parameter estimation in radar systems, Multipath interference, Time series segmentation, Exponential scaling-based importance evaluation EURASIP Journal on Advances in Signal Processing EURASIP Journal on Advances in Signal Processing Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 https://doi.org/10.1186/s13634-023-01053-8 Open Access Correspondence: kangyan@my.swjtu.edu.cn 1 School of Electrical Engineering, Southwest Jiaotong University, No. 999, Xi’an Road, Chengdu 611756, China 2 National Supercomputing Center in Shenzhen, Shenzhen, China 3 School of Electronic Information and Electrical Engineering, Chengdu University, Chengdu, China 4 Department of Electronic Engineering, College of Electronic & Information Engineering, Guangdong Ocean University, Zhanjiang, China 1 Introduction Radar signal processing plays a critical role in detecting and locating targets under diverse environmental conditions [1]. Its efficiency is substantially influenced by the system’s ability to accurately process and interpret received signals. An essential aspect of this process involves the estimation of key parameters that define the radar signals. This task becomes notably challenging due to the phenomenon of multipath interference. This interference arises when radar signals reflect off multiple surfaces, Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 2 of 16 such as the ground, sea, or other structures, before reaching the receiver [2]. This event can result in distorted or misleading information regarding the target’s posi- tion, velocity, and other characteristics, thereby complicating the interpretation of the signals. As illustrated in Fig. 1, the direct and reflected paths, denoted by A →D and A →B →D , respectively, give rise to such multipath interference, where h1 and h2 representing the heights of the radar antenna and the receiver antenna, respec- tively. Moreover, the path lengths of AD , AB , and BD are designated by Rd , R1 , and R2 , respectively. The challenge of multipath interference becomes increasingly evident in complex and dynamic environments, highlighting the need for advanced and robust techniques for parameter estimation. Although there exists considerable research on radar multipath interference signals with a focus on aspects such as simulation modeling [3, 4] and interference elimina- tion [5, 6], studies that specifically address the estimation of parameters for these sig- nals are notably limited. The few existing studies have predominantly used traditional methods to estimate a single parameter of the signal [7, 8]. Although these methods provide partial solutions, they struggle to effectively handle complex and dynamic scenarios due to their dependence on statistical assumptions and linear models. Fur- thermore, their ability to simultaneously estimate multiple parameters is limited, highlighting a notable gap in this crucial area of study. To address this challenge, we propose the deep adaptive temporal network (DAT- Net) is proposed. It consists of a deep learning (DL) model designed for the parameter estimation of radar multipath interference signals. Moreover, DAT-Net incorporates advanced techniques relative to time series segmentation and importance evaluation through the integration of a pruned exact linear time (PELT) approach and an expo- nential scaling-based importance evaluation (ESBIE) algorithm. 1 Introduction Therefore, the main innovations and contributions of this paper are summarized as follows: 1. The introduction of the DAT-Net The DAT-Net is specifically crafted to effectively manage non-stationary time series data and adeptly accommodate the distribution shifts that can arise due to the inherent temporal variability of these data. By strategi- 1. The introduction of the DAT-Net The DAT-Net is specifically crafted to effectively manage non-stationary time series data and adeptly accommodate the distribution shifts that can arise due to the inherent temporal variability of these data. By strategi- Radar Antenna h1 R1 Rd h2 Receiver Antenna sea surface R2 A D B Fig. 1 Schematic diagram illustrating multipath interference effects on radar signal propagation over sea surface Receiver Antenna Radar Antenna Fig. 1 Schematic diagram illustrating multipath interference effects on radar signal propagation over sea surface Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 3 of 16 cally tackling these distribution shifts, the model provides a robust solution for chal- lenges in parameter estimation, ensuring high adaptability and precision even when statistical properties of the data change over time. cally tackling these distribution shifts, the model provides a robust solution for chal- lenges in parameter estimation, ensuring high adaptability and precision even when statistical properties of the data change over time. cally tackling these distribution shifts, the model provides a robust solution for chal- lenges in parameter estimation, ensuring high adaptability and precision even when statistical properties of the data change over time. 2. The integration of the PELT method for effective time series segmentation in DAT- Net. This approach allows to identify the periods of significant divergence in data distribution; thereby, it ensures the model’s adaptability to data relative to different periods and domains. 3. The development of the ESBIE algorithm for dynamic learning of importance weights in DAT-Net. This mechanism provides a better responsive adjustment of the importance weights according to the changes in the distribution distances, ensuring a controlled and stable learning process. 4. DAT-Net showcases the ability to understand inherent nonlinearities in radar mul- tipath interference signals, and thus, it enables the modeling of intricate patterns within the use of data. The efficacy of the method is rigorously validated, with model performance appraised in both single- and multi-parameter estimation contexts, affirming its robust applicability. To sum up, the remaining part of the paper is defined as follows: In Sect. 1 Introduction 2, the related work is presented, whereas the methodology is proposed in Sect. 3. As for Sect. 4, the experimental configuration is introduced. Finally, the conclusion and some future works are shown in Sect. 5. in handling multipath interference Despite the scarcity of research regarding the parameter estimation of radar multipath interference signals, several studies have contributed significantly to the understand- ing and implementation of parameter estimation for conventional radar signals. These established methods, fundamental to radar signal processing, serve as an important starting point reference for our work. For instance, Liu et al. [9] used the cumulative Wigner-Hough transform (CWHT) for estimating parameters of the linear frequency modulation continuous wave (LFMCW) signal, taking advantage of the signal’s periodicity. However, this method was designed for individual LFMCW signals, not for multiple ones. Moreover, Geroleo et al. [10] introduced the periodic Wigner-Ville Hough transform (PWVHT) to detect the LFMCW radar signal and estimate its parameter. This method accommodated multiple pulses within an observation interval at the intercept receiver, extending the accumu- lation of signal energy, and thereby enhancing detection and parameter estimation. In addition, Wen et al. [11] leveraged the focusing capability of the fractional Fourier trans- form (FRFT) to estimate the pulse width and frequency modulation rate of Linear Fre- quency Modulation (LFM) signals. This method yields accurate results even under low signal-to-noise ratio (SNR) conditions. Furthermore, Tang et al. [12] presented a method to estimate the direction of arrival (DOA) using a modified spatial time–frequency dis- tribution (STFD) matrix. This new approach surpassed the limitations of traditional nar- rowband methods when dealing with wideband signals. Finally, Deng et al. [13] proposed Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 4 of 16 a technique for carrier frequency and code period estimation of polyphase-coded radar signals. Their approach, utilizing the Fourier transform and a modified Choi-Williams distribution, was designed to facilitate parameter estimation in environments with low SNRs. As reported, it achieved high estimation accuracy under these challenging conditions. Based on the above discussion, it becomes evident that the existing methods, although their good performance for conventional radar signal parameter estimation, reveal limi- tations when it comes to more complex and dynamic scenarios. These methods often rely on statistical assumptions and linear models, primarily tailored for specific signal types. Therefore, the requirement to build a distinct mathematical model for each sig- nal type presents a significant limitation, particularly when confronting diverse and dynamically varying signal conditions. For instance, Djemal et al. [14] proposed an adap- tive threshold detection approach based on CFAR techniques for radar systems in both homogeneous and non-homogeneous environments. in handling multipath interference As we shift our attention to signals under multipath interference conditions, the scenario becomes even more challenging. These distorted signals display distinct characteristics and behavior compared to con- ventional signals, thus requiring the development of innovative parameter estimation methods. The limitations of existing methods underscore the needs for universal models, such as those crafted using DL techniques. Their benefits include the capability for nonlin- ear modeling, flexibility to handle a wide variety of signals, adaptability to changing sce- narios, and ability to learn and adjust their models based on recorded data. In fact, DL methods have showcased their effectiveness in several signal processing areas and are well-positioned to address the challenges incorporated within multipath interference. The exploration of DL techniques, particularly their potential in achieving robust radar signal parameter estimation under multipath interference conditions, offers an exciting pathway for future research. 2.2 Deep learning for radar signal processinghi The progress in radar signal classification has been significantly influenced by the adop- tion of DL architectures, notably convolutional neural networks (CNNs), owing to their success in image classification tasks. For instance, Sun et al. [15] proposed a technique that leverages a unidimensional convolutional neural network (U-CNN) for radar emit- ter classification, demonstrating competitive accuracy levels. Similarly, Liu et al. [16] utilized a triplet convolutional neural network (T-CNN) to enhance the identification of different modulations of low probability of intercept (LPI) radar signals, particularly effective in harsh electromagnetic environments with low SNRs. To further enhance the performance in radar signal classification, fusion strategies, based on CNN architec- tures, have been explored. An approach, proposed by Akyon et.al [17], employed two independent CNNs to separately process frequency- and phase-related aspects of radar signals. Moreover, efforts have also been made to accelerate CNN’s feature learning. A princi- pal component analysis (PCA)-based CNN architecture, proposed by Ye et al. [18], was developed to reduce the dimensionality of time–frequency distribution (TFD) images. Additionally, techniques, such as convolutional denoising auto-encoder (CDAE) and Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 5 of 16 inception-based deep CNN, have been employed to facilitate the signal recognition and noise reduction in TFD images [19]. Furthermore, deep time series networks have been employed in radar signal pro- cessing. For instance, Zhang et al. [20] utilized recurrent neural networks (RNNs) for theclassify, denoise, and deinterleave pulse streams, aiming to exploit long-term tem- poral patterns to enhance processing outcomes. Furthermore, Apfeld et al. [21] used long short-term memory (LSTM) RNNs to identify multifunction radar emitters. Their approach utilized the frequency and agility of emissions to encode radar pulses into symbolic representations. This enabled the discrimination between similar emitter types by analyzing emission parameters and employing resource management techniques. In another study, Notaro et al. [22] applied RNNs to classify radar emitters, capitalizing on temporal dependencies within pulse streams. For this purpose, they introduced two techniques: per-sequence normalization to improve the temporal pattern extraction, and attribute-specific RNN processing to enhance information efficiency. Their approach outperformed previous methods in terms of accuracy and robustness, especially in noisy environments. These DL-based techniques have shown promising results in radar signal recognition. However, the application of DL methods for radar signal parameter estimation remains limited. 2.2 Deep learning for radar signal processinghi Given that radar signals are typical non-stationary time series data, leveraging deep time series networks for radar signal parameter estimation present an intuitive research direction. 3.1 Problem definition In the context of multi-step time series prediction, we consider the scenario known as temporal covariate shift (TCS) [23]. This concept refers to situations in which the distri- bution of input data, or features, varies across different time intervals, whereas the con- ditional distribution of labels, based on these features, remains consistent. Consider a time series dataset consisting of n segments with corresponding labels. This dataset is denoted as Dtrn = xi, yi n i=1 and is applied for training. It encompasses an unknown quantity K of underlying time periods. Within a constant period i , seg- ments adhere to a specific data distribution, PDi(x, y) . However, when evaluating two distinct periods, i and j (where 1 ≤i = j ≤K ), the input data distributions diverge, i.e., PDi(x) = PDj(x) . Despite this shift, the conditional distributions of labels given the inputs remain consistent, represented as PDi(y\|x) = PDj(y\|x) , and constituting a TCS. Figure 2 provides a visual depiction of this scenario. It displays four distinct periods of the time series data. The first three periods are part of the data training process, each with its unique data distribution, whereas the fourth period represents the test data.h The objectives of this work are twofold: (1) automatically identify the K periods within the training time series data, visually represented by the first three periods in Fig. 2, and (2) construct a predictive model M capable of harnessing the shared characteristics across these periods. This model should provide accurate forecasts for the forthcom- ing r segments, represented as the fourth time interval in Fig. 2 and formally denoted as Dtst = {xj}n+r j=n+1 . Assuming that the test segments belong to a uniform time period, these segments exhibit a distinct input data distribution, separate from that of any the Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 6 of 16 Fig. 2 Multi-period time series data distribution under temporal covariate shift Fig. 2 Multi-period time series data distribution under temporal covariate shift training periods, i.e., PDtst(x) = PDi(x) . However, even with the variation in input data distribution, the conditional distribution of labels given the inputs remains consistent across all periods. It is, expressed as PDtst(y\|x ) = PDi(y\|x ) for any 1 ≤i ≤K. 3.2 Time series segmentation using the PELT approach In order to reach the goal of maximizing shared knowledge extraction from a time series in the presence of temporal covariate shift, this study presents an innovative and effi- cient method to identify periods that exhibit the highest degree of divergence from each other. These periods represent the extreme cases of temporal covariate shift, where the cross-period distributions, denoted as Di and Dj , are the most diverse. As a result, they hold considerable importance for training a predictive model that aims to be robust against distribution shifts, as illustrated in Fig. 2. An easy approach to time series segmentation involves an even division of the data into a fixed number of parts, N , where each part is considered as a minimal-unit period that cannot be further subdivided. Given a set of predefined K values, the optimal K is determined according to a method that maximizes the distribution distance [23]. Never- theless, this approach has significant limitations. Although it is simple, it does not allow for time series data fine-grained segmentation. In addition, it may not always result in the most optimal solution, especially when dealing with large-scale data sets. In response to these limitations, this paper recommends the use of the PELT method. This technique, a state-of-the-art changepoint detection algorithm, is recognized for its computational efficiency, precision, and ability to efficiently segment time series data. It operates based on the primary principle of identifying significant shifts in data dis- tribution, represented as "changepoints.” Moreover, it employs a dynamic programming to achieve computational efficiency while preserving the accuracy of the changepoint detection. The computational efficiency is achieved a pruning rule in the dynamic pro- gramming process, which discards unnecessary computations, hence the name pruned exact linear time. More formally, given a time series data with n time points, the PELT approach mini- mizes the cost function as described in Eq. (1): C(τ) = m i=0 [F(τi+1 −1) −F(τi)] + f (m) C(τ) = m i=0 [F(τi+1 −1) −F(τi)] + f (m) (1) Page 7 of 16 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 In this function, τ1, τ2, ..., τm denote the locations of changepoints, with τ0 = 1 and τm+1 = n + 1 . 3.2 Time series segmentation using the PELT approach Moreover, F(i) symbolizes the cumulative sum of model-fitting costs up to time point i , and f (m) represents a penalty term proportional to the number of change- points m. By adopting the PELT technique, an efficient computation is ensured and the need for pre-specifying a range for K is no more required. This approach can detect periods of vary- ing lengths, yielding to enhance its adaptability to data. The output of PELT, in terms of the identified changepoints, enables segmentation of the time series into K periods. Fur- thermore, the distributions Di and Dj of each period can subsequently be investigated for further analysis. 3.3 Temporal distribution matching enhanced with exponential scaling‑based importance evaluation Once the distinct time periods have been established, a temporal distribution matching (TDM) module is constructed. Its main objective is to extract the shared knowledge across these periods by aligning their distributions. Through this process, the model, symbolized as M , would outperform on unseen test data compared to models that rely solely on local or statistical data. The prediction loss of TDM, denoted as Lpred , in Eq. as follows: The prediction loss of TDM, denoted as Lpred , in Eq. as follows: (2) Lpred(θ) = 1 K K j=1 1 Dj \|Dj\| i=1 ℓ(yj i, M(xj i; θ)) (2) where (xj i, yj i) represents the ith labeled segment from period Dj , ℓ(·, ·) denotes the mean squared error (MSE) loss function, and θ signifies the learnable model parameters. In the proposed model model, M adopts the LSTM technique, a form of deep time series model. The LSTM, through its specialized architecture, effectively processes time series data by capturing long-range temporal dependencies. Its strength lies in learning from and retaining information over extended periods, which renders it particularly suited for our scenario involving time-dependent features across different periods. f However, minimizing solely the prediction loss, as given in Eq. (2), would fosters the pre- dictive knowledge inherent in each individual period and results in failing to reduce the distributional divergence across the different periods. Moreover, this divergence could potentially carry common knowledge valuable to the training model. To address this issue, TDM introduces the concept of an importance evaluation, denoted α, with dimensions α ∈RV . This vector is responsible for assessing the relative significance of the V hidden states within the LSTM, each state being weighted by a normalized α. This approach dynamically reduces the distribution divergence across periods. For a pair of periods, Di, Dj , the loss associated with the temporal distribution match- ing, as expressed in Eq. (3), is expressed as follows: (3) Ltdm(Di, Dj; θ) = V t=1 αt i,jd(ht i, ht j ; θ) (3) where αt i,j represents the distributional importance between periods Di and Dj at state t. where αt i,j represents the distributional importance between periods Di and Dj at state t. Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 8 of 16 Furthermore, calculating all the hidden states in an LSTM model is a straightforward process. 3.3 Temporal distribution matching enhanced with exponential scaling‑based importance evaluation Let δ(·) represent the computation of a subsequent hidden state based on a pre- vious state. The state computation can be expressed as shown in Eq. (4): (4) ht i = δ(xt i , ht−1 i ) ht i = δ(xt i , ht−1 i ) (4) Finally, by merging Eqs. (2) and (3), the comprehensive objective of temporal distribu- tion matching can be defined as in Eq. (5): (5) L(θ, α) = Lpred(θ) + 2 K(K −1) i=j i,j Ltdm(Di, Dj; θ, α) (5) where λ is a trade-off hyper-parameter. The second term calculates the average distribu- tion distances of all pairwise periods. For computational efficiency, we take a mini-batch of Di and Dj to conduct the forward operation in LSTM layers and subsequently concat- enate all hidden features. The resulting TDM can then be performed using Eq. (5). h Therefore, we propose the ESBIE algorithm to learn and identify the importance weights, αt,(n) i,j . The schematic representation of the ESBIE algorithm is displayed in Fig. 3. The initial step involves pre-training the network parameter θ on a fully labeled data from all periods, which leads to get superior hidden state representations that help in the learning process of αt,(n) i,j . We assign the symbol θ0 to this pre-trained parameter. Once θ0 is defined, the ESBIE algorithm is applied to discern the importance of the hid- den states. All weights within each LSTM layer are uniformly initialized and denoted αt,(0) i,j = {1/V }V . To guide the weight updates, we utilize the cross-domain distribution distance. This distance is calculated using the Maximum Mean Discrepancy (MMD) metric [24], recognized for its ability to quantify the disparity between two probability Fig. 3 ESBIE algorithm in temporal distribution matching Fig. 3 ESBIE algorithm in temporal distribution matching Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 9 of 16 distributions, making it an effective choice for high-dimensional spaces. As for the ESBIE algorithm, it follows a nonlinear approach in adjusting weights according to the alterations in distribution distance. When the distribution distance increases, indicating larger divergence, the ESBIE algorithm adjusts the importance weights upwards to work toward reducing this divergence. Conversely, when the distribution distance decreases, indicating a reduced divergence, the importance weights are adjusted downward-. where the scaling function G is defined as follows: (7) G(dt,(n) i,j , dt,(n−1) i,j ) = eηn×(di,jt,(n)−dt,(n−1) i,j ) G(dt,(n) i,j , dt,(n−1) i,j ) = eηn×(di,jt,(n)−dt,(n−1) i,j ) (7) here, ηn = η0 × e−p×n represents the learning rate at the n th epoch, which is exponen- tially decayed from the initial value η0 . Moreover, the decay rate is controlled by the decay constant p , a hyperparameter that is empirically set. In Eq. (6), dt,(n) i,j = D(ht i, ht j ; αt,(n) i,j ) represents the distribution distance at time step t in epoch n . Finally, the weights are normalized after each update using this formula: αt,(n+1) i,j = αt,(n+1) i,j V j=1 αt,(n+1) i,j . Therefore, by applying Eqs. (3) and (6), the importance evaluation can be learned = V j=1 αt,(n+1) i,j . can be learned. Furthermore, the ESBIE mechanism provides a more responsive adjustment of the importance weights to the changes in distribution distances. By integrating a learning rate with exponential decay, this approach allows a more controlled and stable learn- ing process. The learning rate’s decay throughout epochs ensures a gradual fine-tuning phase in which the weights are adjusted more subtly, enhancing the chances of converg- ing to an optimal solution. Leveraging the effectiveness of temporal distribution match- ing across domains, the model’s overall performance is improved when managing data from diverse periods and domains, resulting in a more robust generalization on unseen test data. 3.3 Temporal distribution matching enhanced with exponential scaling‑based importance evaluation The adopted update rules are as follows: (6) αt,(n+1) i,j = αt,(n) i,j × G(dt,(n) i,j , dt,(n−1) i,j ) αt,(n+1) i,j = αt,(n) i,j × G(dt,(n) i,j , dt,(n−1) i,j ) (6) where the scaling function G is defined as follows: 4.1 Dataset: details and specifications This latter, which simulates a flight at an elevation of 8,000 m, follows a circu- lar trajectory while the emitter maintains a linear path from the origin, at an elevation of 3,000 m, directing toward the circular path’s center. Both the transmission and reception elements are simulated as dipole antennas having vertical polarization. The power of the transmitted signal is set at 100 W, with a transmission gain of 20 dB, and a receiver gain of 13 dB. The simulation generates four signal types: LFM, nonlinear frequency modulation (NLFM), binary phase shift keying (BPSK), and quadrature phase shift keying (QPSK). To meet the requirements of the model training, 9000 instances of each signal type are gener- ated, which were divided into sets of 7000 instances for training, 1000 for validation, and 1000 for testing. the receiver. This latter, which simulates a flight at an elevation of 8,000 m, follows a circu- lar trajectory while the emitter maintains a linear path from the origin, at an elevation of 3,000 m, directing toward the circular path’s center. Both the transmission and reception elements are simulated as dipole antennas having vertical polarization. The power of the transmitted signal is set at 100 W, with a transmission gain of 20 dB, and a receiver gain of 13 dB. The simulation generates four signal types: LFM, nonlinear frequency modulation (NLFM), binary phase shift keying (BPSK), and quadrature phase shift keying (QPSK). To meet the requirements of the model training, 9000 instances of each signal type are gener- ated, which were divided into sets of 7000 instances for training, 1000 for validation, and 1000 for testing. 4.1 Dataset: details and specifications For our experiments, we utilized a dataset derived from the “Radar Signal Simulation Platform under Complex Electromagnetic Environment,” located at Southwest China Research Institute of Electronic Equipment. The platform incorporates a parabolic equa- tion method to generate synthetic radar signal data. This method serves as the basis of radar signal processing, aligning and filtering received echo signals with a reference sig- nal, essentially a duplicate of the transmitted signal. This procedure yields a complex sig- nal exhibiting a parabolic trajectory within the time domain. Simulated echo signals are designed to encompass noise and the Doppler effect to mimic real-world radar signals. f A representation of the simulated scenario is displayed in Fig. 4. The setup takes into con- sideration the multipath effects induced by specular reflections from the sea surface. These effects result in distorting the received signal as it traverses different routes before reaching Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 10 of 16 the receiver. This latter, which simulates a flight at an elevation of 8,000 m, follows a circu- lar trajectory while the emitter maintains a linear path from the origin, at an elevation of 3,000 m, directing toward the circular path’s center. Both the transmission and reception elements are simulated as dipole antennas having vertical polarization. The power of the transmitted signal is set at 100 W, with a transmission gain of 20 dB, and a receiver gain of 13 dB. The simulation generates four signal types: LFM, nonlinear frequency modulation (NLFM), binary phase shift keying (BPSK), and quadrature phase shift keying (QPSK). To meet the requirements of the model training, 9000 instances of each signal type are gener- ated, which were divided into sets of 7000 instances for training, 1000 for validation, and 1000 for testing. y x Emitter Start [0, 0, 3000] Receiver Start 20km 50km Receiver Center [300000, 0, 8000] 150m/s 100m/s sea surface Fig. 4 The schematic diagram of the simulation scenario y x Emitter Start [0, 0, 3000] Receiver Start 20km 50km Receiver Center [300000, 0, 8000] 150m/s 100m/s sea surface Fig. 4 The schematic diagram of the simulation scenario Receiver Center Receiver Start 100m/s Fig. 4 The schematic diagram of the simulation scenario Fig. 4 The schematic diagram of the simulation scenario the receiver. 4.3 Application and performance comparison: parameter estimation in radar multipath interference signals In this study, our model is applied to time-domain data from three distinct sets of simu- lated radar multipath interference signals, where each set features distinctive patterns of parameter variation. • In Set 1, all signals—LFM, NLFM, BPSK, and QPSK—have variable bandwidths (BW) while the other parameters remain constant. This requires an estimation of the BW for these signals. • In Set 2, the pulse width (PW) for LFM and NLFM and the number of sub-pulses (NSP) for BPSK and QPSK are varied while the other parameters are kept constant, thus requiring an estimation of PW for LFM and NLFM, and NSP for BPSK and QPSK. • In Set 3 a more complex scenario is proposed. It consists of having two parameters that vary simultaneously. For LFM and NLFM, we aim to estimate both PW and BW, whereas for BPSK and QPSK, both BW and NSP must be estimated. The performance of the model in estimating parameters is evaluated using two sta- tistical measurement approaches: the root mean squared error (RMSE) and the mean absolute error (MAE). The former technique employs a quadratic scoring rule that com- putes the average magnitude of the error by effectively squaring the difference between predicted and observed values before calculating the mean. This technique puts more weight on large errors. Conversely, MAE consists of computing the average absolute dif- ference between predicted and observed values. Moreover, it provides a linear score that can be easily interpretable as it directly averages the absolute error magnitudes. The proposed evaluations involve comparing the DAT-Net model against XGBoost [25], LSTM, LSTNet [26], and AdaRNN [23]. However, in the case of Set 3, the XGBoost model does not provide RMSE and MAE values as it is unable to simultaneously gener- ate two parameter estimates. As a result, these values are represented as "–" in Tables 1 and 2. 4.2 Model parameter settings The model was implemented using the PyTorch deep learning framework, and the exper- iments were executed on a machine equipped with an RTX 3080TI graphics card, using CUDA version 11.7 and the Windows 10 operating system. The PELT method, available in the Python ruptures library, was used for the changepoint detection, It utilized the radial basis function (RBF) model as the cost model. The penalty term was set to 10. In the LSTM model, a two-layer network structure was utilized with a hidden state dimension of 32. The Adam optimizer was used with a learning rate of 0.002. Finally, in the ESBIE algorithm, the initial learning rate η0 was set at 0.2 and the decay con- stant p was equal to the unit. Page 11 of 16 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 4.3 Application and performance comparison: parameter estimation in radar multipath interference signals In more detail, Table 1 (for RMSE) and Table 2 (for MAE) present the results of Table 1 Comparison of RMSE for parameter estimation in radar multipath interference signals across various models Set Parameter(s) Signal Type XGBoost LSTM LSTNet AdaRNN DAT-Net 1 BW LFM 0.2356 0.0346 0.3247 0.0151 0.0051 1 BW NLFM 0.2389 0.0352 0.0524 0.0248 0.0067 1 BW BPSK 0.2935 0.2441 0.2297 0.0459 0.0476 1 BW QPSK 0.2701 0.2356 0.2207 0.0840 0.0475 2 PW LFM 0.1539 0.0680 0.0871 0.0164 0.0087 2 PW NLFM 0.1506 0.0549 0.0788 0.0186 0.0062 2 NSP BPSK 0.2650 0.2371 0.2307 0.1344 0.0342 2 NSP QPSK 0.2564 0.2305 0.2022 0.0432 0.0242 3 BW, PW LFM – 0.1782 0.2362 0.0319 0.0193 3 BW, PW NLFM – 0.2040 0.2384 0.0272 0.0152 3 BW, NSP BPSK – 0.3235 0.3219 0.0944 0.0709 3 BW, NSP QPSK – 0.3142 0.3001 0.0911 0.0706 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 12 of 16 Table 2 Comparison of MAE for parameter estimation in radar multipath interference signals across various models Set Parameter(s) Signal Type XGBoost LSTM LSTNet AdaRNN DAT-Net 1 BW LFM 0.1977 0.0275 0.2727 0.0121 0.0039 1 BW NLFM 0.2043 0.0273 0.0414 0.0196 0.0054 1 BW BPSK 0.2399 0.1910 0.1759 0.0171 0.0188 1 BW QPSK 0.2164 0.1833 0.1756 0.0486 0.0202 2 PW LFM 0.1539 0.0526 0.0685 0.0138 0.0067 2 PW NLFM 0.1215 0.0428 0.0623 0.0149 0.0043 2 NSP BPSK 0.2198 0.1776 0.1826 0.0914 0.0197 2 NSP QPSK 0.2121 0.1768 0.1571 0.0323 0.0163 3 BW, PW LFM – 0.1407 0.1945 0.0267 0.0150 3 BW, PW NLFM – 0.1602 0.1926 0.0231 0.0118 3 BW, NSP BPSK – 0.2667 0.2611 0.0623 0.0429 3 BW, NSP QPSK – 0.2548 0.2429 0.0543 0.0369 Table 2 Comparison of MAE for parameter estimation in radar multipath interference signals across various models these comparisons when performing different simulation scenarios for radar multipath interference signal parameter estimation. When considering the RMSE and MAE metrics, the DAT-Net consistently emerges as the preferred model as it achieves lower error values. This underscores its capability in accurately handling complex temporal dependencies in radar multipath interference signals. Concerning the single-parameter estimation scenarios (simulations regarding Sets 1 and 2), the DAT-Net method continues to demonstrate superior performance. It reflects remarkable accuracy in estimating BW across all signal types in Set 1. 4.3 Application and performance comparison: parameter estimation in radar multipath interference signals However, in the case of BPSK, although DAT-Net performs well, the AdaRNN technique manages to achieve slightly lower RMSE and MAE values. Similarly, in Set 2, DAT-Net effectively estimates the PW for LFM and NLFM signals as well as the NSP for BPSK and QPSK signals, achieving lower errors than its counterparts. Going to scenarios that involve the simultaneous estimation of two parameters (mainly the simulation of Set 3), DAT-Net maintains its exemplary performance, as it consistently achieves lower RMSE and MAE values across all signal types, knowing that XGBoost fails to provide any estimates in this set due to its limitations. As the estimation task shifts from single-parameter scenarios (Sets 1 and 2) to multi- parameter scenarios (Set 3), there is a notable escalation in error metrics is observed across all models. This increase reflects the complexity inherent with multi-parameter estimation. Nonetheless, this increase is notably less severe for DAT-Net compared to the other models. This result shows, once again, robustness ability of DAT-Net to tackle the intricacies of multi-parameter estimation with a remarkable level of error control. Despite the increased complexity, DAT-Net manages to keep error growth in check, highlighting its resilience and adaptability across several estimation scenarios. Upon a detailed examination of the error values, the DAT-Net’s performance with LFM and NLFM signals consistently results in lower RMSE and MAE compared to the results obtained with BPSK and QPSK signals across all sets. This finding could poten- tially reflect the differences in signal structures as well as the DAT-Net’s specific profi- ciency in handling LFM and NLFM signals. Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 13 of 16 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Furthermore, Fig. 5 illustrates the comparative proficiency of the DAT-Net and other models in approximating real values, with a specific emphasis on the BW esti- mation of QPSK signals from Set 1. This visual evaluation is inline with the con- clusions drawn from the error metric analyses, showcasing clearly that DAT-Net provides a better approximation of the actual values compared to the other models. Furthermore, Fig. 5 illustrates the comparative proficiency of the DAT-Net and other models in approximating real values, with a specific emphasis on the BW esti- mation of QPSK signals from Set 1. 4.4 Convergence speed analysis As an advancement over AdaRNN, DAT-Net not only enhances estimation accuracy but also improves the efficiency in model training. Evaluating of model performance goes beyond the estimation precision to encompass the speed at which these accurate estima- tions are achieved. By integrating the ESBIE algorithm, DAT-Net gains a significant advantage in this respect. ESBIE enables a swift decrease in loss during the initial training stages, thereby accelerating the convergence speed. Furthermore, in the later stages of model training, ESBIE contributes to smaller fluctuations in the loss function, promoting stability in the learning process and helping in preventing overfitting. i The improvements in the convergence speed are vividly depicted in Fig. 6a where the graph demonstrates clearly that DAT-Net’s reduces faster the loss compared to AdaRNN during the early phases of training, along with having lower variation in loss during the later learning stages. This enhanced convergence speed and stability, especially notable in the context of large datasets and time-critical tasks, further validates the DAT-Net method position as an improved model over AdaRNN. Finally, as shown in Fig. 6b, the proposed DAT-Net architecture demonstrates compu- tational efficiency improvements over AdaRNN. However, this improvement is offset by longer training times compared to traditional LSTM, as well as XGBoost and LSTNet. The additional time demands of DAT-Net’s series segmentation and TDM components are justified by its superior performance over existing methods across assessed bench- marks. Once the model training is completed, DAT-Net does not significantly elevate computational consumption in the inference phase compared to LSTM. 4.3 Application and performance comparison: parameter estimation in radar multipath interference signals This visual evaluation is inline with the con- clusions drawn from the error metric analyses, showcasing clearly that DAT-Net provides a better approximation of the actual values compared to the other models. The close alignment between the fluctuations in the actual data and the outputs pro- duced by DAT-Net reflects its effective learning and adaptation to model the radar multipath interference signals. Conversely, the plots for the other models, such as XGBoost LSTM LSTNet and AdaRNN reflect greater deviations from the actual The close alignment between the fluctuations in the actual data and the outputs pro- duced by DAT-Net reflects its effective learning and adaptation to model the radar multipath interference signals. Conversely, the plots for the other models, such as XGBoost, LSTM, LSTNet, and AdaRNN, reflect greater deviations from the actual values. These variations highlight the efficiency of DAT-Net in modeling complex temporal dependencies in radar multipath interference signals. Fig. 5 Comparative visualizations of model fits to real values Fig. 5 Comparative visualizations of model fits to real values f d l fi l l Fig. 5 Comparative visualizations of model fits to real values Page 14 of 16 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Funding This work was supported by the Science and Technology on Electronic Information Control Laboratory (No. 6142105190312) and the National Natural Science Foundation of China (No. 62201097). The author promises that the data in this article are available. Author contributions Author contributions KY proposed the model, conducted the experiments, and was responsible for drafting the manuscript. WJ interpreted the results and revised the first draft of the manuscript. YH was instrumental in posing the problem. PS worked on modi- fying grammar issues in the manuscript. ZL was in charge of proofreading the full text of the manuscript. All authors have read and approved the final manuscript. Author contributions KY proposed the model, conducted the experiments, and was responsible for drafting the manuscript. WJ interpreted the results and revised the first draft of the manuscript. YH was instrumental in posing the problem. PS worked on modi- fying grammar issues in the manuscript. ZL was in charge of proofreading the full text of the manuscript. All authors have read and approved the final manuscript. Declarations Competing interests The authors declare that they have no competing interests. Received: 7 July 2023 Accepted: 5 September 2023 5 Conclusionh This paper introduced and thoroughly examined the performance of DAT-Net, a deep learning model specifically designed to manage complex and non-stationarity time series data, while considering radar multipath interference signals. Notably, DAT- Net’s ability to manage various signal types contributes toward addressing the com- plexities frequently encountered in the current research problems. Moreover, the DAT-Net incorporates advanced techniques, such as time series segmentation via the PELT method, and importance evaluation using the ESBIE algorithm. These methods enhance the model’s capability to adapt to shifts in data distribution and offer a robust Fig. 6 Convergence speed and training time Fig. 6 Convergence speed and training time Fig. 6 Convergence speed and training time Fig. 6 Convergence speed and training time Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Page 15 of 16 solution for parameter estimation. The comprehensive comparative analysis achieved in this work underscored the advantages of DAT-Net, demonstrating its superior per- formance in terms of precision over other commonly used models through various scenarios. Specifically, the DAT-Net’s ability to understand and model inherent non- linearities and non-stationarities in radar multipath interference signals distinguishes it from other approaches. Future work could aim to further enhance DAT-Net and broaden its applicability to a wider array of scenarios and datasets. Such improvements might include integrating more advanced machine learning techniques or applying the model to other time- dependent signal processing tasks. Abbreviations DAT-Net Deep Adaptive Temporal Network PELT Pruned exact linear time ESBIE Exponential scaling-based importance evaluation CWHT Cumulative Wigner-Hough transform LFMCW Linear frequency modulation continuous wave PWVHT Periodic Wigner-Ville Hough transform FRFT Fractional Fourier transform LFM Linear frequency modulation DOA Direction-of-arrival STFD Spatial time–frequency distribution CNNs Convolutional neural networks U-CNN Unidimensional convolutional neural network LPI Low probability of intercept SNR Signal-to-noise ratio TFD Time–frequency distribution CDAE Convolutional denoising autoencoder RNNs Recurrent neural networks LSTM Long short-term memory TCS Temporal covariate shift TDM Temporal distribution matching MMD Maximum mean discrepancy NLFM Nonlinear frequency modulation BPSK Binary phase shift keying QPSK Quadrature phase shift keying RBF Radial basis function BW Bandwidths PW Pulse width NSP Number of sub-pulses RMSE Root mean squared error MAE Mean absolute error References Schölkopf, A. Smola, A kernel two-sample test. J. Mach. Learn. Res. 13, 723–773 (2012) 25. T. Chen, C. Guestrin, Xgboost: A scalable tree boosting system, in Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. pp. 785–794 (2016) g y g pp 26. G. Lai, W.C. Chang, Y. Yang, H. Liu, Modeling long-and short-term temporal patterns with deep neural networks, in The 41st international ACM SIGIR conference on research & development in information retrieval. pp. 95–104 (2018) References 1. M.I. Skolnik, Radar handbook. McGraw-Hill Education (2008) 1. M.I. Skolnik, Radar handbook. McGraw-Hill Education (2008) 2. S.L. Wilson, B.D. Carlson, Radar detection in multipath. IEE Proceedings-Radar Sonar Navig. 146, 45–54 (1999) 2. S.L. Wilson, B.D. Carlson, Radar detection in multipath. IEE Proc 2. S.L. Wilson, B.D. Carlson, Radar detection in multipath. IEE Proceedings Radar Sonar Navig. 146, 45 54 (1999) 3. K.H. Craig, M.F. Levy, Parabolic equation modelling of the effects of multipath and ducting on radar systems, in IEE 3. K.H. Craig, M.F. Levy, Parabolic equation modelling of the effec 3. K.H. Craig, M.F. Levy, Parabolic equation modelling of the effects of m Proceedings F (Radar and Signal Processing). (1991) pp. 153–162. IET Proceedings F (Radar and Signal Processing). (1991) pp. 153–162 g g g pp 4. X. Yuan, J. Zhang, L. Yang, Research into modeling and simulation technologies of multi-path effect of sea surface 4. X. Yuan, J. Zhang, L. Yang, Research into modeling and simulation technologies of multi-path effect of sea s radar signal environment. Shipboard Electron. Countermeas. 36, 68–72 (2013) 4. X. Yuan, J. Zhang, L. Yang, Research into modeling and simulation technologies of multi-path effect of sea surfac radar signal environment. Shipboard Electron. Countermeas. 36, 68–72 (2013) g p 5. B.P. Day, A. Evers, D.E. Hack, Multipath suppression for continuous wave radar via slepian sequences. IEEE Trans. Signal Process. 68, 548–557 (2020) 6. K. Abratkiewicz, P.J. Samczyński, R. Rytel-Andrianik, Z. Gajo, Multipath interference removal in receivers of linear frequency modulated radar pulses. IEEE Sens. J. 21, 19000–19012 (2021) q y p 7. W.D. White, Low-angle radar tracking in the presence of multipath. IEEE Trans. Aerosp. Electron. Syst. 835–852 (1974) 8. R.J. Mcaulay, T.P. Mcgarty, Maximum-likelihood detection of unresolved radar targets and multipath. IEEE Trans. Aerosp. Electron. Syst. 821–829 (1974) 9. L. Yong, Z. Guoyi, Z. Xuzhou, Parameter estimation of LFMCW radar signal. J. SIGNAL Process. 30, 848–855 (2014) 10. F.G. Geroleo, M. Brandt-Pearce, Detection and estimation of LFMCW radar signals. IEEE Trans. Aerosp. Electron. Syst. 48, 405–418 (2012) 11. W. Jingyang, Z. Huanyu, W. Yue, Parameters estimation algorithm of LFM pulse compression radar signal. Trans. Beijing Inst. Technol. 32, 746–750 (2012) 12. T. Jianlong, Li. Tao, Y. Shaoquan, Direction-of-arrival estimation based on cross Wigner ville distribution. Syst. Eng. Electron. 26, 714–716 (2004) 13. D. Zhenmiao, Y. Liu, Recognition and parameters estimation of polyphase-coded radar signals. J. Electron. Inf. Tech- nol. 31, 781–785 (2009) 14. References R. Djemal, K. Belwafi, W. Kaaniche, S.A. Alshebeili, A novel hardware/software embedded system based on automatic censored target detection for radar systems. AEU-Int. J. Electron. Commun. 67, 301–312 (2013) 14. R. Djemal, K. Belwafi, W. Kaaniche, S.A. Alshebeili, A novel hardware/software embedded system based on automatic d d f d l j ,i , , , y censored target detection for radar systems. AEU-Int. J. Electron. Commun. 67, 301–312 (2013) 15. J. Sun, G. Xu, W. Ren, Z. Yan, Radar emitter classification based on unidimensional convolutional neural network. IET Radar Sonar Navig. 12, 862–867 (2018) 16. L. Liu, X. Li, Radar signal recognition based on triplet convolutional neural network. EURASIP J. Adv. Signal Process. 2021, 1–16 (2021) 17. F.C. Akyon, Y.K. Alp, G. Gok, O. Arikan, Classification of intra-pulse modulation of radar signals by feature fusion based convolutional neural networks, in 2018 26th European Signal Processing Conference (EUSIPCO). pp. 2290–2294. IEEE (2018) 18. W. Ye, C. Peng, Recognition algorithm of emitter signals based on pca+cnn, in 2018 IEEE 3rd Advanced Information Technology, Electronic and Automation Control Conference (IAEAC). pp. 2410–2414. IEEE (2018) gy pp 19. Z. Qu, W. Wang, C. Hou, C. Hou, Radar signal intra-pulse modulation recognition based on convolutional denoising autoencoder and deep convolutional neural network. IEEE Access. 7, 112339–112347 (2019) autoencoder and deep convolutional neural network. IEEE Acce 20. Z.-M. Liu, S.Y. Philip, Classification, denoising, and deinterleaving of pulse streams with recurrent neural networks. IEEE Trans. Aerosp. Electron. Syst. 55, 1624–1639 (2018) i IEEE Trans. Aerosp. Electron. Syst. 55, 1624–1639 (2018) p y 21. S. Apfeld, A. Charlish, G. Ascheid, Identification of radar emitter type with recurrent neural networks, in 2020 Sensor Signal Processing for Defence Conference (SSPD). pp. 1–5. IEEE (2020) 21. S. Apfeld, A. Charlish, G. Ascheid, Identification of radar emitter type w Signal Processing for Defence Conference (SSPD). pp. 1–5. IEEE (2020) i Signal Processing for Defence Conference (SSPD). pp. 1–5. IEEE (2 g g pp 22. P. Notaro, M. Paschali, C. Hopke, D. Wittmann, N. Navab, Radar emitter classification with attribute-specific recurrent neural networks. arXiv Prepr. arXiv1911.07683. (2019) neural networks. arXiv Prepr. arXiv1911.07683. (2019) p 23. Y. Du, J. Wang, W. Feng, S. Pan, T. Qin, R. Xu, C. Wang, Adarnn: adaptive learning and forecasting of time series, in Proceedings of the 30th ACM international conference on information & knowledge management. pp. 402–411 (2021) 24. A. Gretton, K.M. Borgwardt, M.J. Rasch, B. Competing interests Received: 7 July 2023 Accepted: 5 September 2023 Page 16 of 16 Yan et al. EURASIP Journal on Advances in Signal Processing (2023) 2023:94 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
https://openalex.org/W3131043787	https://oskar-bordeaux.fr/bitstream/20.500.12278/172804/1/sensors-21-01322.pdf	English	null	Emotion Detection for Social Robots Based on NLP Transformers and an Emotion Ontology	Sensors	2,021	cc-by	12,446	Wilfredo Graterol 1 , Jose Diaz-Amado 2,3 , Yudith Cardinale 1,2 , Irvin Dongo 2,4,* , Edmundo Lopes-Silva 3 and Cleia Santos-Libarino 3 Wilfredo Graterol 1 , Jose Diaz-Amado 2,3 , Yudith Cardinale 1,2 , Irvin Dongo 2,4,* , Edmundo Lopes-Silva 3 and Cleia Santos-Libarino 3 Wilfredo Graterol 1 , Jose Diaz-Amado 2,3 , Yudith Cardinale 1,2 , Irvin Dongo 2,4,* , Edmundo Lopes-Silva 3 and Cleia Santos-Libarino 3 1 Departamento de Computación y Tecnología de la Información, Universidad Simón Bolívar, 1 Departamento de Computación y Tecnología de la Información, Universidad Simón Bolívar, 1080 Caracas, Venezuela; 15-10639@usb.ve (W.G.); ycardinale@usb.ve (Y.C.) 1 Departamento de Computación y Tecnología de la Información, Universidad Simón Bolíva 1080 Caracas, Venezuela; 15-10639@usb.ve (W.G.); ycardinale@usb.ve (Y.C.) p p y g , , 1080 Caracas, Venezuela; 15-10639@usb.ve (W.G.); ycardinale@usb.ve (Y.C.) 1080 Caracas, Venezuela; 15-10639@usb.ve (W.G.); ycardinale@usb.ve (Y.C.) 2 Electrical and Electronics Engineering Department, Universidad Católica San Pablo, 04001 Arequipa, Peru; jose_diaz@ifba.edu.br y 2 Electrical and Electronics Engineering Department, Universidad Católica San Pablo, 04001 Arequipa, Peru; jose_diaz@ifba.edu.br 3 Electrical Engineering, Instituto Federal da Bahia, 45078-300 Vitoria da Conquista, Brazil; edmundo_lopes@ifba.edu.br (E.L.-S.); cleialibarino@ifba.edu.br (C.S.-L.) 3 Electrical Engineering, Instituto Federal da Bahia, 45078-300 Vitoria da Conquista, Brazil; edmundo_lopes@ifba.edu.br (E.L.-S.); cleialibarino@ifba.edu.br (C.S.-L.) 4 Estia Institute of Technology, University Bordeaux, 64210 Bidart, France 3 Electrical Engineering, Instituto Federal da Bahia, 45078-300 Vitoria da Conquista, Brazil; edmundo_lopes@ifba.edu.br (E.L.-S.); cleialibarino@ifba.edu.br (C.S.-L.) 4 Estia Institute of Technology, University Bordeaux, 64210 Bidart, France * Correspondence: ifdongo@ucsp.edu.pe; Tel.: +51-951-273-375 edmundo_lopes@ifba.edu.br (E.L. S.); cleialibarino@ifba.edu.br (C.S. L.) 4 Estia Institute of Technology, University Bordeaux, 64210 Bidart, France * Correspondence: ifdongo@ucsp.edu.pe; Tel.: +51-951-273-375 p 4 Estia Institute of Technology, University Bordeaux, 64210 Bidart, France 4 Estia Institute of Technology, University Bordeaux, 64210 Bidart, France * Correspondence: ifdongo@ucsp.edu.pe; Tel.: +51-951-273-375 * Correspondence: ifdongo@ucsp.edu.pe; Tel.: +51-951-273-375 * Correspondence: ifdongo@ucsp.edu.pe; Tel.: +51-951-273-375 Abstract: For social robots, knowledge regarding human emotional states is an essential part of adapting their behavior or associating emotions to other entities. Robots gather the information from which emotion detection is processed via different media, such as text, speech, images, or videos. The multimedia content is then properly processed to recognize emotions/sentiments, for example, by analyzing faces and postures in images/videos based on machine learning techniques or by converting speech into text to perform emotion detection with natural language processing (NLP) techniques. Keeping this information in semantic repositories offers a wide range of possibilities for implementing smart applications. Citation: Graterol, W.; Díaz-Amado, J.; Cardinale, Y.; Dongo, I.; Lopes-Silva, E.; Santos-Libarino, C. Emotion Detection for Social Robots Based on NLP Transformers and an Emotion Ontology. Sensors 2021, 21, 1322. https://doi.org/10.3390/ s21041322 Academic Editor: Carina Soledad González-González Received: 19 December 2020 Accepted: 7 February 2021 Published: 13 February 2021 Academic Editor: Carina Soledad González-González Received: 19 December 2020 Accepted: 7 February 2021 Published: 13 February 2021 Keywords: social robots; natural language processing; ontology; emotion detection; text classiﬁcation Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional clai- ms in published maps and institutio- nal afﬁliations. Wilfredo Graterol 1 , Jose Diaz-Amado 2,3 , Yudith Cardinale 1,2 , Irvin Dongo 2,4,* , Edmundo Lopes-Silva 3 and Cleia Santos-Libarino 3 We propose a framework to allow social robots to detect emotions and to store this information in a semantic repository, based on EMONTO (an EMotion ONTOlogy), and in the ﬁrst ﬁgure or table caption. Please deﬁne if appropriate. an ontology to represent emotions. As a proof-of-concept, we develop a ﬁrst version of this framework focused on emotion detection in text, which can be obtained directly as text or by converting speech to text. We tested the implementation with a case study of tour-guide robots for museums that rely on a speech-to-text converter based on the Google Application Programming Interface (API) and a Python library, a neural network to label the emotions in texts based on NLP transformers, and EMONTO integrated with an ontology for museums; thus, it is possible to register the emotions that artworks produce in visitors. We evaluate the classiﬁcation model, obtaining equivalent results compared with a state-of-the-art transformer-based model and with a clear roadmap for improvement. sensors sensors 1. Introduction Nowadays, the presence of robots are becoming more common in human daily life [1]. Service robots concretely share environments with human beings to actively collaborate with them in speciﬁc daily tasks, such as serving as an assistant to nurses in patient walking and patient sitting tasks in hospital environments [2] or working as a student receptionist at a university [3]. Consequently, human–robot interactions (HRI) need to be revised in order to make the integration of robots into society as seamless and natural as possible. Copyright: © 2021 by the authors. Li- censee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and con- ditions of the Creative Commons At- tribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). An emerging approach to modeling social behaviors for service robots is based on emo- tion recognition. The detected emotion can be used for many purposes, such as dictating a robot’s behavior and accordingly adapting interactions with humans (in real-time) [4–8] or associating the emotion to events or objects in speciﬁc domains (e.g., opinions about com- mercial products, emotions produced by artworks in museums, and emotions produced by plates in restaurants) for further analysis [9,10]. From this perspective, keeping this https://www.mdpi.com/journal/sensors Sensors 2021, 21, 1322. https://doi.org/10.3390/s21041322 Sensors 2021, 21, 1322 2 of 19 2 of 19 information in semantic repositories offers a wide range of possibilities for automatically modeling the robot’s behavior or for supporting the implementation of smart applications. information in semantic repositories offers a wide range of possibilities for automatically modeling the robot’s behavior or for supporting the implementation of smart applications. g pp g p pp The information from which emotion detection is processed can be gathered for robots in real-time via different media as they interact with humans, such as text, speeches, images, or videos [11]. This multimedia content is processed to recognize emotions/sentiments with proper techniques, for example, by analyzing faces and postures in images/videos based on machine learning or by converting speech into text to perform the emotion detection with well-known natural language processing (NLP) approaches [12–15]. In this context, we propose a framework to allow social robots to detect emotions and to store this information in an ontology-based repository. The detected emotions and other information are stored in a semantic repository based on an extensible ontology, called EMONTO (an EMotion ONTOlogy), to represent emotions. 1. Introduction EMONTO can be ex- tended with other speciﬁc domain ontologies representing entities with which the emotion can be related. The detected emotion can be used for many purposes, such as accord- ingly designing the reactions and actions of robots or combining the semantic information of emotions with other ontologies related to a robot’s tasks (e.g., SLAM (Simultaneous Localization And Mapping), navigation, and perception) [16] or with domain-speciﬁc ontologies related to the environment where robots work (e.g., museums, restaurants, and hospitals) [17]. To show the suitability of our approach, we develop a ﬁrst version of this framework focused on emotion detection in text and tested it with a case study of tour-guide robots for museums. The robots receive the text directly from people or by converting speech to text. Then, we used NLP transformers to make sentence embeddings, transforming the text to a vector representation. Thus, the classiﬁcation problem was solved in this new vector space, resulting in emotion labels for each text. The ﬁrst version of the framework relies on (i) a speech-to-text converter based on the Google Application Programming Interface (API) and a Python library; (ii) a neural network to label the emotions in texts based on NLP transformers; and (iii) EMONTO, an ontology to represent emotions integrated with an ontology for museums; thus, it is possible to register the emotions that artworks produce in museum visitors for further analysis. y This ﬁrst version of the framework offers the option for people to communicate with robots directly by text or for robots to perform an analysis on written sources to which they have access, such as social networks. In some situations, text may be the best or only way for people to communicate with robots; for example, in really loud or busy environments where the robots struggles to identify a single person or cannot properly hear voice commands, or for impaired people who cannot properly speak and rely on text messages to communicate their thoughts. Additionally, this approach can be incorporated in a chatbot to guide the responses according to the detected emotions. We describe the stacked ensemble of transformer models to detect emotions from text and experimentally evaluate it. We obtain acceptable results with a clear roadmap for improvement, compared with the NVIDIA-AI [15], a state-of-the-art transformer-based model applied in the same task. 2.1. Social Robots and Emotion Detection from Speech Nowadays, emotion recognition is essential to improving HRI, in particular, for service robots. Emotion recognition allows us to model the social behavior of robots that share spaces with humans, thus making social robots a growing area of robotics where psychology and sociology aspects converge [18]. However, the design and development of robot’s abilities in acting and interacting physically, emotionally, socially, and safely with humans has generally been poorly understood so far [19]. Indeed, in order to socially interact with humans, a robotic system should be able not only to understand user behavior and intentions but also to estimate their emotional state with a high level of acceptability and usability [9]. Thus, much more attention should be paid to an assessment of the feasibility of a social robot in real-life scenarios among different cultures and people [4]. Thus, a growing effort is shown in recent works for natural and seamless integration of robots into society, supported by emotion recognition to model their social behaviors. The sensing capabilities of robots allows us to gather several multimedia content, from which emotions can be recognized. Although our ﬁrst version of the framework considers only voice and text, in the future, we aim to implement the capacity to process all other content. For now, we revise some works dealing with emotion recognition in speech and text. In [5], openSMILE, a framework to extract several speech emotion features was described; it was based on fuzzy C-means to cluster the training set into multiple sub- classes; then, multiple random forest methods make the decision to identify the emotion (i.e., angry, fear, happy, neutral, sad, and surprise) from the selected speech features. In [6], a multi-modal emotion detection model for personal assistant robots was presented. This model was implemented by means of emotion recognition from voice and images, and both outputs were merged to achieve the detected emotion using machine learning techniques. A speech emotion recognition (SER) system based on a combinations of convolution neural network (CNN) and random forest models was presented in [7]. The CNN model was used to extract speech emotion features from the normalized spectogram, and the random forest algorithm performed the classiﬁcation. SER was used on a NAO robot with recorded sound commands, working with four basic emotions: happiness, anger, sadness, and joy. A behavior modulation system (BeMoSys) for social robots was proposed in [8]. 1. Introduction The results show a micro F1 score of 0.6360, a macro F1 score of 0.4995, and a Jaccard index of 0.5076 overall for the model. In summary, the contributions of this research are (i) a general framework for emo- tion recognition in social robots, considering all the sensing capabilities of robots; (ii) EMONTO, the ontology of emotions integrated to the framework that provides the base for implementing complex and smart applications from the detected emotions, related to the environment of the social robot; (iii) a ﬁrst version of the framework, as a proof-of-concept, aimed at emotion recognition from text obtained directly or by converting sound to text; and (iv) an NLP transformer model for emotion recognition in text that behaves similar to a state-of-the-art approach. We present our research organized as follows. Section 2 describes the related work about social robots and emotion detection from speech and from text, and the use of ontologies to store the recognized information. Our framework for emotion detection Sensors 2021, 21, 1322 3 of 19 based on text analysis for social robots is detailed in Section 3. Section 4 presents EMONTO, an extensible ontology that represents emotions and supports the development of smart and interoperable applications. An experimental evaluation of the model to detect emotion from texts is presented in Section 5. We discuss the results obtained in Section 6. Finally, we conclude in Section 7. 2. Related Work In this section, we ﬁrst survey recent studies that show the importance of emotion recognition in improving HRI for social robots. Then, we describe some works in the ﬁeld of emotion detection from text. Finally, we describe recent studies proposing approaches to recognizing emotion supported by ontologies in the context of social robotics. We highlight the relation to or difference of the cited works from our research on these three aspects. 2.1. Social Robots and Emotion Detection from Speech BeMoSys provides robots with the capacity for emotional speech recognition based on signal-processing and machine learning techniques to perform the emotion classiﬁca- tion. It considered ﬁve inputs corresponding to the ﬁve emotions extracted from the speech signal in order to derive the estimated level of happiness. According to this value, the appropriate robot behavior was set. We found few recent works that combine speech and text in the context of robotics for emotion recognition. In [20], a system that combines the analysis of prosodic features of the speech and sentiment analysis on the text (obtained from the speech) to recognize emotion and let robots properly express reactive emotions was presented. Sentiment analysis over the text allows for the identiﬁcation of the polarity (i.e., negative, neutral, and positive), for Sensors 2021, 21, 1322 4 of 19 which the value (−1, 0, or 1) is used in emotion classiﬁcation, combined with an analysis of the prosodic features. The considered categories in this study were embarrassment, unsettling, noticing, remembering, unexpectedness, surprise, hesitation, anxiety, pain, dislike, disappointment, pleasure, and anger. These recent works demonstrate the current trend of voice analysis for emotion recognition in social robots in improving HRI and modeling social behaviors for robots. However, there are still limitations on the quality and precision of the results; thus, more accuracy models should be used, such as text analysis. With our proposed framework, social robots are able to recognize emotions in humans with whom they interact and accordingly to adapt their behaviors and interactions. Hence, our work is a contribution to social robotics. Besides that, converting voice into text represents more beneﬁts is social robotics: robots can analyze writing sources (e.g., social media) as an alternative for speech-impaired people when communicating with robots, does not neglect other multimedia sources, and can be used to improve chatbot applications. 2.2. Emotion Detection from Text For emotion recognition systems, an important aspect to consider is the emotion models that delimit the classiﬁcation process [13]. While according to Cowie [21], in the ﬁeld of psychology, there are many different theories regarding the representation of emotions, within the area of NLP, there are two that stand out as the most used: Ekman’s basic emotions [22] and Plutchik’s wheel of emotions [23]. Ekman’s model consists of six basic emotions: anger, disgust, fear, happiness, sadness, and surprise. Plutchik’s model consists of a multi-dimensional approach, where there are four opposing pairs of axes and emotions are deﬁned as different points along these axis. In this model, an emotion is determined by an emotion axis and its intensity. These axis pairs are joy– sadness, anger–fear, trust–disgust, and surprise–anticipation. Figure 1 shows an extraction of the Plutchik’s model, with these axis and intensity denoted with colors in the concentric circles; from these emotions, many other emotions can be derived as a combination of other emotions and intensities. Most works in emotion detection consider a small subset of these group of emotions. Figure 1. An extract of Plutchik’s wheel of emotions. Figure 1. An extract of Plutchik’s wheel of emotions. 5 of 19 Sensors 2021, 21, 1322 Most common approaches to performing emotion classiﬁcation and general NLP tasks are based on recurrent neural networks (RNN) or some variants of them, such as long short-term memory (LSTM) [24], multiplicative LSTM (mLSTM) [25], or gated recurrent unit (GRU) [26–28]. However, these approaches suffer from a couple of related issues: the complex nature of emotion expression and the shortage of quality data for this task [14]. Finding quality label data for this task is hard to do, at least in order to completely train these types of models. Nevertheless, these limitations have been overcome by an attention-based transformers architecture [29], which is inspired by the encoder–decoder architecture used for sequence to sequence (Seq2seq) tasks. In high-level terms, it consists of two stacks (one of encoders and one of decoders), where the encoder and decoder use some form of self-attention (https://jalammar.github.io/illustrated-transformer/ (accessed on 18 December 2020)) and a feed forward neural network. Each encoder in the stack receives its input from the previous layer, except the ﬁrst one that receives a series of tokens. Then, the ﬁnal encoder in the stack sends its output to every decoder. 2.2. Emotion Detection from Text Then, each decoder uses this and the input from the previous layer in the stack to return its output. Afterward, the output for the ﬁnal decoder is passed to a linear + softmax function to ﬁnally return the output of the network. Recently, it has become more and more common to use transformers for NLP in texts, since they have been proven to be effective in many different tasks (e.g., masked token prediction, next sentence prediction, question answering, machine translation, summarization, natural language inference, and sentiment analysis) [30–32]. This is thanks to the fact that they are trained with unsupervised learning models, diminishing the need for labeled data. Transformers are very effective in transfer learning, allowing researchers to pretrain these transformers with large amounts of general- purpose texts and then to ﬁnetune these models for their speciﬁc tasks with good results, less effort, and labeled data, such as NVIDIA-AI (currently named Megatron-LM), the model proposed in [15], by researchers of NVIDIA. p p y For the classiﬁcation task, the transformer architecture allows us to test different types of methods to approach it. An ensemble is a successful approach in reducing the variance of sub-models by training multiple sub-models to combine the predictions from them [33]. There are many ways to implement an ensemble (e.g., stacking, boosting, and bagging). Recently, many experiments were conducted with ensemble neural networks using transformers. For example, boosting bidirectional encoder representations from transformers (BERT) proposed in [33] is a model to introduce the multi-class boosting into BERT [32]; the classiﬁer for detecting aggression in social media posts by bagging BERT models presented in [34]; and a stacked ensemble with 25 BERT to classify fake news, described in [35]. Currently, there are many available implementations of these transformers, such as the Hugging Face Transformers library [36], which provides a wide variety of implemented pretrained transformers that work with a classic NLP workﬂow; the Simple Transformers library (https://github.com/ThilinaRajapakse/simpletransformers (accessed on 18 Decem- ber 2020)), which is a simpliﬁcation of the previous one, suitable for non-expert researchers; and the Sentence Transformers library [37]. Our classiﬁcation model is based on a transformer architecture, supported on the Simple Transformers library. 2.3. Social Robots, Emotion Detection, and Ontologies 2.3. Social Robots, Emotion Detection, and Ontologies In the domain of social robots and emotion detection, ontologies have been applied in order to represent the emotional knowledge. The authors of [38] proposed a cognitive architecture for an emotion-aware robotic system by using a multilayer perceptron neural network and several features associated with facial expression (e.g., brows, lips, and face muscles motions) and emotions on speech (e.g., words used, syntactic structure, and meaning). They used an upper ontology to cover all common-sense concepts of human states and robotic systems (EmUO) and HTemp ontology to represent the complex context using its n-ary schema. In [39], OntCog that is based on IEEE 1872-2015—IEEE Sensors 2021, 21, 1322 6 of 19 Standard Ontologies for Robotics and Automation, was proposed. The main concept in OntCog was RobotSense that is related to all concepts in this ontology. Other properties such as IRI_Emotion to associate an emotion to a smell were also deﬁned. The same authors propose in [40] an architecture, called cognitive model development environment (CMDE), to capture and process data collected by sensors present in the robot and to store the perception of the environment using an ontology. This ontology is the one proposed in [39]. Other works have used ontologies to store information in the context of robots but for activity detection [41], user-intention recognition [42], object detection [43], etc. y g j These studies used semantic knowledge representation in different aspects of emotion sensing for social robots; however, the ontologies proposed neither represent emotions nor are extensible to be able to relate such emotions to robots’ tasks or to other entities. We propose EMONTO, an extensible ontology of emotions, and we incorporate it into the framework; thus, this semantic information can be used for many proposes in the context of social robotics. 3.1. Human–Robot Interaction In this phase, the social robot interacts with people in the museum, starting conversa- tions with a neutral tone, but eventually, after engaging with the person or group of people, it can adjust its tone to act according to the detected emotions (e.g., excited, energetic, and comprehensive). The robot must attempt to get, via a series of questions, the visitor’s opinion regarding the artworks; name; and, if possible, some unique identiﬁer such as a social media handle. With these data, it is possible to identify registered/frequent visitors (possible facial and posture recognition if images and videos are considered); thus, the robot can remember them and choose how to initiate contact or a different type of conversation. Moreover, data mining in social media is also possible in determining users’ proﬁles. According to the detected emotion in step 4 (see Figure 2), the robot can adapt its behavior (e.g., change the tone to ﬁt the mood of the user and decide whether to continue the conversation). 3.3. Multimedia Processing In this step, the robot converts the multimedia raw data into proper representations for emotion analysis. For example, feature vectors from images and videos, normalized spectograms from audio, and text from speech. For the ﬁrst version of our framework, in this step, the robots can receive text directly or by converting audio to a text format. We developed a speech-to-text algorithm using the available Google API [44] and the SpeechRecognition Python library, which is specialized in speech processing, audio reading, and noise reduction, based on a Hidden Markov Model for speech recognition and an RNN to tackle the sequence problems, where the timing is variable. g For the case of images and videos, features extracted can obey traditional machine learning models as well as more complex models such as multimodal-based ones that are becoming popular for emotion recognition in images and videos. We plan to incorporate such techniques to improve the implementation of our framework. 3.2. Multimedia Capturing During the interaction with the visitor, the robot records and saves all multimedia content captured by its sensors. For the case of speech, the robot keeps in the audio ﬁles all answers and comments it receives from the user. Although, the emotion is detected from the extracted information from the multimedia raw data, keeping the original data (e.g., images, videos, and audio) permits additional analysis, for example to extract prosody features from speech or to extract context information from images, for further experiments. For the ﬁrst version of our framework, in this step, we implemented speech capturing with an ambient noise reduction algorithm to record audio and for speech-to-text conver- sion in order to reduce noises that can interfere with the results. According to the sensors’ capabilities, robots can also gather images and videos and can use facial and body posture recognition to extract the related features for emotion detection. 3. A Framework for Emotion Detection for Social Robots The proposed framework is a process divided into ﬁve sequential steps, as shown in Figure 2: (1) human–robot interaction; (2) multimedia capturing; (3) multimedia processing: image/video processing, audio processing, and text processing (text obtained directly or con- verted from audio); (4) emotion detection; and (5) storage in a semantic repository. From the human–robot interaction (step 1 in Figure 2), multimedia content is gathered from the robot’s sensory capacity; these raw data are saved in a repository (step 2 in Figure 2). The following steps imply transformations of the obtained data from the previous step, starting with the conversion of raw input obtained from the user to proper representations for emotion analysis (e.g., feature vectors and text) (step 3 and step 4 in Figure 2) and ﬁnishing with the population of an emotion ontology that can be used for different purposes (step 5 in Figure 2). This framework can be instantiated in social robots systems, taking advantage of all the sensing capacities of robots. As a proof-of-concept, in this work, we implement the whole pipeline taking into account only text and sound for a service robot in a museum getting feedback from users regarding the artworks. In this section, we detail the steps related to emotion recognition for this ﬁrst implementation of our proposed framework. The next section presents EMONTO and its instantiation. Figure 2. General overview of the framework. Figure 2. General overview of the framework. 7 of 19 Sensors 2021, 21, 1322 3.1. Human–Robot Interaction 3.1. Human–Robot Interaction 3.4. Emotion Detection According to the multimedia data considered, a wide range of emotion detection techniques can be implemented, mostly based on machine learning models. For the ﬁrst version of our framework, in this step, we approached the emotion detec- tion task as a multi-label classiﬁcation problem, modeled with a Transformer architecture and meta-learning. We considered eleven emotions: anger, anticipation, disgust, fear, joy, love, optimism, pessimism, sadness, surprise, and trust. The classiﬁcation task consisted of labeling a text as “neutral or no emotion” or as one or more of the previously mentioned emotions, i.e., a binary vector indicating if each emotion was detected (1) or not (0). Based on the available pretrained transformers, we designed a stacked ensemble architecture, as shown in Figure 3. Each weak learner classiﬁer consists of a transformer, used to create a sentence embedding and a Multi-Layer Perceptron (MLP), with two hidden Sensors 2021, 21, 1322 8 of 19 layers of sizes 500 and 300 (see Figure 4). To avoid overﬁtting of our MLP, we used an early stop in the training, using 10% of the training data to validate in each iteration if the training is still effective. For implementation of the transformers, we used the Sentence Transformers library (https://github.com/UKPLab/sentence-transformers (accessed on 18 December 2020)) [37], since they are speciﬁcally pretrained transformers to perform semantic textual similarity tasks, leading to sentence embeddings. This is a mapping of sentences to a vector space such that sentences with similar semantic meanings are close in this new space. From the Sentence Transformers library, we used the roberta-large-nli-stsb- mean-tokens, distilbert-base-nli-mean-tokens, and bert-large-nli-stsb-mean-tokens transformers to perform the sentence embeddings. Therefore, we use a stacked ensemble with three different weak learners. Figure 3. Stacked ensemble architecture implemented. Figure 4. Structure of each weak learner in our model. Figure 3. Stacked ensemble architecture implemented. Figure 3. Stacked ensemble architecture implemented. Figure 3. Stacked ensemble architecture implemented. Figure 3. Stacked ensemble architecture implemented. Figure 4. Structure of each weak learner in our model. The meta-learning consists of an MLP with one hidden layer of 10 neurons. All MLPs were implemented using sklearn (https://scikit-learn.org/stable/index.html (accessed on 18 December 2020)) [45] and using the default adam solver, the cross-entropy loss function, the RELU activation function, a regularization term set at 0.0001, and a random_state initialized in 1. Previously, word embedding was used for emotion detection in text and for the general NLP task [14]. 3.4. Emotion Detection One common approach to processing a given sentence is to tokenize the sentence with word embeddings and to then pass the sequence of tokens to a RNN. However, with transformers, we have a simple way to directly create embeddings for whole sentences. 4. Emotion Ontology and Instantiation Data provided by the recognition methods are normally stored for analysis, which can generate a complex network of knowledge. The ontologies are well-deﬁned and standard models to represent the knowledge managed by these applications. In this work, we propose EMONTO, an ontology to represent emotions and to support the development of smart and interoperable applications. 4.1. EMONTO: An Extensible Emotion Ontology EMONTO is an extensible ontology that represents emotions under different catego- rization proposals. We adopted some concepts from several emotion ontologies [46,47] to provide compatibility. Figure 5 shows the main classes of EMONTO. The central 9 of 19 Sensors 2021, 21, 1322 9 of 19 class is emo:Emotion, which has a category (hasCategory) according to a Category class. The current version of EMONTO considers archetypal [46], Douglas Cowie [46], and Robert Plutchik [48] emotion categorizations, which group emotions into 6, 25, and 56 (8 basic emo- tions) values, respectively. Nevertheless, any other categorization model can be integrated as a subclass of emo:Category. class is emo:Emotion, which has a category (hasCategory) according to a Category class. The current version of EMONTO considers archetypal [46], Douglas Cowie [46], and Robert Plutchik [48] emotion categorizations, which group emotions into 6, 25, and 56 (8 basic emo- tions) values, respectively. Nevertheless, any other categorization model can be integrated as a subclass of emo:Category. Figure 5. EMONTO: an extensible emotion ontology. Figure 5. EMONTO: an extensible emotion ontology. EMONTO has emo:Event as a class that connects the emo:Object, emo:Person, and emo:Emotion entities. An emotional Event is produced by (isProducedBy) a Person and is caused by (isCausedBy) an Object (e.g., artworks, candidates, and plates). An Event can produce several Emotions. The entities emo:Object and emo:Person are general classes that can connect other ontologies, such as museum and artwork ontologies [17] as Object or user- proﬁle ontologies [49] as Person. The idea is to make EMONTO extensible and ﬂexible to be easily adopted in scenarios where data related to the recognized emotions need to be stored for further analysis. The ontology provides the modality (emo:Modality) of the information used to recognize the emotion (e.g., emo:Gesture, emo:Face, emo:Posture), and the type of annotator (emo:AutomaticAnnotator and emo:HumanAnnotaton). Moreover, a datatype property emo:hasIntensity is associated to the category to express the level of conﬁdence (a ﬂoat value between 0.0 and 1.0). Figure 5. EMONTO: an extensible emotion ontology. EMONTO has emo:Event as a class that connects the emo:Object, emo:Person, and emo:Emotion entities. An emotional Event is produced by (isProducedBy) a Person and is caused by (isCausedBy) an Object (e.g., artworks, candidates, and plates). An Event can produce several Emotions. The entities emo:Object and emo:Person are general classes that can connect other ontologies, such as museum and artwork ontologies [17] as Object or user- proﬁle ontologies [49] as Person. 4.2. Instantiation of our Emotion Ontology 4.2. Instantiation of our Emotion Ontology Once the emotions are recognized by our multi-modal recognition method, an Event is created to represent the emotional event (<E0001 rdf:type emo:Event>), associated to some datatype properties (e.g., <E0001 emo:createdAt 1607984678>). Then, the Event is associated to either new instances of Person and Object (<E0001 emo:isProducedBy P0001> and <0001 emo:isCausedBy O0001>) or to existing values. Person and Object should be also recognized by using other models that can be combined with our multi-modal method in the proper data acquisition phase. For example, by applying facial recognition to identify registered or new users in the system and object detection to recognize speciﬁc objects (e.g., artworks in museums and plates in restaurants). This work is focused on emotion recognition; Person and Object detection is beyond the scope of this research. g j y p According to the results of the emotion recognition method, one or more emotion en- tities are created (<EM0001 rdf:type emo:Emotion>, <EM0002 rdf:type emo:Emotion>, <EM0003 rdf:type emo:Emotion>, etc.) and associated to the Event (<E0001 emo:produces EM0001>, <E0001 emo:produces EM0002>, <E0001 emo:produces EM0003>). Each emo- tion has a category (<EM0001 emo:hasCategory C0001>, and <EM0002 emo:hasCategory C- 0002>, <EM0003 emo:hasCategory C0003>) which can be archetypal, Douglas Cowie, or Robert Plutchik classiﬁcations (<C0001 rdf:type ArchetypalCategory>, <C0002 rdf:- type ArchetypalCategory>, and <C0003 rdf:type ArchetypalCategory>). The datatype properties emo:hasIntensity and emo:emotionValue are added to the category (<C0001 emo:hasIntensity 0.13> and <C0001 emo:emotionValue "anger">) with the values ob- tained by the emotion recognition method (e.g., 0.13 as intensity and “anger” as emotion). Modality and annotator are also instantiated. A pseudocode of creating triples is shown in Algorithm 1. First, libraries related to Resource Description Framework (RDF) man- agement have to be import (line 1), then a new graph, which contains the RDF triples, is created (line 2). Namespaces of the ontology are added (lines 3–5). From line 6 to line 17, new RDF triples are added to the graph. Algorithm 1: Creating RDF triples. Algorithm 1: Creating RDF triples. 1 import RDF libraries 2 g = Graph() 3 EMO = Namespace(“http://www.emonto.org/ (accessed on 18 December 2020)”) //Creating a Namespace. 4 g.add_namespace(“emo”, EMO) //Adding the namespace EMO. 5 g.add_namespace(“foaf”, FOAF) //Adding the namespace FOAF, which is already defined in the libraries. 6 g.add_triple((EMO.E0001, RDF.type, EMO.Event)) //Creating an Event. 7 g.add_triple((EMO.P0001, RDF.type, FOAF.Person)) //Creating a Person. 8 g.add_triple((EMO.O0001, RDF.type, EMO.Object)) //Creating a Object. 9 g.add_triple((EMO.EM0001, RDF.type, EMO.Emotion)) //Creating an Emotion 1. 4.1. EMONTO: An Extensible Emotion Ontology The idea is to make EMONTO extensible and ﬂexible to be easily adopted in scenarios where data related to the recognized emotions need to be stored for further analysis. The ontology provides the modality (emo:Modality) of the information used to recognize the emotion (e.g., emo:Gesture, emo:Face, emo:Posture), and the type of annotator (emo:AutomaticAnnotator and emo:HumanAnnotaton). Moreover, a datatype property emo:hasIntensity is associated to the category to express the level of conﬁdence (a ﬂoat value between 0.0 and 1.0). EMONTO has emo:Event as a class that connects the emo:Object, emo:Person, and emo:Emotion entities. An emotional Event is produced by (isProducedBy) a Person and is caused by (isCausedBy) an Object (e.g., artworks, candidates, and plates). An Event can produce several Emotions. The entities emo:Object and emo:Person are general classes that can connect other ontologies, such as museum and artwork ontologies [17] as Object or user- proﬁle ontologies [49] as Person. The idea is to make EMONTO extensible and ﬂexible to be easily adopted in scenarios where data related to the recognized emotions need to be stored for further analysis. The ontology provides the modality (emo:Modality) of the information used to recognize the emotion (e.g., emo:Gesture, emo:Face, emo:Posture), and the type of annotator (emo:AutomaticAnnotator and emo:HumanAnnotaton). Moreover, a datatype property emo:hasIntensity is associated to the category to express the level of conﬁdence (a ﬂoat value between 0.0 and 1.0). 10 of 19 10 of 19 Sensors 2021, 21, 1322 4.2. Instantiation of our Emotion Ontology 10 g.add_triple((EMO.EM0002, RDF.type, EMO.Emotion)) //Creating an Emotion 2. 11 g.add_triple((EMO.EM0003, RDF.type, EMO.Emotion)) //Creating an Emotion 3. 12 g.add_triple((EMO.E0001, EMO.produces, EMO.EM0001)) //Associating E0001 to EM001. 13 g.add_triple((EMO.E0001, EMO.produces, EMO.EM0002)) //Associating E0001 to EM002. 14 g.add_triple((EMO.E0001, EMO.produces, EMO.EM0003)) //Associating E0001 to EM003. 15 ... 16 g.add_triple((EMO.C0001, EMO.hasIntensity, Literal(0.13))) //Intensity value. 17 g.add_triple((EMO.C0001, EMO.emotionValue, Literal(“anger”))) //Recognized emotion. 18 ... 4.3. Querying Our Emotion Ontology The semantic repository enables the possibility of implementing simple and advanced queries in order to retrieve speciﬁc values or to infer new knowledge, in both cases for data analysis. For example, a simple query can retrieve information about the emotions produced by a speciﬁc artwork. The pseudocode presented in Algorithm 2 retrieves the emotions produced by the artwork The_Starry_Night: in line 1, libraries related to the RDF Sensors 2021, 21, 1322 11 of 19 11 of 19 manipulation are imported; then, a graph should be initialized to read the RDF triples (lines 2 and 3); the namespaces used in the ontology have to be added (line 4); and ﬁnally, following the SPARQL syntax, a query is performed (lines 5–15). manipulation are imported; then, a graph should be initialized to read the RDF triples (lines 2 and 3); the namespaces used in the ontology have to be added (line 4); and ﬁnally, following the SPARQL syntax, a query is performed (lines 5–15). Algorithm 2: Obtaining the artwork The_Starry_Night. Algorithm 2: Obtaining the artwork The_Starry_Night. 1 import RDF libraries 2 g = Graph() 3 g.read(“database_emotions.ttl”, format=“ttl”) 4 g.add_namespace(“emo”, “http://www.emonto.org/ (accessed on 18 December 2020)”) 5 qres = g.query(“SELECT ?emotion_l 6 WHERE { 7 ?emotion a emo:Emotion ; 8 ?emotion emo:emotionValue 9 ?emotion_l . 10 ?event a emo:Event ; 11 ?event emo:produces ?emotion ; 12 ?event emo:isCausedBy 13 ?object. 14 ?object a emo:Object ; 15 ?object rdfs:label “The_Starry_Night” . 16 }”) Other information such as the artworks that produce the emotion “surprise” can be also obtained (see Algorithm 3). Algorithm 3: Obtaining the emotion surprise. Algorithm 3: Obtaining the emotion surprise. Algorithm 3: Obtaining the emotion surprise. 1 “SELECT ?object_l 2 WHERE 3 ?emotion a emo:Emotion ; 4 ?emotion emo:emotionValue “surprise”. 5 ?event a emo:Event ; 6 ?event emo:produces ?emotion ; 7 ?event emo:isCausedBy ?object. 8 ?object a emo:Object ; 9 ?object rdfs:label ?object_l . 10 ” In these examples, the Object entity is an artwork, but an object could also be plates in restaurants, movies, or any entity that produce emotions. The extensibility of EMONTO allows researchers to integrate any speciﬁc domain ontology. 5. Experimental Evaluation of the NLP Transformer Model In this section, we show the performance of our transformer-based method compared with the NVIDIA-AI model [15], which is also based on a transformer architecture. We detail the procedures for obtaining adequate data for training and testing, some important implementation details, and the comparative results. We used the dataset for SemEval-2018 Task 1 [28], specifically the dataset E-c. This dataset contains tweets in English, each with its own ID and contents, that are labeled with a vector of eleven binary numbers, corresponding to the eleven considered emotions (anger, anticipation, disgust, fear, joy, love, optimism, pessimism, sadness, surprise, and trust). If the value at position k is 1, the tweet contains the emotion k; otherwise, the value is 0. Table 1 illustrates some samples taken from the dataset in which the ﬁrst text presents the emotions anger, fear, and sadness. 12 of 19 Sensors 2021, 21, 1322 Table 1. Examples of tagged text in the dataset. Table 1. Examples of tagged text in the dataset. Text Text Target Has anyone else had a bad experience with Poun... [1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0] #America ﬁnding #gratitude amidst the sadness... [0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0] Let us not burden our remembrances with a heav... [0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0] @AdsByFlaherty And you’re cheerfully defending... [1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0] The irony and hilarity of making Taylor Lautne... [0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0] Training of the model was performed with the 2018-E-c-En-train ﬁle that contains 6838 tweets labeled as indicated previously, and all the result metrics were calculated using 2018-E-c-En-test-gold that contains 3259 tweets. Training was performed using all 6838 samples of the 2018-E-c-En-train dataset, since validation was done on the 2018-E-c- En-test-gold dataset; thus, there was no need to split the data used for training. In Table 2, we present the number of positive values per emotion among the samples of the dataset; there is some class imbalance as the surprise and trust labels are poorly represented. How- ever, since this is a multi-label classiﬁcation, this is not a trivial issue to handle and we ended up ﬁnding better results without upsampling the data. Table 2. Number of occurrences of the tested emotions in the samples of the training dataset. Table 2. 5. Experimental Evaluation of the NLP Transformer Model Number of occurrences of the tested emotions in the samples of the training dataset. Anger Anticipation Disgust Fear Joy Love Optimism Pessimism Sadness Surprise Trust 1101 425 1099 485 1442 516 1143 375 960 170 153 Anger Anticipation Disgust Fear Joy Love Optimism Pessimism Sadness Surprise Trust 1101 425 1099 485 1442 516 1143 375 960 170 153 For metrics, as per the ofﬁcial competition rules, we used the Jaccard index to measure accuracy as this is a multi-label classiﬁcation task, based on Equation (1), where T is the set of tweets in the dataset and where, for any given tweet t, Gt is the set or the expected labels for t and Pt is the set of predicted labels for t. Accuracy = 1 \|T\| ∑ t∈T \|Gt ∩Pt\| \|Gt ∪Pt\| (1) (1) Algorithm 4 shows the general process of training: the implementations needed for both the transformers and the MLPs are imported (line 1 and line 2, respectively); they can be either user-made libraries or publicly available ones; then, the training samples as well as their expected labels are loaded (line 4 and line 5); the transformer models that will be used for each of the weak learners are also loaded (line 7); afterward, the training of the weak learners is done (lines 8–15); once the training cycle is performed, all predictions made by the weak learners are combined (line 17) to be used to train the meta-learner MLP; and this ﬁnal classiﬁer is created (line 19) and trained (line 20). This algorithm is available in Python (https://colab.research.google.com/drive/16Ja5MUFb_ A0RmaBVKvcjYxGSmvmSTtMN?usp=sharing (accessed on 18 December 2020)). 13 of 19 Sensors 2021, 21, 1322 Algorithm 4: Training process. 5. Experimental Evaluation of the NLP Transformer Model 1 include transformers; //Import the implementation of transformers being used 2 include MLP; //Import MLP that will be used 3 // Load the data that will be used for training 4 Data x_train = data_read(training_samples); 5 Data y_train = data_read(training_results); 6 // Deﬁne the set of transformers to use (non empty set) 7 Set models = Set of transformers in ensemble; 8 List train_embeddings = []; 9 // For each transformer, train the weak learner associated 10 foreach transformer model in models do 11 MLP clf; 12 List embbedding = model.generate_embeddings(x_train); 13 clf.train(embbedding,y_train); 14 // Store the created embeddings created by the weak learner 15 train_embeddings.append(clf.predict(embbedding)) 16 // Concatenate the outputs of each sample from the weak learners 17 train_embeddings = concatenate_elementwise(train_embeddings) 18 // Use that to train the ﬁnal meta-learning 19 MLP clf; 20 clf.train(train_embeddings, y_train) Algorithm 4: Training process. 1 include transformers; //Import the implementation of transformers 2 include MLP; //Import MLP that will be used 3 // Load the data that will be used for training 4 Data x_train = data_read(training_samples); 5 Data y_train = data_read(training_results); 6 // Deﬁne the set of transformers to use (non empty set) 7 Set models = Set of transformers in ensemble; 8 List train_embeddings = []; 9 // For each transformer, train the weak learner associated 10 foreach transformer model in models do 11 MLP clf; 12 List embbedding = model.generate_embeddings(x_train); 13 clf.train(embbedding,y_train); 14 // Store the created embeddings created by the weak learner 15 train_embeddings.append(clf.predict(embbedding)) 16 // Concatenate the outputs of each sample from the weak learners 17 train_embeddings = concatenate_elementwise(train_embeddings) 18 // Use that to train the ﬁnal meta-learning 19 MLP clf; 20 clf.train(train_embeddings, y_train) In Figure 6, we show the accuracy evolution plot of the final meta-learning. This network ﬁnished training with only two epochs, which is expected given the difference in size of the number of free parameters in the network and the size of the training size. Figure 6. Validation curves for the ﬁnal meta-learning. Table 3 shows the obtained results using the sklearn built-in functions. We include the results of F1 score, precision, and recall from our model, both the micro average (i.e., calculating the number of positives and negatives globally among all classes) and the macro average (i.e., calculating the metric for each label and then ﬁnding their unweighted mean (https://scikit-learn.org/stable/modules/generated/sklearn.metrics.f1_score.html) (accessed on 18 December 2020)). 5. Experimental Evaluation of the NLP Transformer Model The best results are in precision, which means that the model is good at avoiding false positives. The F1 scores show an overall ability for detecting true positives, up to 76%. However, it struggles when it comes to accurately detecting negative values, as shown by the recall scores. This statement is further conﬁrmed by looking at the results of the confusion matrix in Table 4, that shows that the amount of false positives produced is consistently lower than the number of false negatives produced. Figure 6. Validation curves for the ﬁnal meta-learning. Figure 6. Validation curves for the ﬁnal meta-learning. Figure 6. Validation curves for the ﬁnal meta-learning. Table 3 shows the obtained results using the sklearn built-in functions. We include the results of F1 score, precision, and recall from our model, both the micro average (i.e., calculating the number of positives and negatives globally among all classes) and the macro average (i.e., calculating the metric for each label and then ﬁnding their unweighted mean (https://scikit-learn.org/stable/modules/generated/sklearn.metrics.f1_score.html) (accessed on 18 December 2020)). The best results are in precision, which means that the model is good at avoiding false positives. The F1 scores show an overall ability for detecting true positives, up to 76%. However, it struggles when it comes to accurately detecting negative values, as shown by the recall scores. This statement is further conﬁrmed by looking at the results of the confusion matrix in Table 4, that shows that the amount of false positives produced is consistently lower than the number of false negatives produced. Sensors 2021, 21, 1322 14 of 19 The relatively big difference between macro and micro scores may be caused by a class imbalance in our training set. Table 2 shows that the labels surprise and trust have around one tenth of the samples that the most represented one (i.e., love) has. Thus, the values from the confusion matrix in Table 4 show that these poorly represented labels are the ones with the least amount of positive labels predicted, meaning that our model is not learning any pattern to classify them and is labeling most of the samples as not containing those emotions. Accuracy on the other hand seems to indicate that our model, on average, strays around 47% from the expected emotions per instance. Table 3. Overall results of the the model on the 2018-E-c-En-test-gold dataset. 6. Discussion The ﬁrst version of our framework, as a proof-of-concept, demonstrates the feasibility and suitability of a robot system able to recognize emotions from interactions with humans, which can be used to adapt the robot’s behavior and to instantiate a semantic repository, for further analysis. This experience also gives the opportunity to extract its current limitations and some lessons learned. 5. Experimental Evaluation of the NLP Transformer Model Average Type F1 Score Precision Recall Accuracy Micro 0.664 0.761 0.589 0.535 Macro 0.481 0.650 0.432 ts of the confusion matrix for each emotion. We show the percentage of true negatives (TN), Table 3. Overall results of the the model on the 2018-E-c-En-test-gold dataset. Table 4. Tabulated results of the confusion matrix for each emotion. We show the percentage of true negatives (TN), false positives (FP), false negatives (FN), and true positives (TP) per label. ble 4. Tabulated results of the confusion matrix for each emotion. We show the percentage of true neg lse positives (FP), false negatives (FN), and true positives (TP) per label. Anger Anticip. Disgust Fear Joy Love Opt. Pessim. Sadness Surprise Trust TN 58.852 85.947 58.515 83.431 50.967 80.853 55.784 86.683 63.025 94.692 95.121 FP 7.364 1.013 7.763 1.688 4.787 3.314 9.144 1.810 7.518 0.092 0.184 FN 9.481 12.059 9.604 6.935 10.954 8.315 10.341 9.880 12.010 5.002 4.664 TP 24.310 0.982 24.118 7.947 33.292 7.518 24.732 1.626 17.367 0.216 0.031 Compared with the NVIDIA-AI model, the results presented in Table 5 show that our model is behind the other by at most 7% across all evaluated metrics. This leaves us at a good starting point as. Unlike the NVIDIA-AI model which was ﬁne tuned for the tweet dataset, ﬁne-tuning our Transformers for this task and possibly changing our MLP classiﬁers for more reﬁned models could lead to better results. We explore these possibilities further in the Discussion section. Table 5. Comparison of our model with NVIDIA-AI. Model Accuracy Micro F1 Macro F1 Stacked Ensemble (ours) 0.535 0.664 0.481 NVIDIA-AI 0.577 0.690 0.561 6 Discussion Table 5. Comparison of our model with NVIDIA-AI. 6.2. Extending the Framework’s Functionalities Most social robots have the ability to record video and to capture images, along with the ability to record audio. Hence, the framework can be extended to be able to perform a multi-modal classiﬁcation, considering different modalities, such as text, voice, face expressions, and body postures and gestures. Although this modiﬁcation implies the design of a more complex classiﬁcation model, the results would be improved and offer a wide range of applicability in the area of social robotics. As developed, our framework allows for a straightforward way to add different media to the learning process, as it would be a matter of simply creating a classiﬁer for it and adding it to the stack, making it simple to continue development on it. This framework is mainly useful for two purposes: enhancing social behavior for service robots and collecting data for more complex analysis. The ontological repository enables enlarging the possibilities to improve both purposes. It can be used as direct feedback to record internal statistics of what emotions are evoked by entities (e.g., pieces of art in the museum, plates in a restaurant, and candidates in elections). That way, the robot can adjust its tone and interactions to ﬁt the mood according to the speciﬁc entity, instead of using a default tone. g Moreover, the capability of social robots in analyzing texts enable the opportunity to analyze social media to infer what types of people come to the place and what types of stimuli they are attracted to. Thus, with the gathered data and the inference capability of ontologies, the knowledge can be used for decision making. For example, in the museum case, we can obtain a clear rating of artworks, which can help the museum managers decide what pieces to display at any given moment or to decide how to dispose exhibitions. As it is, this version of our framework can complement chatbot systems by guiding responses according to the detected emotion. 6.1. Improvements for the Classiﬁcation Model The classiﬁcation method can be improved in several aspects. One of them is in the transformers. The ones we used were ﬁne-tuned for semantic textual similarity of tweets, in which a lot of emojis and abbreviations are present. Conversations in museums are quite different; thus, a more appropriate dataset, with more appropriate texts, can be used to ﬁne-tune the transformers. As explained in Section 5, an imbalance in the dataset generates some problems when it came to training the model, as some emotions could not be correctly detected. While in single-label classiﬁcation problems, there have been many techniques developed Sensors 2021, 21, 1322 15 of 19 15 of 19 to mitigate this, in the multi-label case, there are not many solutions available to solve this. For example, one common technique in the single-label case is to randomly upsample the data. In the multi-label case, doing this could generate more instances of the other labels, generating even more imbalance. However, there have been some algorithms developed in recent years to approach this problem that were not used in this instance of our framework, such as multilabel synthetic minority over-sampling technique (MLSMOTE), that we plan to try in the future. y The ensemble method and the classiﬁers used for the weak learners could be also improved. As mentioned before, both boosting and bagging are other ensemble methods that have shown good performance in other tasks when used with transformers [33,34]; thus, it would be worth evaluating these methods for our model. We used MLP for the classiﬁers, as they tend to perform well in some classiﬁcation tasks and are easy to use for multi-label classiﬁcation. Nevertheless, there have been works attempting to classify sentences with other architectures, such as CNNs [50]; hence, this is another part of the current model that could be tweaked. 6.3. HRI Improved with Emotion Detection In some scenarios for service robots, they have to choose the person to interact with and start the interaction; thus, a selection algorithm is needed. We suggest a selection algorithm tending to choose people that are alone or in “small" groups (“small", in this case is implementation dependent), since this makes it easier for social robots to attract their attention. Selecting a small group of people should also help the speech-to-text component of the framework as it would mean less chatter noise around the robot when it records the person’s speech. Once the person or group of persons is selected, the robot should try to start a conversation. If the robot fails to engage a conversation with the user because, for example, the user ignored it, then it can repeat this step until it ﬁnds a person to engage with. Once this happens, a conversation is started. Initially, the robot should start with a neutral Sensors 2021, 21, 1322 16 of 19 tone for all engaged persons, but once it has obtained enough data regarding the emotions people have within its context, it can adjust itself to make the interaction as natural as possible. For example, suppose this framework is implemented in social robots in an amusement park that has a roller coaster and a haunted house. Initially, all robots would start with neutral tones, but eventually, after engaging with many excited people, the ones on the roller coaster, they will adjust their tone to act excited and energetic as well. Meanwhile, after engaging with many scared people, the ones in the haunted house, they should start to act afraid as well. 6.4. EMONTO Extensibility The output of the classiﬁcation process strongly depends on the emotion represen- tation model. For example, Ekman’s basic emotions model [22] supports a one versus all classiﬁcation problem, while Plutchik’s wheel of emotions [23] implies a multi-label classiﬁcation problem for a subset of possible emotions or a polarity problem among each of the axes of the wheel. EMONTO supports both approaches and can be extended with other emotion models without affecting the implementation of the framework. Moreover, the entities Person and Object of EMONTO can be replaced by entities deﬁned in user-proﬁle ontologies for Person and speciﬁc domain ontologies representing entities to which the emotions can be associated (e.g., artworks, museums, and dish) for Object. Moreover, EMONTO can be extended with other ontologies that model the speciﬁc task of the service robots (e.g., SLAM ontologies). Hence, EMONTO is ﬂexible and extensible for any scenario in social robotics. Author Contributions: Conceptualization: Y.C., J.D.-A., and I.D.; data curation: W.G. and E.L.-S.; methodology: Y.C., J.D.-A., I.D., and W.G.; software: W.G. and E.L.-S.; validation, W.G.; investigation: W.G. and E.L.-S.; writing—original draft preparation: W.G.; writing—review and editing: Y.C., J.D.-A., I.D., W.G., E.L.-S., and C.S.-L. All authors have read and agreed to the published version of the manuscript. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 2014, 6, 954–967. [CrossRef] 10 B t l F P tti V O t l b d ff ti d l t i t k i th i l ti b I f P M 2016 pp 9. Cavallo, F.; Limosani, R.; Manzi, A.; Bonaccorsi, M.; Esposito, R.; Di Rocco, M.; Pecora, F.; Teti, G.; Safﬁotti, A.; Dario, P. Development of a socially believable multi-robot solution from town to home. Cogn. Comput. 2014, 6, 954–967. [CrossRef] 10. Bertola, F.; Patti, V. Ontology-based affective models to organize artworks in the social semantic web. Inf. Process. Manag. 2016, 52 139–162 [CrossRef] 11. Tsiourti, C.; Weiss, A.; Wac, K.; Vincze, M. Multimodal integration of emotional signals from voice, body, and context: Effects of (in) congruence on emotion recognition and attitudes towards robots. Int. J. Soc. Robot. 2019, 11, 555–573. [CrossRef] 12. Sailunaz, K.; Dhaliwal, M.; Rokne, J.; Alhajj, R. Emotion detection from text and speech: A survey. Soc. Netw. Anal. Min. 2018, 8, 28. [CrossRef] 11. Tsiourti, C.; Weiss, A.; Wac, K.; Vincze, M. Multimodal integration of emotional signals from voice, body, and context: Effects of (in) congruence on emotion recognition and attitudes towards robots. Int. J. Soc. Robot. 2019, 11, 555–573. [CrossRef] ( ) g g J , , [ ] 12. Sailunaz, K.; Dhaliwal, M.; Rokne, J.; Alhajj, R. Emotion detection from text and speech: A survey. Soc. Netw. Anal. Min. 2018, 8, 28. [CrossRef] Martínez-Barco, P. Emotion Detection from text: A Survey. In Proceedings of the Workshop on Natural Langu L.; Martínez-Barco, P. Emotion Detection from text: A Survey. In Proceedings of the Workshop on Natural La on Systems Research Working Days (JISIC), Quito, Ecuador, 20–24 October 2014; pp. 37–43. [CrossRef] 13. Canales, L.; Martínez-Barco, P. Emotion Detection from text: A Survey. In Proceedings of the Workshop on Natural Language 5th Information Systems Research Working Days (JISIC), Quito, Ecuador, 20–24 October 2014; pp. 37–43. [CrossRef] Systems Research Working Days (JISIC), Quito, Ecuador, 20–24 October 2014; pp. 37–43. [CrossRef] y g y pp 14. Seyeditabari, A.; Tabari, N.; Zadrozny, W. Emotion Detection in Text: A Review. arXiv 2018, arXiv:1806 A.; Tabari, N.; Zadrozny, W. Emotion Detection in Text: A Review. arXiv 2018, arXiv:1806.00674. 15. Kant, N.; Puri, R.; Yakovenko, N.; Catanzaro, B. Practical Text Classiﬁcation with Large Pre-Trained Language Models. arXiv 2018, arXiv:1812.01207. 16. Cornejo-Lupa, M.A.; Ticona-Herrera, R.P.; Cardinale, Y.; Barrios-Aranibar, D. References In Emotion Recognition and Understanding for Emotional Human-Robot Interaction Systems; Springer: Berlin/Heidelberg, Germany, 2021; 3. Karar, A.; Said, S.; Beyrouthy, T. Pepper Humanoid Robot as a Service Robot: A Customer Approach. In Proceedings of the 2019 3rd International Conference on Bio-engineering for Smart Technologies (BioSMART), Paris, France, 24–26 April 2019; pp. 1–4. 4. Li, D.; Rau, P.P.; Li, Y. A cross-cultural study: Effect of robot appearance and task. Int. J. Soc. Robot. 2010, 2, 175–186. [CrossRef] 5. Chen, L.; Wu, M.; Pedrycz, W.; Hirota, K. Two-Layer Fuzzy Multiple Random Forest for Speech Emotion Recognition. In Emotion Recognition and Understanding for Emotional Human-Robot Interaction Systems; Springer: Berlin/Heidelberg, Germany, 2021; pp. 77–89. 5. Chen, L.; Wu, M.; Pedrycz, W.; Hirota, K. Two-Layer Fuzzy Multiple Random Forest for Speech Emotion Recognition. In Emotion Recognition and Understanding for Emotional Human-Robot Interaction Systems; Springer: Berlin/Heidelberg, Germany, 2021; pp. 77–89. pp 6. Castillo, J.C.; Castro-González, Á.; Alonso-Martín, F.; Fernández-Caballero, A.; Salichs, M.Á. Emotion detection and regulation from personal assistant robot in smart environment. In Personal Assistants: Emerging Computational Technologies; Springer: Berlin/Heidelberg, Germany, 2018; pp. 179–195. g y pp 7. Zheng, L.; Li, Q.; Ban, H.; Liu, S. Speech emotion recognition based on convolution neural network combined with random forest. In Proceedings of the 2018 Chinese Control And Decision Conference (CCDC), Shenyang, China, 9–11 June 2018; pp. 4143–4147. 8 Lytridis C ; Vrochidou E ; Kaburlasos V Emotional speech recognition toward modulating the behavior of a social robot 7. Zheng, L.; Li, Q.; Ban, H.; Liu, S. Speech emotion recognition based on convolution neural network combined with random forest. In Proceedings of the 2018 Chinese Control And Decision Conference (CCDC), Shenyang, China, 9–11 June 2018; pp. 4143–4147. 8. Lytridis, C.; Vrochidou, E.; Kaburlasos, V. Emotional speech recognition toward modulating the behavior of a social robot. In Proceedings of the JSME Annual Conference on Robotics and Mechatronics (Robomec), Hiroshima, Japan, 5–8 June 2018; pp. 1A1–B14. g ( ), y g, , J ; pp 8. Lytridis, C.; Vrochidou, E.; Kaburlasos, V. Emotional speech recognition toward modulating the behavior of a social robot. In Proceedings of the JSME Annual Conference on Robotics and Mechatronics (Robomec), Hiroshima, Japan, 5–8 June 2018; pp. 1A1–B14. pp 9. Cavallo, F.; Limosani, R.; Manzi, A.; Bonaccorsi, M.; Esposito, R.; Di Rocco, M.; Pecora, F.; Teti, G.; Safﬁotti, A.; Dario, P. Development of a socially believable multi-robot solution from town to home. Cogn. Comput. Data Availability Statement: Data available in a publicly accessible repository. Acknowledgments: The authors thank Jesamin Zevallos, Percy Maldonado, and Anabel Paredes who participated in an initial version of EMONTO as well as Betzabeth Jimenez for her participation in the elaboration of the images. Funding: This research was supported by the FONDO NACIONAL DE DESARROLLO CIENTÍFICO, TECNOLÓGICO Y DE INNOVACIÓN TECNOLÓGICA—FONDECYT as an executing entity of CONCYTEC under grant agreement No. 01-2019-FONDECYT-BM-INC.INV in the project RUTAS: Robots for Urban Tourism Centers, Autonomous and Semantic-based. Funding: This research was supported by the FONDO NACIONAL DE DESARROLLO CIENTÍFICO, TECNOLÓGICO Y DE INNOVACIÓN TECNOLÓGICA—FONDECYT as an executing entity of CONCYTEC under grant agreement No. 01-2019-FONDECYT-BM-INC.INV in the project RUTAS: Robots for Urban Tourism Centers, Autonomous and Semantic-based. Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data available in a publicly accessible repository. References 1. Müller, C. Automation Strategies Drive 12% Increase in Number of Robots at Work Globally. 2020. Available online: https: //ifr.org/post/automation-strategies-drive-12-increase-in-number-of-robots-at-work-globally (accessed on 18 December 2020). 2. Abubakar, S.; Das, S.K.; Robinson, C.; Saadatzi, M.N.; Logsdon, M.C.; Mitchell, H.; Chlebowy, D.; Popa, D.O. ARNA, a Service robot for Nursing Assistance: System Overview and User Acceptability. In Proceedings of the 2020 IEEE 16th International Conference on Automation Science and Engineering (CASE) Hong Kong China 20 21 August 2020; pp 1408 1414 1. Müller, C. Automation Strategies Drive 12% Increase in Number of Robots at Work Globally. 2020 //ifr org/post/automation strategies drive 12 increase in number of robots at work globally (access 1. Müller, C. Automation Strategies Drive 12% Increase in Number of Robots at Work Globally. 2020. Available online: https: //ifr.org/post/automation-strategies-drive-12-increase-in-number-of-robots-at-work-globally (accessed on 18 December 2020). 1. Müller, C. Automation Strategies Drive 12% Increase in Number of Robots at Work Globally. 2020. Available online: https: //ifr.org/post/automation-strategies-drive-12-increase-in-number-of-robots-at-work-globally (accessed on 18 December 2020). 2. Abubakar, S.; Das, S.K.; Robinson, C.; Saadatzi, M.N.; Logsdon, M.C.; Mitchell, H.; Chlebowy, D.; Popa, D.O. ARNA, a Service robot for Nursing Assistance: System Overview and User Acceptability. In Proceedings of the 2020 IEEE 16th International Conference on Automation Science and Engineering (CASE), Hong Kong, China, 20–21 August 2020; pp. 1408–1414. g p g g y ( ) 2. Abubakar, S.; Das, S.K.; Robinson, C.; Saadatzi, M.N.; Logsdon, M.C.; Mitchell, H.; Chlebowy, D.; Popa, D.O. ARNA, a Service robot for Nursing Assistance: System Overview and User Acceptability. In Proceedings of the 2020 IEEE 16th International Conference on Automation Science and Engineering (CASE), Hong Kong, China, 20–21 August 2020; pp. 1408–1414. g g g g g pp 3. Karar, A.; Said, S.; Beyrouthy, T. Pepper Humanoid Robot as a Service Robot: A Customer Approach. In Proceedings of the 2019 3rd International Conference on Bio-engineering for Smart Technologies (BioSMART), Paris, France, 24–26 April 2019; pp. 1–4. 3. Karar, A.; Said, S.; Beyrouthy, T. Pepper Humanoid Robot as a Service Robot: A Customer Approach. In Proceedings of the 2019 3rd International Conference on Bio-engineering for Smart Technologies (BioSMART), Paris, France, 24–26 April 2019; pp. 1–4. 4. Li, D.; Rau, P.P.; Li, Y. A cross-cultural study: Effect of robot appearance and task. Int. J. Soc. Robot. 2010, 2, 175–186. [CrossRef] 5. Chen, L.; Wu, M.; Pedrycz, W.; Hirota, K. Two-Layer Fuzzy Multiple Random Forest for Speech Emotion Recognition. 19. Sauer, J.; Schmutz, S.; Sonderegger, A.; Messerli, N. Social stress and performance in human-machine interaction: A neglected research ﬁeld. Ergonomics 2019, 62, 1377–1391. [CrossRef] 7. Conclusions The need for interaction between machines and humans is becoming more common in people’s daily lives. The effort to improve these relationships through the interpretation of social behaviors are more and more frequent among research developed in the area of social robotics. The interpretation of the feelings of a human being is an important tool that a robot must know how to use to enact correct behavior in a social environment. Our framework is a clear contribution in this area, since it allows robots to interpret a person’s basic feelings and emotion recognition algorithms. Besides that, the information gathered by the robots and the outputs of the classiﬁcation process can ﬁnally be organized in an ontology, leading to more complex analyses and uses of the data. The results obtained with a ﬁrst implementation of our framework, as a proof-of-concept, show that the integration of speech-to-text and emotion detection algorithms with an ontology was a success despite the fact that there is still the possibility of being improved. Our future research is focused on improvements of the framework, starting with the suggestions from Section 6 to provide a robust architecture in the community to develop third-party applications in social robotics. We realize the need to incorporate more interpretation characteristics that can complement the detection of a person’s feelings in our framework, such as detection of the face, posture, and context in which the person is involved. Thus, a new classiﬁcation model that integrates all these characteristics to have more veracity in detecting the feeling will be developed. Once this extension is done, we could make another comparison to Megatron-LM to put in perspective how much value these new features add to our framework. In addition, the framework could be part of an autonomous navigation system applied to a social robot, which would complement decision-making in these processes as well as assuming better socially acceptable behavior towards humans. Author Contributions: Conceptualization: Y.C., J.D.-A., and I.D.; data curation: W.G. and E.L.-S.; methodology: Y.C., J.D.-A., I.D., and W.G.; software: W.G. and E.L.-S.; validation, W.G.; investigation: W.G. and E.L.-S.; writing—original draft preparation: W.G.; writing—review and editing: Y.C., J.D.-A., I.D., W.G., E.L.-S., and C.S.-L. All authors have read and agreed to the published version of the manuscript. Sensors 2021, 21, 1322 17 of 19 17 of 19 References HuggingFace’s Transformers: State-of-the-art Natural Language Processing arXiv 2020 arXiv:1910 03771 g y pp 36. Wolf, T.; Debut, L.; Sanh, V.; Chaumond, J.; Delangue, C.; Moi, A.; Cistac, P.; Rault, T.; Louf, R.; Funtowicz, M.; et al. HuggingFace’s Transformers: State-of-the-art Natural Language Processing. arXiv 2020, arXiv:1910.03771. 36. Wolf, T.; Debut, L.; Sanh, V.; Chaumond, J.; Delangue, C.; Moi, A.; Cistac, P.; Rault, T.; Louf, R.; Funtow Transformers: State-of-the-art Natural Language Processing. arXiv 2020, arXiv:1910.03771. ansformers: State-of-the-art Natural Language Processing. arXiv 2020, arXiv:1910.03771. ers, N.; Gurevych, I. Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. In Proceedin rence on Empirical Methods in Natural Language Processing Hong Kong China 3 7 November 2019 37. Reimers, N.; Gurevych, I. Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing, Hong Kong, China, 3–7 November 2019. 38. Ayari, N.; Abdelkawy, H.; Chibani, A.; Amirat, Y.Y. Towards Semantic Multimodal Emotion Recognition for Enhancing Assistive Services in Ubiquitous Robotics. In Proceedings of the AAAI 2017 Fall Symposium Series, Arlington, VA, USA, 9–11 November 2017; pp. 2–9. pp 39. Azevedo, H.; Romero, R.A.F.; Ribeiro Belo, J.P. Reducing the gap between cognitive and robotic systems. In Proceedings of the 2017 26th IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN), Lisbon, Portugal, 28–31 August 2017; pp. 1049–1054. [CrossRef] g pp 40. Azevedo, H.; Belo, J.P.; Romero, R. Using Ontology as a Strategy for Modeling the Interface Between the Cognitive and Robotic Systems. J. Intell. Robot. Syst. 2020, 99, 431–449. [CrossRef] 41. Mojarad, R.; Attal, F.; Chibani, A.; Fiorini, S.R.; Amirat, Y. Hybrid Approach for Human Activity Recognition by Ubiquitous Robots. In Proceedings of the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Madrid, Spain, 1–5 October 2018; pp. 5660–5665. [CrossRef] pp 42. Jeon, H.; Kim, T.; Choi, J. Ontology-Based User Intention Recognition for Proactive Planning of Intelligent Robot Behavior. In Proceedings of the 2008 International Conference on Multimedia and Ubiquitous Engineering (MUE 2008), Busan, Korea, 24–26 April 2008; pp. 244–248. [CrossRef] 43. Fukuda, H.; Mori, S.; Kobayashi, Y.; Kuno, Y.; Kachi, D. Object recognition based on human description ontology for service robots. In Proceedings of the IECON 2014—40th Annual Conference of the IEEE Industrial Electronics Society, Dallas, TX, USA, 29 October–1 November 2014; pp. 4051–4056. [CrossRef] pp 44. Shakhovska, N.; Basystiuk, O.; Shakhovska, K. Development of the Speech-to-Text Chatbot Interface Based on Google API. References In Proc ti l W k h S ti E l ti (S E l 2018) N O l LA USA 5 6 J 2018 mad, S.M.; Bravo-Marquez, F.; Salameh, M.; Kiritchenko, S. SemEval-2018 Task 1: Affect in Tweets. In Proceed onal Workshop on Semantic Evaluation (SemEval-2018), New Orleans, LA, USA, 5–6 June 2018. Mohammad, S.M.; Bravo-Marquez, F.; Salameh, M.; Kiritchenko, S. SemEval-2018 Task 1: Affect in Tweets. International Workshop on Semantic Evaluation (SemEval-2018), New Orleans, LA, USA, 5–6 June 2018. 29. Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A.N.; Kaiser, L.; Polosukhin, I. Attention Is All You Need. arXiv 2017, arXiv:1706.03762 30. Raffel, C.; Shazeer, N.; Roberts, A.; Lee, K.; Narang, S.; Matena, M.; Zhou, Y.; Li, W.; Liu, P.J. Exploring the Limits of Transfer Learning with a Uniﬁed Text-to-Text Transformer. arXiv 2020, arXiv:1910.10683. g 31. McCann, B.; Keskar, N.S.; Xiong, C.; Socher, R. The Natural Language Decathlon: Multitask Learning as Question Answering. arXiv 2018, arXiv:1806.08730. 32. Devlin, J.; Chang, M.W.; Lee, K.; Toutanova, K. BERT: Pre-training of Deep Bidirectional Transformers for Language Understand- ing. arXiv 2019, arXiv:1810.04805. g 33. Huang, T.; She, Q.; Zhang, J. BoostingBERT:Integrating Multi-Class Boosting into BERT for N arXiv:2009.05959. .; Zhang, J. BoostingBERT:Integrating Multi-Class Boosting into BERT for NLP Tasks. arXiv 2020 33. Huang, T.; She, Q.; Zhang, J. BoostingBERT:Integrating Multi-Class Boosting into BERT for NLP Tasks. arXiv 2020, arXiv:2009.05959. arXiv:2009.05959. 34. Risch, J.; Krestel, R. Bagging BERT models for robust aggression identiﬁcation. In Proceedings of the Second Workshop on Trolling, Aggression and Cyberbullying, Marseille, France, 11–16 May 2020; pp. 55–61. 34. Risch, J.; Krestel, R. Bagging BERT models for robust aggression identiﬁcation. In Proceedings of t Trolling, Aggression and Cyberbullying, Marseille, France, 11–16 May 2020; pp. 55–61. g gg y y g y 35. Liu, S.; Liu, S.; Ren, L. Trust or Suspect? An Empirical Ensemble Framework for Fake News Classiﬁcation. In Proceedings of the 12th ACM International Conference on Web Search and Data Mining, Melbourne, Australia, 11–15 February 2019; pp. 11–15. 35. Liu, S.; Liu, S.; Ren, L. Trust or Suspect? An Empirical Ensemble Framework for Fake News Classiﬁcation. In Proceedings of the 12th ACM International Conference on Web Search and Data Mining, Melbourne, Australia, 11–15 February 2019; pp. 11–15. 36. Wolf, T.; Debut, L.; Sanh, V.; Chaumond, J.; Delangue, C.; Moi, A.; Cistac, P.; Rault, T.; Louf, R.; Funtowicz, M.; et al. References A Survey of Ontologies for Simultaneous Localization and Mapping in Mobile Robots. ACM Comput. Surv. (CSUR) 2020, 53, 1–26. 17. Pinto-De la Gala, A.; Cardinale, Y.; Dongo, I.; Ticona-Herrera, R. Towards an Ontology for Urban Tourism. In Proceedings of the 36th Annual ACM Symposium on Applied Computing, Gwangju, Korea, 22–26 March 2021; ACM: New York, NY, USA, 2021. 18. Cavallo, F.; Semeraro, F.; Fiorini, L.; Magyar, G.; Sinˇcák, P.; Dario, P. Emotion modelling for social robotics applications: A review. J. Bionic Eng. 2018, 15, 185–203. [CrossRef] 19. Sauer, J.; Schmutz, S.; Sonderegger, A.; Messerli, N. Social stress and performance in human-machine interaction: A neglected research ﬁeld. Ergonomics 2019, 62, 1377–1391. [CrossRef] Sensors 2021, 21, 1322 18 of 19 18 of 19 20. Li, Y.; Ishi, C.T.; Ward, N.; Inoue, K.; Nakamura, S.; Takanashi, K.; Kawahara, T. Emotion recognition by combining prosody and sentiment analysis for expressing reactive emotion by humanoid robot. In Proceedings of the 2017 Asia-Paciﬁc Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC), Kuala Lumpur, Malaysia, 12–15 December 2017; pp. 1356–1359. pp 21. Cowie, R.; Cornelius, R.R. Describing the Emotional States That Are Expressed in Speech. Speech Commun. 2003, 40, 5–32. [CrossRef] [ ] 22. Ekman, P. Basic Emotions. In Handbook of Cognition and Emotion; John Wiley & Sons, Ltd.: Hoboken, NJ, USA, 1999; Chapter 3, pp. 45–60. [CrossRef] 23. Plutchik, R. Emotions: A general psychoevolutionary theory. Approaches Emot. 1984, 1984, 197–219. g p y y y pp 24. Hochreiter, S.; Schmidhuber, J. Long Short-Term Memory. Neural Comput. 1997, 9, 1735–1780. [CrossRef] [PubMed] 25. Krause, B.; Lu, L.; Murray, I.; Renals, S. Multiplicative LSTM for sequence modelling. arXiv 2017, arXiv 26. Cho, K.; van Merrienboer, B.; Gulcehre, C.; Bahdanau, D.; Bougares, F.; Schwenk, H.; Bengio, Y. Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation. arXiv 2014, arXiv:1406.1078. 26. Cho, K.; van Merrienboer, B.; Gulcehre, C.; Bahdanau, D.; Bougares, F.; Schwenk, H.; Bengio, Y. Learni using RNN Encoder-Decoder for Statistical Machine Translation. arXiv 2014, arXiv:1406.1078. 27. Cho, K.; van Merrienboer, B.; Bahdanau, D.; Bengio, Y. On the Properties of Neural Machine Translation: Encoder-Decoder Approaches. arXiv 2014, arXiv:1409.1259. pp 28. Mohammad, S.M.; Bravo-Marquez, F.; Salameh, M.; Kiritchenko, S. SemEval-2018 Task 1: Affect in Twe International Workshop on Semantic Evaluation (SemEval-2018), New Orleans, LA, USA, 5–6 June 201 mmad, S.M.; Bravo-Marquez, F.; Salameh, M.; Kiritchenko, S. SemEval-2018 Task 1: Affect in Tweets. 47. Lin, R.; Amith, M.T.; Liang, C.; Duan, R.; Chen, Y.; Tao, C. Visualized Emotion Ontology: A model for representing visual cues of emotions. BMC Med. Inform. Decis. Mak. 2018, 18, 101–113. [CrossRef] pp g p p p 50. Kim, Y. Convolutional Neural Networks for Sentence Classiﬁcation. arXiv 2014, arXiv:1408.5882. f [ ] 48. Plutchik, R. A general psychoevolutionary theory of emotion. In Theories of Emotion; Elsevier: Amsterdam, The Netherlands, 1980; pp. 3–33. pp 49. Katifori, A.; Golemati, M.; Vassilakis, C.; Lepouras, G.; Halatsis, C. Creating an Ontology for the User Proﬁle: Method and Applications. In Proceedings of the AI* AI Workshop RCIS, Ouarzazate, Morocco, 23–26 April 2007; pp. 407–412. g p p pp volutional Neural Networks for Sentence Classiﬁcation. arXiv 2014, arXiv:1408.5882. , ; , ; g, ; , ; , ; , gy p g emotions. BMC Med. Inform. Decis. Mak. 2018, 18, 101–113. [CrossRef] 48. Plutchik, R. A general psychoevolutionary theory of emotion. In Theories of Emotion; Elsevier: Amsterdam, The Netherlands, 1980; pp. 3–33. References In Proceedings of the MoMLeT 2019, Shatsk, Ukraine, 2–4 June 2019; pp. 212–221. uaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, P.; Weiss, R.; Dubourg, V. Machine Learning in Python. J. Mach. Learn. Res. 2011, 12, 2825–2830. 45. Pedregosa, F.; Varoquaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, et al. Scikit-learn: Machine Learning in Python. J. Mach. Learn. Res. 2011, 12, 2825–2830. g y 46. Grassi, M. Developing HEO Human Emotions Ontology. In Biometric ID Management and Multimodal Communication; Fierrez, J., Ortega-Garcia, J., Esposito, A., Drygajlo, A., Faundez-Zanuy, M., Eds.; Springer: Berlin/Heidelberg, Germany, 2009; pp. 244–251. Sensors 2021, 21, 1322 19 of 19 19 of 19
https://openalex.org/W3128599480	https://www.mdpi.com/2227-9032/9/3/287/pdf?version=1615173528	English	null	Primary School Teachers’ Experiences of Sexual Education in Spain. A Qualitative Study.	null	2,021	cc-by	9,939	Citation: Plaza-del-Pino, F.J.; Soliani, I.; Fernández-Sola, C.; Molina-García, J.J.; Ventura- Miranda, M.I.; Pomares-Callejón, M.Á.; López-Entrambasaguas, O.M.; Ruiz-Fernández, M.D. Primary School Teachers’ Perspective of Sexual Education in Spain. A Qualitative Study. Healthcare 2021, 9, 287. https://doi.org/10.3390/ healthcare9030287 Keywords: sexual education; school; teachers’ perspective; qualitative research Received: 3 February 2021 Accepted: 2 March 2021 Published: 5 March 2021 Article Primary School Teachers’ Perspective of Sexual Education in Spain. A Qualitative Study Citation: Plaza-del-Pino, F.J.; Soliani, I.; Fernández-Sola, C.; Molina-García, J.J.; Ventura- Miranda, M.I.; Pomares-Callejón, M.Á.; López-Entrambasaguas, O.M.; Ruiz-Fernández, M.D. Primary School Teachers’ Perspective of Sexual Education in Spain. A Qualitative Study. Healthcare 2021, 9, 287. https://doi.org/10.3390/ healthcare9030287 Academic Editor: Jitendra Singh Received: 3 February 2021 Accepted: 2 March 2021 Published: 5 March 2021 healthcare healthcare healthcare Article Primary School Teachers’ Perspective of Sexual Education in Spain. A Qualitative Study Fernando Jesús Plaza-del-Pino 1 , Isabelle Soliani 2, Cayetano Fernández-Sola 1,3,* , Joaquín Jesús Molina-García 4, María Isabel Ventura-Miranda 1 , María Ángeles Pomares-Callejón 1, Olga María López-Entrambasaguas 5 and María Dolores Ruiz-Fernández 1 1 Department of Nursing, Physiotherapy and Medicine, University of Almeria, 04120 Almeria, Spain; ferplaza@ual.es (F.J.P.-d.-P.); mvm737@ual.es (M.I.V.-M.); mpc630@ual.es (M.Á.P.-C.); mrf757@ual.es (M.D.R.-F.) 1 Department of Nursing, Physiotherapy and Medicine, University of Almeria, 04120 Almeria, Spain; ferplaza@ual.es (F.J.P.-d.-P.); mvm737@ual.es (M.I.V.-M.); mpc630@ual.es (M.Á.P.-C.); mrf757@ual.es (M.D.R.-F.) 2 Almeria Colegio Internacional, 04007 Almeria, Spain; isabellesoliani@gmail.com 3 2 Almeria Colegio Internacional, 04007 Almeria, Spain; isabellesoliani@gmail.com 3 Facultad de Ciencias de la Salud, Universidad Autónoma de Chile, Santiago 7500000, Chile 4 Facultad de Ciencias de la Educación, University of Almeria, 04120 Almeria, Spain; joaquinmolinag@gmail.com 5 Departament of Nursing, University of Jaén, 23071 Jaén, Spain; omlopez@ujaen.es * Correspondence: cfernan@ual.es; Tel.: +34-950-21-4570 g 4 Facultad de Ciencias de la Educación, University of Almeria, 04120 Almeria, Spain; joaquinmolinag@gmail.com j q g g 5 Departament of Nursing, University of Jaén, 23071 Jaén, Spain; omlopez@ujaen.es 5 Departament of Nursing, University of Jaén, 23071 Jaén, Spain; omlopez@ujaen.es * Correspondence: cfernan@ual.es; Tel.: +34-950-21-4570 5 Departament of Nursing, University of Jaén, 23071 Jaén, Spain; omlopez@ujaen.es * Correspondence: cfernan@ual.es; Tel.: +34-950-21-4570 * Correspondence: cfernan@ual.es; Tel.: +34-950-21-4570 Abstract: Sexual education is a part of the teaching-learning process that addresses cognitive, psy- chological, physical and social aspects of sexuality. The purpose of sexual education is to provide people with knowledge, abilities, attitudes and values that will help them to have good sexual health, well-being and dignity. The objective of this study was to explore the perspective of primary school teachers regarding Sexual Education in school. A descriptive qualitative study was designed based on content thematic analysis. Fifteen open-ended interviews with primary school teachers were carried out, followed by inductive data analysis using ATLAS.ti software. Two key themes emerged from the analysis: “In search of a comprehensive approach to Sexual Education” and “Barriers to Sexual Education in schools: From the lack of training to fear of the families”. We conclude that despite the efforts to implement a comprehensive approach to Sexual Education that recognises sexuality as a right, primary school teachers face difﬁculties in delivering Sexual Education in schools due to a lack of training and the fear that parents will reject their children being spoken to about sexuality. 1. Introduction However, there are obstacles to SE, such as parents’ reluctance to talk openly about sexuality with their children [11,12], peer pressure [5,13] and the heavy inﬂuence of the media’s representation of sexuality [11]. p y Despite its proven beneﬁts [4,14,15], SE in schools remains an unresolved issue in many Western societies [16–18]. There is ongoing controversy regarding the ideal setting for this type of education to take place (i.e., at school, or home) [19] as well as its content, format and nature in relation to school [12,14,16]. In the school setting, the debate focuses on whether to include SE as a subject [1] or keep it as a transversal content [20]. Literature in the ﬁeld reports numerous innovative programmes that tend to fail due to primary school teachers’ fear of discussing the topic openly and ﬁrmly [18,21]. Some studies suggest a lack of knowledge, which is occasionally compensated for by the assistance of healthcare professionals [22] or medicine students [23]. In terms of content, some authors cite contents linked to health concerns related to sexuality—such as unwanted pregnancies [14,24] or the excessive weighting of reproductive health in Science [21]—which results in certain topics being excluded, such as sexuality as a right, [1] sexual diversity [25] or gender issues. In an attempt to overturn these shortcomings, various innovative programmes have been developed that have a holistic vision of SE in schools that goes beyond sexual reproduction, with the aim of addressing the body, feelings, etc. [21]. Likewise, there are SE programmes based on life skills [26] or positive values that offer students the conﬁdence and knowledge required to protect themselves and make informed decisions [15,22,23]. In Asturias, Spain, a programme developed by external experts had a temporary effect on protective behaviours, thus suggesting that one-off interventions have a limited impact [27]. p gg g p In some countries and various states of the United States, SE is obligatory in public schools and is delivered by primary school teachers or nurses [28,29]. In Europe, Finland and Sweden, SE is part of the school curriculum, which has led to positive developments in preventing risk and promoting sexual health, although society and students alike have deemed it insufﬁcient, inefﬁcient and unadapted [10,30]. 1. Introduction The United Nations Educational, Scientiﬁc and Cultural Organization (UNESCO) deﬁnes comprehensive sexuality education as part of the teaching-learning process that addresses cognitive, psychological, physical and social aspects of sexuality [1]. The purpose of this process is to provide people with knowledge, abilities, attitudes and values that will help them to have good sexual health, well-being and dignity; establish social and sexual relations based on respect; analyse the repercussions of one’s decisions on one’s own well-being and that of others; and understand how to protect one’s rights [1]. Moreover, the World Health Organization (WHO) deﬁnes sexual health as a state of physical, emotional, mental and social well-being in regard to sexuality, not just as the absence of illness or of dysfunction or weakness [2]. This “social” connotation of the concept of sexual health requires a positive and respectful approach to sexuality that takes into account the community and family context [3–5] and includes sexual relations, sexual pleasure and safe sexual experiences, free from coercion, discrimination and violence [2]. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). p Recent studies in Spain have brought to light the shortcomings of affective sexual education amongst school students, who access information about sexuality through social networks and internet pornography [6]. This can contribute to the early onset of sexual https://www.mdpi.com/journal/healthcare Healthcare 2021, 9, 287. https://doi.org/10.3390/healthcare9030287 Healthcare 2021, 9, 287 2 of 12 activity [7] and a distorted perception of relationships, contributing to the imitation of pre- dominant chauvinist and violent male attitudes reﬂected in the pornography industry [6]. activity [7] and a distorted perception of relationships, contributing to the imitation of pre- dominant chauvinist and violent male attitudes reﬂected in the pornography industry [6]. p g p y y The introduction of Sexual Education (SE) in schools has been linked to the decrease in behaviours that reﬂect inequality [8], fewer unstable relationships [9], positive develop- ment in terms of risk prevention and the promotion of sexual health [10] and a reduced association of sexual relations with other high-risk behaviours—such as drug consump- tion [9]. 2.2. Participants and Setting This study was carried out in four public schools in southern Spain. The schools had six years of primary education (1st to 6th grade of primary) with an average of two classes per year group and 60 primary school teachers (average of 15 per school). The participants were chosen using convenience sampling, with the following inclusion criteria: to be primary school teachers with teaching experience in ES, thus excluding those without experience in regulated education. Prior to the study, the participants were contacted by the researchers who carried out the interviews and were invited to participate in the study. The ﬁnal sample was composed of 15 active primary school teachers. Four teachers did not want to be interviewed, claiming a lack of time on the proposed dates, thus resulting in them being excluded from the process. The average age of the interviewees was 43.4 years (standard deviation 11.85) with average teaching experience of 17.8 years (standard deviation 10.41). The sociodemographic data can be found in Table 1. Table 1. Sociodemographic data of participants. Interview Code Age Gender Years of Experience E1 45 Woman 18 E2 60 Woman 34 E3 38 Man 10 E4 40 Woman 19 E5 35 Woman 11 E6 30 Man 9 E7 29 Man 4 E8 43 Woman 16 E9 51 Woman 26 E10 33 Woman 21 E11 48 Woman 17 E12 62 Man 40 E13 25 Woman 1 E14 55 Man 20 E15 57 Man 22 Table 1. Sociodemographic data of participants. 2.1. Design This is a descriptive qualitative study based on content thematic analysis [38]. This design is based on a naturalist research paradigm that aims to understand phenomena from the point of view of their protagonists [39]. 1. Introduction In the United Kingdom, Sexual Education in schools is not a homogenous issue; it varies from a lack of regulation in Scotland to being primarily focused on biological aspects (e.g., anatomy, infections) in England [31]. The Dutch SE programme is considered the most comprehensive and liberal in the world. It aims to foster self-sufﬁciency, self-respect and respect for others [32]. In Spain, teachers suggest working on coeducation as a cross-curricular content [33]. p gg g Concern surrounding pregnancy and sexually transmitted diseases continues to per- meate SE programmes, which tend to deﬁne sexual health as reproductive health [11,32]. This leads to parents having the idea that SE should promote abstinence as the best option for young people [34]. Sexually transmitted infections are a serious problem for sexual health due to the way in which sexual relations are taking place [35]. However, there is in- creasing emphasis placed on other aspects such as coeducation or studies about stereotypes. In fact, research conducted in Spain highlights the persistence of sexist behaviours and stereotypes amongst students in elementary education [33] and sexist prejudice amongst future primary school teachers [36]. Although there are studies that focus on the results of teaching innovation [21] as well as how to work with parents on the topic of SE [37], there is a scarcity of literature that delves into the perspectives of primary school teachers regarding SE in schools in Spain [33]. Exploring the perspectives of primary school teachers could contribute to understanding their own attitudes, prejudice, limitations and tools for teaching SE in schools [27,36]. The objective of this study is, therefore, to explore the perspective of primary school teachers regarding SE in schools in Spain. 3 of 12 Healthcare 2021, 9, 287 2.3. Data Collection Data Analysis ATLAS.ti software (ATLAS.ti Scientiﬁc Software Development GmbH, Berlin, Ger- many) was used for data analysis, integrating the transcriptions, the codifying and cate- gorization system and memos in a project. Thematic analysis was carried out following an inductive strategy (the themes are not based on a literary review but instead emerge from the data), in compliance with the steps described by Mayring (2000) and adapted for psychoeducational and health research [39,40]. First step: Selecting the object of analysis within a communication framework. The research group started by deﬁning their theoretical, professional or scientiﬁc stance [40]. In our study we used Pender’s Model of Health Promotion (MHP) based on the importance of developing the cognitive process in order to change behaviour [41] as our theoretical lens. The MHP allows for the understanding of human conduct related to health, pointing towards healthy behaviours that are not only based on preventing sexually transmitted infections or pregnancies but also, in line with the WHO [2], address pleasure and safe sexual experiences that are free from coercion, discrimination and violence. Second step: Preliminary analysis. In this stage, the transcripts were read with the aim of gaining a general understanding and holistic vision of the texts. Third step: Defining the units of analysis. Through a careful re-reading of each document, significant sentences or extracts were chosen as units of analysis. We aimed to select units of analysis sufﬁciently large to be considered as a whole but small enough to be a relevant meaning unit during the analysis process [42]. A total of 197 quotes were selected as units of analysis and were codiﬁed in the following phase. Fourth step: Codiﬁcation: establishing rules of analysis and classiﬁcation codes. The ‘open coding’ function of ATLAS.ti software was used to assign codes to the units of analysis, giving them meaning. The “insert comment code” function was applied to deﬁne the usage rules of each code as well as the requirements for successive codiﬁcation of units of analysis with each code. Table 2. Interview protocol. Stage of the Interview Subject Content and Example of Questions Introduction Our intention I am a member of a research group about Sexual education in schools. Knowing your perspectives could help and be useful to propose improvements in SE in Spanish schools. Information and ethical issues We need to record the conversation in order for the research team to analyze the data. 2.3. Data Collection Only the research team will have access to the recordings. Participation is totally voluntary and you can leave the study at any time you wish. Your identity will be protected, and your name and personal data will not be revealed. Consent Verbal acceptance of the participants and signing of the corresponding document. Beginning Introductory question As a teacher, tell me about the importance of teaching students about sexuality to you. Development Conversation guide Tell me about the contents that you think sexual education includes. Which content should be included in your opinion? How do you implement sexual education? Tell me about the advantages and disadvantages or obstacles of a subject about sexual education in schools, and as cross-curricular content. Closing Final question Is there anything else you would like to tell me? Appreciation Thank you for taking part. Your testimony will be used for the research study. We remain at your disposal if you need anything. You will receive the study upon completion. Table 2. Interview protocol. Content and Example of Questions 2.3. Data Collection Data collection took place between June and July 2017, via open-ended in-depth interviews. Participants were recruited via an email invitation, in which the conditions and objectives of the study were explained. Subsequently, an appointment was made with the individuals who gave consent to participate in the study. In order to ensure natural and open-ended interviews in which the participants felt they could express themselves openly, the researchers received training and practised the interview techniques before conducting the interviews. An interview protocol was used, providing information about the objective, ethical issues, consent and questions to guide the conversation (Table 2). The average duration of the interviews was 44 min. The interviews were recorded on a digital device with prior consent from the interviewee. They were transcribed immediately after taking place so that they could be revised and enhanced with the researcher’s notes. 4 of 12 Healthcare 2021, 9, 287 , , Table 2. Interview protocol. Stage of the Interview Subject Content and Example of Questions Introduction Our intention I am a member of a research group about Sexual education in schools. Knowing your perspectives could help and be useful to propose improvements in SE in Spanish schools. Information and ethical issues We need to record the conversation in order for the research team to analyze the data. Only the research team will have access to the recordings. Participation is totally voluntary and you can leave the study at any time you wish. Your identity will be protected, and your name and personal data will not be revealed. Consent Verbal acceptance of the participants and signing of the corresponding document. Beginning Introductory question As a teacher, tell me about the importance of teaching students about sexuality to you. Development Conversation guide Tell me about the contents that you think sexual education includes. Which content should be included in your opinion? How do you implement sexual education? Tell me about the advantages and disadvantages or obstacles of a subject about sexual education in schools, and as cross-curricular content. Closing Final question Is there anything else you would like to tell me? Appreciation Thank you for taking part. Your testimony will be used for the research study. We remain at your disposal if you need anything. You will receive the study upon completion. 2.4. 2.4. Data Analysis ATLAS.ti software (ATLAS.ti Scientiﬁc Software Development GmbH, Berlin, Ger- many) was used for data analysis, integrating the transcriptions, the codifying and cate- gorization system and memos in a project. Thematic analysis was carried out following an inductive strategy (the themes are not based on a literary review but instead emerge from the data), in compliance with the steps described by Mayring (2000) and adapted for psychoeducational and health research [39,40]. p y First step: Selecting the object of analysis within a communication framework. The research group started by deﬁning their theoretical, professional or scientiﬁc stance [40]. In our study we used Pender’s Model of Health Promotion (MHP) based on the importance of developing the cognitive process in order to change behaviour [41] as our theoretical lens. The MHP allows for the understanding of human conduct related to health, pointing towards healthy behaviours that are not only based on preventing sexually transmitted infections or pregnancies but also, in line with the WHO [2], address pleasure and safe sexual experiences that are free from coercion, discrimination and violence. Second step: Preliminary analysis. In this stage, the transcripts were read with the aim of gaining a general understanding and holistic vision of the texts. Third step: Defining the units of analysis. Through a careful re-reading of each document, significant sentences or extracts were chosen as units of analysis. We aimed to select units of analysis sufﬁciently large to be considered as a whole but small enough to be a relevant meaning unit during the analysis process [42]. A total of 197 quotes were selected as units of analysis and were codiﬁed in the following phase. Fourth step: Codiﬁcation: establishing rules of analysis and classiﬁcation codes. The ‘open coding’ function of ATLAS.ti software was used to assign codes to the units of analysis, giving them meaning. The “insert comment code” function was applied to deﬁne the usage rules of each code as well as the requirements for successive codiﬁcation of units of analysis with each code. Healthcare 2021, 9, 287 5 of 12 nits of Fifth step: Categorization. The codes were analysed semantically to explore their meaning and group them into themes and subthemes, using the ATLAS.ti function of “group codes” and “link codes”. In order to deﬁne the emerging themes, conceptual and explanatory memos were created. g g p , g “group codes” and “link codes”. 2.5. Rigor Reliab various res Reliability was achieved through researcher’s triangulation throughout the process; various researchers revised the entire coding and categorization system, discarding the themes and subthemes for which there was not general consensus. To increase the internal validity, all opinions and experiences were represented, and the transcripts and analysis were returned to some participants for them to conﬁrm the content and the researchers’ interpretations. To improve the trustworthiness of the study, was used a checklist that includes different phases of preparation, organisation and reporting of the study [42]. various researchers revised the entire coding and categorization system, discarding the themes and subthemes for which there was not general consensus. To increase the internal validity, all opinions and experiences were represented, and the transcripts and analysis were returned to some participants for them to confirm the content and the researchers’ interpretations. To improve the trustworthiness of the study, was used a checklist that includes different phases of preparation, organisation and reporting of the study [42]. 2.6. Ethical Issues 2.4. Data Analysis In order to define the emerging themes, conceptual and explanatory memos were created. Sixth step: The final integration of the findings. In qualitative analysis, the wording and revision of the conclusions are analysed. In our final report, we used themes, memos Sixth step: The ﬁnal integration of the ﬁndings. In qualitative analysis, the wording and revision of the conclusions are analysed. In our ﬁnal report, we used themes, memos and codes, allowing us to achieve robust results with the support of all previously carried out work. y p , , and codes, allowing us to achieve robust results with the support of all previously carried out work. 2.5. Rigor 2.6. Ethical Issues The study wa E I E The study was approved by the University of Almeria’s Ethical and Research Com- mittee (ENFISMED15-17). All participants were informed of the objective of the study, they participated voluntarily, and they gave written consent to participate in the study. They had the possibility to end the interview at any given moment. In order to ensure anonymity and conﬁdentiality, the participants were identiﬁed using an alphanumeric code. All data have been handled in accordance with European and Spanish data protection laws. mittee (ENFISMED15-17). All participants were informed of the objective of the study, they participated voluntarily, and they gave written consent to participate in the study. They had the possibility to end the interview at any given moment. In order to ensure anonymity and confidentiality, the participants were identified using an alphanumeric code. All data have been handled in accordance with European and Spanish data protec- tion laws. 3. Results 3. Results Two themes emerged from the interview analysis that allowed us to understand primary school teachers’ perspectives regarding SE (Figure 1: Conceptual map of emerg- ing themes). Results are presented following the structure: Name of the theme, Deﬁni- tion/explanation of the theme. Name of the Subtheme, explanation, representative quote. The sequence explanation–quotation is repeated throughout each subtheme until the latter was fully explained. Two themes emerged from the interview analysis that allowed us to understand pri- mary school teachers’ perspectives regarding SE (Figure 1: Conceptual map of emerging themes). Results are presented following the structure: Name of the theme, Definition/ex- planation of the theme. Name of the Subtheme, explanation, representative quote. The sequence explanation–quotation is repeated throughout each subtheme until the latter was fully explained. Figure 1. Conceptual map of emerging themes. Key: [] is part of; => is cause of; <> opposed to; } is property of. Figure 1. Conceptual map of emerging themes. Key: [] is part of; => is cause of; <> opposed to; } is property of. ure 1. Conceptual map of emerging themes. Key: [] is part of; => is cause of; <> opposed to; } is property of. re 1. Conceptual map of emerging themes. Key: [] is part of; => is cause of; <> opposed to; } is property of It’s a complicated issue . . . I’m a bit afraid of addressing it with children. (E11) It’s a complicated issue . . . I’m a bit afraid of addressing it with children. (E11) The interviewees believe that the ideal timing to receive affective sexual education is during adolescence. Whilst topics of basic anatomy and emotions, not explicitly related to sexuality, are addressed with very young children, SE is totally absent from the primary school stage. It is only in teenage years that the issue is dealt with again in high school through speciﬁc programmes such as the Andalusian Health Ministry’s Health Education Programme ‘Forma joven’. Programme ‘Forma joven’. ( . . . ) At a later age, from 10 to 14, they change dramatically, they start having their ﬁrst sexual encounters and I think that is when you should start talking to them about sexuality. (E14) The main theme that emerged from the analysis of the participants’ comments regard- ing teaching sexuality to children is coitus and its negative consequences. Primary school teachers are concerned about sexually transmitted diseases, pregnancy and issues related to contraception, as conﬁrmed by two participants: [I talk to them] about sexual relations, how both members must want to take part, but most of all the importance of being careful given that they could face problems They shouldn’t [I talk to them] about sexual relations, how both members must want to take part, but most of all, the importance of being careful, given that they could face problems. They shouldn’t make decisions lightheartedly but rather contemplate their actions carefully (E3) [I talk to them] about sexual relations, how both members must want to take part, but most of all, the importance of being careful, given that they could face problems. They shouldn’t p of all, the importance of being careful, given that they could face problems. They shouldn’t make decisions lightheartedly but rather contemplate their actions carefully. (E3) Basically, the contents that allow us to drum into them the consequences of not using protection: disease, pregnancy etc. (E2) When referring to the contents relating to sexuality, primary school teachers believe that their inclusion in the Science curriculum sufﬁce, even though they only refer to reproduction and anatomical matters. I would include sexual organs, safe and consensual sexual practices, prevention of disease and pregnancy . . . I don’t know, things like that. It’s a complicated issue . . . I’m a bit afraid of addressing it with children. (E11) (E4) It is already included in the Science curriculum, there is an entire unit dedicated to reproduction and the reproductive system. (E13) 3.1.2. The Blueprint for Affective Sexual Education Based on Rights 3.1. In Search of a Comprehensive Approach to Sexual Education For the participants of the study, SE is an important factor in improving students’ quality of life given that they are sexual beings from birth. Educating individuals on this topic from a young age can not only contribute to their full and satisfying exercise of sexuality in future adult life but also to their possession of more tolerant attitudes based on equality. Moreover, the participants consider that sexuality not only addresses genital and Healthcare 2021, 9, 287 6 of 12 reproductive matters but also encompasses a holistic vision of body image, how we have relationships with our peers and our feelings during those relationships. reproductive matters but also encompasses a holistic vision of body image, how we have relationships with our peers and our feelings during those relationships. reproductive matters but also encompasses a holistic vision of body image, how we have relationships with our peers and our feelings during those relationships. 3.1.1. The Predominance of Preventative Sexual and Reproductive Education Aimed at Young People 3.1.1. The Predominance of Preventative Sexual and Reproductive Education Aimed at Young People Primary school teachers do not address the contents of SE with their students in a natural way. They admit that, according to Spanish legislation, it is a transversal topic similar to education in road safety or anti-discrimination, for example. The majority of the interviewees avoid terms such as coitus, masturbation, etc., referring to them in an evasive or euphemistic way. For example, two participants avoided the topic in the following way: SE (blushing), well . . . is information about . . . that topic [avoiding the term “sex”], SE (blushing), well . . . is information about . . . that topic [avoiding the term “sex”], right? (E1) It’s complicated, we already have a heavy workload to have to then educate children in a topic for which we have not been prepared, to be honest (worried face). (E9) It’s complicated, we already have a heavy workload to have to then educate children in a topic for which we have not been prepared, to be honest (worried face). (E9) Furthermore, in schools, aspects related to sexuality are included in a set of transversal contents such as healthy eating, gender equality and education for peace. Added to the lack of speciﬁc training in the ﬁeld, it becomes more difﬁcult to evaluate these contents. The thing is that it’s easy to say “look, let’s put this topic into the transversal contents”, but they are contents that cannot be evaluated, unlike Maths. This means that if the matter comes up, we address it but if not, we don’t. (E08) Ultimately, transversal contents get lost in different subjects and they don’t end up being addressed. (E11) Ultimately, transversal contents get lost in different subjects and they don’t end up being addressed. (E11) 3.1.2. The Blueprint for Affective Sexual Education Based on Rights (E10) They copy everything they see the most dangerous thing is that they are watching They copy everything they see . . . the most dangerous thing is that they are watching pornography on their mobiles at an ever-younger age and they think that relationships are like that. (E12) They copy everything they see . . . the most dangerous thing is that they are watching pornography on their mobiles at an ever-younger age and they think that relationships are like that. (E12) 3.2. Barriers to Sexual Education in Schools: From the Lack of Training to Fear of the Families One of the most common limitations, according to the teachers, is the lack of training on the topic of sexuality. This is noted both at university education level, in which childhood sexual development is not addressed, as well as in a lack of training on the speciﬁc topic itself. Yet, there is training in the use of interactive whiteboards, the bilingual curriculum and similar topics. Moreover, the teachers expressed their apprehension regarding the students’ families, not only in having to speak to them clearly and decisively but also out of fear of the potential action families could take against them. This, in turn, leads to SE programmes failing. 3.1.2. The Blueprint for Affective Sexual Education Based on Rights Some participants understand that SE in schools should not merely address biological and reproductive contents but should instead have a more holistic vision of sexuality, including sexuality as a human right (diversity, respect, consensuality, etc.), communica- tion (e.g., interpersonal relationships, safe sexual experiences) and the emotional aspects (e.g., pleasure). That’s it, I’d talk about prevention but also about how to get involved in relationships, about respect . . . (E15) That’s it, I’d talk about prevention but also about how to get involved in relationships, about respect . . . (E15) I would focus on Sexual Identiﬁcation, introduce feelings, empathy towards different sexual tendencies ( . . . ), something that would help them to learn and respect everyone, however they may be. (E6) 7 of 12 Healthcare 2021, 9, 287 The teachers are also very concerned about gender equality, given that it is of high importance nowadays. They are of the opinion that we should dismantle stereotypes seen in the media, especially on television or in social media. In class or at break time, children tend to act in a way that reﬂects the clichés seen in the media. The teachers feel obliged to tackle the issue. . . . one must promote respect. They see how women are treated as objects on TV programmes and then they demonstrate these attitudes in class. For us it’s really hard to ﬁght against what they see on TV. (E10) They copy everything they see . . . the most dangerous thing is that they are watching pornography on their mobiles at an ever-younger age and they think that relationships are like that. (E12) . . . one must promote respect. They see how women are treated as objects on TV programmes and then they demonstrate these attitudes in class. For us it’s really hard to ﬁght against what they see on TV. (E10) They copy everything they see . . . the most dangerous thing is that they are watching pornography on their mobiles at an ever-younger age and they think that relationships are like that. (E12) . . . one must promote respect. They see how women are treated as objects on TV programmes and then they demonstrate these attitudes in class. For us it’s really hard to ﬁght against what they see on TV. 3.2.1. Lack of Training in a Comprehensive Approach to Sexuality The interviewees repeatedly refer to a shared complaint; the absence of training in SE throughout their university degrees, stating that university programmes are full of contents involving procedural knowledge, leaving little time to address contents more speciﬁcally related to SE. Two participants highlighted this in the following way: In my degree, they didn’t even tell me that I had to address this topic with the kids in class. (E11) In my degree, they didn’t even tell me that I had to address this topic with the kids in class. (E11) 4. Discussion The objective of this study was to explore the perspectives of primary school teachers regarding SE in schools in Spain. A qualitative approach to the study has allowed us to study these perspectives in detail using a wide and varied sample [36]. Using MHP [41] as the study’s lens, allows us to focus on SE in schools as a way of teaching health education that incorporates wellbeing and not just the absence or prevention of disease [2]. This facili- tates a more comprehensive approach, preferred by students [12], which includes emotional and social aspects such as preventing bullying or promoting healthy relationships [15]. p p g y g p g y p Our results suggest that SE refers to a complex transversal content in Spanish schools, that is often excluded completely or only addressed in Science. Furthermore, a recent qualitative synthesis suggests that sex is a powerful topic that can give rise to intense emotions, reactions and feelings; however, it is taught in the same way as other subjects in schools [31]. By contrast, in recent years, the United States is launching more inclusive programmes that explore sexual health in terms of its physical, social, emotional, intellec- tual and spiritual dimensions [23,24]. This study suggests that, despite the obstacles that primary school teachers face, they try to deliver SE based on rights. For this reason, using external experts (for example, sexologists), could be of great help [27]. p p g g p In accordance with current literature on the topic, our study highlights that the common approach to this type of education is based on abstinence, prevention and repro- duction [14,21], and that it is awkward, unhelpful and often uses scare tactics [43]. Various authors have concluded that SE based on abstinence does not reduce sexual activity itself or when young people start to engage in it [12,44]. Moreover, society’s view of sexuality differentiates gender roles. For example, virginity is seen as a virtue in women whereas it is considered a ﬂaw in a man [45]. However, it is possible to modify traditional stances due to socio-cultural changes related to gender identity and ideology [46]. A comprehensive approach to SE does not only prevent disease and unwanted pregnancies but also prevents homophobic bullying, domestic violence and child abuse [15]. Furthermore, it promotes healthy relationships, socioemotional learning and media literacy [15,47]. 3.2.2. Avoiding the Topic of Sexuality out of Fear of the Families The teachers express their apprehension of having to address the topic of SE due to the families. Added to the lack of training they receive on the subject, there is the constant scrutiny of parents and the potential outcomes their opinions can have on the teachers. You sometimes don’t do anything because the families are on your back observing your every move. On many occasions, a family has come and said, “who are you to say that to my child?” If that happens in subjects like English language, imagine if you start tackling issues related to sexuality. (E7) You sometimes don’t do anything because the families are on your back observing your every move. On many occasions, a family has come and said, “who are you to say that You sometimes don’t do anything because the families are on your back observing your every move. On many occasions, a family has come and said, “who are you to say that You sometimes don’t do anything because the families are on your back observing your every move. On many occasions, a family has come and said, “who are you to say that to my child?” If that happens in subjects like English language, imagine if you start tackling issues related to sexuality. (E7) y y f y y y to my child?” If that happens in subjects like English language, imagine if you start tackling issues related to sexuality. (E7) Normally, if a topic related to sexuality arises in class, whether it is due to a conﬂict in the playground or simply from questions in class, the teachers avoid responding out of fear of what the families might say. 8 of 12 Healthcare 2021, 9, 287 I steer clear of those topics. If it comes up in class, I try to change the subject and avoid answering, just because of what they might say at home later. (E9) I steer clear of those topics. If it comes up in class, I try to change the subject and avoid answering, just because of what they might say at home later. (E9) This leads us to leave those topics for certain school subjects, ( . . . ) we don’t have a lot of time anyway. (E1) 4. Discussion In this respect, other experiences suggest the need to take into account the SE received by parents [11], religious beliefs [5,55] and the possibility of resistance against education in this area [56]. A further cause of concern regarding SE are children with speciﬁc education needs (SEN). Literature suggests that parents with autistic children are not aware of whether their children have experienced orgasms [57]. There is a notable increase in quality of life amongst these children after receiving SE, as it enables them to understand differences in identity and sexual practices [58]. However, this concern did not arise in our study, perhaps due to the context in which the study took place or to the participants’ lack of SEN students. Both in this research study and others [35] it is claimed that increasing knowledge of sexual health and sexuality leads to a more aware approach towards the understanding and risks of sexual encounters. Literature supports the notion that affective sexual education teaches children to protect themselves from mistreatment, in so much as they are aware of appropriate and inappropriate behaviours. They also advocate that SE gives children the knowledge required to increase protective behaviours against abuse [59]. Likewise, incor- rect and aversive beliefs about SE can be changed [35] not only in schools but throughout life, given that childhood and adolescence are not the only stages in life that one should receive formal SE as it should continue into adulthood [29]. This study suggests including aspects related to sexuality across all subjects and not only in Science. This would con- tribute to a holistic approach to SE based on rights and not just on prevention. For example, subjects like Spanish Language could explore non-sexist language or History could look into the historical causes for gender discrimination. g Some of the limitations of this study are related to the preparation phase, speciﬁcally the sampling section [42]. The sample is appropriate and was composed of participants with knowledge of the research topic. However, the sample was selected in public schools in a middle-class urban setting. This leads us to believe that extending this study to primary school teachers in private or semi-private schools as well as other more varied settings could lead to a more robust set of results. It is a heterogeneous sample in terms of age (range: 25–60 years) and professional experience (1–40 years). 4. Discussion The participants of the study argue that the lack of training and the fear of the families are barriers to a comprehensive and open-minded approach to sexuality in schools. Our study was inﬂuenced by the values and beliefs prevalent in Spanish society. Within this context, there has also been a change in values related to sexuality, concurring with other studies that allude to socio-political changes with important repercussions on SE [48]. Likewise, in the United Kingdom the topic has evolved from immoral to normal over a 70-year period [23,49]. On the other hand, our results reﬂected a concern amongst primary school teachers regarding gender equality, as they ﬁnd themselves in the difﬁcult position of breaking down stereotypes learned from the media, social networks and early age pornography consumption, as stated by Ballester and Orte [6]. Our study reﬂects that primary school teachers often lack the determination required to speak about sexuality openly, which is in line with other studies that afﬁrm that SE can fail due solely to the fear of speaking frankly [21]. According to young people in several studies, schools struggle to accept that some young people are sexually active, leading to a lack of relevant content for them in SE [31]. The information era allows children to be exposed to various sources that reﬂect nudity, identity and sexual conduct continuously, underlining the importance of addressing these issues with adapted language but in an open way [50]. Moreover, it is necessary to include content regarding the possibilities, limitations and risks of social media in relation to sexual activity and behaviours [47]. Healthcare 2021, 9, 287 9 of 12 9 of 12 A vital datum from this study is the lack of SE training amongst teachers, leading to shortcomings when addressing this subject in schools. The teachers are aware of this deﬁcit and stress the importance of speciﬁc training in sexuality, which is a stance supported by various authors who also value the need for teachers to receive speciﬁc SE training [51–53] in programmes directed at all levels, including pre-adolescence [54]. The latter is perhaps the most important age to educate children on sexuality, with the support of their families. One of the key barriers to SE alluded to in this study is the tendency to avoid talking about sexuality out of fear of families with opposing views to teachers speaking to their children about sexuality. 4. Discussion This has provided us with a diversity of perspectives, but a more homogenous sample would have allowed us to have a deeper understanding of the difﬁculties associated with each individual’s experience or the generational difference when talking about sex. New studies could compare perspectives and experiences of young and older teachers in terms of speaking about sexuality with their pupils. The primary school teachers have identiﬁed their perspectives regarding SE in schools and the key obstacles that they face. Research is therefore needed into how to make them more competent. 5. Conclusions Primary school teachers conclude that SE in schools has an excessively preventative approach. Nonetheless, they recognise an attempt to implement a more comprehensive SE that includes affective sexual education and sexuality as a human right. The teachers are faced with difﬁculties when addressing SE in schools due to a lack of speciﬁc training in the area and fear of parents rejecting their children being spoken to about sexuality. The primary school teachers have highlighted the key barriers that they face. It is necessary to continue researching their strengths, capabilities and how to make Healthcare 2021, 9, 287 10 of 12 10 of 12 them more competent in delivering SE in schools. Understanding formative experiences in other countries and surroundings could also be useful. them more competent in delivering SE in schools. Understanding formative experiences in other countries and surroundings could also be useful. It is essential to develop speciﬁc SE training programmes for teachers and to establish educational programmes in Spanish schools that support an integrated, holistic approach to sexual health based on equality. Furthermore, it is vital to create awareness campaigns regarding the need for SE in schools given that it is a fundamental element in developing individuals who are able to make free and healthy choices concerning their relationships. These campaigns should be directed at the general population to break the taboos regarding SE for children. This could be done within institutions (e.g., governments) or from the schools themselves through parent meetings. Practical implications: as SE is a cross-curricular competence in the Spanish education system, it is vital that teachers address it from different perspectives. Leaving it solely to Science teachers risks SE becoming too focused on prevention. The participants stated a lack of training in the subject, which is why the support of specialists from the sexology ﬁeld, for example, would be of great beneﬁt in the design of strategies and materials for SE in schools. Author Contributions: Conceptualization, J.J.M.-G., O.M.L.-E. and C.F.-S.; methodology, C.F.-S.; for- mal analysis, M.D.R.-F., F.J.P.-d.-P. and J.J.M.-G.; investigation, J.J.M.-G. and I.S.; data curation, M.I.V.- M., M.Á.P.-C.; writing—original draft preparation, J.J.M.-G., C.F.-S. and O.M.L.-E.; writing—review and editing, F.J.P.-d.-P., M.D.R.-F., M.Á.P.-C., M.I.V.-M. and I.S.; visualization, J.J.M.-G. and I.S.; project administration, C.F.-S. All authors have read and agreed to the published version of the manuscript. Funding: This research has received funding from the Health Science Research Group CTS-451 of the University of Almeria. 5. Conclusions Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Nursing Science, Physio- therapy and Medicine Department, University of Almeria (code ENFISMED15-17, date of approval 22 February 2017). Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Nursing Science, Physio- therapy and Medicine Department, University of Almeria (code ENFISMED15-17, date of approval 22 February 2017). Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: Data of this study are audio recordings and their transcripts (conﬁden- tially), which are stored in an ATLAS.ti software Project. Acknowledgments: The authors thank the participants for sharing their experiences and opinions with us. We thank the Health Science Research Group CTS-451 of the University of Almeria. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. 8. Shin, K.R.; Park, H.; Cha, C. Sex education during the school-aged years inﬂuences sexual attitudes and sexual health in college: A comparative study from Korea. Nurs. Health Sci. 2011, 13, 328–334. [CrossRef] [PubMed] , ; , g f y p , ; , p , 7. Hald, G.M.; Kuyper, L.; Adam, P.C.; de Wit, J.B. Does Viewing Explain Doing? Assessing the Association Between Sexually Explicit Materials Use and Sexual Behaviors in a Large Sample of Dutch Adolescents and Young Adults. J. Sex. Med. 2013, 10, 2986–2995. [CrossRef] [PubMed] 4. Brömdal, A.; Zavros-Orr, A.; Lisahunter; Hand, K.; Hart, B. Towards a whole-school approach for sexuality education in supporting and upholding the rights and health of students with intersex variations. Sex Educ. 2020, 1–16. [CrossRef] 5. Nadeem, A.; Cheema, M.K.; Zameer, S. Perceptions of Muslim parents and teachers towards sex education in Pakistan. Sex Educ. 2021, 21, 106–118. [CrossRef] 6. Ballester, L.; Orte, C. Nueva Pornografía y Cambios en las Relaciones Interpersonales, 1st ed.; Octaedro: Barcelona, Spain, 2019. References 2011, 18, 303–304. 22. Keogh, S.C.; Stillman, M.; Leong, E.; Awusabo-Asare, K.; Sidze, E.; Monzón, A.S.; Motta, A. Measuring the quality of sexua llejero, M.L.; Torres, B. Sex education: Sex and gender in Primary School textbooks. Rev. Educ. 2011, 18, 303 eogh, S.C.; Stillman, M.; Leong, E.; Awusabo-Asare, K.; Sidze, E.; Monzón, A.S.; Motta, A. Measuring the q 22. Keogh, S.C.; Stillman, M.; Leong, E.; Awusabo-Asare, K.; Sidze, E.; Monzón, A.S.; Motta, A. Measurin education implementation at the school level in low- and middle-income countries. Sex Educ. 2019, 20 education implementation at the school level in low- and middle-income countries. Sex Educ. 2019, 20, 119–137. [CrossRef] 23. Harvey, A. Sexpression:UK: The university students delivering sex and relationships education to schools. BMJ 2020, 370, m2816. [CrossRef] 23. Harvey, A. Sexpression:UK: The university students delivering sex and relationships education to schools. BMJ 2020, 370, m2816. [CrossRef] 24. Elia, J.P.; Tokunaga, J. Sexuality education: Implications for health, equity, and social justice in the United States. Health Educ. 2015, 115, 105–120. [CrossRef] 25. Galaz, C.; Troncoso, L.; Morrison, R. Miradas Críticas sobre la Intervención Educativa en Diversidad Sexual. Rev. Lat. Educ. Inclusiva. 2016, 10, 93–111. [CrossRef] 26. Lee, G.Y.; Lee, D.Y. Effects of a life skills-based sexuality education programme on the life-skills, sexuality knowledge, self- management skills for sexual health, and programme satisfaction of adolescents. Sex Educ. 2018, 19, 519–533. [CrossRef] g ( p ) 28. Brewin, D.; Koren, A.; Morgan, B.; Shipley, S.; Hardy, R.L. Behind Closed Doors: School Nurses and Sexual Education. J. Sch. Nurs. 2014, 30, 31–41. [CrossRef] [PubMed] 29. Graf, A.S.; Patrick, J.H. Foundations of life-long sexual health literacy. Health Educ. 2015, 115, 56–70. [CrossRef] 30. Lindroth, M. Sex education and young people in group homes: Balancing risks, rights and resilienc Sex Educ. 2014, 14, 400–413. [CrossRef] ucation and young people in group homes: Balancing risks, rights and resilience in sexual health promotion 00–413. [CrossRef] 31. Pound, P.; Langford, R.; Campbell, R. What do young people think about their school based sex and relationship education? A qualitative synthesis of young people’s views and experiences. BMJ Open 2016, 6, e011329. [CrossRef] [PubMed] 32. Naezer, M.; Rommes, E.; Jansen, W. Empowerment through sex education? Rethinking paradoxical policies. Sex Educ. 2017, 17, 712–728. [CrossRef] qualitative synthesis of young people s views and experiences. BMJ Open 2016, 6, e011329. [CrossRef] [PubMed] 32. Naezer, M.; Rommes, E.; Jansen, W. Empowerment through sex education? References 1. Organización de las Naciones Unidas para la Educación, la Ciencia y la Cultura (UNESCO). Orientaciones técnicas internacionales sobre educación en sexualidad: Un enfoque basado en evidencia. Edición revisada. Educación 2018, 2030, 1–6. 1. Organización de las Naciones Unidas para la Educación, la Ciencia y la Cultura (UNESCO). Orientaciones técnicas internacionales sobre educación en sexualidad: Un enfoque basado en evidencia. Edición revisada. Educación 2018, 2030, 1–6. q 2. World Health Organization (WHO). Gender and Reproductive Rights, Glossary, Sexual Health; World Health Organization (WHO): Geneva, Switzerland, 2002. 3. Mehrotra, P.; Schensul, S.L.; Saggurti, N.; Burleson, J.A.; Maitra, S. The WHO concept of sexual health applied to an urban por community in Mumbai, India. Int. J. Sex Health 2015, 27, 57–68. [CrossRef] 4. Brömdal, A.; Zavros-Orr, A.; Lisahunter; Hand, K.; Hart, B. Towards a whole-school approach for sexuality education in supporting and upholding the rights and health of students with intersex variations. Sex Educ. 2020, 1–16. [CrossRef] 5. Nadeem, A.; Cheema, M.K.; Zameer, S. Perceptions of Muslim parents and teachers towards sex education in Pakistan. Sex Educ. 2021, 21, 106–118. [CrossRef] 4. Brömdal, A.; Zavros-Orr, A.; Lisahunter; Hand, K.; Hart, B. Towards a whole-school approach for sexuality education in supporting and upholding the rights and health of students with intersex variations. Sex Educ. 2020, 1–16. [CrossRef] 6. Ballester, L.; Orte, C. Nueva Pornografía y Cambios en las Relaciones Interpersonales, 1st ed.; Octaedro: Barcelona, Spain, 2019. g f y p p 7. Hald, G.M.; Kuyper, L.; Adam, P.C.; de Wit, J.B. Does Viewing Explain Doing? Assessing the Association Between Sexually Explicit Materials Use and Sexual Behaviors in a Large Sample of Dutch Adolescents and Young Adults. J. Sex. Med. 2013, 10, 2986–2995. [CrossRef] [PubMed] 11 of 12 11 of 12 Healthcare 2021, 9, 287 9. Lago, A.M.; Esteiro, M.P.; Mosquera, L.; Bouza, E.T. Una manera diferente de abordar la sexualidad, la contracepción y la prevención de infecciones de transmisión sexual desde la escuela en la Costa da Morte/A differentapproach to sexuality, contraception and prevention of sexuallytransmittedinfectionsfromschoolon Costa da Morte. Enferm. Glob. 2015, 39, 137–154. p p y f 10. Helmer, J.; Senior, K.; Davison, B.; Vodic, A. Improving sexual health for young people: Making sexu Sex Educ. 2014, 15, 158–171. [CrossRef] y f K.; Davison, B.; Vodic, A. Improving sexual health for young people: Making sexuality education a priority 58–171. [CrossRef] , , [ ] 11. Allen, K.R.; Lavender-Stott, E.S. References Family Contexts of Informal Sex Education: Young Men’s Perceptions of First Sexual Images. Fam. Relat. 2015, 64, 393–406. [CrossRef] 12. Zaw, P.P.T.; McNeil, E.; Oo, K.; Liabsuetrakul, T.; Htay, T.T. Abstinence-only or comprehensive sex education at Myanmar schools: Preferences and knowledge among students, teachers, parents and policy makers. Sex Educ. 2021, 21, 75–90. [CrossRef] 13. Bay-Cheng, L.Y. The Trouble of Teen Sex: The construction of adolescent sexuality through school-based sexuality education. Sex Educ. 2003, 3, 61–74. [CrossRef] 14. Goldfarb, E.S.; Lieberman, L.D. Three Decades of Research: The Case for Comprehensive Sex Education. J. Adolesc. Health 2021, 68, 13–27. [CrossRef] [PubMed] 15. Kantor, L.M.; Lindberg, L.D.; Tashkandi, Y.; Hirsch, J.S.; Santelli, J.S. Sex Education: Broadening the Deﬁnition of Relevant Outcomes. J. Adolesc. Health 2021, 68, 7–8. [CrossRef] [PubMed] 16. Peter, C.R.; Tasker, T.B.; Horn, S.S. Parents’ attitudes toward comprehensive and inclusive sexuality sexual health topics and forms of curricula. Health Educ. 2015, 115, 71–92. [CrossRef] Tasker, T.B.; Horn, S.S. Parents’ attitudes toward comprehensive and inclusive sexuality education: Belief h topics and forms of curricula. Health Educ. 2015, 115, 71–92. [CrossRef] 17. Ellis, S.J.; Bentham, R.M. Inclusion of LGBTIQ perspectives in school-based sexuality education in Aotearoa/New Zealand: An exploratory study. Sex Educ. 2020, 1–15. [CrossRef] p y y 18. Astle, S.; McAllister, P.; Emanuels, S.; Rogers, J.; Toews, M.; Yazedjian, A. College students’ suggestions for improving sex education in schools beyond ‘blah blah blah condoms and STDs’. Sex Educ. 2021, 21, 91–105. [CrossRef] 19. Meschke, L.L.; Peter, C.R. Hmong American Parents’ Views on Promoting Adolescent Sexual Health. Am. J. Sex. Educ. 2014, 9, 308–328. [CrossRef] 20. Rubba, F.; Scamardo, M.S.; Di Mauro, F.; Egidio, R.; Avati, P.; Passione, A.; Giulierini, P.; Continisio, I.G.; Iervolino, A.; Triassi, M. If the dolphin come to help: Education and Health without Boundaries for transversal competences. Eur. J. Public Health 2020, 30, 166–742. [CrossRef] 21. Pellejero, M.L.; Torres, B. Sex education: Sex and gender in Primary School textbooks. Rev. Educ. 2011, 18, 303–304. 22 Keogh S C Stillman M Leong E Awusabo Asare K Sidze E Monzón A S Motta A Measuring the qualit of se M.L.; Torres, B. Sex education: Sex and gender in Primary School textbooks. Rev. Educ. 2011, 18, 303–304. 21. Pellejero, M.L.; Torres, B. Sex education: Sex and gender in Primary School textbooks. Rev. Educ. 2011, 21. Pellejero, M.L.; Torres, B. Sex education: Sex and gender in Primary School textbooks. Rev. Educ. References Rethinking paradoxical policies. Sex Educ. 2017, 17, 712–728. [CrossRef] [ ] 33. Barea, E.M.G.; Marín, Y.R. Estereotipos de género en la infancia. Pedagog. Soc. Rev. Interuniv. 2020, 36, 125–138. [CrossRef] 34. Walcott, C.M.; Chenneville, T.; Tarquini, S. Relationship between recall of sex education and college students’ sexual attitudes and behavior. Psychol. Sch. 2011, 48, 828–842. [CrossRef] y 35. Cortés-Alfaro, A.; García-Roche, R.; Lazo-González, Z.; Bolaños, J.C.; Ochoa-Soto, R.; Pérez-Sosa, D. Intervención en el conocimiento, creencias y percepción de riesgo sobre infecciones de transmisión sexual y el vih/sida. Med. Electr. 2014, 18, 11–18. 36. Heras-Sevilla, D.; Ortega-Sánchez, D. Evaluation of Sexist and Prejudiced Attitudes Toward Homosex Teachers: Analysis of Related Variables. Front. Psychol. 2020, 11, 2334. [CrossRef] 37. Grossman, J.M.; Frye, A.; Charmaraman, L.; Erkut, S. Family homework and school-based sex education: Delaying early adolescents’ sexual behavior. J Sch. Health 2013, 83, 810–817. [CrossRef] [PubMed] J 38. Mayring, P. Qualitative content analysis. Forum Qual. Soc. Res. 2000, 1, 20. 38. Mayring, P. Qualitative content analysis. Forum Qual. Soc. Res. 2000, 1, 20. 12 of 12 12 of 12 Healthcare 2021, 9, 287 39. Fernández-Sola, C.; Granero-Molina, J.; Hernández-Padilla, J.M. Comprender Para Cuidar: Avances en Investigación Cualitativa en Ciencias de la Salud; Editorial Universidad Almería: Almería, Spain, 2020. p 40. Cáceres, P. Análisis cualitativo de contenido: Una alternativa metodológica alcanzable. Psicoperspectivas. Individuo Soc. 2003, 2, 53–82. 41. Pender, N.J.; Murdaugh, C.L.; Parsons, M.A. Health Promotion in Nursing Practice, 5th ed.; Pearson/Prentice Hall: Upper Saddle River, NJ, USA, 2006. 42. Elo, S.; Kääriäinen, M.; Kanste, O.; Pölkki, T.; Utriainen, K.; Kyngäs, H. Qualitative Content Analysis. SAGE Open 2014, 4, 2158244014522633. [CrossRef] 43. Astle, S.; Leonhardt, N.; Willoughby, B. Home Base: Family of Origin Factors and the Debut of Vaginal Sex, Anal Sex, Oral Sex, Masturbation, and Pornography Use in a National Sample of Adolescents. J. Sex Res. 2020, 57, 1089–1099. [CrossRef] 44. Trenholm, C.; Devaney, B.; Fortson, K.; Clark, M.; Quay, L.; Wheeler, J. Impacts of abstinence educatio risk of pregnancy, and risk of sexually transmitted diseases. J. Policy Anal. Manag. 2008, 27, 255–276. [C 45. Saeteros, C.; Piñero, I.C.; Saeteros, R.C.; Pérez Piñero, C.J.; Sanabria, C.G. Vivencias de la sexualidad en estudiantes universitarios. Rev. Cub. Salud. Publica 2013, 39, 915–928. 46. Grose, R.G.; Grabe, S.; Kohfeldt, D. Sexual Education, Gender Ideology, and Youth Sexual Empowerment. J. Sex Res. 2013, 51, 742–753. [CrossRef] 47. Ochoa-Pineda, A.C.; Aranda-Torres, C.J. References Sexting: Signo de Identidad Juvenil en la Sociedad Digital; Universidad Almería: Almería, Spain, 2020. 48. Toledano-Sierra, M.; Ruiz-Fernández, M.D.; Hernández-Padilla, J.M.; Granero-Molina, J.; Pomares-Callejón, M. Ángeles; Jiménez- Lasserrotte, M.D.M.; Fernández-Sola, C. Female sexuality during an era of political repression in Spain. A qualitative study on the survivors of Francoism. Women’s Stud. Int. Forum 2020, 79, 102343. [CrossRef] 49. Iyer, P.; Aggleton, P. Seventy years of sex education inHealth Education Journal: A critical review. Health Educ. J. 2014, 74, 3–15. [CrossRef] 50. Shtarkshall, R.A.; Santelli, J.S.; Hirsch, J.S. Sex Education and Sexual Socialization: Roles for Educators and Parents. Perspect. Sex. Reprod. Health 2007, 39, 116–119. [CrossRef] [PubMed] 51. Franco-Morales, F.; Correa-Molina, E.; Venet, M.; Perez-Bedoya, S. The relationship between attitudes towards and knowledge of sexual diversity /La relación entre actitudes y conocimientos sobre diversidad sexual. Cult. Educ. 2016, 28, 500–538. [CrossRef] 52. Manzano-Pauta, D.; Jerves-Hermida, E. Sex Education: Teachers’ Perceptions in the City of Cuenca 2013–2014. Rev. Electr. Educ. 2018, 22, 82–96. [CrossRef] 53. Alonso, D.M.; Giménez, A.M.; Fernández-Rio, F.J. Análisis y valoración del proceso de incorporación de las Competencias Básicas en Educación Primaria. Rev. Investig. Educ. 2015, 33, 233–246. [CrossRef] 54. Preston, M.J. ‘They’re just not mature right now’: Teachers’ complicated perceptions of gender and anti-queer bullying. Sex Educ. 2016, 16, 22–34. [CrossRef] 55. Spadt, S.K.; Rosenbaum, T.Y.; Dweck, A.; Millheiser, L.; Pillai-Friedman, S.; Krychman, M. CME Information: Sexual Health and Religion: A Primer for the Sexual Health Clinician (CME). J. Sex. Med. 2014, 11, 1606–1619. [CrossRef] 56. Alekseeva, E.G.; Krasnopolskaya, I.; Skokova, Y. Introducing sexual education to Russian schools. Health Educ. 2015, 115, 7–37. [CrossRef] 57. Dewinter, J.; Vermeiren, R.; Vanwesenbeeck, I.; Van Nieuwenhuizen, C. Parental Awareness of Sexual E Boys With Autism Spectrum Disorder. J. Autism Dev. Disord. 2015, 46, 713–719. [CrossRef] [PubMed] Boys With Autism Spectrum Disorder. J. Autism Dev. Disord. 2015, 46, 713–719. [CrossRef] [PubMed] 58. Barnett, J.P.; Maticka-Tyndale, E. Qualitative Exploration of Sexual Experiences Among Adults on the Autism Spectrum: Implications for Sex Education. Perspect. Sex. Reprod. Health 2015, 47, 171–179. [CrossRef] [PubMed] y p 58. Barnett, J.P.; Maticka-Tyndale, E. Qualitative Exploration of Sexual Experiences Among Adults on the Autism Spectrum: Implications for Sex Education. Perspect. Sex. Reprod. Health 2015, 47, 171–179. [CrossRef] [PubMed] p p p 59. Zwi, K.; Woolfenden, S.; Wheeler, D.; O’Brien, T.; Tait, P.; Williams, K. School-Based Education Programmes for the Prevention of Child Sexual Abuse. Campbell Syst. Rev. References 2007, 3, 1–40. [CrossRef] 59. Zwi, K.; Woolfenden, S.; Wheeler, D.; O’Brien, T.; Tait, P.; Williams, Child Sexual Abuse. Campbell Syst. Rev. 2007, 3, 1–40. [CrossRef]
https://openalex.org/W4280546388	https://www.frontiersin.org/articles/10.3389/fpubh.2022.814112/pdf	English	null	COVID-19 Impact on Household Food Security in Urban and Peri-Urban Areas of Hyderabad, India	Frontiers in public health	2,022	cc-by	11,538	COVID-19 Impact on Household Food Security in Urban and Peri-Urban Areas of Hyderabad, India Ravula Padmaja 1†, Swamikannu Nedumaran 1, Padmanabhan Jyosthnaa 1†, Kasala Kavitha 1†, Assem Abu Hatab 2,3 and Carl-Johan Lagerkvist 2* 1 Enabling Systems Transformation, International Crops Research Institute for the Semi-Arid Tropics, Patancheru, India, 2 Department of Economics, Swedish University of Agricultural Sciences, Uppsala, Sweden, 3 Department of Economics and Rural Development, Arish University, Arish, Egypt Ravula Padmaja 1†, Swamikannu Nedumaran 1, Padmanabhan Jyosthnaa 1†, Kasala Kavitha 1†, Assem Abu Hatab 2,3 and Carl-Johan Lagerkvist 2* 1 Enabling Systems Transformation, International Crops Research Institute for the Semi-Arid Tropics, Patancheru, India, 2 Department of Economics, Swedish University of Agricultural Sciences, Uppsala, Sweden, 3 Department of Economics and Rural Development, Arish University, Arish, Egypt This paper investigates the impact of the COVID-19 pandemic on food security and on coping-strategies in urban and peri-urban areas of the Hyderabad, India. Household survey data were collected before (October 2018) and during (January 2021) the onset of the pandemic. Results from logistic regression with the standarized Food Insecurity Expecience Scale (FIES) as dependent variable reveal that close to 40% of the households surveyed experienced a deterioration in food security status during the pandemic. In particular, we ﬁnd that food security is closely related to the sector of employment in which the primary income- earning member of a household is engaged. To mitigate the impact of the pandemic on their food security, our sampled households adopted a variety of consumption-smoothing strategies including availing credit from both formal and informal sources, and liquidating their savings. Compared to households with severe or moderate level of food insecurity, households facing a mild level of food insecurity relied on stored food as a strategy to smoothen consumption in response to the income shock imparted by the pandemic. In addition, the results indicate that urban households, who adopted similar coping strategies as those adopted by peri-urban households, tended to be more food-insecure. Finally, the duration of unemployment experienced during the pandemic signiﬁcantly inﬂuenced the status of household food security. These ﬁndings can inform the formulation of immediate and medium-term policy responses, including social protection policies conductive to mitigating the impacts of the COVID-19 pandemic and ameliorating the governance of urban food security during unexpected events and shocks. Keywords: food security, pandemic (COVID-19), livelihood, coping strategies, urban, peri-urban, Hyderabad (India), India Edited by: Tarek Ben Hassen, Qatar University, Qatar Reviewed by: Adriana Radosavac, Univerzitet Privredna akademija Novi Sad, Serbia Zeljko Vasko, University of Banja Luka, Bosnia and Herzegovina Correspondence: Carl-Johan Lagerkvist carl-johan.lagerkvist@slu.se †These authors have contributed equally to this work Specialty section: This article was submitted to Public Health and Nutrition, a section of the journal Frontiers in Public Health Received: 12 November 2021 Accepted: 23 March 2022 Published: 13 May 2022 ORIGINAL RESEARCH published: 13 May 2022 doi: 10.3389/fpubh.2022.814112 ORIGINAL RESEARCH published: 13 May 2022 doi: 10.3389/fpubh.2022.814112 Edited by: Tarek Ben Hassen, Qatar University, Qatar INTRODUCTION The lockdown froze economic activity across the country and delivered a large aggregate supply and demand shock to the economy. The consequences have been unprecedented in scale and intensity. Livelihoods were devastated due to the inability to maintain job security, food production was compromised and supply chains were distrupted. The adverse eﬀects of a countrywide lockdown combined with weak political, economic, and social interventions had extended beyond income shocks and aﬀected household food security (11, 12). Loss of employment, curtailed contracts and reduced wages exacerbated food insecurity risk (13). Food security and ﬁnancial security are fundamentally interconnected but there is sparse literature showing this connection. Income volatility has been gaining attention within the broader literature of economic well-being, and qualitative research suggests plausible association with the considerable challenges of meeting household food needs (32–35). Income shock and expenditure shock are strongly associated with food insecurity. Similarly, extended periods of unemployment increase the risks of food insecurity (36). These unfolding COVID-19 impacts on food security in developing countries have a strong territorial/spatial dimension (14), as regions have been heterogeneously aﬀected in the short-run, and the medium- and long-term impact will vary signiﬁcantly across regions. One of the greatest challenges facing the world’s rapidly-growing urban population is how to access suﬃcient, aﬀordable, and nutritious food (15). In particular, densely populated and deprived urban areas were the reportedly hardest hit than other areas (16). For many urban households, especially those living in poorer communities, labor is the most important asset. The fact that the majority of workers in such communities tend to work in the informal sector, earn a variable income and have little or no access to private or social insurance makes access to suﬃcient food a crucial issue (17). It is in this context that the global economic slowdown trigerred by the COVID-19 pandemic, as well as the disease itself, has exacerbated existing societal inequalities in most countries (7). Thus, COVID- 19 impacts on food security in developing countries should be understood in the light of the rapid urbanization processes that many developing countries have been experiencing in recent decades. In this respect, evidence suggests that the burgeoning challenges posed by increased urbanization to the economic and social futures of developing countries through its eﬀects on the resilience of food systems to unexpecteded shocks, such as disease outbreaks and other nature-induced changes. INTRODUCTION The consequences have been unprecedented in scale and intensity. Livelihoods were devastated due to the inability to maintain job security, food production was compromised and supply chains were distrupted. The adverse eﬀects of a countrywide lockdown combined with weak political, economic, and social interventions had extended beyond income shocks and aﬀected household food security (11, 12). Loss of employment, curtailed contracts and reduced wages exacerbated food insecurity risk (13). Economic lockdown and conﬁnement measures implemented due to the pandemic have impacted employment across sectors within and between countries asymetrically (4–8). It has resulted in increased in unemployment rates, work from home arrangements and aﬀected labor force participation. The overall economic downturn globally forced companies or ﬁrms to downsize their businesses, in some cases even complete shutdown, which got translated as reduced work hours for partial pay or losing their jobs entirely for many employees. The segments of the workforce most likely to be impacted are the most vulnerable groups, less educated low–wage workers, and those with non-standard contracts (temporary contracts, self-employed) (9, 10) and exacerbate the labor market iniquities. Literature on impacts of the pandemic on various sectors has been emerging since the onset of the global COVID 19 pandemic. Research has focused on impacts of the Covid 19 pandemic on social and associated psychological and health impacts due to the restrictions of social and physical mobility of people (18, 19) on the positive and negative environmental impacts (20–22), agriculture,supply chains and food systems (1, 23–26). Extensive focus has been dedicated to observing the potential impacts of the pandemic on various economic indicators such as global poverty, government expenditures, budget deﬁcits, employments etc. limited only to global and national scale (27–31). Contextualized data on the insidious growth and extensive impact of the COVID-19 pandemic on individuals and households (micro scale) is still emerging. The varying eﬀects of COVID on the diﬀerent economic strata needs to be assessed thoroughly on various economic parameters like livelihoods and income, access to markets etc. to build evidence that can support policy formation to develop robust coping strategies that ensure income smoothing and consumption. p y q In India, the sudden nationwide lockdown imposed by the national government from March 24, 2020 to May 31,2020 was one of the most extensive and stringent COVID-19 lockdowns in the world. In thus clamping down, the government’s singular focus was on saving lives, not livelihoods. Frontiers in Public Health \| www.frontiersin.org INTRODUCTION Padmaja R, Nedumaran S, Jyosthnaa P, Kavitha K, Abu Hatab A and Lagerkvist C-J (2022) COVID-19 Impact on Household Food Security in Urban and Peri-Urban Areas of Hyderabad, India. Front. Public Health 10:814112. doi: 10.3389/fpubh.2022.814112 On top of the direct health impacts of the COVID-19, the pandemic has disrupted food supply chains in developing countries, destabilized food prices and created profound negative eﬀects on food security (1). In particular, the measures that governments in developing countries adopted to contain the spead of the virus have caused disruptions in transportation, manufacturing and service provisioning, which subsequently increased unemployment and caused an income loss estimated May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org COVID-19 Impact on Household Food Security Padmaja et al. at USD 220 billion (2). These losses will reverberate across societies and impact education, human rights, and in most cases, basic food security and nutrition (3). of social services, limited access to basic amenities and the resulting public health risks. The relationship between food security, food systems and sustainability needs to be given engaged consideration in the urban areas. Understanding this relationship is crucial because urban poverty and food insecurity are interrelated. However, there has been a lack of or limited systematic analysis of how urbanization aﬀects contemporary food insecurity risk. basic food security and nutrition (3). Economic lockdown and conﬁnement measures implemented due to the pandemic have impacted employment across sectors within and between countries asymetrically (4–8). It has resulted in increased in unemployment rates, work from home arrangements and aﬀected labor force participation. The overall economic downturn globally forced companies or ﬁrms to downsize their businesses, in some cases even complete shutdown, which got translated as reduced work hours for partial pay or losing their jobs entirely for many employees. The segments of the workforce most likely to be impacted are the most vulnerable groups, less educated low–wage workers, and those with non-standard contracts (temporary contracts, self-employed) (9, 10) and exacerbate the labor market iniquities. In India, the sudden nationwide lockdown imposed by the national government from March 24, 2020 to May 31,2020 was one of the most extensive and stringent COVID-19 lockdowns in the world. In thus clamping down, the government’s singular focus was on saving lives, not livelihoods. The lockdown froze economic activity across the country and delivered a large aggregate supply and demand shock to the economy. INTRODUCTION That is, urbanization is often associated with poverty, overburdening The combined eﬀect of food price and income shocks arising from global crises such as the COVID-19 pandemic has been suggested as the likely cause of a sharp increase in hunger and poverty in low-income countries (37). There are reasons to expect that the pandemic has deeply altered food environments. First, the way people engage and interact with the food system to acquire, prepare and consume food has changed due to the lockdowns and the subsequent supply chain disruptions. Although most households in urban regions are net buyers of food, higher food prices are likely to have reduced household access to staple food. Secondly, the economy-wide negative impact of the pandemic and the subsequent lockdown which resulted in a loss of jobs across the country, has likely further limited households’ ability to purchase food at higher prices (38). This only reinforces the need to understand the lockdown’s impact on household food security status and coping mechanisms in the face of income shortfalls and food price shocks (39). With this background, this study adds to the emerging literature on the impact of COVID-19 on food security in developing countries in several important ways. Leveraging on May 2022 \| Volume 10 \| Article 814112 2 COVID-19 Impact on Household Food Security Padmaja et al. FIGURE 1 \| The grid of locations selected on the basis of GIS data for a study of the COVID-19 lockdown’s impact on household food security in Hyderabad, India [based on Gumma et al. (40)]. FIGURE 1 \| The grid of locations selected on the basis of GIS data for a study of the COVID-19 lockdown’s impact on household food security in Hyderabad, India [based on Gumma et al. (40)]. FIGURE 1 \| The grid of locations selected on the basis of GIS data for a study of the COVID-19 lockdown’s impact on household food security in Hyderabad, India [based on Gumma et al. (40)]. level in the context of the pandemic and the coping strategies employed to smoothen consumption. 1Food Insecurity Experience Scale \| Voices of the Hungry \| Food and Agriculture Organization of the United Nations (fao.org). INTRODUCTION the multiple point data availability spanning across 4 years between 2018 and 2021 from a larger project the study makes a unique contribution to the emerging literature to understand the dynamics in the food security status in the aftermath of the pandemic and the phased lock down at a micro level using the food security status of the households prior to the pandemic in 2018 as benchmark capturing the spatial diﬀerences among urban and peri-urban households. We also attempt to assess the eﬀect of pandemic on food security mediated through impact on changes in labour force participation and associated income loss. 2A sub-administrative division commonly used in India and some Asian countries. 3Anganwadis are rural child care centers in India. They were started by the Indian government in 1975 as part of the Integrated Child Development Services (ICDS) program to combat child hunger and malnutrition. Each center is managed by an anganwadi teacher. 4The elected head of a village assembly or gram sabha. 5ASHA = Accredited Social Health Activist. 6Census and Survey Processing System: https://www.csprousers.org/. Study Location The selection of the study area for the present study was based on the GIS/remote-sensing analysis by Gumma et al. (40), which assessed urban expansion and other land-use and land-cover changes in Hyderabad from 2005 to 2016. Using the outer ring road of Hyderabad as a boundary of the city (Figure 1) and following the method of Gumma et al. (40), we identiﬁed four quadrants/grids, each having similar features, on the map of Hyderabad: two grids in peri-urban areas and two in urban areas (Figure 1). In this paper, we seek to reach the following interrelated objectives: • To analyze the impact of COVID-19 on the livelihoods of households residing in urban and peri-urban areas. • To understand, by employing the Food Insecurity Experience Scale (FIES)1, the dynamics of food security at the household The population data of each mandal2 falling within the grid, fully or partly, were collected from the District Census Handbook May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 3 COVID-19 Impact on Household Food Security Padmaja et al. TABLE 1 \| Grid-wise proportionate sampling framework. Grid number Population Category Sample proportion (%) 10 209,524 Rural-Urban 16.51 17 347,141 Urban-Urban 27.36 19 230,543 Rural-Urban 18.17 23 461,156 Rural-Urban 36.34 TABLE 1 \| Grid-wise proportionate sampling framework. some additional modules. Data from these two rounds is not considered for this study. After the onset of the COVID-19 pandemic in March 2020, from March 24 to May 31, 2020 strict lockdown restrictions were imposed nationwide in India on movement of goods and people. During the pandemic, there were three phases of withdrawal (unlock) of restrictions. The ﬁrst unlock covered the period from June 1 to July 31, 2020 when certain essential services were restored and limited movement of people was allowed. The second unlock covered the period from August 1 to September 30, 2020 when there was a gradual opening up of the economy for a restricted time during the day and curfews were restricted to the late evening and night. The third unlock covered the period from October 1 to November 30, 2020, which saw the economy starting to get back to normalcy with restrictions on businesses and movement completely removed for all practical purposes. 2011, and the proportion of geographic area contributed to the grid by each mandal was calculated. Household Selection It should be highlighted that this study is part of a larger project in which a longitudinal panel of data was to be collected in four rounds (between 2018 to 2021) with the aim of identifying the status and implications of urbanization on food and nutrition security. The selection of the households was done in consultation with the local government workers (Anganwadi teachers3, sarpanches4, and ASHA5 workers) and based on the sampling strategy illustrated in the previous sub-section and presented in Table 1. Study Location The mandal population in the respective grid area was proportional to its geographical area in the grid. The proportion of the geographical area of a mandal within the grid was multiplied by the total population of the mandal. Using this method, the total population of each grid and its contribution to the total sample were calculated (Table 1, Appendix 1). Refer to Supplementary Materials for appendices. y During December 2020 to January 2021, a telephonic/remote survey was conducted on the same sample households as in the ﬁrst round to understand the impact of COVID-19 pandemic on household food security. The telephonic survey covered the post-outbreak lockdown and three phases of withdrawal (unlock) of restrictions. The mobility restrictions due to the pandemic imposed by the Government of Telangana did not enable personal face to face interviews during this period. Out of the 660 households that were interviewed in round one, only 325 households could be interviewed through the telephonic survey. Audio recorded consent was taken from the households after the objectives of the survey were explained to the respondents. Data was recorded on tablet computers running KoBo Toolbox software7. 7https://www.kobotoolbox.org/ 8Refers to the nature of primary employment of the main income earner of the household: self-employment, salaried work which includes private and public sector, casual work, etc. 8Refers to the nature of primary employment of the main income earner of the household: self-employment, salaried work which includes private and public sector, casual work, etc. 7https://www.kobotoolbox.org/ 8 Data Collection The pre- and post-pandemic survey questionnaires including the standardized Food Insecurity Experience Scale (FIES) are presented in Appendix 2 and in Appendix 3, respectively. We would also like to particularly mention the following: Prior to data collection in the ﬁrst round, a written approval for the survey was taken from the local administration of the Greater Hyderabad Municipal Corporation. The formal approval letters helped our personnel gain access to the chosen locations and elicit the cooperation of people in the community. As part of the ethical consideration, prior written consent was also taken from the respondents before each interview with the households. • The number of unemployed days for a household was calculated on the basis of the number of days of participation in the labor force during March—November 2020. The respondents self reported the availability of employment or non-employment during this period. The ﬁrst round of data collection took place before the onset of the pandemic (October, 2018-February, 2019). In this round, 660 households were selected on the basis of the criteria laid down for this project as explained in the above section. The enumerators recorded the data on tablet computers using CsPro software6 For this present study, this round forms our baseline data against which we are measuring the changes. The second and third rounds of data collection for the project was carried out during June 2019 and November 2019, respectively. These rounds had some common modules from round one and also • The self-reported actual income and the approximate range of income received by the household during February 2020 was taken as the baseline to assess changes in income and income class. Income received in February 2020 served as the baseline as it was the closest proxy to liquid cash available within the household to meet immediate expenses during the lockdown. We used 240 days as the benchmark ﬁgure for the purpose of our computation (of the number of unemployed days) as it was the maximum number of days that a primary income earner could have been employed across all types of employment8 during the May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 4 COVID-19 Impact on Household Food Security Padmaja et al. above period. Thus, unemployment percentage was calculated from the number of days out of 240 that a household reported its primary earner as being out of employment. Background Characteristics of Households by Food Insecurity State The majority of households in our survey experienced a deterioration in their food security status during the pandemic. Households with moderate food insecurity status had the highest FIES score as well as number of unemployed days and lowest income. Around 25% of the households experienced mild food insecurity, and 17% experienced moderate food insecurity. Around 15% of the households reported an improvement in their food security status, while about 40% said it had worsened. Though the pandemic was a covariate shock, its idiosyncratic nature is evident from its diﬀerential impact on household food insecurity even in cases where the primary income earner belonged to the same class of worker. Overall, workers from peri-urban households bore a greater brunt of the impact than their counterparts in urban areas, as is evident from the higher incidence of worse food insecurity (both mild and moderate) among such households (Tables 2A,B, Figure 5). g However, while urban households found it less diﬃcult to get back to work, they experienced greater income reduction compared to their counterparts in the peri-urban areas. Income loss was particularly steeper for households engaged in a self- employed enterprise (with and without employees of their own). This could be attributed to the dampening of overall demand due to income and job losses as a consequence of the pandemic. The impact of dampening of overall demand is evident from the higher loss of income suﬀered by self-employed urban households with one or more employees compared to similar households in the peri-urban areas. Between urban and peri- urban areas, income losses sustained by other categories of households were comparable (Figure 3). Frontiers in Public Health \| www.frontiersin.org Data Collection Households were categorized into three groups in terms of change in income status—“improved,” “reduced,” and “maintained status quo”—relative to the income range reported by them for the month of February 2020. The categorizing households by food insecurity level was done as per Ballard et al. (41). Household food insecurity was the outcome variable of interest in these models, whose covariates were categorical variables coded as 1 if the household was food-insecure, mildly food-insecure, or moderately food-insecure, and 0 if it was food-secure. The other categorical variables included in the model were area of residence, type of employment (and interaction between these two variables), coping strategies adopted, income, and number of unemployed days. reduction in income status, perhaps due to a pay cut. About 5% of households earning regular salaries improved their income status. These were households whose primary earners had jobs related to health sectors or had got into jobs oﬀering a higher salary. The pandemic and the restrictive measures taken by the government had a diﬀerential impact on diﬀerent classes of workers; salaried workers having secure employment were the least aﬀected in both urban and peri-urban areas. A comparison of the maximum incomes received by households during the lockdown and three phases of unlock with corresponding income data gathered in our pre-pandemic survey (October 2018 to February 2019) showed that most of the households that drew their income from regular salaried work with secure employment managed to maintain their income level or even saw a slight increase. In both urban and peri-urban areas, households whose income came from casual work in non-farm employment experienced a decline in income compared to the pre-lockdown period. The number of such households was slightly higher in the peri-urban areas (Figure 4). Economic Loss From the information provided by each household on labor force participation by its primary income earner, we computed the number of unemployed days endured by the household during the lockdown and three phases of unlock, a period spanning from March 24 to November 30, 2020 (In our sample of households, we found that there was none that had more than one employed member at any time during this period.). During the lockdown period, the majority of households in our sample-except a small proportion-were not able to participate in the work force. In the subsequent three unlock phases, households located in the urban areas found it relatively easier to get back into the job market than peri-urban households. Accordingly, the percentage of unemployed days was relatively higher for peri-urban households compared to urban households (Figure 2). Determinants of Household Food Insecurity in the Post-pandemic Period Based on a logistic regression model approach, Table 3 shows that income is negatively associated with all forms of food insecurity. We found that income and the type of employment of the primary income earner of a household are the determinants of a change in the food security status of a household. Since income stability is largely dependent on the nature of primary employment of a household (self-employment, regular salaried work, casual work, etc.), being employed in the private sector is associated with a lesser likelihood of household food insecurity. Similarly, residing in an urban area was associated with an increased likelihood of household food insecurity. Self-employed households living in urban areas seemed to face an increased risk of being food-insecure, a ﬁnding that could be attributed to the Analyzing the data for changes in household income status,8 we found that, in urban as well as peri-urban areas, the majority of households that depended on casual work, which was not related to agriculture or allied activities or were self-employed, experienced a reduction in income status. The number of such households reporting a lower income status during the pandemic was higher in the peri-urban areas. On the other hand, the majority of households that had a primary income earner in a regular salaried job did not experience a change in income status; however, 24% of such households did experience a May 2022 \| Volume 10 \| Article 814112 5 COVID-19 Impact on Household Food Security Padmaja et al. FIGURE 2 \| Unemployment (%) experienced by households in urban and peri-urban areas of Hyderabad, India during the COVID-19 lockdown and three-phased unlock (removal of restrictions). large income losses such households have suﬀered during and after the lockdown (Table 3). The negative coeﬃcients on savings could be attributed to the possibility that these households have lesser savings to tide over a food insecurity situation for long. Once the savings are spent, these households ﬁnd themselves food-insecure (Table 3). The positive coeﬃcients on loans show that access to ﬁnance was employed as a coping strategy by food-insecure households in both urban and peri- urban contexts. The positive association between self-employed urban households and food insecurity is plausibly due to the supply chain disruptions that impacted their business and the overall dampening of demand due to reduction in income. Frontiers in Public Health \| www.frontiersin.org Dynamics of Household Food Security between private job holders who have greater income stability and also the inclination and opportunity to save more money and resources relative to, for example, non-farm workers. This proposition indeed ﬁnds resonance in the negative coeﬃcients we found for urban households employed in the private sector. However, this beneﬁt did not accrue to self-employed households, which endured deterioration of food security as they suﬀered large reductions in income besides having less access to savings (Table 4). In assessing the food security status of households during the pandemic, we used data that we had collected before the onset of the pandemic (pre-pandemic food security status) as a benchmark against which to understand the changes that occurred during the 5-week-long lockdown and the three-phased relaxation of restrictions. We found that the pre-pandemic food security status of a household based on the assessment using the ﬁrst round of data collected in 2018 was a major determinant of its food security status in the lockdown-unlock period as well. This ﬁnding highlights the need to prioritize and target the already vulnerable households and ensure that they are covered by the assistance programs and social safety net schemes launched by various agencies in the aftermath of the outbreak. Savings are positively associated with improvement in household food security, while also being negatively associated with deterioration in food security status. This eﬀect points to a likely correlation Determinants of Household Food Insecurity in the Post-pandemic Period The results reﬂect that those employed in the private sector were less likely to experience food insecurity. Urban households were found to be at greater risk of food insecurity than peri- urban households with similar employment status. This possibly was due to the drastic reduction in job opportunities during the lockdown and the higher cost of living in the urban areas. Urban households, despite having better access to ﬁnancial resources/services such as loans and savings, found it diﬃcult to cope with the stress of food insecurity as can be seen from the positive association between urban dwelling and food insecurity. May 2022 \| Volume 10 \| Article 814112 6 Padmaja et al. COVID-19 Impact on Household Food Security FIGURE 3 \| Change in income experienced by various categories of workers during the lockdown and unlock phases of the COVID-19 pandemic in urban and peri-urban areas of Hyderabad, India. FIGURE 3 \| Change in income experienced by various categories of workers during the lockdown and unlock phases of the COVID-19 pandemic in urban and peri-urban areas of Hyderabad, India. Frontiers in Public Health \| www.frontiersin.org Coping Strategies We found that households with diﬀerent food insecurity status employed diﬀerent coping strategies—depending on their access to such options. Households in both urban and peri-urban areas employed similar coping strategies to deal with the stress to their food security. Households in a state of mild food insecurity tended to fall back on their store of food as a coping mechanism. Households who were in a more intense state of food insecurity May 2022 \| Volume 10 \| Article 814112 7 Padmaja et al. COVID-19 Impact on Household Food Security FIGURE 4 \| Change in household income status (improved/reduced/status quo) relative to pre-pandemic income levels experienced by different categories of households (categorized by type of employment) in urban and peri-urban areas of Hyderabad, India. FIGURE 4 \| Change in household income status (improved/reduced/status quo) relative to pre-pandemic income levels experienced by different categories of households (categorized by type of employment) in urban and peri-urban areas of Hyderabad, India. TABLE 2A \| Food insecurity status of sample households before and after COVID-19 outbreak in March 2020 in Hyderabad, India. Pre-pandemic Post-pandemic Peri-urban Urban Peri-urban Urban Mild Moderate Secure Mild Moderate Secure Mild Moderate Secure Mild Moderate Secure No.of households 22 19 151 12 13 108 49 33 109 34 21 76 FIES score 2.04 5.42 0.00 2.33 7.07 0.00 2.18 5.15 0.00 2.32 4.76 0.00 Proportion 6.76 5.84 46.46 3.69 4.00 33.23 15.07 10.15 33.53 10.46 6.46 23.38 %Change 8.31 4.31 −12.93 6.77 2.46 −9.85 FIES score is the number of afﬁrmative responses of the households to the Food Insecurity Experience Scale administered. LE 2A \| Food insecurity status of sample households before and after COVID-19 outbreak in March 2020 in Hyderabad, India. May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 8 COVID-19 Impact on Household Food Security Padmaja et al. (see columns “moderate” and “FIES” in Table 3) tended to take loans or liquidate into savings to cope with the stress (Table 3). Households in the “mild food insecurity” category have relatively more access to stored food compared to other households. Similarly, those in the “moderate food insecurity” category have relatively more access to ﬁnancial resources in the form of loans— from both formal and informal sources—but have the lowest access to savings. Coping Strategies Households that experienced an improvement in their food security status have relatively high access to ﬁnancial resources both in the form of loans and savings besides being protected by various social safety schemes. On the other hand, households that experienced a deterioration in their food security status have better access to loans compared to savings (Table 5). TABLE 2B \| Improvement/deterioration in household food security (in terms of FIES score) due to impact of COVID-19 outbreak in March 2020 in Hyderabad, India. Improved (n =45)† Deteriorated (n = 123)ℓ FIES score 0.82 3.47 Unemployed days 136 140 Family Size 5.00 4.00 †Improved: Household whose food security status has improved in the pandemic period compared to the pre pandemic period. ℓDeteriorated: Household whose food security status has deteriorated in the pandemic period compared to the pre pandemic period. TABLE 2B \| Improvement/deterioration in household food security (in terms of FIES score) due to impact of COVID-19 outbreak in March 2020 in Hyderabad, India Households whose food security status deteriorated (Table 2B) had the highest number of unemployed days on average. Households experiencing “moderate food insecurity” have less access to savings like those in the “deteriorated” category. More than 70% of the households in the “moderate” and “deteriorated” categories borrowed money to cope with the exogenous shock of the pandemic. Households in the “mild food insecurity” category employed stored food as a coping strategy to a greater extent as they have that tendency to store food more FIGURE 5 \| Household food insecurity status by type of employment in urban and peri-urban areas of Hyderabad. May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 9 COVID-19 Impact on Household Food Security Padmaja et al. TABLE 3 \| Determinants of household food insecurity in urban and pei-urban locations during the COVID-19 pandemic (March-November 2020). Coping Strategies Insecure Mild Moderate FIES Intercept 5.137* 2.846 0.878 (2.047) (2.076) (2.529) Income −0.554* −0.433 −0.315 −0.432** (0.207) (0.211) (0.259) (0.171) Unemployed days 0.004* 0.001 0.007 0.005* (0.002) (0.002) (0.003) (0.002) Place (Base category: Peri-Urban) Urban −1.406* −1.726* 0.337 −0.870 (0.77) (0.896) (1.013) (0.688) Occupation (Base category: Casual Non-farm worker) Farm −1.101 0.120 −1.444 (0.974) (0.972) (0.893) Private sector −0.975** −0.299 −1.134* −1.014** (0.479) (0.501) (0.667) (0.43) Public sector −0.156 −0.267 0.562 0.213 (0.684) (0.719) (0.875) (0.609) Self-employed −0.877* −0.0173 −1.219* −0.828* (0.484) (0.494) (0.674) (0.429) Others −0.022 −0.953 1.079 0.440 (1.005) (1.171) (1.048) (0.847) Place × Occupation Urban × Farm Urban × Private sector 0.925 0.914 0.505 0.926 (0.684) (0.774) (0.917) (0.628) Urban × Public sector 0.540 1.289 −0.789 −0.034 (1.045) (1.105) (1.453) (0.923) Urban × Self-employed 1.581 2.075* −0.637 0.691 (0.748) (0.796) (1.103) (0.65) Urban × Others −0.069 1.936 −0.974 (1.639) (1.747) (1.53) Coping strategies Place × Savings Peri-Urban × Yes −1.291* −0.347 −1.982* −1.324* (0.368) (0.384) (0.591) (0.343) Urban × Yes −1.082 0.214 −2.518* −1.313* (0.488) (0.547) (0.796) (0.477) Place × Loans Peri-Urban ×Yes 0.676* 0.221 1.186* 0.718 (0.386) (0.389) (0.624) (0.353) Urban × Yes 1.118 0.548 1.122* 0.962 (0.435) (0.462) (0.59) (0.388) Place × stored food Peri-Urban × Yes 0.373 0.787 −0.493 −0.079 (0.374) (0.377) (0.524) (0.332) Urban × Yes 0.809* 1.157 −0.336 0.323 (0.483) (0.551) (0.669) (0.453) Observations 316 316 306 316 Pseudo R2 0.176 0.071 0.287 0.099 Akaike’s Crit 393.099 370.362 237.258 871.183 Standard errors are in parentheses. p < 0.01, p < 0.05, p < 0.1. Logit estimates reported. The dependent variable in column 2 is 1 if household is food- insecure and 0 if there is no food insecurity. The dependent variable in column 3 is 1 if household faces mild food insecurity and 0 if no food insecurity, or faces moderate or severe food insecurity. The dependent variable in column 4 is 1 if the household faces d t d f d i it d 0 if th i f d i it ild f d TABLE 4 \| Dynamics of household food insecurity ‘as assessedin terms of FIES scores. Coping Strategies Insecure Mild Moderate FIES Intercept 5.137 2.846 0.878 (2.047) (2.076) (2.529) Income −0.554* −0.433 −0.315 −0.432 (0.207) (0.211) (0.259) (0.171) Unemployed days 0.004* 0.001 0.007 0.005* (0.002) (0.002) (0.003) (0.002) Place (Base category: Peri-Urban) Urban −1.406* −1.726* 0.337 −0.870 (0.77) (0.896) (1.013) (0.688) Occupation (Base category: Casual Non-farm worker) Farm −1.101 0.120 −1.444 (0.974) (0.972) (0.893) Private sector −0.975** −0.299 −1.134* −1.014** (0.479) (0.501) (0.667) (0.43) Public sector −0.156 −0.267 0.562 0.213 (0.684) (0.719) (0.875) (0.609) Self-employed −0.877* −0.0173 −1.219* −0.828* (0.484) (0.494) (0.674) (0.429) Others −0.022 −0.953 1.079 0.440 (1.005) (1.171) (1.048) (0.847) Place × Occupation Urban × Farm Urban × Private sector 0.925 0.914 0.505 0.926 (0.684) (0.774) (0.917) (0.628) Urban × Public sector 0.540 1.289 −0.789 −0.034 (1.045) (1.105) (1.453) (0.923) Urban × Self-employed 1.581 2.075* −0.637 0.691 (0.748) (0.796) (1.103) (0.65) Urban × Others −0.069 1.936 −0.974 (1.639) (1.747) (1.53) Coping strategies Place × Savings Peri-Urban × Yes −1.291* −0.347 −1.982* −1.324* (0.368) (0.384) (0.591) (0.343) Urban × Yes −1.082 0.214 −2.518* −1.313* (0.488) (0.547) (0.796) (0.477) Place × Loans Peri-Urban ×Yes 0.676* 0.221 1.186* 0.718 (0.386) (0.389) (0.624) (0.353) Urban × Yes 1.118 0.548 1.122* 0.962 (0.435) (0.462) (0.59) (0.388) Place × stored food Peri-Urban × Yes 0.373 0.787 −0.493 −0.079 (0.374) (0.377) (0.524) (0.332) Urban × Yes 0.809* 1.157 −0.336 0.323 (0.483) (0.551) (0.669) (0.453) Observations 316 316 306 316 Pseudo R2 0.176 0.071 0.287 0.099 Akaike’s Crit 393.099 370.362 237.258 871.183 Standard errors are in parentheses. p < 0.01, p < 0.05, p < 0.1. Logit estimates reported. The dependent variable in column 2 is 1 if household is food- insecure and 0 if there is no food insecurity. The dependent variable in column 3 is 1 if TABLE 4 \| Dynamics of household food insecurity ‘as assessedin terms of FIES scores. TABLE 3 \| Determinants of household food insecurity in urban and pei-urban locations during the COVID-19 pandemic (March-November 2020). Occupation (Base category: Casual Non-farm worker) p p p Logit estimates reported. Dependent variable in column 2 is 1 if household experienced improvement in food security status and 0 if there was no change in food security status, or faced deterioration in food security status. Dependent variable in column 3 is 1 if household experienced deterioration in food security status and 0 if there was no change in food security status, or experienced improvement in food security status. Standard errors are in parentheses. *p < 0.01, p < 0.05, p < 0.1. Frontiers in Public Health \| www.frontiersin.org Standard errors are in parentheses. Coping Strategies Improved Deteriorated Intercept −0.388 2.712 (2.702) (1.985) Income 0.715 −0.422* (0.622) (0.213) Unemployed days −0.006 0.004* (0.009) (0.002) Precovid score 3.063* −0.431* (0.694) (0.108) Region (base category: Peri-Urban) Urban 6.557 −0.989 (5.687) (0.789) Occupation Farm −0.799 −0.098 (27.111) (1.013) Private sector 2.676 −1.15 (1.638) (0.501) Public sector −0.072 −0.788 (2.182) (0.73) Self-employed 1.516 −1.219 (1.955) (0.509) Others 4.739 0.635 (5.263) (1.247) Region × Occupation Urban× Farm Urban × Private sector −7.095 1.244* (4.468) (0.71) Urban × Public sector −2.366 1.298 (3.558) (1.077) Urban× Self-employed −2.548 1.728 (4.28) (0.781) Urban × Others −14.31 (165.654) Coping strategies Region × Savings Peri-Urban × Yes 6.212* −1.349* (2.201) (0.389) Urban × Yes 4.338 −1.36 (4.763) (0.553) Region × Loans Peri-Urban × Yes −0.787 0.781* (1.279) (0.41) Urban × Yes −3.05 0.99 (1.99) (0.451) Region × stored loans Peri-Urban × Yes 0.480 0.380 (1.076) (0.396) Urban × Yes −3.141 0.267 (4.791) (0.522) Observations 316 312 Pseudo R2 0.825 0.211 Akaike’s crit 84.11 365.35 Standard errors are in parentheses. p < 0.01, p < 0.05, p < 0.1. Logit estimates reported. Dependent variable in column 2 is 1 if household experienced improvement in food security status and 0 if there was no change in food security status TABLE 3 \| Determinants of household food insecurity in urban and pei-urban locations during the COVID-19 pandemic (March-November 2020). Economic Impacts and Impacts on Food Security There has been a severe contraction of labor demand in India that has materialized unevenly across diﬀerent occupations and skill levels. According to the Center for Monitoring Indian Economy,9 the labor force participation rate fell to an all-time low in March 2020, and the unemployment rate rose sharply. Employment slumped by 9 million jobs, going from 443 million in January 2020 to 434 million in March 2020. This decline was the result of a fall of 15 million (from 411 million to 396 million) in the number of employed people and a 6 million rise (from 32 million to 38 million) in the number of unemployed people in March 2020. However, not all non-farm workers were equally aﬀected by the lockdown. The diﬀerential impact observed across the informal sector could be attributed to the two separate branches of employment that exist within the informal sector: One group comprises those who operate/work informally within informal enterprises and those outside of informal enterprises Valodia et al. (52). The second group includes casual laborers, domestic servants, and other forms of labor-intensive employment, who tend to rely on cash income. Those in the latter group may be the most at risk of food insecurity as they face the challenge of inconsistent income. Households with members employed in the formal sector, where income is regular, do not have the same risk of food insecurity (53). As the COVID-19 lockdown progressed, 80% of India’s informal workers lost their jobs (42). The livelihoods of daily workers, street vendors, small enterprises, and retail traders came to a complete stop for various reasons (43). More than 50% of informal workers and their families are estimated to have been pushed into poverty due to the reduction in labor incomes triggered by the lockdown (44). In the urban regions of the state of Telangana—whose capital is Hyderabad—the labor force participation rate and the greater unemployment rate were 40.47 and 5.20%, respectively, between September and December 2020 (45). Being the major city in Telangana, Hyderabad was an interesting case to examine the trends in labor force participation in and around the city. The trends observed in our study can be attributed to the restrictions imposed on physical mobility in the city when the lockdown was announced on March 24, 2020. Economic Impacts and Impacts on Food Security For people who live in the peri-urban areas around the city and commute to work in the non-farm sector everyday, this cut oﬀ access to employment. The mass layoﬀs and the closure of many small businesses in and around the city further added to these diﬃculties. The results of our study indicate that the lockdown’s impact on livelihoods was more severe in peri-urban areas than in urban areas. This is consistent with the general consensus in India that the pandemic hit small businesses, daily-wage earners, and low-wage earners, leaving them with no jobs or reduced incomes (46). These ﬁndings are also in line with the macro trend observed at the national level: the national unemployment rate increased to 26% in April 2020, before easing to 19% in the subsequent months (45). The ﬁndings of our study show that the type of employment in which the primary income earner of a household is engaged and the income it earns for the household are the determinants of any change in the food security status of the household. The sensitivity of food consumption to income changes, particularly labor income, is well-established in literature (54, 55). Liquid savings and access to credit inﬂuence heterogeneous consumption responses to income shocks (56). Our results support these ﬁndings as households that are primarily employed in sectors that provide better job security and greater stability in income relative to non-farm work which is characterized by high levels of insecurity both in terms of job and income. As a result, they are less likely to face more intense levels of food insecurity (moderate category). Overall, our results support the ﬁndings of Egger et al. (57), who report that both negative income shock and income level aﬀect the predicted probability of households facing food insuﬃciency or insecurity. It is reasonable to expect that income losses across sectors in the urban areas resulted from the closure of businesses across sectors, especially small businesses. Such a major impact on livelihoods leads to reduced economic access to resources and services and thereby increases the risk of food insecurity. In addition, food accessibility in urban areas is largely the result of food aﬀordability (47). 9https://unemploymentinindia.cmie.com/ Coping Strategies May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org Frontiers in Public Health \| www.frontiersin.org 10 COVID-19 Impact on Household Food Security Padmaja et al. than other kind of households. Nearly 60% of the households experiencing “mild food insecurity” lacked access to savings. local markets (48). Reduced purchasing power due to a drop in household incomes impacts their food security. Besides reduction in income, urban consumers were also aﬀected by the disruption of long-distance food supply chains, which resulted in product non-availability and higher retail/online prices. There was an 8% decline in the availability of fruits and vegetables and a 14% decline in the availability of edible oils in three of India’s metropolitan cities immediately after the lockdown was imposed. The prices of major food items— constituting more than 25% of the urban food consumption basket—spiked immediately after the lockdown was announced in March 2020: pulses by 6%, edible oils by 3.5%, potatoes by 15% and tomatoes by 28% (49, 50). Economic access to food was constrained due to reduced income, restrictions on physical movement and restricted consumer access to aﬀordable food markets, as reﬂected in the Composite Consumption Behavior Change Index (CCBCI) (51). Economic Impacts and Impacts on Food Security The increase in staple cereal prices due to the COVID-19 outbreak in Asia has started to impact prices in Regular salaried employment such as in the private sector helps households improve their food security status gradually compared to households primarily employed in the non-farm sector as casual labor. Most of the households in our sample went from being food-secure in the pre-pandemic period to being food-insecure since the outbreak in March 2020. Relatively, private sector jobs oﬀer higher job security and other optional beneﬁts than jobs in the non-farm sector. Private sector jobs also oﬀer better income stability compared to non-farm jobs. With relatively higher income, households are better equipped to May 2022 \| Volume 10 \| Article 814112 11 Frontiers in Public Health \| www.frontiersin.org COVID-19 Impact on Household Food Security Padmaja et al. TABLE 5 \| Household access to coping strategies based on FIES scores. Percent households with access Percent households without access Coping strategy Mild Moderate Improved Deteriorated Mild Moderate Improved Deteriorated Loans 66.27 80.36 64.44 71.54 33.73 19.64 35.56 28.46 Savings 43.27 16.07 71.11 30.08 56.63 83.93 28.89 69.92 Stored food 57.83 41.07 28.99 52.03 42.17 58.93 71.11 47.97 Gov Aid 86.75 87.5 93.33 87.8 13.25 12.50 6.67 12.20 Mild and Moderate denote actual food insecurity status of the household in the pandemic period.Improved and deteriorsated denote the dynamics in the food security status of the household in the pandemic period compared to the pre pandemic period. TABLE 5 \| Household access to coping strategies based on FIES scores. make savings, which can be utilized to withstand an unexpected shock corroborating with ﬁndings of Gjerston, 2016 (58). Our ﬁndings support similar ﬁndings by Kesar et al. (59), who reported that self-employed households were better oﬀin terms of food security, especially in urban areas despite facing job losses. Furthermore, our results also corroborate the ﬁndings of literature, our results show that households that are able to overcome short-term liquidity constraints by borrowing seem to smoothen food consumption and are less likely to be food- insecure, implying that they can mitigate the risk of becoming food-insecure (58, 60–63). coping strategies to tackle the stress to their food security caused due to an income shock (76). Since a large portion of the total household income goes toward food purchase and consumption in urban areas, these ﬁndings are expected. Economic Impacts and Impacts on Food Security The monthly percapita food consumption expenditure in the urban regions of the study location (Hyderabad district) as per the 68th national sample survey stood at INR 1196.78 which was higher than the state avaerage in 2011–12. The inﬂationary pressure as reﬂected by the CPI and food inﬂation during the lockdown and the subsequent months of phased withdrawl were high at 6.6 and 9.1%, respectively, in 2020–21 in the country which had a huge impact on aﬀordability. The impacts of high food inﬂation on consumption during the period were reﬂected in the CCBCI as elucidated in the above section on economic impacts. Our results on increased prevalence of food insecurity in the post pandemic period resonates and corroborates with ﬁndings of Srivastava and Sivaramane (77) who estimate reduction in the overall food expenditure between 4.98 and 21.34% compared to the pre pandemic period. Households that were already food-insecure before the pandemic experienced higher levels of food insecurity on a relative scale. The ﬁndings of Gaintan-Rossi et al. (64) support our ﬁnding that economic shocks more strongly aﬀect households that were already vulnerable prior to the shock. Though households did use savings to cope with the exogenous shock of the pandemic, they probably did not have suﬃciently large savings to sustain their food security for an extended period. Once the savings dried up, they were not in a position to access adequate food. CONCLUSIONS Our ﬁndings also show that there exists a relationship between low food security and households whose members are employed as casual workers (65). This link could be attributed to the implicit challenges present in the informal sector stemming from the absence of formal regulations. The multiple challenges faced by those employed in the informal sector are well-documented in the literature (66–69). The diﬃculties faced by a household in getting an assured employment that provides a sustained livelihood is one of the primary challenges in achieving food security in the urban areas. Food accessibility in urban areas is largely a result of food aﬀordability (48). The COVID-19 pandemic has negatively impacted food systems and food security in the Hyderabad region of India. The impact of the pandemic on the households has been heterogenous.Households have experienced both idiosyncratic and covariate shocks (78, 79). The ﬁndings of our study reﬂect the transitory nature of food security in the region as a result of the shock. Household food security dynamics are largely inﬂuenced by the sector in which the main earning member of a household is employed, income, and access to diﬀerent coping mechanisms. The use of coping strategies seems to depend on their availability and accessibility to the household. Our results reaﬃrm the signiﬁcance of employment oppurtunities and savings for the poor to circumvent unexpected shocks.Our study thus underscores the need for policy that promotes saving behavior among the poor in addition to ensuring them secure employment oppurtunities that provides stable income. These results conﬁrm the importance of savings for poor households and underline the crucial role policies can play to support savings and ensure stable incomes through secure employment opportunities. Ex-ante or forward-looking risk and vulnerability analyses are essential for targeting and implementing risk Frontiers in Public Health \| www.frontiersin.org May 2022 \| Volume 10 \| Article 814112 DATA AVAILABILITY STATEMENT The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh. 2022.814112/full#supplementary-material The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation upon request. ACKNOWLEDGMENTS The authors gratefully acknowledge the assistance of enumerators who participated in data collection. The authors thank the community members/respondents who took part in the study. Coping Strategies The coping strategies observed by our study corroborate with patterns observed in other developing countries and Low and Middle Income Countries (LMICs). Households have used formal and informal borrowing as a strategy to meet immediate expenses (70–73). Lack of access to such a coping strategy is reported to increase the likelihood of a household being food- insecure (74, 75). Also, the fact that most of the households’ food security has deteriorated implies that they also relied on and employed both food-based coping strategies of reducing the quantity and quality of food consumption and ﬁnancial May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 12 COVID-19 Impact on Household Food Security Padmaja et al. ETHICS STATEMENT mitigation interventions. Since our study is based on cross- sectional data, the ﬁndings of the study are not generalizable to a wider context. We recommend such assessments in the future which will help improve the preparedness of society and communities to cope with such unprecedented situations through targeted eﬀorts. The studies involving human participants were reviewed and approved by ICRISAT. The patients/participants provided their written informed consent to participate in this study. FUNDING Another limitation of the study is its inability to apply a gender lens to this understanding of food security during the pandemic. There were cases when women could not be contacted during our telephonic survey as many of them did not have access to a personal mobile phone, or their husbands were not interested in their wives participating in the telephonic survey. Further, women we contacted also said their care responsibilities had increased due to the pandemic and hence they were not available to participate in the survey. A gendered understanding would have brought out interesting insights on the diﬀerential impacts of the pandemic. This study was supported by funding (Grant Number: 2016- 00350) from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS). This study was supported by funding (Grant Number: 2016- 00350) from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS). This study was supported by funding (Grant Number: 2016- 00350) from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS). AUTHOR CONTRIBUTIONS Using remote/telephonic survey as a methodology for data collection during the pandemic period is not unproblematic. Even though these surveys are cost eﬀective compared to face to face interviews, the most important issue is the reduction in the response rate, i.e., the loss in the sample due to either telephone not working, change in the contact numbers given earlier, and repondents refusal to take the survey. Other issues with phone- based that were observed include infrastructural constraints in some settings (e.g., electricity and mobile connectivity issues), length of the interview and concerns around ensuring participant privacy. Thus, even though phone-based surveys are considered suitable in many settings, these limitations appear and have to be addressed. RP: conceptualization, methodology, data curation, investigation, formal analysis, writing—original draft, and writing—review and editing. SN: conceptualization, methodology, writing—original draft, and writing—review and editing. PJ and KK: data curation, investigation, formal analysis, and writing—original draft. C-JL and AA: conceptualization, methodology, writing—review and editing, and funding acquisation. All authors contributed to the article and approved the submitted version. REFERENCES Eighth Edition Updated Estimates and Analysis Return to Workplace and Vaccination : Part I. Labour Market Developments in 2020 – 21 : Increasing Disparities. Geneva (2021). 12. Pereira M, Oliveira AM. Poverty and food insecurity may increase as the threat of COVID-19 spreads. Public Health Nutr. (2020) 23:3236–40. doi: 10.1017/S1368980020003493 32. Edin K, Boyd M, Mabli J, Ohls J, Worthington J, Greene S, et al. SNAP Food Security In-Depth Interview Study. (2013). Available online at: https://fns- prod.azureedge.us/sites/default/ﬁles/SNAPFoodSec.pdf 13. Reeves A, Loopstra R, Tarasuk V. Wage-Setting Policies, Employment, and Food Insecurity: A Multilevel Analysis of 492 078 People in 139 Countries. Washington, DC (2020). 33. Dynan K, Elmendorf D, Sichel D. The evolution of household income volatility. B. E J Econom Anal Pol. (2012) 12. doi: 10.1515/1935-1682.3347 14. OECD. The impact of the COVID-19 crisis on regional and local governments : main ﬁndings from the joint CoR-OECD survey. OECD Reg Dev Pap. (2020) 5:1–38. doi: 10.1787/fb952497-en 34. Morduch J, Schneider R. The ﬁnancial diaries How American Families Cope in a World of Uncertainty. Princeton University Press. (2017) 1–248. doi: 10.1515/9781400884599 15. Crush J. Hungry Cities of the Global South (rep. (2016). p. 1–24. Available online at: https://scholars.wlu.ca/hcp (accessed August 22, 2021) 35. Rauscher E, Elliott W. Wealth as security: growth curve analyses of household income and net worth during a recession. J Fam Econ Issues. (2016) 37:29–41. doi: 10.1007/s10834-015-9442-7 16. Chiwona-Karltun L, Amuakwa-Mensah F, Wamala-Larsson C, Amuakwa- Mensah S, Abu Hatab A, Made N, et al. COVID-19: from health crises to food security anxiety and policy implications. Ambio. (2021) 50:794–811. doi: 10.1007/s13280-020-01481-y 36. Bartfeld J, Collins JM. Food insecurity, ﬁnancial shocks, and ﬁnancial coping strategies among households with elementary school children in wisconsin. J Consum Aﬀ. (2017) 51:519–48. doi: 10.1111/joca.12162 37. FAO. Food Security and the Financial Crisis. (2009). Available online at: http:// www.fao.org/3/k6360e/k6360e.pdf (accessed July 15, 2021) 17. Szabo S. Urbanisation and food insecurity risks: assessing the role of human development. Ecol Econ. (2015) 44:28–48. doi: 10.1080/13600818.2015.1067292 38. Floro MS, Bali Swain R. Food security, gender, and occupational choice among urban low-income households. World Dev. (2013) 42:89–99. doi: 10.1016/j.worlddev.2012.08.005 18. Xiong C, Hu S, Yang M, Luo W, Zhang L. Mobile device data reveal the dynamics in a positive relationship between human mobility and COVID-19 infections. Proc Natl Acad Sci USA. (2020) 117:27087–9. doi: 10.1073/pnas.2010836117 39. Akter S, Basher SA. REFERENCES The impacts of food price and income shocks on household food security and economic well-being: evidence from rural Bangladesh. Glob Environ Chang. (2014) 25:150–62. doi: 10.1016/j.gloenvcha.2014.02.003 19. Hwang TJ, Rabheru K, Peisah C, Reichman W, Ikeda M. Loneliness and social isolation during the COVID-19 pandemic. Int Psychogeriatr. (2020) 32:1217–20. doi: 10.1017/S1041610220000988 40. Gumma MK, Mohammad I, Nedumaran S, Whitbread A, Lagerkvist CJ. Urban sprawl and adverse impacts on agricultural land: a case study on Hyderabad, India. Remote Sens. (2017) 9:1136. doi: 10.3390/rs9111136 20. Rume T, Islam SMD-U. Environmental eﬀects of COVID-19 pandemic and potential strategies of sustainability. Heliyon. (2020) 6:e04965. doi: 10.1016/j.heliyon.2020.e04965 41. Ballard TJ. KAW& CC. The Food Insecurity Experience Scale: Development of a Global Standard for Monitoring Hunger Worldwide. Technical Paper. Rome (2013). 21. Saadat S, Rawtani D, Mustansar C. Since January 2020 Elsevier has Created a COVID-19 Resource Centre With Free Information in English and Mandarin on the Novel Coronavirus COVID-19. The COVID-19 Resource Centre is Hosted on Elsevier Connect, the Company’s Public News and Information (2020). Available online at: https://www.binasss.sa.cr/agocovid/4.pdf 42. Action Aid Association (India). Workers in the Time of COVID-19:Round I of the National Study on Informal Workers. (2020). p. 1–69. Available online at: https://www.actionaidindia.org/wp-content/uploads/2020/08/Workers-in- the-time-of-Covid-19_ebook1.pdf (accessed July 26, 2021) the-time-of-Covid-19_ebook1.pdf (accessed July 26, 2021) 22. Shakil MH, Munim ZH, Tasnia M, Sarowar S. COVID-19 and the environment: a critical review and research agenda. Sci Total Environ. (2020) 745:141022. doi: 10.1016/j.scitotenv.2020.141022 43. Shubhangi G. India’s Informal Workers Under the Lockdown: A Snapshot. (2020). p. 3–12. Available online at: https://sprf.in/wp-content/uploads/ 2021/02/Indias-Informal-Workers-Under-the-Lockdown-SPRF-2021.pdf (accessed July 26, 2021) 23. Hobbs JE. Food supply chains during the COVID-19 pandemic. Can J Agric Econ Can d’agroeconomie. (2020) 68:171–6. doi: 10.1111/cjag.12237 24. Stephens EC, Martin G, van Wijk M, Timsina J, Snow V. Editorial: impacts of COVID-19 on agricultural and food systems worldwide and on progress to the sustainable development goals. Agric Syst. (2020) 183:102873. doi: 10.1016/j.agsy.2020.102873 44. ILO. COVID-19 Crises and the Informal Economy Immediate Responses and Policy Challenges. (2020). Available online at: https://www.ilo.org/wcmsp5/ groups/public/@ed_protect/@protrav/@travail/documents/brieﬁngnote/ wcms_743623.pdf (accessed July 15, 2021) 25. Seleiman MF, Selim S, Alhammad BA, Alharbi BM, Juliatti FC. Will novel coronavirus (COVID-19) pandemic impact agriculture, food security and animal sectors? \| Qual será o impacto da pandemia do novo coronavirus (COVID-19) na agricultura, segurança alimentar e setores de produção animal. Biosci J. (2020) 36:1315–26. doi: 10.14393/BJ-v36n4a2020-54560 45. Unemployment. Available online at: https://unemploymentinindia.cmie.com/ (accessed July 15, 2021) 46. REFERENCES 5. Reichelt M, Makovi K, Sargsyan A. The impact of COVID-19 on gender inequality in the labor market and gender-role attitudes. Eur Soc. (2020) 23:S228–245. doi: 10.1080/14616696.2020.1823010 1. Abu Hatab A, Krautscheid L, Boqvist S. Covid-19, livestock systems and food security in developing countries: a systematic review of an emerging literature. Pathogens. (2021) 10:1–19. doi: 10.3390/pathogens100 50586 6. Adams-Prassl A, Boneva T, Golin M, Rauh C. Inequality in the impact of the coronavirus shock: evidence from real time surveys. SSRN. (2020) 189:1–33. doi: 10.2139/ssrn.3590881 2. Pak A, Adegboye OA, Adekunle AI, Rahman KM, McBryde ES, Eisen DP. Economic consequences of the COVID-19 outbreak: the need for epidemic preparedness. Front Public Heal. (2020) 8:241. doi: 10.3389/fpubh.2020. 00241 7. Ashford NA, Hall RP, Arango-Quiroga J, Metaxas KA, Showalter AL. Addressing inequality: the ﬁrst step beyond COVID-19 and towards sustainability. Sustainability. (2020) 12:5404. doi: 10.3390/su12135404 8. Brodeur A, Clark A, Fleche S, Powdthavee N. Discussion paper series COVID- 19, lockdowns and well-being : evidence from google trends COVID-19, lockdowns and well-being : evidence from google trends. IZA Discuss Inst Labor Econ. (2020) 193:1–8. doi: 10.2139/ssrn.3596670 3. United Nations Development Programme. COVID-19: Looming Crisis in Developing Countries Threatens to Devastate Economies and Ramp Up Inequality. United Nations Development Programme. Available online at: https://www.undp.org/press-releases/covid-19-looming-crisis-developing- countries-threatens-devastate-economies-and-ramp (accessed August 22, 2021) 9. Pouliakas K, Branka J. EU Jobs at Highest Risk of Covid-19 Social Distancing: Will the Pandemic Exacerbate Labour Market Divide? Available online at: https://papers.ssrn.com/abstract=3608530 (accessed November 1, 2021) 4. Béland L-P, Brodeur A, Wright T. The Short-Term Economic Consequences of Covid-19: Exposure to Disease, Remote Work and Government Response. Germany: IZA Institute of Labor Economics (2020). p. 1–92. 10. Fana M, Torrejón Pérez S, Fernández-Macías E. Employment impact of Covid-19 crisis: from short term eﬀects to long terms prospects. J Ind Bus Econ. (2020) 47:391–410. doi: 10.1007/s40812-020-00168-5 May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 13 COVID-19 Impact on Household Food Security Padmaja et al. shaping Africa’s Economic Future. Washington, DC: World Bank (2020). doi: 10.1596/978-1-4648-1568-3 shaping Africa’s Economic Future. Washington, DC: World Bank (2020). doi: 10.1596/978-1-4648-1568-3 11. Rasul G, Nepal AK, Hussain A, Maharjan A, Joshi S, Lama A, et al. Socio-economic implications of COVID-19 pandemic in South Asia: emerging risks and growing challenges. Front Sociol. (2021) 6:629693. doi: 10.3389/fsoc.2021.629693 31. X ILO Monitor. COVID-19 and the World of Work. REFERENCES Science. (2011) 332:1278–84. doi: 10.1126/science.1200138 63. Zaki M. Access to short-term credit and consumption smoothing within the pay cycle. (2016). Available online at: https://papers.ssrn.com/sol3/papers.cfm? abstract_id=2741001 78. Pradhan KC, Mukherjee S. Covariate and idiosyncratic shocks and coping strategies for poor and non-poor rural households in India. J Quant Econ. (2017) 16:101–27. doi: 10.1007/s40953-017-0073-8 79. Ansah K, Gershon I, Gardebroek C, Ihle R. Shock interactions, coping strategy choices and household food security. Clim. Develop. (2021) 5:414–26. doi: 10.1080/17565529.2020.1785832 64. Gaitán-Rossi P, Vilar-Compte M, Teruel G, Pérez-Escamilla R. Food insecurity measurement and prevalence estimates during the COVID-19 pandemic in a repeated cross-sectional survey in Mexico. Public Health Nutr. (2021) 24:412–21. doi: 10.1017/S1368980020004000 65. Crush J, Frayne B. No. 01: The Invisible Crisis: Urban Food Security in Southern Africa. Available online at: https://scholars.wlu.ca/afsun (accessed July 15, 2021) Conﬂict of Interest: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. 66. Tacoli C. Food (in)security in rapidly urbanizing, low-income contexts. Handb Urban Food Secur Glob South. (2020) 14:23–33. doi: 10.4337/9781786431516.00007 Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their aﬃliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. 67. Bargain O, Kwenda P. Earnings structures, informal employment, and self- employme: new evidence from Brazil, Mexico, and South Africa. Rev Income Wealth. (2011) 1–23. doi: 10.2139/ssrn.1545138 68. Sparks DL, Barnett ST. The informal sector in Sub-Saharan Africa: out of the shadows to foster sustainable employment and equity? Int Bus Econ Res J. (2010) 9:1. doi: 10.19030/iber.v9i5.563 Copyright © 2022 Padmaja, Nedumaran, Jyosthnaa, Kavitha, Abu Hatab and Lagerkvist. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. 69. Maxwell D, Levin C, Armar-Klemesu M, Ruel M, Morris S, Ahiadeke C. REFERENCES J Econ Perspect. (2018) 32:167–94. doi: 10.1257/jep.32.3.167 71. Josephson A, Kilic T, Michler JD. Socioeconomic impacts of COVID- 19 in low-income countries. Nat Hum Behav 2021 55. (2021) 5:557–565. doi: 10.1038/s41562-021-01096-7 55. Krueger D, Perri F. How do Households Respond to Income Shocks? University of Pennsylvania, Pennsylvania, PA, United States (2009). Available online at: http://www.fperri.net/PAPERS/howdohouseholds_latest.pdf 72. Tefera T, Tefera F. Determinants of households food security and coping strategies for food shortfall in mareko district, guraghe zone Southern Ethiopia. J Food Secur. (2014) 2:92–9. doi: 10.12691/jfs-2-3-4 56. Baker SR. Debt and the consumption response to household income shocks. SSRN Electron J. (2015) 2541142:1–460. doi: 10.2139/ssrn.2541142 57. Egger D, Miguel E, Warren SS, Shenoy A, Collins E, Karlan D, et al. Falling living standards during the COVID-19 crisis: quantitative evidence from nine developing countries. Sci Adv. (2021) 7:eabe0997. doi: 10.1126/sciadv.abe0997 73. Kansiime MK, Tambo JA, Mugambi I, Bundi M, Kara A, Owuor C. COVID- 19 implications on household income and food security in Kenya and Uganda: ﬁndings from a rapid assessment. World Dev. (2021) 137:105199/ doi: 10.1016/j.worlddev.2020.105199 58. Gjertson L. Emergency Saving and Household Hardship. J Fam Econ Issues. (2014) 37:1–17. doi: 10.1007/s10834-014-9434-z 74. Huang J, Guo B, Kim Y. Food insecurity and disability: do economic resources matter? Soc Sci Res. (2010) 39:111–24. doi: 10.1016/j.ssresearch.2009. 07.002 59. Kesar S, Abraham R, Lahoti R, Nath P, Basole A. Pandemic, informality, and vulnerability: Impact of COVID-19 on livelihoods in India. Canad J Develop. (2021). 42:145–64. doi: 10.1080/02255189.2021.1890003 75. Guo B. Household assets and food security: evidence from the survey of program dynamics. J Fam Econ Issues. (2010) 32:98–110. doi: 10.1007/s10834-010-9194-3 60. Fitzpatrick K. Discussion Paper Series DP 2013-07 Use of Alternative Financial Services and Childhood Food Insecurity Financial Services and Food Insecurity Among Households with Children. (2013). Available online at: www.ukcpr.org (accessed July 15, 2021) 76. Farzana FD, Rahman AS, Sultana S, Raihan MJ, Haque MA, Waid JL, et al. Coping strategies related to food insecurity at the household level in Bangladesh. PLoS ONE. (2017) 12:e0171411. doi: 10.1371/journal.pone.0171411 61. Fitzpatrick K, Coleman-Jensen A. Food on the Fringe: Food insecurity and the use of payday loans. Soc Serv Rev. (2014) 88:553–93. doi: 10.1086/679388 77. Srivastava SK, Sivaramane N. Income-induced eﬀects of COVID-19 on the food consumption pattern of Indian households. Agric Econ Res Rev. (2020) 33:15–24. doi: 10.5958/0974-0279.2020.00014.2 62. Karlan D, Zinman J. Microcredit in theory and practice: using randomized credit scoring for impact evaluation. REFERENCES Sharma GD, Mahendru M. Lives or livelihood: insights from locked- down India due to COVID19. Soc Sci Humanit Open. (2020) 2:100036. doi: 10.1016/j.ssaho.2020.100036 47. Blekking J, Waldman K, Tuholske C, Evans T. Formal/informal employment and urban food security in Sub-Saharan Africa. Appl Geogr. (2020) 114:102131. doi: 10.1016/j.apgeog.2019.102131 26. Ivanov D. Predicting the impacts of epidemic outbreaks on global supply chains: a simulation-based analysis on the coronavirus outbreak (COVID- 19/SARS-CoV-2) case. Transp Res Part E Logist Transp Rev. (2020) 136:101922. doi: 10.1016/j.tre.2020.101922 48. Development Bank A. ADB Brief 139: Food Security in Asia and the Paciﬁc amid the COVID-19 Pandemic. Manila (2020). 27. Nicola M, AlsaﬁZ, Sohrabi C, Kerwan A, Al-Jabir A, Iosiﬁdis C, et al. The socio-economic implications of the coronavirus pandemic (COVID-19): a review. Int J Surg. (2020) 78:185. doi: 10.1016/j.ijsu.2020.04.018 49. Narayanan S, Saha S. Urban food markets and the lockdown in India. SSRN Electron J. (2020) 3599102:1–25. doi: 10.2139/ssrn.3599102 50. Mahajan K, Tomar S. COVID-19 and supply chain disruption: evidence from food markets in India†. Am J Agric Econ. (2021) 103:35–52. doi: 10.1111/ajae.12158 28. Sumner A, Ortiz-Juarez E, Holy C. Precarity and The Pandemic: COVID- 19 and Poverty Incidence, Intensity, and Severity in Developing Countries. WIDER Working Paper, No. 2020/77. The United Nations University World Institute for Development Economics Research (UNU-WIDER), Helsinki, Finland (2020). Available online at: https://www.wider.unu.edu/sites/default/ ﬁles/Publications/Working-paper/PDF/wp2020-77.pdf 51. Cariappa AGA, Acharya KK, Adhav CA, R S, Ramasundaram P. Pandemic led food price anomalies and supply chain disruption: evidence from COVID-19 incidence in India. SSRN Electron J. (2020). doi: 10.2139/ssrn.3680634 52. Valodia I, Lebani L, Skinner C, Devey R. Low-waged and informal employment in South Africa. Transform Crit Perspect South Africa. (2006) 60:90–126. doi: 10.1353/trn.2006.0007 29. Security F, In N, Africa E. Urban focus 3. (2020) 30. Zeufack AG, Calderon C, Kambou G, Djiofack CZ, Kubota M, Korman V, et al. Africa’s Pulse, No. 21, Spring 2020 : An analysis of issues May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 14 COVID-19 Impact on Household Food Security Padmaja et al. the evidence from developing countries. Curr Res Environ Sustain. (2020) 2:100014. doi: 10.1016/j.crsust.2020.100014 the evidence from developing countries. Curr Res Environ Sustain. (2020) 2:100014. doi: 10.1016/j.crsust.2020.100014 53. Crush J, Frayne B, Pendleton W. The crisis of food insecurity in African cities. J. Hunger Environ. Nutr. (2012) 7:271–92. doi: 10.1080/19320248.2012.702448 g 54. Kaplan G, Violante GL. Microeconomic heterogeneity and macroeconomic shocks. Frontiers in Public Health \| www.frontiersin.org REFERENCES Urban Livelihoods and Food and Nutrition Security in Greater Accra, Ghana in Collaboration with Noguchi Memorial Institute for Medical Research. Washington, DC: World Health Organization (2000). 70. Workie E, Mackolil J, Nyika J, Ramadas S. Deciphering the impact of COVID- 19 pandemic on food security, agriculture, and livelihoods: a review of 70. Workie E, Mackolil J, Nyika J, Ramadas S. Deciphering the impact of COVID- 19 pandemic on food security, agriculture, and livelihoods: a review of May 2022 \| Volume 10 \| Article 814112 Frontiers in Public Health \| www.frontiersin.org 15
https://openalex.org/W3097237895	https://digital.csic.es/bitstream/10261/229120/1/highealth.pdf	English	null	High-depth African genomes inform human migration and health	Nature	2,020	cc-by	23,687	High-depth African genomes inform human migration and health Ananyo Choudhury1, Shaun Aron1, Laura R. Botigué2, Dhriti Sengupta1, Gerrit Botha3, Taoufik Bensellak4, Gordon Wells5,6, Judit Kumuthini5,6, Daniel Shriner7, Yasmina J. Fakim8,9, Anisah W. Ghoorah9, Eileen Dareng10,11, Trust Odia12, Oluwadamilare Falola12, Ezekiel Adebiyi12,13, Scott Hazelhurst1,14, Gaston Mazandu3, Oscar A. Nyangiri15, Mamana Mbiyavanga3, Alia Benkahla16, Samar K. Kassim17, Nicola Mulder3, Sally N. Adebamowo18,19, Emile R. Chimusa20, Donna Muzny21, Ginger Metcalf21, Richard A. Gibbs21,22, TrypanoGEN Research Group, Charles Rotimi7, Michèle Ramsay1,23, H3Africa Consortium, Adebowale A. Adeyemo7 ✉, Zané Lombard23 ✉ & Neil A. Hanchard22 ✉ https://doi.org/10.1038/s41586-020-2859-7 Received: 10 May 2019 Accepted: 7 August 2020 Published online: 28 October 2020 Open access Check for updates https://doi.org/10.1038/s41586-020-2859-7 The African continent is regarded as the cradle of modern humans and African genomes contain more genetic variation than those from any other continent, yet only a fraction of the genetic diversity among African individuals has been surveyed1. Here we performed whole-genome sequencing analyses of 426 individuals— comprising 50 ethnolinguistic groups, including previously unsampled populations—to explore the breadth of genomic diversity across Africa. We uncovered more than 3 million previously undescribed variants, most of which were found among individuals from newly sampled ethnolinguistic groups, as well as 62 previously unreported loci that are under strong selection, which were predominantly found in genes that are involved in viral immunity, DNA repair and metabolism. We observed complex patterns of ancestral admixture and putative-damaging and novel variation, both within and between populations, alongside evidence that Zambia was a likely intermediate site along the routes of expansion of Bantu-speaking populations. Pathogenic variants in genes that are currently characterized as medically relevant were uncommon—but in other genes, variants denoted as ‘likely pathogenic’ in the ClinVar database were commonly observed. Collectively, these findings refine our current understanding of continental migration, identify gene flow and the response to human disease as strong drivers of genome-level population variation, and underscore the scientific imperative for a broader characterization of the genomic diversity of African individuals to understand human ancestry and improve health. the Khoe and San (KS) languages3. This broad classification remains as a framework, although several language families contain independent groups; the Khoe and San families, for example, are contestedly grouped together as KhoeSan in the literature despite having distinct histories (Supplementary Note 1.1). A list of affiliations appears at the end of the paper. Article Article Article a Source of data West Africa Central Africa Southern Africa East Africa North Africa TrypanoGEN H3A Baylor 1000G SAHGP AGVP Mali GambiaGuinea Côte d’Ivoire Sierra Leone GhanaBenin Nigeria Cameroon Kenya Zambia Botswana South Africa SOT, XHS Democratic Republic of the Congo Burkina Faso Ethiopia Uganda MAL CIV WGR SSG BRN CAM FNB WGR BOT BSZ DRC UNS, UBS b Principal component 2 (0.6% of variance) 0.025 0 –0.025 –0.050 Principal component 1 (0.8% of variance) 0 0.025 0.050 0.075 –0.025 SOT MSL ESN GWD WOL SOM ORO ZUL BSZ LWK YRI XHS WGR BAG FNB MAL DRC UNS UBS CIV SSG GUM AMH CAM BOT BRN SOT MSL ESN GWD WOL SOM ORO ZUL BSZ LWK YRI XHS WGR BAG FNB MAL DRC UNS UBS CIV SSG GUM AMH CAM BOT BRN WGR BRN MAL UBS CAM BSZ UNS BOT West Africa Central Africa Southern Africa East Africa Fig. 1 \| H3Africa WGS data. a, Geographical regions and populations of origin for H3Africa WGS data. The size of the circles indicates the relative number of sequenced samples from each population group (before quality control; Supplementary Methods Table 1). Samples with WGS data from the 1000 Genomes Project and the African Genome Variation Project are included for comparison (grey circles). CAM includes 25 individuals who are homozygous for the sickle mutation (HbSS); MAL includes unaffected individuals with a family history of neurological disease; BOT comprises children who are HIV-positive; BRN included only female participants. 1000G, 1000 Genome Project; AGVP, African Genome Variation Project. Maps were created using R43. b, Principal component analysis of African WGS data showing the first two principal components. New populations used in this study are indicated by Studying the genomes of peoples across the breadth of Africa pre- sents unique opportunities for understanding the population demog- raphy of human disease. The Human Heredity and Health in Africa (H3Africa) Consortium was conceived to redress the dearth of genomics research in Africa8, and, to date, supports 48 projects across 34 coun- tries. An important mandate of H3Africa is to characterize genetic diversity across Africa to facilitate a framewok for genomic research. To this end, we analysed whole-genome sequencing (WGS) data gener- ated in 426 individuals from ongoing H3Africa studies, including 314 high-depth (average depth of coverage, 30×) and 112 medium-depth (average depth of coverage, 10×) whole-genome sequences, encom- passing 50 ethnolinguistic groups from 13 countries across Africa (Fig. Insights into migration and admixture A major focus of our survey was to bridge the gaps in African population-scale WGS data by including samples from understudied geographical regions. To examine this, we first contextualized our understudied populations alongside previous African WGS efforts2,4,17 using principal component analysis (Fig. 1b and Supplementary Fig. 1). As previously observed4, the first principal component separated the NS, AA, and to some extent, east African NC speakers (Bantu-speaking individuals from Uganda (UBS)) from other NC speakers. The second principal component placed the remaining NC speakers along a cline from west to south (Fig. 1b). Individuals from Mali (MAL), which include some non-NC-speaking groups, were the notable exception, showing more dispersion between individuals. West African populations—the Fon from Benin (FNB), Gur speakers from west Africa (WGR), Soussou from Guinea (SSG), people living in Côte d’Ivoire (CIV) and MAL— projected in proximity to, but often independently of, other west Afri- can NC speakers such as the Yoruba (YRI) and Esan (ESN) from Nigeria2. Eastern (UBS) and southern African (Botswana (BOT)) NC speakers clustered with previously studied populations from their respective geographical regions. Fig. 1 \| H3Africa WGS data. a, Geographical regions and populations of origin for H3Africa WGS data. The size of the circles indicates the relative number of sequenced samples from each population group (before quality control; Supplementary Methods Table 1). Samples with WGS data from the 1000 Genomes Project and the African Genome Variation Project are included for comparison (grey circles). CAM includes 25 individuals who are homozygous for the sickle mutation (HbSS); MAL includes unaffected individuals with a family history of neurological disease; BOT comprises children who are HIV-positive; BRN included only female participants. 1000G, 1000 Genome Project; AGVP, African Genome Variation Project. Maps were created using R43. b, Principal component analysis of African WGS data showing the first two principal components. New populations used in this study are indicated by crosses. Population abbreviations are as described in the 1000 Genomes and H3Africa Projects as provided in Supplementary Methods Table 1 and Supplementary Table 22. Shaded background elipses relate to the geographical regions as shown in a. Five population groups—the Berom of Nigeria (BRN), individals from CAM, individuals from the Democratic Republic of the Congo (DRC), Bantu speakers from Zambia (BSZ) and NS speakers from Uganda (UNS)—showed distinctive principal component localization. BRN sam- ples localized independently of other west African populations, tending towards east African populations (Fig. 1b). Article 1a and Supplementary Methods Table 1). Some of these groups are studied here for the first time, providing a unique overview of the diverse landscape of African genomic variation. Source of data West Africa Central Africa Southern Africa East Africa North Africa TrypanoGEN H3A Baylor 1000G SAHGP AGVP Mali GambiaGuinea Côte d’Ivoire Sierra Leone GhanaBenin Nigeria Cameroon Kenya Zambia Botswana South Africa SOT, XHS Democratic Republic of the Congo Burkina Faso Ethiopia Uganda MAL CIV WGR SSG BRN CAM FNB WGR BOT BSZ DRC UNS, UBS a Our analyses focused on single-nucleotide variants (SNVs) in samples from three African resources: the H3Africa Consortium (here H3A-Baylor; http://www.h3africa.org/)8, the Southern African Human Genome Programme (SAHGP)17 and the Trypanosomiasis Genomics Network of the H3Africa Consortium (TryopanoGEN)18 (Methods). High depth of coverage WGS data from H3A-Baylor encom- passing 314 individuals from west (Burkina Faso, Ghana, Mali, Nigeria, Benin and Cameroon), central (Zambia) and south (Botswana) African countries (Supplementary Methods Table 1) were used for analyses of rare and novel variation, and to identify selective sweeps. The full data- set was used for analyses of population ancestry and admixture, and to assess medically relevant variation. A total of 41,645,936 high-quality SNVs were identified across all groups; of these, 31,160,639 were found in the dataset that had a high depth of coverage (HC-WGS dataset). b b Principal component 2 (0.6% of variance) 0.025 0 –0.025 –0.050 Principal component 1 (0.8% of variance) 0 0.025 0.050 0.075 –0.025 SOT MSL ESN GWD WOL SOM ORO ZUL BSZ LWK YRI XHS WGR BAG FNB MAL DRC UNS UBS CIV SSG GUM AMH CAM BOT BRN SOT MSL ESN GWD WOL SOM ORO ZUL BSZ LWK YRI XHS WGR BAG FNB MAL DRC UNS UBS CIV SSG GUM AMH CAM BOT BRN WGR BRN MAL UBS CAM BSZ UNS BOT West Africa Central Africa Southern Africa East Africa High-depth African genomes inform human migration and health The Bantu languages are the most widely spo- ken in sub-Saharan Africa, and this broad dispersion has been ascribed to a series of migrations across the continent over the past 5,000 years9. These migration events and subsequent admixture with resident populations have had a pivotal role in shaping the genomic landscape of Africa, as it involved adaptations to new exposures and experiences. The signatures of these adaptations are evident in patterns of allelic variation associated with key physiological traits or prevalent communicable diseases, exem- plified by variations in HBB10, LCT11, APOL112 and G6PD13. Recent studies have identified signatures of selection that reflect the importance of new variation, introduced by admixture, to traits as diverse as diet, height, blood pressure and skin pigmentation4,9,14–16. Advances in genomics have empowered the interrogation of the human genome across global populations2, with the resulting studies demon- strating that Africa harbours the most genetic variation and diversity3,4. These studies provided insights into medically relevant genetic loci and aided in the interpretation of the pathogenicity of genetic vari- ants5, advancing precision medicine for all populations6. To date, only a limited number of the around 2,000 African ethnolinguistic groups have been genetically characterized, predominantly using genotyping arrays, which contained a limited number of variants that are common in European populations. The population distribution of novel, rare and medically relevant variation among African individuals thus remains largely unknown, which adversely affects our understanding of the genetic contributions to Mendelian and complex diseases7,8. Classically, sub-Saharan African populations have been described in the context of four major language families: Afro-Asiatic (AA), Nilo-Saharan (NS), Niger–Congo (NC)—which includes the Bantu language family—and Nature \| Vol 586 \| 29 October 2020 \| 74 741 742 \| Nature \| Vol 586 \| 29 October 2020 Insights into migration and admixture Dates are shown in terms of number of generations (1 generation = 29 years). Maps were created using R43. YRI CAM DRC ANG BSZ UBS,UNS LWK BOT b a XHS European West African Central African East African South African RFF KS CEU JOL BRN CAM DRC UBS UNS KAL SOM BSZ BOT SOT ZUL SEBantu BabongoE Bakoya Bezan Mbuti Biaka Ju\|’hoansi GGK Nama FNB WGR CIV SSG MAL consistent with our current understanding of admixture patterns across the continent, showing KS gene flow in BOT; AA speaker gene flow in UBS, and varying degrees of gene flow from rainforest foragers (RFF; ethnolinguistically diverse peoples distributed across the forested regions of central Africa) in central-west African populations (DRC and CAM) (Fig. 2a and Extended Data Fig. 1). In addition, these analyses also revealed two interesting and unreported admixture events—RFF gene flow in UNS and gene flow from NS speakers in the BRN (Fig. 2a and Extended Data Fig. 1). Further analyses using additional population datasets (Supplementary Fig. 5 and Supplementary Note 1.2) suggested that the distinction between UNS and other NS-speaker populations could be due to increased gene flow from RFF as well as the absence of AA admixture. Traces of east African ancestry, originating from waves of trans-Sahelian migrations in the past few thousand years, have been reported in other populations, including the Hausa of Nigeria3; how- ever, the observation of east African gene flow, possibly from Chad (Supplementary Fig. 6, Supplementary Table 2 and Supplementary Note 1.2), in the largest autochthonous central Nigerian population (BRN) is highly unique. b a BRN UBS UNS LWK DRC ANG BOT CAM 72 59 99 84 28 25 57 53 70 50 KSHG AA RFF NS 20 25 50 30 60 90 70 50 Number of generations ago c c Analyses based on masking of non-NC ancestry in BOT highlighted the contribution of non-NC gene flow to the genetic distance between populations (Supplementary Note 1.3 and Supplementary Fig. 7). By contrast, BSZ, unlike their geographical neighbours from Angola, the DRC and Botswana, did not show evidence for any major gene flow from non-NC speakers, such as the RFF or KS groups (Fig. 2a and Extended Data Fig. 1). Similarly low levels of local group admixture have been noted for Bantu speakers from Malawi16 and Mozambique21. Insights into migration and admixture Individuals from CAM and the DRC, consistent with the geographical proximity of these two coun- tries, localized together, forming an independent central-west African group (Fig. 1b). UNS individuals localized independently of another NS population (Gumuz (GUZ)4) (Fig. 1b). Similarly, BSZ were found to be largely separate from other NC-speaking groups (Fig. 1b). Even in the context of a much wider representation of African populations from array-based data (Supplementary Fig. 2), some of these populations retained their distinctive localization patterns, demonstrating the extensive genomic diversity of the sampled populations. Accordingly, we observed a statistically significant positive correlation between geo- graphical and genetic distances of NC-speaking populations (R = 0.96; t-statistic; P < 1 × 10−9) (Methods and Supplementary Fig. 3). We used an unsupervised clustering approach implemented in ADMIXTURE19 to evaluate whether gene flow from non-NC speakers differentiated our study populations (Fig. 2a, Extended Data Fig. 1 and Supplementary Fig. 4). Admixture events were then evaluated further using f3-statistics20 (Supplementary Table 1). Results were 742 \| Nature \| Vol 586 \| 29 October 2020 742 YRI CAM DRC ANG BSZ UBS,UNS LWK BOT b a XHS European West African Central African East African South African RFF KS BRN UBS UNS LWK DRC ANG BOT CAM 72 59 99 84 28 25 57 53 70 50 KSHG AA RFF NS 20 25 50 30 60 90 70 50 Number of generations ago c CEU JOL BRN CAM DRC UBS UNS KAL SOM BSZ BOT SOT ZUL SEBantu BabongoE Bakoya Bezan Mbuti Biaka Ju\|’hoansi GGK Nama FNB WGR CIV SSG MAL Fig. 2 \| Population admixture and genetic ancestry among African populations. a, Admixture plot showing select African populations based on WGS and array data for K = 10. b, Proposed movement during the Bantu migration, showing the populations that were used for inference. Blue line shows the migration patterns inferred by genetic distance estimates with Zambia (BSZ) as an intermediate staging ground for Bantu migrations further east (red–teal arrow) and south (red–yellow arrow). The dotted black line shows the previously proposed late-split route9; the dotted blue–green line through the DRC indicates an alternative model of migration. GGK, Gǀwi, Gǁana and baKgalagadi. c, Key admixture dates (in generations) in populations of interest based on MALDER results. The colour of each circle represents the admixture date for NC components in each population group (KS, AA, RFF and NS). Insights into migration and admixture Blue line shows the migration patterns inferred by genetic distance estimates with Zambia (BSZ) as an intermediate staging ground for Bantu migrations further east (red–teal arrow) and south (red–yellow arrow). The dotted black line shows the previously proposed late-split route9; the dotted blue–green line through the DRC indicates an alternative model of migration. GGK, Gǀwi, Gǁana and baKgalagadi. c, Key admixture dates (in generations) in populations of interest based on MALDER results. The colour of each circle represents the admixture date for NC components in each population group (KS, AA, RFF and NS). Dates are shown in terms of number of generations (1 generation = 29 years). Maps were created using R43. Our attempts to estimate the dates of some key admixture events22 showed that KS gene flow in southern Africa and RFF gene flow in CAM were largely in agreement with previous studies4,23 (Fig. 2c and Sup- plementary Table 2). The date range for RFF admixture in UNS was similar to that for CAM and consistent with previous surveys23, hinting at a possibility for range alteration of RFF populations both east and west of the central rainforest around 60–70 generations ago (Fig. 2c). Previous studies on trans-Sahelian migration to west Africa have sug- gested two distinct waves of migration: one more than 100 generations ago (2,900 years ago) and a more-recent wave in the last 35 generations (1,015 years ago)24. On the basis of a variety of east African proxy popu- lations, we estimated that admixture in BRN occurred approximately 50 to 70 generations ago (1,500–2,000 years ago) (Fig. 2c). These distinct dates are suggestive of a previously unknown demographic event, either at the local level, or possibly at a wider geographical scale (Supplementary Note 1.5). boundaries of Botswana, Cameroon and Mali (Supplementary Note 1.6, Supplementary Table 3 and Supplementary Fig. 9), and distinctively long segments of runs of homozygosity among MAL individuals (Supplementary Note 1.7 and Supplementary Figs. 10, 11). Analyses of uniparental markers (mitochondria and Y chromosome) identified a predominance of certain uniparental haplogroups in BOT (L0d), BRN (L3) and MAL (E1b1b) (Supplementary Note 1.8 and Supplementary Fig. 12) that further underlies the complex ancestral contributions to these groups. Insights into migration and admixture A recent attempt to reconstruct the route of Bantu migration across central Africa concluded that populations from Angola were the best source of Bantu speakers for east and south African Bantu speakers, and suggested a westerly route of Bantu-speaker migration via Angola9. Inclusion of the DRC and BSZ populations in our dataset, therefore, enabled us to further investigate this route (Fig. 2b). Principal com- ponent analysis and identity-by-descent sharing demonstrated that BSZ is genetically closer to both UBS and BOT compared to other cen- tral African populations (Supplementary Fig. 8). Moreover, formal admixture tests supported BSZ as the most likely central African source population for Bantu-speaker ancestries in east and south Africa (Sup- plementary Table 1). Furthermore, the degree of identity-by-descent sharing between population groups suggested that populations from Angola were the closest central or central-west African population to BSZ (Supplementary Fig. 8). Taken together, these estimates lead us to posit that Zambia was an intermediate site in the likely route of Bantu migration to both east and south Africa (Fig. 2b). An orthogonal approach using admixture graphing also supported this hypothesis (Supplementary Note 1.4). Fig. 2 \| Population admixture and genetic ancestry among African Fig. 2 \| Population admixture and genetic ancestry among African populations. a, Admixture plot showing select African populations based on WGS and array data for K = 10. b, Proposed movement during the Bantu migration, showing the populations that were used for inference. Blue line shows the migration patterns inferred by genetic distance estimates with Zambia (BSZ) as an intermediate staging ground for Bantu migrations further east (red–teal arrow) and south (red–yellow arrow). The dotted black line shows the previously proposed late-split route9; the dotted blue–green line through the DRC indicates an alternative model of migration. GGK, Gǀwi, Gǁana and baKgalagadi. c, Key admixture dates (in generations) in populations of interest based on MALDER results. The colour of each circle represents the admixture date for NC components in each population group (KS, AA, RFF and NS). Dates are shown in terms of number of generations (1 generation = 29 years). Maps were created using R43. Fig. 2 \| Population admixture and genetic ancestry among African populations. a, Admixture plot showing select African populations based on WGS and array data for K = 10. b, Proposed movement during the Bantu migration, showing the populations that were used for inference. Revealing further genomic variation In general, SNV discovery in the H3A-Baylor populations correlated with sample size, with between 12 and 20 million SNVs identified per population. Variant calling also revealed a total of 190,555 potentially multiallelic sites (Methods), most of which (more than 90%, n = 189,900) included three alleles in the dataset; the remainder of sites (n = 655) were biallelic in the dataset, but had a third allele in the reference genome Additional distinctive trends in the demographic history of some of these populations were observed, such as extensive variation in inter- and intra-ethnolinguistic groups within the defined geopolitical Nature \| Vol 586 \| 29 October 2020 \| 743 743 Article 20 15 5 10 1 2 3 2 3 4 Number of novel SNVs (×103) a Novel variant discovery (×106) b c Novel variant discovery (×105) Population BOT WGR BRN BBC FNB BSZ MAL Population BOT WGR BRN BBC FNB BSZ MAL 0 10 8 10 11 11.5 12 12.5 11 12 13 10 12 14 16 15 20 25 30 d R = 0.9; P = 1.4 × 10–09 CAM Number of novel SNVs (×103) Proportion of KS ancestry BRN e Proportion of RFF ancestry f FNB MAL R = 0.93; P = 2.6 × 10–11 R = 0.58; P = 0.003 WGR BSZ BOT g R = 0.22; P = 0.28 Proportion of non-NC ancestry 0.55 0.20 0.15 0.10 0.05 0.4 0.3 0.2 0.1 0.60 0.65 0.70 0.75 0.07 0.08 0.09 0.10 0.11 0.12 0.13 Proportion of EA ancestry Fig. 3 \| Novel variation in the H3Africa dataset. a, Novel variants per individual in each population (n = 24 biological independent samples randomly chosen from each group to match the smallest used dataset). Shading within a population reflects self-identified ethnolinguistic affiliations (Supplementary Table 3). b, c, The number of additional total (b) and common (c) variants discovered in each population starting with those identified in BOT. d–g, Correlation (Pearson, line of best fit is shown in green) between the number of novel SNVs and proportion of KS in BOT (d), RFF in CAM (e), non-NC in MAL (f) and east African (EA) ancestry in BRN (g). Revealing further genomic variation Given the current overrepresentation of individuals with central and west African ancestry (for example, YRI) in publicly available genomic databases, we also assessed whether novel variant discovery might be improved in ancestries that are not as well-represented. We found a strong correlation between variant discovery and the proportion of non-central-west African ancestry among our populations, particularly KS (R = 0.9, P = 1.4 × 10−9) and RFF (R = 0.93, P = 2.6 × 10−11) ancestries (Fig. 3d–g). We also analysed the non-coding regions detected as CLR outliers (Supplementary Note 2.2), focusing on long contiguous stretches of signals (Supplementary Table 5), outliers that coincided with genome- wide association study (GWAS) signals (Supplementary Table 9 and Supplementary Fig. 14), and signals that overlapped with expres- sion quantitative trait loci (Supplementary Figs. 15–17). We detected evidence for signatures of selection at several regulatory loci, including signals that are linked to traits such as chronic kidney disease, uterine fibroids and blood cell indices (Supplementary Note 2.2). Different methods of detecting selection will identify different types and timescales of selection. When we compared regions with outlier CLR sores with those detected using an orthogonal method28 based on the integrated haplotype homozygosity score—which detects more-recent sweeps29—we found that around 20% of the CLR-identified genes were also detected in the analysis based on the integrated hap- lotype homozygosity score, including the immune genes ERCC2, C5AR1 and TRAF2 (Supplementary Table 10). Loci detected using both Revealing further genomic variation 20 15 5 10 1 2 3 2 3 4 Number of novel SNVs (×103) a Novel variant discovery (×106) b c Novel variant discovery (×105) Population BOT WGR BRN BBC FNB BSZ MAL Population BOT WGR BRN BBC FNB BSZ MAL 0 d CAM BRN e f FNB MAL WGR BSZ BOT g 10 8 10 12 14 16 15 20 25 30 d R = 0.9; P = 1.4 × 10–09 Number of novel SNVs (×103) Proportion of KS ancestry e 0.0 0.4 0.3 0.2 0.1 d 10 8 10 11 12 13 10 12 14 16 15 20 25 30 d R = 0.9; P = 1.4 × 10–09 Number of novel SNVs (×103) Proportion of KS ancestry e Proportion of RFF ancestry f R = 0.93; P = 2.6 × 10–11 0.20 0.15 0.10 0.05 0.4 0.3 0.2 0.1 g 11 12 R = 0.58; P = 0.003 g Proportion of non-NC ancestry 0.55 0.60 0.65 0.70 0.75 11 11.5 12 12.5 g R = 0.22; P = 0.28 0.07 0.08 0.09 0.10 0.11 0.12 0.13 Proportion of EA ancestry Proportion of non-NC ancestry 0.55 0.60 0.65 0.70 0.75 Fig. 3 \| Novel variation in the H3Africa dataset. a, Novel variants per individual in each population (n = 24 biological independent samples randomly chosen from each group to match the smallest used dataset). Shading within a population reflects self-identified ethnolinguistic affiliations (Supplementary Table 3). b, c, The number of additional total (b) and common (c) variants discovered in each population starting with those identified in BOT. Fig. 3 \| Novel variation in the H3Africa dataset. a, Novel variants per discovered in each population starting with those identified in BOT. d–g, Correlation (Pearson, line of best fit is shown in green) between th number of novel SNVs and proportion of KS in BOT (d), RFF in CAM (e), n discovered in each population starting with those identified in BOT. d–g, Correlation (Pearson, line of best fit is shown in green) between th discovered in each population starting with those identified in BOT. d–g, Correlation (Pearson, line of best fit is shown in green) between the number of novel SNVs and proportion of KS in BOT (d), RFF in CAM (e), non-NC in MAL (f) and east African (EA) ancestry in BRN (g). Fig. 3 \| Novel variation in the H3Africa dataset. Revealing further genomic variation a, Novel variants per individual in each population (n = 24 biological independent samples randomly chosen from each group to match the smallest used dataset). Shading within a population reflects self-identified ethnolinguistic affiliations (Supplementary Table 3). b, c, The number of additional total (b) and common (c) variants d–g, Correlation (Pearson, line of best fit is shown in green) between the number of novel SNVs and proportion of KS in BOT (d), RFF in CAM (e), non-NC in MAL (f) and east African (EA) ancestry in BRN (g). P < 0.001; Supplementary Table 5), which are suggestive of recent selec- tion, were detected in each of the six HC-WGS populations. Collectively, these regions mapped to 107 genes, of which 62 (58%) were novel, and 45 were identified previously4,9 (Supplementary Tables 5–7). Almost half of these novel selected loci were outliers in two or more populations, which is perhaps a result of ascertaining predominantly NC popula- tions (Supplementary Tables 5–7). However, there was still discern- ible heterogeneity in selective pressure between populations—only 13 loci were detected as outliers in more than four of the six populations (Fig. 4a), and some signals (for example, signals that overlap with ART3 and MAMDC4; Fig. 4a) were only detected as outliers in one or two of the sampled populations. Functional annotation of putatively selected genes revealed that these genes were predominantly associated with immune-related functions (Supplementary Table 8), inclusive of genes such as C5AR1 and MYH10 (bacterial infection); ARHGEF1, ERCC2 and TRAF2 (viral infection) and IFNGR2 (both viral and bacterial infection). In addition, some of the previously characterized selection signals, such as APOL1 and LARGE1, were observed at a more liberal threshold of P < 0.01 (Extended Data Fig. 2a). (GRCh37). Multiallelic sites can provide unique insights into human migration and disease; however, more consistent and accurate annota- tion of such sites is required to capitalize on this potential25. In total, around 3.4 million SNVs in the H3A-Baylor dataset had not been previously reported (Methods and Supplementary Table 4). These novel SNVs accounted for 2–5% of all SNVs in each population, and, at the single-population level, the vast majority (88%) occurred once. Given the modest per-population sample sizes, however, some of these singleton variants are likely to be common at the population level, a view supported by the observation that 9–20% of population-singleton SNVs were shared with at least one other population (Supplementary Fig. 13a). Revealing further genomic variation Individuals from CAM had the fewest novel SNVs, whereas indi- viduals from BOT and MAL had the most, both in absolute number and when normalized to the fewest number of sampled individuals (Fig. 3a). T d t i h th th di f l i t i d t t In total, around 3.4 million SNVs in the H3A-Baylor dataset had not been previously reported (Methods and Supplementary Table 4). These novel SNVs accounted for 2–5% of all SNVs in each population, and, at the single-population level, the vast majority (88%) occurred once. Given the modest per-population sample sizes, however, some of these singleton variants are likely to be common at the population level, a view supported by the observation that 9–20% of population-singleton SNVs were shared with at least one other population (Supplementary Fig. 13a). Individuals from CAM had the fewest novel SNVs, whereas indi- viduals from BOT and MAL had the most, both in absolute number and when normalized to the fewest number of sampled individuals (Fig. 3a). To determine whether the discovery of novel variants in our dataset was saturated, we plotted the cumulative number of novel variants discovered, using BOT as our starting population. With each addi- tional population, novel variant discovery did not reach a definite plateau (Fig. 3b), even after removing singleton novel SNVs—more than 6,000 novel SNVs were still observed between the last two popula- tions (FNB and MAL; Fig. 3c). Given the current overrepresentation of individuals with central and west African ancestry (for example, YRI) in publicly available genomic databases, we also assessed whether novel variant discovery might be improved in ancestries that are not as well-represented. We found a strong correlation between variant discovery and the proportion of non-central-west African ancestry among our populations, particularly KS (R = 0.9, P = 1.4 × 10−9) and RFF (R = 0.93, P = 2.6 × 10−11) ancestries (Fig. 3d–g). p ( g ) To determine whether the discovery of novel variants in our dataset was saturated, we plotted the cumulative number of novel variants discovered, using BOT as our starting population. With each addi- tional population, novel variant discovery did not reach a definite plateau (Fig. 3b), even after removing singleton novel SNVs—more than 6,000 novel SNVs were still observed between the last two popula- tions (FNB and MAL; Fig. 3c). Identifying new signatures of selection 22 METTL22 MDMAD4 COX7A2 WDPDP HMCN2 CD79A SNX24 FRRS1 ITSN2 ART3 HPSE F11R FNB BOT CAM BRN WGR BSZ 100 50 0 100 50 0 100 50 0 100 50 0 100 50 0 100 50 0 Fig. 4 \| Selection and medically relevant variants in African populations. differentiated by the most commonly associated inheritance pattern of the a APOL1 G2 APOL1 high-risk variants HLA-B5701 HbS APOL1 G1 Southern Africa West Africa Central Africa East Africa North Africa SSG CIV MAL WGR BRN CAM DRC UBS UNS BSZ BOT SOT XHS FNB d c Density Allele frequency 0 2 4 6 All Dominant Recessive ZEB1:10-31750140-A-C LDLRAP1:1-25890188-C-T GATA4:8-11606312-T-C 0 0.25 0.50 0.75 1.00 Proportion of total 0 0.04 0.06 Number of alleles 0 2 4 6 8 10 12 14 16 Chr. 3 Chr. 4 Chr. 5 Chr. 6 Chr. 7 Chr. 8 Chr. 9 Chr. 10 Chr. 11 Chr. 12 Chr. 13 Chr. 14 Chr. 15 Chr. 16 Chr METTL22 MDMAD4 COX7A2 HMCN2 SNX24 ART3 HPSE CAM BRN WGR BSZ 50 0 100 50 0 100 50 0 100 50 0 Allele frequency 0.50 Fig. 4 \| Selection and medically relevant variants in African populations. a, Circular Manhattan plot showing the CLR score distribution in 10-kb windows in the six HC-WGS populations (Supplementary Tables 5, 6). Loci with CLR scores > 49.5 (corresponding to a P < 0.001) are shown as red dots. Genes within regions with significant outlier scores in four or more groups (FRRS1, ITSN2, WDPCP, SNX24, METTL22 and HMCN2) or two or fewer groups (ART3, F11R, CD79A, COX7A2, HPSE and MAMDC4) are highlighted. b, Burden of pathogenic (class 5) ClinVar SNVs in H3Africa cohort. c, Density plot of frequencies of pathogenic and likely pathogenic ClinVar SNVs (n = 262) Fig. 4 \| Selection and medically relevant variants in African populations. differentiated by the most commonly associated inheritance pattern of the monogenic disease gene in cases in which a gene has been implicated; three variants with allele frequency > 5% are shown, illustrated as gene name:chromosome-base pair position-reference allele-variant allele. d, Distribution of disease alleles common to Africa across H3Africa populations. The map was created using R43. In each population, the corresponding bar graphs show the relative proportions of the specific disease-associated alleles (Supplementary Table 21). HbS in CAM and FNB are omitted as they include individuals with homozygous sickle cell disease (HbSS). Fig. Identifying new signatures of selection 4 \| Selection and medically relevant variants in African populations. a, Circular Manhattan plot showing the CLR score distribution in 10-kb windows in the six HC-WGS populations (Supplementary Tables 5, 6). Loci with CLR scores > 49.5 (corresponding to a P < 0.001) are shown as red dots. Genes within regions with significant outlier scores in four or more groups (FRRS1, ITSN2, WDPCP, SNX24, METTL22 and HMCN2) or two or fewer groups (ART3, F11R, CD79A, COX7A2, HPSE and MAMDC4) are highlighted. b, Burden of pathogenic (class 5) ClinVar SNVs in H3Africa cohort. c, Density plot of frequencies of pathogenic and likely pathogenic ClinVar SNVs (n = 262) d, Distribution of disease alleles common to Africa across H3Africa populations. The map was created using R43. In each population, the corresponding bar graphs show the relative proportions of the specific disease-associated alleles (Supplementary Table 21). HbS in CAM and FNB are omitted as they include individuals with homozygous sickle cell disease (HbSS). approaches included 10 previously reported genes and 11 genes for which selective sweeps are, to our knowledge, reported for the first time in this study (Supplementary Table 10). scan—all had KS ancestry that was substantially higher than the mean genome-wide distribution of KS ancestry (+3 s.d.). Although we were unable to detect any major differences in KS ancestry at the gene-set level for either CLR- or PBS-based outlier genes, the high KS ancestry in the aforementioned genes bolsters reports of adaptive introgression in southern African selection signals9. To identify selection signals that are unique to southern African (proxied by BOT), central African (proxied by CAM) and west Afri- can (proxied by WGR) populations, we used an approach based on the population branch statistic (PBS)30. This analysis identified three genes that are involved in metabolism (MRAP2, ARSK and GPD2) among those uniquely selected in BOT (Extended Data Fig. 2b, Supplementary Note 2.3 and Supplementary Tables 11, 12) and genes that are involved in DNA maintenance among those unique to WGR (C12orf65 and FAN1; Extended Data Fig. 2b and Supplementary Table 13) and CAM (FZR1, TDP1 and KCTD1; Supplementary Note 2.3 and Supplementary Table 14). We also found evidence for preferential gene flow from KS among the selection signals in BOT (Extended Data Fig. 2c and Supplementary Table 15, 16). CLR selection outliers GNL1, MYH10 and SMC1B, as well as TIGD3 and VDAC3—both of which were identified in the PBS-based Identifying new signatures of selection Adaptive selection of genomic loci in response to dietary, environ- mental and infectious-disease exposures, have been well-described across the continent26. The distribution of composite likelihood ratio (CLR)27 statistic scores for 10-kb windows across the genome in the six populations is summarized in Fig. 4a. Outlier windows (CLR score > 49.5; 744 \| Nature \| Vol 586 \| 29 October 2020 APOL1 G2 APOL1 high-risk variants HLA-B5701 HbS APOL1 G1 Southern Africa West Africa Central Africa East Africa North Africa SSG CIV MAL WGR BRN CAM DRC UBS UNS BSZ BOT SOT XHS FNB d quency minant cessive GATA4:8-11606312-T-C 0 0.75 1.00 Proportion of total population 0 0.04 0.06 0.08 0.10 Number of alleles 0 2 4 6 8 10 12 14 16 b Chr. 1 Chr. 2 Chr. 3 Chr. 4 Chr. 5 Chr. 6 Chr. 7 Chr. 8 Chr. 9 hr. 21 Chr. 22 MDMAD4 COX7A2 WDPDP HMCN2 SNX24 FRRS1 ITSN2 ART3 HPSE F11R FNB BOT CAM BRN WGR BSZ 100 50 0 100 50 0 100 50 0 100 50 0 100 50 0 100 50 0 African populations. tribution in 10-kb ary Tables 5, 6). Loci with hown as red dots. Genes r more groups (FRRS1, r fewer groups (ART3, hted. b, Burden of c, Density plot of Var SNVs (n = 262) differentiated by the most commonly associated inheritance pattern of the monogenic disease gene in cases in which a gene has been implicated; three variants with allele frequency > 5% are shown, illustrated as gene name:chromosome-base pair position-reference allele-variant allele. d, Distribution of disease alleles common to Africa across H3Africa populations. The map was created using R43. In each population, the corresponding bar graphs show the relative proportions of the specific disease-associated alleles (Supplementary Table 21). HbS in CAM and FNB are omitted as they include individuals with homozygous sickle cell disease (HbSS). APOL1 G2 APOL1 high-risk variants HLA-B5701 HbS APOL1 G1 Southern Africa West Africa Central Africa East Africa North Africa SSG CIV MAL WGR BRN CAM DRC UBS UNS BSZ BOT SOT XHS FNB d c Density Allele frequency 0 2 4 6 All Dominant Recessive ZEB1:10-31750140-A-C LDLRAP1:1-25890188-C-T GATA4:8-11606312-T-C 0 0.25 0.50 0.75 1.00 Proportion of total population 0 0.04 0.06 0.08 0.10 Number of alleles 0 2 4 6 8 10 12 14 16 b a Chr. 1 Chr. 2 Chr. 3 Chr. 4 Chr. 5 Chr. 6 Chr. 7 Chr. Identifying new signatures of selection 8 Chr. 9 Chr. 10 Chr. 11 Chr. 12 Chr. 13 Chr. 14 Chr. 15 Chr. 16 Chr. 17 Chr. 18 Chr. 19 Chr. 20 Chr. 21 Chr. 22 METTL22 MDMAD4 COX7A2 WDPDP HMCN2 CD79A SNX24 FRRS1 ITSN2 ART3 HPSE F11R FNB BOT CAM BRN WGR BSZ 100 50 0 100 50 0 100 50 0 100 50 0 100 50 0 100 50 0 Fig. 4 \| Selection and medically relevant variants in African populations. a, Circular Manhattan plot showing the CLR score distribution in 10-kb windows in the six HC-WGS populations (Supplementary Tables 5, 6). Loci with CLR scores > 49.5 (corresponding to a P < 0.001) are shown as red dots. Genes within regions with significant outlier scores in four or more groups (FRRS1, ITSN2, WDPCP, SNX24, METTL22 and HMCN2) or two or fewer groups (ART3, F11R, CD79A, COX7A2, HPSE and MAMDC4) are highlighted. b, Burden of pathogenic (class 5) ClinVar SNVs in H3Africa cohort. c, Density plot of differentiated by the most commonly associated inheritance pattern of the monogenic disease gene in cases in which a gene has been implicated; three variants with allele frequency > 5% are shown, illustrated as gene name:chromosome-base pair position-reference allele-variant allele. d, Distribution of disease alleles common to Africa across H3Africa populatio The map was created using R43. In each population, the corresponding bar graphs show the relative proportions of the specific disease-associated allele (Supplementary Table 21). HbS in CAM and FNB are omitted as they include APOL1 G2 APOL1 high-risk variants HLA-B5701 HbS APOL1 G1 Southern Africa West Africa Central Africa East Africa North Africa SSG CIV MAL WGR BRN CAM DRC UBS UNS BSZ BOT SOT XHS FNB d c Density Allele frequency 0 2 4 6 All Dominant Recessive ZEB1:10-31750140-A-C LDLRAP1:1-25890188-C-T GATA4:8-11606312-T-C 0 0.25 0.50 0.75 1.00 Proportion of total population 0 0.04 0.06 0.08 0.10 Number of alleles 0 2 4 6 8 10 12 14 16 b a Chr. 1 Chr. 2 Chr. 3 Chr. 4 Chr. 5 Chr. 6 Chr. 7 Chr. 8 Chr. 9 Chr. 10 Chr. 11 Chr. 12 Chr. 13 Chr. 14 Chr. 15 Chr. 16 Chr. 17 Chr. 18 Chr. 19 Chr. 20 Chr. 21 Chr. Article The site frequency spectra of variants that are predicted to be dam- aging and or likely to be benign in our populations (Supplementary Fig. 18) were consistent with expectations of purifying selection, but also revealed a substantive number of shared and common putative loss-of-function (LOF) variants (Extended Data Fig. 3b and Supplemen- tary Fig. 18) of likely relevance to variant curation efforts (Extended Data Fig. 3c, Supplementary Fig. 19 and Supplementary Note 3). In addition to its role in rare Mendelian disease, putative protein-damaging varia- tion has also been associated with common multifactorial diseases32. Given the burden of infectious diseases on the African continent, we explored the relationship between putative LOF variation in genes implicated in specific infectious diseases and regional differences in dis- ease mortality as an available proxy for disease outcome. We calculated a putative LOF burden ratio for each population using putative LOF vari- ants in genes designated as ‘directly’ (n = 181) and ‘indirectly’ (n = 1,842) implicated in influenza (Methods and Supplementary Table 18), and plotted this against the 2016 country-specific influenza mortality rates from the World Health Organization (WHO)33. We observed a modest inverse correlation (R2 = 0.33, Pearson’s correlation) between the puta- tive LOF burden ratio and country-specific influenza mortality rates (Extended Data Fig. 3d) that was significantly different from random (mean R2 = 0.165, s.e.m. = 0.006, Pearson; P = 1.45 × 10−102, Wilcoxon signed-rank test; Extended Data Fig. 3d). Among west African groups, the resulting correlation was even more striking (R2 = 0.99; random iteration mean R2 = 0.235, s.e.m. = 0.008; P = 9.33 × 10−302, Wilcoxon signed-rank test; Extended Data Fig. 3d, e). A strong correlation was also seen with mortality associated with infection with human immunode- ficiency virus (HIV) (R2 = 0.501, random mean R2 = 0.187, s.e.m. = 0.006; P = 6.86 × 10−209, Wilcoxon signed-rank test), but not with infection with malaria (R2 = 0.120; random mean R2 = 0.165, s.e.m. = 0.006; P = 0.99, Wilcoxon signed-rank test) or hepatitis C virus (HCV) (R2 = 0.0002; ran- dom mean R2 = 0.174, s.e.m. = 0.006; P = 1.00, Wilcoxon signed-rank test) (Extended Data Fig. 3e). Although environmental and socioeconomic factors remain important contributors to country-reported mortality, these data provide further support for a host genetic contribution to outcomes in some infectious diseases. Article were masked (Methods). Among the remaining HDVs (n = 2,497), more than 40% (n = 1,106) were observed between BOT (southern Africa) and MAL (northwest Africa) (Supplementary Table 17), and this geographi- cal separation also generated the most divergent allele frequencies. Some of these HDVs probably also reflect the high proportion and historically deeper KS ancestry among BOT31. Empirically, 275 HDVs were located within 50 kb of genome-wide significant single-nucleotide polymorphisms (P < 5 × 10−8) in the NHGRI-EBI GWAS Catalog, and these primarily mapped to genes implicated in cardiometabolic traits, such as systolic blood pressure and type 2 diabetes (Methods and Extended Data Fig. 3a). Even though the vast majority of GWAS have been con- ducted in populations of European ancestry, GWAS hits proximal to HDVs were from studies that included participants of diverse ancestries (Extended Data Fig. 3a). unique variants annotated as pathogenic or likely pathogenic (level 4), around 21% (54 out of 262) had a minor allele frequency (MAF) > 0.05 in at least one HC-WGS population (Fig. 4c and Supplementary Table 20) and, of these, 13 (4.9%) had a MAF < 0.05 across all population groups in gnomAD (Extended Data Fig. 4a and Supplementary Note 4). We interpret this finding as most suggestive of variant misclassifications in ClinVar and other databases. Finally, we surveyed the frequency of select, well-described, disease-associated alleles commonly found in populations of Afri- can descent (Methods). Consistent with expectations for a locus that protects against malaria mortality, estimates of the G6PD A− 202A and 376G alleles were generally consistent with the distribution of endemic malaria across the continent. Eleven single-nucleotide poly- morphisms in G6PD associated with protection against severe malaria in heterozygotic female participants35 also showed similarly divergent frequencies across our populations (Extended Data Fig. 4b). Similarly, the common sickle cell disease mutation (rs334; HbS; MIM 603903) was found at typically high allele frequencies in malaria-endemic west and east African populations36 (Fig. 4d and Supplementary Table 21). HbS frequencies ranged from 10% (BRN) to 19% (SSG), but was nearly absent from south African (BOT, XHS and SOT) genomes. Notably, HbS showed widely divergent allele frequencies in the two populations from Uganda, with the Bantu-speaking (UBS) population having one of the highest frequencies (20%) whereas the allele was not observed in the UNS population. Article APOL1 G1 and G2 alleles in the homozygous or compound heterozy- gous state (G1/G1, G1/G2 or G2/G2) confer protection against Trypa- nosoma brucei gambiense infection but also increase susceptibility to nephropathy in non-trypanosomiasis endemic areas, especially in the presence of HIV infection12. G1 and G2 allele frequencies varied widely across our populations, and were the highest among west African groups (for example, the G1 frequency was 43% in CIV and 34% in FNB), but were significantly lower elsewhere on the continent (Fig. 4d). This is consistent with previous reports of geographical correlation with the prevalence of sleeping sickness12. The frequency of G1 was also highly differentiated between the two Uganda population groups (UBS, 14%; UNS, 2%)37, and this was in contrast to the frequencies of G2 in the two groups (UBS, 10.6%; UNS, 11.5%). The overall frequency of the APOL1 risk genotype (that is, the recessive state for the diplotype: G1/G1, G1/ G2 or G2/G2) varied substantially among the groups in which it was present (Fig. 4d and Supplementary Table 21), suggesting that the risk of APOL1-associated nephropathy is appreciably high across the continent, with the highest burden in west Africa. Previously, the Maasai in Kinyawa, Kenya38 were the only African population noted to have an appreciable frequency of HLA-B5701, which mediates hypersensitivity to the antiretroviral drug abacavir. In our expanded dataset, the allele was absent from west African popula- tions, but observed in BSZ (1.5%), UNS (3.3%) and XHS (6.3%) (Supple- mentary Table 21), at frequencies typically observed in populations of European or Asian ancestry. Highly differentiated variants in African genomes The complex population structures and variable selection pressures observed in our dataset are known to promote differentiation in allele frequencies between populations. We therefore sought to identify highly differentiated variants (HDVs) with substantially different (more than 40%) allele frequencies between H3A-Baylor populations (Methods). For this analysis, HDVs across the β-globin gene cluster (chromosome 11) and the HLA region (chromosome 6), both of which are known to have extensive linkage disequilibrium between markers, Nature \| Vol 586 \| 29 October 2020 \| 745 745 Online content Alongside immune genes, we also observed positive selection in genes associated with DNA repair, reproduction, and carbohydrate and lipid metabolism, as well as geographical-region-specific positive selection in genes such as PLAT and SERPINA1 within NC speakers. 6. Posey, J. E. et al. Insights into genetics, human biology and disease gleaned from family based genomic studies. Genet. Med. 21, 798–812 (2019). 7. Landry, L. G., Ali, N., Williams, D. R., Rehm, H. L. & Bonham, V. L. Lack of diversity in genomic databases is a barrier to translating precision medicine research into practice. Health Aff. 37, 780–785 (2018). 8. H3Africa Consortium. Enabling the genomic revolution in Africa. Science 344, 1345–1346 (2014). 9. Patin, E. et al. Dispersals and genetic adaptation of Bantu-speaking populations in Africa and North America. Science 356, 543–546 (2017). 10. Hanchard, N. et al. Classical sickle beta-globin haplotypes exhibit a high degree of long-range haplotype similarity in African and Afro-Caribbean populations. BMC Genet. 8, 52 (2007). 11. Ranciaro, A. et al. Genetic origins of lactase persistence and the spread of pastoralism in Africa. Am. J. Hum. Genet. 94, 496–510 (2014). 12. Genovese, G. et al. Association of trypanolytic ApoL1 variants with kidney disease in 12. Genovese, G. et al. Association of trypanolytic ApoL1 variants with kidney dis African Americans. Science 329, 841–845 (2010). 12. Genovese, G. et al. Association of trypanolytic Apo African Americans. Science 329, 841–845 (2010). African Americans. Science 329, 841–845 (2010). 13. Sabeti, P. C. et al. Detecting recent positive selection in the human genome from haplotype structure. Nature 419, 832–837 (2002). 13. Sabeti, P. C. et al. Detecting recent positive selection in the haplotype structure. Nature 419, 832–837 (2002). haplotype structure. Nature 419, 832–837 (2002). 14. Schlebusch, C. M. et al. Genomic variation in seven Khoe-San groups reveals adaptation and complex African history. Science 338, 374–379 (2012). The combined effect of ancestral events and exposure to infec- tious agents on the diversity and variation of African genomes was perhaps best exemplified by the stark allele frequency differences observed between the UNS and UBS groups from Uganda. Despite their shared geography, the two groups varied significantly in the frequencies of three of the four medically relevant variants surveyed, including HbS and G6PD alleles. These two loci are known to be protec- tive against severe malaria, which is endemic in Uganda. Online content Genomic evidence for shared common ancestry of East African hunting-gathering populations and insights into local adaptation. Proc. Natl Acad. Sci. USA 116, 4166–4175 (2019). 16. Skoglund, P. et al. Reconstructing prehistoric African population structure. Cell 171, 59–71 (2017). 17. Choudhury, A. et al. Whole-genome sequencing for an enhanced understanding of genetic variation among South Africans. Nat. Commun. 8, 2062 (2017). 18. Ilboudo, H. et al. Introducing the TrypanoGEN biobank: a valuable resource for the elimination of human African trypanosomiasis. PLoS Negl. Trop. Dis. 11, e0005438 (2017). 19. Alexander, D. H., Novembre, J. & Lange, K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 19, 1655–1664 (2009). 20. Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012). 21. Semo, A. et al. Along the Indian Ocean coast: genomic variation in Mozambique provides new insights into the Bantu expansion. Mol. Biol. Evol. 37, 406–416 (2020). 22. Loh, P.-R. et al. Inferring admixture histories of human populations using linkage disequilibrium. Genetics 193, 1233–1254 (2013). 23. Patin, E. et al. The impact of agricultural emergence on the genetic history of African rainforest hunter-gatherers and agriculturalists. Nat. Commun. 5, 3163 (2014). 24. Shriner, D. & Rotimi, C. N. Genetic history of Chad. Am. J. Phys. Anthropol. 167, 804–812 (2018). 25 Campbell I M et al Multiallelic positions in the human genome: challenges for genetic 3. Tishkoff, S. A. et al. The genetic structure and history of Africans and African Americans. Science 324, 1035–1044 (2009). . Gurdasani, D. et al. The African Genome Variation Project shapes medical genetics in Africa. Nature 517, 327–332 (2015). 5. Lek, M. et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285–291 (2016). Viral epidemics, including outbreaks of HIV, Ebola and Lassa fever, have been reported across Africa. Against this backdrop, our obser- vations of selected loci that overlap genes that are important to viral infection support the potential for a hitherto undescribed role for resistance and/or susceptibility to viral infections in shaping the genomes of human populations across Africa. This was partially bol- stered by strong correlations between putative LOF variants in genes implicated in influenza and HIV and their respective disease mortalities, although the latter observations require replication and confirmatory analyses in cohorts of individuals with the diseases. Online content Historically, differences in HbS frequencies between these groups were attributed to the relatively recent spread of malaria to NS populations; how- ever, our results, and other recent findings41, suggest that recent migration from northern regions, where malaria is less common (UNS), as opposed to from malaria-endemic western regions (UBS) is another plausible explanation42. A similar argument can be made for the APOL1 high-risk G1 and G2 alleles, which have been shown to provide protection against specific Trypanosoma species12. These alleles are commonly observed in trypanosomiasis-endemic regions such as Uganda and west Africa (UBS), but less so in northeastern geographies including Sudan (UNS). Similarly, HLA-B5701 was previ- ously only observed among northeast African populations; its high frequency among the UNS, yet absence from the UBS, probably also reflects this ancestral divergence. and complex African history. Science 338, 374–379 (2012) 15. Scheinfeldt, L. B. et al. Genomic evidence for shared common ancestry of East African hunting-gathering populations and insights into local adaptation. Proc. Natl Acad. Sci. USA 116, 4166–4175 (2019). 16. Skoglund, P. et al. Reconstructing prehistoric African population structure. Cell 171, 59–71 (2017). 17. Choudhury, A. et al. Whole-genome sequencing for an enhanced understanding of genetic variation among South Africans. Nat. Commun. 8, 2062 (2017). genetic variation among South Africans. Nat. Commun. 8, 2062 (2017 18. Ilboudo, H. et al. Introducing the TrypanoGEN biobank: a valuable resource for the elimination of human African trypanosomiasis. PLoS Negl. Trop. Dis. 11, e0005438 (2017). 19. Alexander, D. H., Novembre, J. & Lange, K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 19, 1655–1664 (2009). 20. Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012). 20. Patterson, N. et al. Ancient admixture in human history. Genetics 1 21. Semo, A. et al. Along the Indian Ocean coast: genomic variation in Mozambique provides new insights into the Bantu expansion. Mol. Biol. Evol. 37, 406–416 (2020). 22. Loh, P.-R. et al. Inferring admixture histories of human populations using linkage disequilibrium. Genetics 193, 1233–1254 (2013). 23. Patin, E. et al. The impact of agricultural emergence on the genetic history of African rainforest hunter-gatherers and agriculturalists. Nat. Commun. 5, 3163 (2014). 24. Shriner, D. & Rotimi, C. N. Genetic history of Chad. Am. J. Phys. Anthropol. 167, 804–812 (2018). 25. Campbell, I. M. et al. Multiallelic positions in the human genome: challenges for genetic analyses. Hum. Mutat. 37, 231–234 (2016). 26. Campbell, M. Discussion To provide a more global context for medically relevant genetic varia- tion, we annotated our dataset with the American College of Medical Genetics and Genomics (ACMG) Secondary Findings gene panel (ACMG 2.0) of reportable variants in 59 genes34. Only eight individuals carried any reportable ACMG variants, and these were limited to singleton vari- ants (one per participant) in TINF2, KCNQ1 and RYR1 (associated with dominantly inherited disorders), as well as ATP7B and PKP2 (associated with recessively inherited disorders). By contrast, almost everyone in our HC-WGS cohort carried at least one variant designated as ‘patho- genic’ (level 5) in the ClinVar Database (v.20181028) (http://www.ncbi. nlm.nih.gov/clinvar/), with each person carrying a median of 7 (range, 2–14) alleles (Fig. 4b and Supplementary Table 19). Among the 262 This study represents one of the most-extensive studies of high- depth-sequenced African genomes to date. We deliberately focused on SNVs—which could be confidently inferred—but a similar wealth of diversity and novelty is likely to be found within other variant classes. Despite drawing individuals from ongoing genomics studies on the continent, care was taken to avoid possible systematic biases and the patterns of variation are thus expected to be largely representative of the respective groups. Our results reveal a genetic continuum of NC speaker populations across the continent and extend our current understanding of the routes, timing and extent of the Bantu migration—the defining demo- graphic event of African genetic diversity. The proposed route overlaps 746 \| Nature \| Vol 586 \| 29 October 2020 746 with the spread of the Kalundu pottery tradition, which has also been associated with the Bantu expansion in these regions39,40. However, the estimated dates for the spread of Kalundu pottery predate our admixture dates, and the association between this tradition and Bantu migration has been questioned, leaving the proposed parallels between the archaeological and genetic migrations unresolved. with the spread of the Kalundu pottery tradition, which has also been associated with the Bantu expansion in these regions39,40. However, the estimated dates for the spread of Kalundu pottery predate our admixture dates, and the association between this tradition and Bantu migration has been questioned, leaving the proposed parallels between the archaeological and genetic migrations unresolved. Online content C. & Tishkoff, S. A. African genetic diversity: implications for human demographic history, modern human origins, and complex disease mapping. Annu. Rev. Genomics Hum. Genet. 9, 403–433 (2008). 27. Pavlidis, P., Živkovic, D., Stamatakis, A. & Alachiotis, N. SweeD: likelihood-based detection of selective sweeps in thousands of genomes. Mol. Biol. Evol. 30, 2224–2234 (2013). Our findings indicate that the implementation and use of genome- level sequence data in Africa will require a broadly ascertained and comprehensive compendium of variation, alongside high-level curation of variants. African genome variation is likely to be a bet- ter representation of variant distribution for both African diaspora and global populations, and, therefore, a full repertoire of African genomic variation could provide a better genomic reference for both medical and population genetics. The data generated have facilitated the development of a microarray genotyping chip and imputation panel, and are being made available to researchers in the field (see Methods, ‘Data availability’) as important springboards for future studies of demography, migration, ancestry and genetic variation in Africa. p g 28. Szpiech, Z. A. & Hernandez, R. D. selscan: an efficient multithreaded program to perform EHH-based scans for positive selection. Mol. Biol. Evol. 31, 2824–2827 (2014). EHH-based scans for positive selection. Mol. Biol. Evol. 31, 2824–2827 (201 29. Vitti, J. J., Grossman, S. R. & Sabeti, P. C. Detecting natural selection in genomic data. Annu. Rev. Genet. 47, 97–120 (2013). 30. Yi, X. et al. Sequencing of 50 human exomes reveals adaptation to high altitude. Science 329, 75–78 (2010). 31. Retshabile, G. et al. Whole-exome sequencing reveals uncaptured variation and distinct ancestry in the southern African population of Botswana. Am. J. Hum. Genet. 102, 731–743 (2018). 32. Lim, E. T. et al. Distribution and medical impact of loss-of-function variants in the Finnish founder population. PLoS Genet. 10, e1004494 (2014). 32. Lim, E. T. et al. Distribution and medical impact of loss of function variants in the Finnish founder population. PLoS Genet. 10, e1004494 (2014). 33. World Health Organization. WHO Influenza (Seasonal): Fact Sheet https://www.who.int/ news-room/fact-sheets/detail/influenza-(seasonal) (2016). 33. World Health Organization. WHO Influenza (Seasonal): Fact Sheet https://www.who.int/ news-room/fact-sheets/detail/influenza-(seasonal) (2016). 34. Kalia, S. S. et al. Recommendations for reporting of secondary findings in clinical exome and genome sequencing, 2016 update (ACMG SF v2.0): a policy statement of the American College of Medical Genetics and Genomics. Genet. Med. 19, 249–255 (2017). . Nielsen, R. et al. Tracing the peopling of the world through genomics. Nature 541, 302–310 (2017). Online content Detecting recent positive selection in the human genome from haplotype structure. Nature 419, 832–837 (2002). 14. Schlebusch, C. M. et al. Genomic variation in seven Khoe-San groups reveals adaptation and complex African history. Science 338, 374–379 (2012). 15. Scheinfeldt, L. B. et al. Genomic evidence for shared common ancestry of East African hunting-gathering populations and insights into local adaptation. Proc. Natl Acad. Sci. USA 116, 4166–4175 (2019). 16. Skoglund, P. et al. Reconstructing prehistoric African population structure. Cell 171, 59–71 (2017). 17 Choudhury A et al Whole genome sequencing for an enhanced understanding of 1. Nielsen, R. et al. Tracing the peopling of the world through genomics. Nature 541, 302–310 (2017). 2. The 1000 Genomes Project Consortium. A global reference for human genetic variation. Nature 526, 68–74 (2015). 3. Tishkoff, S. A. et al. The genetic structure and history of Africans and African Americans. Science 324, 1035–1044 (2009). 4. Gurdasani, D. et al. The African Genome Variation Project shapes medical genetics in Africa. Nature 517, 327–332 (2015). 5. Lek, M. et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285–291 (2016). 6. Posey, J. E. et al. Insights into genetics, human biology and disease gleaned from family based genomic studies. Genet. Med. 21, 798–812 (2019). 7. Landry, L. G., Ali, N., Williams, D. R., Rehm, H. L. & Bonham, V. L. Lack of diversity in genomic databases is a barrier to translating precision medicine research into practice. Health Aff. 37, 780–785 (2018). 8. H3Africa Consortium. Enabling the genomic revolution in Africa. Science 344, 1345–1346 (2014). 9. Patin, E. et al. Dispersals and genetic adaptation of Bantu-speaking populations in Africa and North America. Science 356, 543–546 (2017). 10. Hanchard, N. et al. Classical sickle beta-globin haplotypes exhibit a high degree of long-range haplotype similarity in African and Afro-Caribbean populations. BMC Genet. 8, 52 (2007). 11. Ranciaro, A. et al. Genetic origins of lactase persistence and the spread of pastoralism in Africa. Am. J. Hum. Genet. 94, 496–510 (2014). 12. Genovese, G. et al. Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science 329, 841–845 (2010). 13. Sabeti, P. C. et al. Detecting recent positive selection in the human genome from haplotype structure. Nature 419, 832–837 (2002). 14. Schlebusch, C. M. et al. Genomic variation in seven Khoe-San groups reveals adaptation and complex African history. Science 338, 374–379 (2012). 15. Scheinfeldt, L. B. et al. Online content Any methods, additional references, Nature Research reporting sum- maries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author con- tributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41586-020-2859-7. Nigeria, in keeping with its tremendous linguistic diversity, is cur- rently the best-represented African country in terms of genomic data. Our observations of substantial NS admixture in the Berom, and both HDVs and novel variation in both NC and non-NC speakers, suggest that Nigerian populations in existing public databases are not only likely to underrepresent the genomic diversity of Nigeria, but are almost cer- tainly poor general proxies of African continental groups. Additional deep sequencing in multiple African populations will be needed to provide a more-comprehensive compendium of variation across the continent. 1. Nielsen, R. et al. Tracing the peopling of the world through genomics. Nature 541, 302–310 (2017). 2. The 1000 Genomes Project Consortium. A global reference for human genetic variation. Nature 526, 68–74 (2015). 3. Tishkoff, S. A. et al. The genetic structure and history of Africans and African Americans. Science 324, 1035–1044 (2009). 4. Gurdasani, D. et al. The African Genome Variation Project shapes medical genetics in Africa. Nature 517, 327–332 (2015). 5. Lek, M. et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285–291 (2016). 6. Posey, J. E. et al. Insights into genetics, human biology and disease gleaned from family based genomic studies. Genet. Med. 21, 798–812 (2019). 7. Landry, L. G., Ali, N., Williams, D. R., Rehm, H. L. & Bonham, V. L. Lack of diversity in genomic databases is a barrier to translating precision medicine research into practice. Health Aff. 37, 780–785 (2018). 8. H3Africa Consortium. Enabling the genomic revolution in Africa. Science 344, 1345–1346 (2014). 9. Patin, E. et al. Dispersals and genetic adaptation of Bantu-speaking populations in Africa and North America. Science 356, 543–546 (2017). 10. Hanchard, N. et al. Classical sickle beta-globin haplotypes exhibit a high degree of long-range haplotype similarity in African and Afro-Caribbean populations. BMC Genet. 8, 52 (2007). 11. Ranciaro, A. et al. Genetic origins of lactase persistence and the spread of pastoralism in Africa. Am. J. Hum. Genet. 94, 496–510 (2014). 12. Genovese, G. et al. Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science 329, 841–845 (2010). 13. Sabeti, P. C. et al. © The Author(s) 2020 1Sydney Brenner Institute for Molecular Bioscience, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa. 2Center for Research in Agricultural Genomics (CRAG), Plant and Animal Genomics Program, CSIC-IRTA-UAB-UB, Barcelona, Spain. 3Computational Biology Division and H3ABioNet, Department of Integrative Biomedical Sciences, IDM, University of Cape Town, Cape Town, South Africa. 4System and Data Engineering Team, Abdelmalek Essaadi University, ENSA, Tangier, Morocco. 5Centre for Proteomic and Genomic Research (CPGR), Cape Town, South Africa. 6South African National Bioinformatics Network, University of the Western Cape, Bellville, South Africa. 7Center for Research on Genomics and Global Health, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA. 8Department of Agriculture and Food Science, Faculty of Agriculture, University of Mauritius, Reduit, Mauritius. 9Department of Digital Technologies,Faculty of Information, Communication & Digital Technologies, University of Mauritius, Reduit, Mauritius. 10Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK. 11Institute of Human Virology Nigeria, Abuja, Nigeria. 12Covenant University Bioinformatics Research (CUBRe), Covenant University, Ota, Nigeria. 13Department of Computer and Information Sciences, Covenant University, Ota, Nigeria. 14School of Electrical and Information Engineering, University of the Witwatersrand, Johannesburg, South Africa. 15College of Veterinary Medicine, Animal Resources and Biosecurity, Makerere University, Kampala, Uganda. 16Laboratory of Bioinformatics, Biomathematics and Biostatistics (BIMS), Institute Pasteur of Tunis, Tunis, Tunisia. 17Medical Biochemistry and Molecular Biology Department, Faculty of Medicine, Ain Shams University, Abbaseya, Cairo, Egypt. 18Department of Epidemiology and Public Health, University of Maryland School of Medicine, University of Maryland Baltimore, Baltimore, MD, USA. 24Institute of Human Virology and Greenebaum Cancer Center, University of Maryland School of Medicine, Baltimore, MD, USA. 25Navrongo Health Research Centre, Navrongo, Ghana. 26Clinical Research Unit of Nanoro, Institut de Recherche en Sciences de la Sante, Bobo-Dioulasso, Burkina Faso. 27Faculty of Medicine and Odontostomatology, University of Science, Techniques and Technologies of Bamako (USTTB), Bamako, Mali. 28Service de Neurologie, Centre Hospitalier Universitaire du Point “G”, Bamako, Mali. 29Neurogenetics Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA. 30Botswana-Baylor Children’s Clinical Centre of Excellence, Gaborone, Botswana. 31Medical Microbiology, College of Health Sciences, Makerere University, Kampala, Uganda. 32Baylor College of Medicine Children’s Foundation, Kampala, Uganda. 33Department of Pathology and Immunology, Baylor College of Medicine, Houston, TX, USA. 34Department of Biological Sciences, University of Botswana, Gaborone, Botswana. 35University of Ghana Medical School, Accra, Ghana. 36The University of Kansas School of Medicine, Kansas City, KS, USA. Online content Nature \| Vol 586 \| 29 October 2020 \| 747 Nature \| Vol 586 \| 29 October 2020 \| 7 747 TrypanoGEN Research Group TrypanoGEN Research Group 41. Mulindwa, J. et al. High levels of genetic diversity within Nilo-Saharan populations: implications for human adaptation. Am. J. Hum. Genet. 107, 473–486 (2020). Mulindwa, J. et al. High levels of genetic diversity within Nilo-Saharan popu 42. Shiroya, O. J. E. The Lugbara states — politics, economics and warfare in the eighteenth and nineteenth centuries. TransAfrican J. Hist. 10, 125–183 (1981). Enock Matovu15, Bruno Bucheton39, Christiane Hertz-Fowler40,41, Mathurin Koffi42, Annette Macleod43, Dieudonne Mumba-Ngoyi44, Harry Noyes40, Oscar A. Nyangiri15, Gustave Simo45 & Martin Simuunza46 42. Shiroya, O. J. E. The Lugbara states — politics, economics and warfa and nineteenth centuries. TransAfrican J. Hist. 10, 125–183 (1981). 43. R Core Team. R: A Language and Environment for Statistical Computing. http://www. R-project.org/ (R Foundation for Statistical Computing, 2017). 748 \| Nature \| Vol 586 \| 29 October 2020 Article 35. Manjurano, A. et al. African glucose-6-phosphate dehydrogenase alleles associated with protection from severe malaria in heterozygous females in Tanzania. PLoS Genet. 11, e1004960 (2015). 19University of Maryland Greenebaum Comprehensive Cancer Center, University of Maryland School of Medicine, University of Maryland Baltimore, Baltimore, MD, USA. 20Division of Human Genetics, Department of Pathology, Faculty of Health Sciences, Institute for Infectious, Disease and Molecular Medicine, University of Cape Town, Cape Town, South Africa. 21Human Genome Sequencing Center, Baylor College of Medicine, Houston, TX, USA. 22Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas, USA. 23Division of Human Genetics, National Health Laboratory Service, and School of Pathology, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa. Lists of members and their affiliations appear at the end of the paper. ✉e-mail: adeyemoa@mail.nih.gov; zane.lombard@wits.ac.za; hanchard@bcm.edu 36. Howes, R. E., Battle, K. E., Satyagraha, A. W., Baird, J. K. & Hay, S. I. G6PD deficiency: global distribution, genetic variants and primaquine therapy. Adv. Parasitol. 81, 133–201 (2013). 37. Kimuda, M. P. et al. No evidence for association between APOL1 kidney disease risk alleles and human African trypanosomiasis in two Ugandan populations. PLoS Negl. Trop. Dis. 12, e0006300 (2018). 38. Rotimi, C. N. & Jorde, L. B. Ancestry and disease in the age of genomic medicine. N. Engl. J. Med. 363, 1551–1558 (2010). 39. Phillipson, D. W. Iron Age history and archaeology in Zambia. J. Afr. Hist. 15, 1–25 (1974). 39. Phillipson, D. W. Iron Age history and archaeology in Zambia. J. Afr. Hist. 15, 1–25 (1974). 40. Schlebusch, C. M. & Jakobsson, M. Tales of human migration, admixture, and selection in Africa Annu Rev Genomics Hum Genet 19 405–428 (2018) 39. Phillipson, D. W. Iron Age history and archaeology in Zambia. J. Afr. Hist. 15, 1 25 (1974). 40. Schlebusch, C. M. & Jakobsson, M. Tales of human migration, admixture, and selection in Africa. Annu. Rev. Genomics Hum. Genet. 19, 405–428 (2018). H3Africa Consortium Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Ananyo Choudhury1, Shaun Aron1, Laura Botigué2, Dhriti Sengupta1, Gerrit Botha3, Taoufik Bensellak4, Gordon Wells5,6, Judit Kumuthini6, Daniel Shriner7, Yasmina J. Fakim8,9, Anisah W. Ghoorah9, Eileen Dareng10,11, Trust Odia12, Oluwadamilare Falola12, Ezekiel Adebiyi12,13, Scott Hazelhurst1,14, Gaston Mazandu3, Oscar A. Nyangiri15, Mamana Mbiyavanga3, Alia Benkahla16, Samar K. Kassim17, Nicola Mulder3, Sally N. Adebamowo18,19, Emile R. Chimusa20, Charles Rotimi7, Michèle Ramsay1,23, Adebowale A. Adeyemo7, Zané Lombard23, Neil A. Hanchard22, Clement Adebamowo11,18,19,24, Godfred Agongo1,25, Romuald P. Boua1,26, Abraham Oduro25, Hermann Sorgho26, Guida Landouré27,28, Lassana Cissé28, Salimata Diarra28,29, Oumar Samassékou27, Gabriel Anabwani30, Mogomotsi Matshaba30, Moses Joloba31, Adeodata Kekitiinwa32, Graeme Mardon22,33, Sununguko W. Mpoloka34, Samuel Kyobe31, Busisiwe Mlotshwa34, Savannah Mwesigwa31, Gaone Retshabile34, Lesedi Williams34, Ambroise Wonkam20, Ahmed Moussa4, Dwomoa Adu35, Akinlolu Ojo36, David Burke37, Babatunde O. Salako38, Enock Matovu15, Bruno Bucheton39, Christiane Hertz-Fowler40,41, Mathurin Koffi42, Annette Macleod43, Dieudonne Mumba-Ngoyi44, Harry Noyes40, Oscar A. Nyangiri15, Gustave Simo45, Martin Simuunza46, Philip Awadalla47,48, Vanessa Bruat48 & Elias Gbeha48 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Quality control before variant calling and BAM file augmentation. Quality control before variant calling and BAM file augmentation. Before variant calling, the percentage of aligned reads was found to be 90% or greater in each sample. The resulting BAM files were merged on a per-sample basis and these sample-level BAM files were recalibrated using GATK45. This process consisted of a per-sample realignment of reads around known and discovered insertions and deletions (indels) using the known indels from the gold datasets of the Mills Devine and 1000 Genomes Project (1000G), as well as the low-coverage dataset of 1000G phase 1. GATK (v.3.3-0) RealignerTargetCreator and Indel- Realigner were used for alignment, in addition to base quality score recalibration with GATK BaseRecalibrator and PrintReads (using known variant sites from dbSNP v.138 and the same indels used in lo- cal realignment). SAMtools was used to generate a base quality score and MD tag (that is, a string describing the mismatching positions of a read to the reference used for reference-free SNV and indel calling), which helped to improve calling quality. Additional quality checks for cross-sample contamination were performed using VerifyBamID from 1000G Omni2.5 VCF, requiring that the calculated FREEMIX was less than 0.05. One H3A-Baylor sample did not pass these quality controls and was not included in the downstream analysis. The TrypanoGEN Collaborative Centre of the H3Africa Consortium. The TrypanoGEN project is a collaborative centre funded by the Well- come Trust to study the host and parasite genetics and genomics of trypanosomiasis across Africa18. Ethical approval for the use of samples for genomic studies of trypanosomiasis was obtained in participating countries and informed consent for genomic studies and sharing of data with researchers working on other diseases was obtained from participating individuals. A total of 300 individuals were recruited from five countries (Uganda, Zambia, the DRC, Cameroon and Côte d’Ivoire) of which 200 had been sequenced for inclusion in this project (Supplementary Methods Table 2). DNA was extracted from blood samples in the respective countries of collection with the exception of samples from Guinea and Côte d’Ivoire, for which DNA extraction was performed at CIRDES in Burkina Faso. WGS was subsequently per- formed on an Illumina Hiseq 2500 to an average depth of 10× using the Illumina TruSeq PCR-free kit at the Centre for Genomic Research, University of Liverpool, UK. Variant annotation. Variant annotations were obtained using SnpEff (version 4.3-3) (Supplementary Table 23) with human genome build GRCh37.75 (October 2016). Samples, datasets and sequencing platforms The primary analysis datasets were derived from three sources: The H3Africa Consortium (referred to as H3A-Baylor)8, The TrypanoGEN Collaborative Centre of the H3Africa Consortium (TrypanoGEN)18 and the SAHGP17. All samples were collected after appropriate approvals had been obtained from local Ethics Boards and Committees in each of the represented countries, and participants gave informed consent (see Supplementary Note 5 for details). Sequencing was performed on Illumina platforms and—after alignment and baseline quality control to account for the different platforms and coverage depths (Supplemen- tary Methods Table 2)—data were combined to create a single merged file (Supplementary Methods Fig. 1) to facilitate downstream analyses. Data reporting No statistical methods were used to predetermine sample size. The experiments were not randomized and the investigators were not blinded to allocation during experiments and outcome assessment. Methods Johannesburg (protocol number: M120223). Three groups of partici- pants were enrolled and venous blood was collected using EDTA tubes. Inclusion criteria were as follows: male, over the age of 18 years, four grandparents who speak the same language as the participant, not known to be related to the other participants in the study and willing to provide broad informed consent (including consent to share data and DNA for future studies approved by the HREC (Medical)). Two main Bantu-speaking ethnolinguistic groups were included: the So- tho (Sotho–Tswana-speaking individuals; n = 8) were recruited from in and around the town of Ventersburg in the Free State Province and the Xhosa-speaking individuals (Nguni language; n = 7) were recruited from the Eastern Cape Province. One individual spoke Zulu (Nguni language) and was from Johannesburg. The DNA samples were normal- ized to around 60 ng μl−1 and approximately 5 μg DNA was submitted to the Illumina Service Centre in San Diego, USA, for sequencing on the Illumina HiSeq 2000 instrument (around 100-bp paired-end reads, about 314-bp insert size) with a minimum read depth of coverage of 30×. Johannesburg (protocol number: M120223). Three groups of partici- pants were enrolled and venous blood was collected using EDTA tubes. Inclusion criteria were as follows: male, over the age of 18 years, four grandparents who speak the same language as the participant, not known to be related to the other participants in the study and willing to provide broad informed consent (including consent to share data and DNA for future studies approved by the HREC (Medical)). Two main Bantu-speaking ethnolinguistic groups were included: the So- tho (Sotho–Tswana-speaking individuals; n = 8) were recruited from in and around the town of Ventersburg in the Free State Province and the Xhosa-speaking individuals (Nguni language; n = 7) were recruited from the Eastern Cape Province. One individual spoke Zulu (Nguni language) and was from Johannesburg. The DNA samples were normal- ized to around 60 ng μl−1 and approximately 5 μg DNA was submitted to the Illumina Service Centre in San Diego, USA, for sequencing on the Illumina HiSeq 2000 instrument (around 100-bp paired-end reads, about 314-bp insert size) with a minimum read depth of coverage of 30×. 37Department of Human Genetics, University of Michigan Medical School, Ann Arbor, MI, USA. 38College of Medicine, University of Ibadan, Ibadan, Nigeria. 39Institut de Recherche pour le Développement, Montpellier, France. 40Institute of Integrative Biology, University of Liverpool, Liverpool, UK. 41Wellcome Trust, London, UK. 42Jean Lorougnon Guede University, Daloa, Côte d’Ivoire. 43Institute of Biodiversity Animal Health and Comparative Medicine, University of Glasgow, Glasgow, UK. 44Institut National de Recherche Biomedicale, Kinshasa, Democratic Republic of Congo. 45Faculty of Science, University of Dschang, Dschang, Cameroon. 46School of Veterinary Medicine, University of Zambia, Lusaka, Zambia. 47Department of Molecular Genetics, University of Toronto, Toronto, Ontario, Canada. 48Ontario Institute for Cancer Research, Toronto, Ontario, Canada. 748 \| Nature \| Vol 586 \| 29 October 2020 Data processing and merging See Supplementary Table 23 for a full list of references for databases and software used. The H3Africa Consortium. The principal investigators of each of the 19 projects funded during the first 5-year funding cycle of H3Africa were invited to submit samples for WGS, provided the existing consent for recruited individuals included the broad use of samples for WGS, and the project had existing ethics approval for such a study. In addition, a broader request was made to consortium researchers with other samples from African populations that would be similarly eligible for inclusion, even if they were not recruited through a formal H3Africa project. A total of 519 samples from 8 projects were submitted for con- sideration. Submitted samples were predominantly ascertained from control datasets recruited from the respective studies, with the excep- tion of samples from case-only studies in Cameroon, Botswana, Mali and Benin (Fig. 1 and Supplementary Methods Table 1). These samples were prioritized to include population and ethnolinguistic groups that had not previously been sequenced. Samples were shipped to the Human Genome Sequencing Center at Baylor College of Medicine, Houston, USA, under signed material transfer agreements from each project. A total of 348 samples were prepared using the TruSeq Nano DNA Library Prep Kits and underwent WGS on an Illumina X-Ten to a minimum depth of coverage of 30×. The resulting dataset was labelled H3A-Baylor (Supplementary Methods Table 2). Alignment and pre-processing of reads. Raw FASTQ reads generated by sequencing were mapped to the human reference genome GRCh37 (also known as hs37d5) using the BWA-MEM algorithm of the BWA soft- ware package44. Optical and PCR duplicate reads were marked with Picard MarkDuplicates on a per-sample basis45 and reads were sorted by coordinate using SAMtools v.0.1.1946. Article As the TrypanoGEN dataset had a lower depth of coverage (around 10×), a minimum confidence threshold at which variants were called and included was set to 10; for both of the datasets with a high depth of coverage (SAHGP and H3A-Baylor), a minimum confidence threshold at which variants were called and included was set to 30. Joint variant calling was done on each dataset using GenotypeGVCFs in GATK. Variant filtering of autosomal genes. Variant quality score recalibra- tion (VQSR) was performed for each dataset separately. SNVs were filtered using VariantRecalibrator and ApplyRecalibration in GATK. For SNVs, we used overlapping sites from HapMap III and 1000G phase 1 Omni2.5 sites as truth and training sets (prior probabilities of 15 and 12, respectively for HapMap III and 1000G phase 1 Omni2.5 sites). High-confidence 1000G phase 1 SNVs were used as an additional training set (prior probability of 10). dbSNP v.138 was used as a set of known sites (prior probability of 2). To build the VQSR Gaussian mix- ture model, we used annotations at each site related to coverage (QD (QualByDepth) and DP—where DP is the approximate read depth after filtering reads with poor mapping quality and bad mates and QD is the variant confidence normalized by the unfiltered depth for the variant allele); strand bias (FS (FisherStrand) and SOR (StrandOddsRatio)— where FS is a Phred-scaled P value using Fisher’s exact test and SOR is the odds ratio of a 2 × 2 contingency table of positive/negative strand and reference/alternative allele), mapping quality (MQ, MQRankSum and ReadPosRankSum—where MQ is the root mean square of the mapping qualities), which serves to average across reads and samples; MQRank- Sum is the Z-score from a Wilcoxon rank-sum test of alternative versus reference mapping qualities; and ReadPosRankSum is the Z-score from a Wilcoxon rank-sum test of alternative versus reference read-position biases) and likelihood-based Hardy–Weinberg equilibrium tests (In- breedingCoeff). The resulting receiver operating characteristic curves were filtered by the variant quality score log odds ratios calculated by VariantRecalibrator, and all SNVs below the VQSLOD threshold of 99.5% were removed. To facilitate faster joint calling between high- and low-coverage datasets, a union of all of the high-coverage SNV sites of each individual dataset was created. Article information and genotype likelihoods from initial calls using Mar- vin47 (Supplementary Table 23), with default parameters to perform the genotype refinement at sites shared by multiple individuals. This approach has been used for the 1000G2. We conducted further qual- ity control on autosomal chromosomes by removing individuals and sites with high missingness (>5%); this resulted in the removal of 17 samples and around 500,000 sites. Checks for heterozygosity or relat- edness were left for downstream analysis. Owing to the unavailability of haplotype scaffold panels, we independently phased and inferred haplotypes without reference haplotypes using both Eagle2.048 and SHAPEIT247(Supplementary Table 23). We also enabled SHAPEIT2 to produce the graph structures on which to generate the final phased haplotypes; this resulted in a pair of phased haplotypes per dataset. For each pair of haplotypes, we compared sites discordant between haplotypes generated by SHAPEIT2 (91.2%) versus Eagle2.0 (98.6%) and the VCF file before phasing. Because the estimated switch error in phasing was lower in Eagle (0.26%) than SHAPEIT2 (0.71%), we opted to use the Eagle phase panel as the default for downstream analyses. information and genotype likelihoods from initial calls using Mar- vin47 (Supplementary Table 23), with default parameters to perform the genotype refinement at sites shared by multiple individuals. This approach has been used for the 1000G2. We conducted further qual- ity control on autosomal chromosomes by removing individuals and sites with high missingness (>5%); this resulted in the removal of 17 samples and around 500,000 sites. Checks for heterozygosity or relat- edness were left for downstream analysis. Owing to the unavailability of haplotype scaffold panels, we independently phased and inferred haplotypes without reference haplotypes using both Eagle2.048 and SHAPEIT247(Supplementary Table 23). We also enabled SHAPEIT2 to produce the graph structures on which to generate the final phased haplotypes; this resulted in a pair of phased haplotypes per dataset. For each pair of haplotypes, we compared sites discordant between haplotypes generated by SHAPEIT2 (91.2%) versus Eagle2.0 (98.6%) and the VCF file before phasing. Because the estimated switch error in phasing was lower in Eagle (0.26%) than SHAPEIT2 (0.71%), we opted to use the Eagle phase panel as the default for downstream analyses. call per-sample SNVs and indels from SAHGP, H3A-Baylor and Trypano- GEN datasets in gVCF mode. Article As SNVs in the individual datasets were already ‘known’ from previous alignments, joint variant calling was done across all the datasets only at the union using Genotype gVCF in GATK to create a merged final dataset (Supplementary Methods Fig. 1). For the sex and mitochondrial chromosomes, the X chromo- some followed the same VQSR filter approach as the autosomes; Y and mitochondrial chromosomes, however, were not filtered using the VQSR model. The pseudoautosomal regions on the X chromosome were called as diploid and other X chromosome regions were called haploid in female participants. The Y chromosome pseudoautosomal regions were called diploid and the remaining regions were called haploid in male participants. The mitochondrial chromosomes were called as diploid for ease of processing. All downstream analyses were carried out on biallelic sites only, but we did investigate multiallelic sites to gain an appreciation of their relative abundances and patterns of variation. Multiallelic variation is a largely unexplored topic in genome surveys, in part because it is dif- ficult to discern between true multiallelic sites and sequencing errors. For this reason, we imposed fairly stringent measures to conservatively call such sites. We focused on multiallelic sites from the high cover- age H3A-Baylor dataset, and discarded multiallelic sites embedded in repetitive regions or regions adjacent to known copy-number vari- ation, as well as those for which the third allele was observed in fewer than six reads. Quality control before variant calling and BAM file augmentation. We used the SnpEff default parameters including the -lof argument to annotate for LOF and nonsense-mediated decay predictions. We also included annotations for variant labels from dbSNP (v.150), for clinical importance from ClinVar Database (v.20181028) and for GWAS hits from the GWAS catalogue (v.2019-10- 14) (Supplementary Table 23). In addition to functional annotations, variants were also annotated for allele frequencies in the 1000G, ExAC (r.2.0.1), gnomAD (v.2.0.2), African Genome Variation Project (AGVP)4, SAHGP and TryopanoGEN datasets. SAHGP. SAHGP is a multi-ethnic project to investigate the genomic diversity of the peoples of Southern Africa and build genomics ca- pacity in that region (https://sahgp.sanbi.ac.za/)17. The use of SAHGP samples for this study was approved by the Human Research Ethics Committee (HRECΞMedical) of the University of the Witwatersrand, Variant discovery. The choice of software for variant calling of both low- and high-depth of coverage WGS data was based on the evaluations of the AGVP4. We used HaplotypeCaller (Supplementary Table 23) to Data analyses The resulting ‘clean’ dataset was analysed by teams arranged around four main study areas: (1) population structure and admixture; (2) signatures of selection; (3) rare variation; and (4) medically relevant genes and variants. Each of these study areas are described in more detail below. Studies of selection and rare variation were limited to the 314 individuals in the H3A-Baylor dataset that had sequencing data with a high depth of coverage to improve calling confidence for rare sequence variants. The full dataset was used for the remaining analyses (population structure and admixture, and medically relevant variants). See Supplementary Table 23 for a full list of references for databases and software used. Population structure and admixture The output of merge-ibd-segments was used to compute the average pairwise IBD sharing between the different NC groups by using the previously described expression57. Admixture analysis. For the admixture analysis, in addition to the African WGS data, we merged our joint dataset with existing African genotyping array datasets (Supplementary Table 22). We carried out quality control and pruning of the merged dataset as described above. We ran ADMIXTURE (v.1.3.0)19 50 times with a random seed for each value of K from 2 to 15. We generated ADMIXTURE cross-validation error estimates to determine the optimal value of K. Admixture runs were merged and summarized using both the FullSearch and Greedy algorithm and G pairwise similarity statistic in CLUMPP (v.1.1.2)52. The results of the greedy algorithm are shown. The Genesis software was used for PCA and admixture visualization of this dataset. Admixture masking. To identify the contribution of non-Bantu- speaking ancestry in the observed population structure, we estimated non-NC local ancestry in BOT using an estimate from RFMix_v258. Phas- ing of the dataset was done using the Sanger imputation server (https:// imputation.sanger.ac.uk/) and the African Genome Resource reference panel. Data from the Juǀ’hoansi, the Gǀwi, Gǁana and baKgalagadi, the ǂKhomani and Karretjie populations14, were used as the KS source, YRI as the Bantu-speaking source and CEU as the Eurasian source (parameters used: -forward-backward -e 2). Regions with high KS ancestry (>20%, and at least 25 SNPs) in BOT were identified and masked from the full dataset, after which the PCA was regenerated. Similarly, we identified and masked regions with more than 20% east African ancestry in BRN (identified using Tubu from Chad59 as the east African source popula- tion) and repeated the PCA. As the number of SNVs included in the analy- sis had the potential to affect the principal component estimates, we thinned the whole dataset to 150,000, linkage-disequilibrium-pruned SNVs for comparisons with the masked datasets. Procrustes and FST analyses. We used FST to estimate the pairwise dis- tance between the various African populations (Supplementary Fig. 3), and then implemented PROCRUSTES53 using an in-house-generated script (see ‘Code availability’) to evaluate the correlation between geographical distances and FST-based genetic distances. Population structure and admixture Primary datasets. The primary datasets for the population structure and admixture analyses included WGS data generated from the merged H3A-Baylor–TryopanoGEN–SAHGP dataset (hereafter referred to as the joint dataset; Supplementary Table 22). Sample-level quality control. The initial dataset consisted of 564 individuals, and these were subjected to additional quality control: we excluded individuals with >1% missing data (n = 86—all from the Tryp- anoGEN dataset) and identified duplicate samples and familial-related samples using the identity-by-descent (IBD) approach in PLINK (v.1.90, http://www.cog-genomics.org/plink/1.9/)49. IBD was calculated af- ter removing SNVs in strong linkage disequilibrium by pruning SNVs with r2 > 0.15 within a window of 1,000 bp. A total of 93 individuals with PiHAT > 0.25 with at least one other sample (Mali, 57; Uganda, 2; Zambia, 19; Mossi, 3; Fon, 12) were identified; a subset of 51 unrelated individuals were randomly selected for inclusion from the pairs of related individuals (Mali, 26; Uganda, 2; Zambia, 12; Mossi, 2; Fon, 9). Multiallelic variants and haplotype phasing. After curation of in- dividual (unphased) VCF files from each dataset and the subsequent merged dataset, data for each chromosome were independently phased to provide two haplotypes per individual. We first opted to decom- pose multiallelic variants in the VCF file before phasing as follows: (1) decomposing the VCF file such that variants with multiple alleles were expanded into distinct variant records—one record for each reference/ alternative allele combination; and (2) normalizing the decomposed VCF file so that variants were represented using the most-parsimonious alleles from the human genome reference (GRCh37). Combining (1) and (2) resulted in some genotypes being split over two VCF records, such as missing/alt1, missing/alt2 or ref/alt1 and ref/alt2. To improve the accuracy of low-pass and low-coverage whole-genome data, we leveraged population linkage disequilibrium, haplotype SNV level quality control. Of the initial 41,645,936 SNVs, we removed 1,801,483 with call rate <99%; 84,758 SNVs with a significant deviation from Hardy–Weinberg equilibrium (P < 1 × 10−6); 25,352,806 SNVs with f3-statistic coupled with a corresponding high negative z-score was considered to be supporting evidence for the admixture event. f3-statistic coupled with a corresponding high negative z-score was considered to be supporting evidence for the admixture event. MAF < 0.01; and all A/T and C/G (ambiguous) SNVs, to facilitate merg- ing with additional datasets (see below). After sample- and SNV-level quality control, a total of 426 individuals and 14,406,889 SNVs were available for analyses. Admixture dating. Population structure and admixture To provide further insights into the historical con- text of the admixture events, we attempted to date some of the events based on the four regional datasets described in the previous section. We used MALDER (v.1.0) (https://github.com/joepickrell/malder)55—a modified version of ALDER22—which is able to predict the occurrence of multiple admixture events in a test population. We tested for specific admixture events between our joint population dataset and additional reference populations informed by the results from the admixture analysis. The minimum genetic distance to start curve fitting was set to 0.005 cM to account for short-range linkage disequilibrium between African populations, together with the (Rutgers v.3) recombination ge- netic map. Significant results were assessed based on the amplitude of the fitted linkage disequilibrium curves and the corresponding z-scores. Additional datasets. The cleaned joint dataset was merged with data from four additional African datasets (Supplementary Table 22). We ap- plied the same individual and SNV quality control parameters as above to each of the additional datasets, retaining SNVs that were common to all datasets in subsequent merging. To maximize the number of SNVs available for each of the subsequent analyses, we merged a variety of different populations for the various studies outlined below. Principal component analysis. For the initial principal component analysis (PCA), we merged all currently available WGS data for African populations regardless of depth of coverage (Supplementary Table 22). We performed linkage disequilibrium pruning on our merged dataset using PLINK (v.1.90)49 to remove correlated (r2 > 0.15) single nucleotide polymorphisms (SNPs) in a 1,000-SNP window, advancing by 10 SNPs at a time. The pruned dataset contained 1,013,758 SNPs and 1,253 in- dividuals with a genotype call rate of 99.9%. We used the smartPCA program from EIGENSOFT50 to perform PCA on the pruned dataset and the Genesis software51 for PCA visualization (Supplementary Table 23). IBD sharing distance. To investigate the distribution of IBD segments shared between the different NC populations, we used the program Re- fined IBD56 in Beagle 4.1 (Supplementary Table 23). The merged dataset (consisting of 590,914 SNPs and 396 individuals) used in this analysis included select populations from Trypanogen, H3A-Baylor, and two previously published studies9,14. Default refined IBD parameters were used to estimate the shared IBD segments between pairs of individuals. The IBD segments were further filtered by implementing the program merge-ibd-segments, to remove breaks and short gaps in IBD segments (>0.6 cM in length). Rare and novel variation To minimize the effect of false SNV discovery related to the low depth of coverage and biases that arise from use of different datasets, analyses of rare and novel variation were carried out using only the data with a high depth of coverage (HC-WGS) found in the H3A-Baylor dataset. Comparisons among populations were made between countries rather than regions or ethnic groups. For each variant, we recorded the ances- tral allele, its derived allele frequency both at the dataset level and for the seven populations, its predicted effect on biological function following sequence ontology terms, and the predicted effect of the mutation using SNPEff v.4.372 (Supplementary Table 23). CLR scores. CLR scores were calculated using SweepFinder (imple- mented in SWEED)27 for 10 kb non-overlapping sliding windows in each population. Genomic regions which have been previously shown to produce false-positive hits in WGS data (a custom list based on https:// sites.google.com/site/anshulkundaje/projects/blacklists and a pre- viously published study67) along with a 1-Mb flanking sequence on either side were excluded to minimize the effect of sequencing-related artefacts. To identify a threshold for identifying extreme outliers, we randomly sampled 10,000 10-kb regions from the 6 populations. On the basis of the distribution of CLR scores in this set, we identified CLR scores > 49.5 to correspond a P-value cut-off of P < 0.001 and took this as our significance threshold. Rare variant definition. We defined a rare variant as a SNV with a de- rived allele frequency of ≤0.01; in most populations this corresponded to a single variant event. Allele counts were orientated as ancestral or derived rather than reference or alternative to avoid biases that arise from the construction of the reference genomes. Owing to differences in sample size across populations, we subsampled 24 individuals for each population and then built the relative site frequency spectrum for each effect category. Integrated haplotype homozygosity scores. Integrated haplotype homozygosity scores (iHS) for SNVs with MAF > 0.05 were estimated using SelScan28 in each population. For each population, the scores were then normalized across 40 allele frequency bins. As advised in previous analyses68,69, instead of focusing on maximum iHS variants, we aimed to identify genomic regions with the highest fraction of extreme iHS-containing variants. For this, based on the background distribu- tion of the normalized iHS scores in all of the populations we identified \|iHS\| > 2.6 to correspond to P < 0.01. Article 100 SNVs with at least one SNV per 50 kb on average and a maximum of 1 heterozygous call and 5 missing calls. A window size of 100 kb was used to scan for ROHs across the genome. Following a previously published approach62, the ROH segments—depending on genomic length—were separated into three classes: short ROHs (<500 kb, class A), which most likely represent homozygosity for ancient haplotypes; intermediate ROHs (500 kb–1.5 Mb, class B), which are most likely the result of distant relatedness within a population; and long ROHs (>1.5 Mb, class C), which are suggestive of assortative mating. To provide a comparison of the ROH distribution across Africa in addition to the seven populations from our study, five African populations (YRI, LWK, ESN, MSL and GWD (see Supplementary Table 22 for definitions)) from the 1000G dataset and populations (BAG, ZUL and Ethiopian) from AGVP were included in the analysis. As the number of samples for both 1000G and AGVP datasets were around 100 per population, we randomly downsampled each population for these two datasets to 50 individuals per population. We generated an additional dataset with the modified PLINK param- eter set (--homozyg-kb 300 and --homozyg-window-het 3) for better homogenization of the combined datasets with low and high depths of coverage. FHAT1 and FHAT2 were also estimated using PLINK v.1.9 with default parameters. African population and BOT as the representative south African popu- lation. FST scores for exonic SNVs with MAF > 0.01 in the dataset were estimated between pairs of the representative populations as well as with CHB (downsampled to 50 individuals) from the 1000G dataset using VCFtools70. We then used a previously published method30 to es- timate population PBS between WGR and BOT, between BOT and CAM, and between WGR and CAM. The SNVs with highest branch lengths (P < 0.001) in a population compared to the other populations (one at a time) were considered as signals. Integration with GTEx. Analyses for the integration with the Genotype-Tissue Expression (GTEx) dataset were performed as follows. Chromosomal positions for selected loci falling in non-coding regions were intersected with significant cis-expression quantitative trait loci (cis-eQTLs) of 49 tissues in the GTEx project by downloading version 8 of the per-tissue cis-eQTL data from the GTEx portal71. Non-coding outlier regions were each annotated for the number of eQTLs contained for each tissue (range, 0–2). Article Non-coding regions with CLR scores below the fifth centile (n = 14,088), were then used as a ‘neutral’ (non-selected) background control dataset. For each tissue, a subset of regions equal to the number of non-coding outlier regions (n = 152) was randomly selected and the number of contained eQTLs tabulated. This process was repeated 1,000 times for each tissue to generate a quantitative distribution of eQTL overlaps. Then, for this initial iteration, a t-score was calculated for each tissue. To rank the tissues, we then repeated the initial iteration 1,000 times to generate a distribution of t-scores for each tissue. The same process was used to generate non-coding outlier region–eQTL distributions for each tissue in each population. Population structure and admixture We used the function distVincentyEllipsoid in the R package geosphere (v.1.5-7) to estimate great circle distances between the geographical midpoints of those countries for which individuals were recruited from across the country, and regional or city midpoints for recruitments that were limited to a specific geographical location. FST was estimated using smartPCA for the joint dataset in EIGENSOFT50, principal coordinate analysis was performed on the geographical and genetic distance ma- trices, and then the test statistic was constructed as described pre- viously53. Finally, we established the null distribution by randomly permuting the labels on the FST matrix. For the test of all populations, we generated 1 x 107 permutations; for the test of only NC speakers, we generated 1 × 109 permutations. Permuted distributions were used to derive empirical P values. Admixture model testing. We next tested the admixture model. We used qpGraph20 to test various alternative models of gene flow to iden- tify the best possible central African NC-speaker population for ad- mixture in southern and east African populations (see Supplementary Note 1.3 for details.) Mitochondrial and Y-chromosome haplogroups. We then analysed the mitochondrial and Y-chromosome haplogroups. Haplogrep260 was used to identify mitochondrial haplotypes for each individual. Y-chromosome haplogroup analysis was done using the AMY-tree al- gorithm and tool61. For each sample, the variants detected from the WGS VCF files were extracted and converted into the correct format before input into the AMY-tree program. Testing for the presence of admixture. We next tested for the pres- ence of admixture. To further explore and corroborate the admixture events observed in the previous approaches we used a formal test for admixture using the f3-statistic. As we were interested in exploring admixture events on a regional scale, and to maintain the highest reso- lution in terms of the number of SNPs, we generated four datasets for this analysis. The datasets represented population groups from west, central-west, east and south Africa; f3-statistics were generated for all possible combinations of populations using both TreeMix threepop (v.1.13)54 and Admixtools qp3pop (v.1.0)20. In both cases, a negative Runs of homozygosity. We also investigated runs of homozygosity (autozygosity). For the identification of runs of homozygosity (ROHs), PLINK v.1.949 was used with the following parameters: a minimum of Signatures of selection Datasets. Our dataset included samples with different sequencing depths, which can adversely affect nucleotide diversity and allele frequency estimates63; therefore, the identification of signatures of selection was limited to samples from the H3A-Baylor dataset with a high depth of coverage. Samples that appeared as outliers for each population in the PCA and outliers in the full dataset were removed. Similarly, related individuals identified using the method described above were also excluded. Owing to a smaller sample size and high within-group diversity, the MAL group was excluded from this analysis. Genes in selected regions were identified using the Ensembl database64 and assessed for (predicted) functional impact using the Ensembl, OMIM65 and GeneCards databases66 (Supplementary Table 23). Rare and novel variation This class of variants includes those in the penultimate and last exons of genes (predicted to escape nonsense-mediated decay), but are most similar to vari- ants labelled as putative LOF variants in the literature. These vari- ants were then mapped to their respective genes using BioMart80. Downstream analysis was performed using online databases of dis- ease–gene associations, including DisGenet81, eDGAR82, OMIM65 and CTDbase83. Highly differentiated variants. Highly differentiated variants (HDVs) were identified by first calculating the derived allele frequency for variants for which the ancestral state was known and then assessing pairwise differences in allele frequency between populations. Al- though a 25% frequency difference threshold has been used to identify HDVs across different continents1, we opted for a more-conservative threshold of a 40% difference in the derived allele frequency, con- sistent with the modest population sample sizes. We visualized the distribution of HDVs by grouping the difference in derived allele fre- quencies between any two populations into bins representing 0.1 frequency and plotting the number of variants that fell in each bin. We also created a table of the total number of HDVs between each pairwise population comparison and the number of HDVs once we removed variants linked to HBB on chromosome 11 and HLA on chro- mosome 6 (two regions with a large number of HDVs related to known selection and strong linkage disequilibrium). To infer the biological relevance of HDVs, the GWAS Catalog75 available on 14 October 2019 was downloaded (https://www.ebi.ac.uk/gwas/api/search/down- loads/alternative) under hg37 and formatted using Bedtools76. We then intersected HDV positions with SNP positions of genome-wide significant (P < 5 × 10−8), replicated GWAS hits within 50 kb of the HDV. The results were visualized using Circa (http://omgenomics.com/ circa/) (Supplementary Table 23). Overlapping GWAS hits and their corresponding experimental trait factors were then further analysed for biological relevance (see below). For the putative LOF burden analysis, lists of genes that are associated with or that influence influenza, HIV, malaria and hepatitis C infections were extracted using GeneCards66 (Supplementary Table 18). These lists are largely populated by genes involved in the host transcriptional response to infection. For each population group, we first determined the number of putative LOF variants in human genes directly implicated in the infection as defined by genecards (direct genes). Medically relevant variants The full dataset (both low- and high-coverage WGS data) was used to catalogue medically relevant variants. Medically actionable variants. SNVs were first annotated against the list of genes included in the ACMG recommendations for report- ing of incidental findings in clinical exome- and genome-sequencing data34,84. These recommendations are based on the recognition that in whole-exome sequencing or WGS, incidental or secondary findings that are unrelated to the indication for ordering the sequencing but of potential relevance to patient care may be found. We estimated the burden of ACMG gene panel version 2.0 variants, including the total number of alleles observed, median and range of number of alleles per sample, number (%) of samples that carried at least one allele, and genes in which variants were identified. GWAS Catalog experimental trait factor analysis. We next performed an experimental trait factor analysis using the GWAS Catalog. Genetic variants influence changes in phenotype and physiology in different population groups. As these groups often share similar environment conditions, we expect changes that are essential to adaptation to be conserved, even though this may be constrained by genetic capacity. At a molecular level, this could lead to the divergence (variation) or conver- gence (conservation) of protein function or structure between groups. Experimental Factor Ontology (EFO) annotations were retrieved from the European Bioinformatics Institute (EBI) website (https://www.ebi. ac.uk/ols/ontologies/efo) and those mapping to GWAS SNVs within 50 kb of HDVs were extracted from https://www.ebi.ac.uk/gwas/docs/ file-downloads. Clinical annotation of variants. SNVs were also annotated for clini- cal importance using the NCBI ClinVar database (http://www.ncbi. nlm.nih.gov/clinvar/; 20181028 version)85. ClinVar provides a freely available report archive of relationships between medically relevant variants and phenotypes. The terms for clinical significance reported by ClinVar are those recommended by the ACMG. These range from ‘0-Uncertain significance’ to ‘5-Pathogenic’ and include codes such as ‘255-Other’. In the present study, variants of interest were those clas- sified by ClinVar as ‘Pathogenic’, which corresponds to ‘5-Pathogenic’ in the ASN.1 set of terms. Frequencies were estimated for the whole dataset and by country. Similar to the analysis of ACMG variants, the burden of these variants was also computed, including the total num- ber of alleles observed, median (range) number of alleles per sample, number (%) of samples that carried at least one allele and genes in which variants were identified. Rare and novel variation For each 10-kb window that was scanned for the CLR analysis in a population, we measured the percent- age of SNPs with outlier iHS scores (\|iHS\| > 2.6). The top 1% of windows with the highest percentage of outlier iHS score were considered to be outlier windows for each population. Novel variants. By definition, novel variants have not been previously discovered and are generally rare and often private; however, given the dearth of data from Africa, such variants may well be common in a population or set of populations. For the purposes of this analysis, we focused on the high-coverage H3A-Baylor dataset and defined a variant as ‘novel’ if the variant was not present in the dbSNP v.15073, ExAC5 or gnomAD v.2.0.274 databases, or the TryopanoGEN18 or SAHGP17 datasets or if the variant was not identified in the AGVP4 dataset. We also excluded SNVs that were fixed in all populations but for which the derived allele was inferred to be the reference allele, as possible technical artefacts (incorrect ancestral status). To visualize the contribution of novel variant discovery for each population, we plotted the number of unique novel variants identi- fied per individual within each population, as well as the number of population-specific novel variants for each population. Novel variant PBS analysis. For the PBS analysis, we used WGR as the representa- tive west African population, CAM as the representative central-west discovery was also represented as a cumulative function, in which we sequentially plotted the number of novel variants that were discovered each time a new population was included. To discriminate between rare and common variant discovery, we also plotted the cumulative number of novel variants discovered using common novel variants only (that is, all novel variants except the ones that were observed only once). Correlation between novel variant discovery and ancestry was calculated for the KS ancestry in BOT, RFF ancestry in CAM, east African ancestry in Berom and non-NC ancestry in MAL. Individual-based ancestry proportions were obtained from ADMIXTURE K = 6 results and Pearson’s correlations were determined in R. Putative LOF variation. For the analysis of putative LOF variants, we used a local pipeline—ALOFT79—to provide annotations for puta- tive protein-damaging variants in protein-coding genes, including stop-loss, stop-gain and canonical splice sites. Rare and novel variation The majority (75.3%) of putative LOF variants in influenza-associated direct genes were apparent population-singleton variants, with only around 13% being shared between populations (Supplementary Table 18). After extensive benchmarking (Supplementary Methods Figs. 2, 3), we chose to normalize the putative LOF burden in direct genes by dividing by the number of putative LOF variants in ‘indirect’ genes associated with the same infection. This was done to account for potential differences in mutation rate and any uncharacterized gene biases between popula- tions. The putative LOF ratio for each population was then plotted against the mortality rates reported by WHO in 201633 for the infec- tion being surveyed. The correlation between the two values was then calculated, and its significance measured against 1,000 iterations of calculated ratios using a set of random genes similar in data size to the direct gene list for that disease. Article into dbSNP. The H3Africa Initiative is committed to providing research data generated by the H3Africa research projects to the entire research community. H3Africa research seeks to promote fair collaboration between scientists in Africa and those from elsewhere. The H3Africa Consortium Data Sharing, Access and Release Policy outlines a policy framework that places a firm focus on African leadership and capac- ity building as guiding principles for African genomics research. The policy and related documents are available here: https://h3africa.org/ index.php/consortium/consortium-documents/. Variants of clinical importance to African populations. We also de- fined the burden of variants of particular clinical importance to African populations. Population burden and inter-population differentiation were determined for genetic variants related to (1) sickle cell anae- mia (HBB); (2) trypanosomiasis and end-stage renal disease (APOL1); (3) glucose-6-phosphate dehydrogenase deficiency (G6PD); and (4) response to antiretroviral therapy with abacavir (HLA-B5701). Sickle cell anaemia is an autosomal recessive disorder determined principally by a missense mutation in the HBB gene (rs334; HbS). The disorder is most common in Africa, in particular west and central Africa, and remains one of the most-important monogenic disorders of clinical and public health relevance on the continent. Two cohorts from the H3A-Baylor dataset included individuals with homozygous (HbSS) sickle cell disease (CAM and FNB), and these were excluded from frequency estimates of the HbS allele. Code availability Code for the implementation of PROCRUSTES is available at https:// github.com/dshriner/Procrustes, licensed under the GNU General Public License v.3.0. Variants in APOL1 are associated with various forms of kidney dis- ease, primarily in individuals of African ancestry and is protective against severe disease caused by infection with Trypanosoma brucei. The two major risk variants (haplotypes) are labelled G1 (defined by rs73885319, often in complete linkage disequilibrium with rs60910145 (T > G)) and G2 (a 6-bp indel, rs143830837). The risk genotype is the recessive state for the diplotype that includes two risk variants: that is, G1/G1, G1/G2, G2/G2. 44. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25, 1754–1760 (2009). 44. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25, 1754–1760 (2009). 45. McKenna, A. et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 20, 1297–1303 (2010). 46. Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009). 47. O’Connell, J. et al. A general approach for haplotype phasing across the full spectrum of relatedness. PLoS Genet. 10, e1004234 (2014). 48. Loh, P. R., Palamara, P. F. & Price, A. L. Fast and accurate long-range phasing in a UK Biobank cohort. Nat. Genet. 48, 811–816 (2016). 48. Loh, P. R., Palamara, P. F. & Price, A. L. Fast and accurate long-range phasing in a UK Biobank cohort. Nat. Genet. 48, 811–816 (2016). 49. Chang, C. C. et al. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience 4, 7 (2015). 49. Chang, C. C. et al. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience 4, 7 (2015). G6PD deficiency is an X-linked red cell enzymopathy that increases the risk of haemolysis in affected individuals. It is an important risk fac- tor for neonatal jaundice and drug-related haemolysis. Male individuals have the disorder if they carry a G6PD mutation on their X chromosome (that is, hemizygous), whereas most affected female individuals carry mutations on both X chromosomes (homozygous). Although there are over 180 known mutations in the gene, including several deleterious mutations, most of these mutations are rare or have a low frequency. Reporting summary 61. Van Geystelen, A., Decorte, R. & Larmuseau, M. H. D. AMY-tree: an algorithm to use whole genome SNP calling for Y chromosomal phylogenetic applications. BMC Genomics 14, 101 (2013). g genome SNP calling for Y chromosomal phylogenetic applications. BMC Genomics 14, 101 (2013). Further information on research design is available in the Nature Research Reporting Summary linked to this paper. 62. Pemberton, T. J. et al. Genomic patterns of homozygosity in worldwide human populations. Am. J. Hum. Genet. 91, 275–292 (2012). 63. Fumagalli, M. Assessing the effect of sequencing depth and sample size in population genetics inferences. PLoS ONE 8, e79667 (2013). 63. Fumagalli, M. Assessing the effect of sequencing de genetics inferences. PLoS ONE 8, e79667 (2013). Medically relevant variants To elucidate potential divergent EFO annotations associated with the identified HDVs, we computed Kappa Statistic similarity scores between EFO annotations based on the population in which the associated SNV was found77. These similarity scores were computed on a collapsed EFO at level 4 with the root of the ontology assumed to be located at the level 0, and using only ‘process’, ‘material property’ and ‘material entity’ upper level concepts. Although a similarity score threshold of 0.3 or 0.4 has been often used to identify convergent (>0.3) or divergent (<0.3) annotations, we opted for a stricter threshold of 0.2 based on the similarity score dataset of all EFO annotation pairs. Finally, we mapped different divergent EFO annotations to their associated proteins to identify enriched biological processes and molecular functions that reflect genomic variations among different population groups using the ClueGO software78. Code availability The current analysis focused on the common variants definitively associated with G6PD deficiency in sub-Saharan Africa, namely, the 202A/376G G6PD A allele (that is, the G6PD A-deficiency states associ- ated with either rs1050828 (c.202G>A) or rs1050829 (c.376A>G). We also examined a set of 11 G6PD variants recently shown to be associated with protection from severe malaria34. 50. Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis. PLoS Genet. 2, e190 (2006). 50. Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis. PLoS Genet. 2, e190 (2006). 51. Buchmann, R. & Hazelhurst, S. Genesis PCA and Admixture Plot Viewer. Version 0.2.6 http://www.bioinf.wits.ac.za/software/genesis (2014). 52. Jakobsson, M. & Rosenberg, N. A. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23, 1801–1806 (2007). 52. Jakobsson, M. & Rosenberg, N. A. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23, 1801–1806 (2007). 53. Wang, C. et al. Comparing spatial maps of human population-genetic variatio Procrustes analysis. Stat. Appl. Genet. Mol. Biol. 9, 13 (2010). 54. Pickrell, J. K. & Pritchard, J. K. Inference of population splits and mixtures f genome-wide allele frequency data. PLoS Genet. 8, e1002967 (2012). 55. Pickrell, J. K. et al. Ancient west Eurasian ancestry in southern and eastern Africa. Proc. Natl Acad. Sci. USA 111, 2632–2637 (2014). 56. Browning, B. L. & Browning, S. R. Improving the accuracy and efficiency of identity-by-descent detection in population data. Genetics 194, 459–471 (2013). 57. Atzmon, G. et al. Abraham’s children in the genome era: major Jewish diaspora populations comprise distinct genetic clusters with shared Middle Eastern ancestry. Am. J. Hum. Genet. 86, 850–859 (2010). The rs2395029 SNP in HLA-B*5701 mediates an adverse allergic response to HIV therapy with abacavir and is found in up to 5% of indi- viduals of European ancestry. 58. Maples, B. K., Gravel, S., Kenny, E. E. & Bustamante, C. D. RFMix: a discriminative modeling approach for rapid and robust local-ancestry inference. Am. J. Hum. Genet. 93, 278–288 (2013). For each of the vignette variants, overall African frequencies and frequencies by country or ethnolinguistic grouping were estimated. 59. Haber, M. et al. Chad genetic diversity reveals an African history marked by multiple 59. Haber, M. et al. Code availability Chad genetic diversity reveals an African history marked by multiple Holocene Eurasian migrations. Am. J. Hum. Genet. 99, 1316–1324 (2016). ber, M. et al. Chad genetic diversity reveals an African history marked by m 60. Weissensteiner, H. et al. HaploGrep 2: mitochondrial haplogroup classification in the era of high-throughput sequencing. Nucleic Acids Res. 44, W58–W63 (2016). of high-throughput sequencing. Nucleic Acids Res. 44, W58–W63 (20 44. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25, 1754–1760 (2009). Data availability ‘TrypanoGEN: an integrated approach to the identification of genetic determinants of susceptibility to trypanosomiasis’, was funded by the Wellcome Trust (099310/Z/12/Z). L.R.B. was supported by the CERCA Programme/Generalitat de Catalunya and by the Spanish Ministry of Economy and Competitiveness, through the ‘Severo Ochoa Programme for Centres of Excellence in R&D’ 2016–2019 (SEV-2015-0533). N.M. (principal investigator), S.A., G.B., G.W., J.K., Y.J.F., T.O., O.F., E.A., S.H., G. Mazandu, M. Mbiyvanga, A.B., S.K.K., E.R.C. and A. Moussa are funded by the NIH H3ABioNet grant under award number U24HG006941. The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the African Academy of Sciences, the National Institutes of Health or the Wellcome Trust. 74. Karczewski, K. J. et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581, 434–443 (2020). 74. Karczewski, K. J. et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581, 434–443 (2020). 75. MacArthur, J. et al. The new NHGRI-EBI Catalog of published genome-wide association studies (GWAS Catalog). Nucleic Acids Res. 45, D896–D901 (2017). 76. Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010). features. Bioinformatics 26, 841–842 (2010). 77. Mazandu, G. K., Chimusa, E. R., Mbiyavanga, M. & Mulder, N. J. A-DaGO-Fun: an adaptable Gene Ontology semantic similarity-based functional analysis tool. Bioinformatics 32, 477–479 (2016). 78. Bindea, G. et al. ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks. Bioinformatics 25, 1091–1093 (2009). 79. Balasubramanian, S. et al. Using ALoFT to determine the impact of putative loss-of-function variants in protein-coding genes. Nat. Commun. 8, 382 (2017). 80. Smedley, D. et al. The BioMart community portal: an innovative alternative to large, centralized data repositories. Nucleic Acids Res. 43, W589–W598 (2015). Author contributions Study design: M.R., C.R., Z.L. and A.A.A. Manuscript oversight and editing: N.A.H. (team leader), Z.L., S.K.K. and S.N.A. Data processing and quality control: N.M. (team leader), G.B., G. Mazandu, M. Mbiyavanga and E.R.C. Population genetics: S.A. (team leader), A.C., D. Sengupta, D. Shriner, S.H., E.A., T.O., E.D. and O.F. Signatures of selection: A.C. (team leader), D. Sengupta and T.B. Novel and rare variation: L.R.B. (team leader), N.A.H., T.B., O.A.N., G.B. and M. Mbiyavanga. Medically relevant variation: A.A.A. (team leader), Y.J.F., A.B., N.M., J.K., G.W., N.A.H., A.W.G. and T.B. Data generation: R.A.G., G. Data availability Metcalf and D.M. Data providers: The African Collaborative Center for Microbiome and Genomics Research (ACCME): C.A. (principal investigator), S.N.A. and A.A.A; TrypanoGEN: E.M. (principal investigator), D.M.-N., M.K., G.S., B.B., M.S., C.H.-F., H.N. and A. Macleod. AWI-Gen: M.R. (principal investigator), H.S., R.P.B., G. Agongo and A. Oduro. H3Africa Kidney Disease Research Network: A. Ojo, D.A. (principal investigators), B.O.S. and D.B. Awadalla lab: P.A. (principal investigator), E.G. and V.B. Exploring Perspectives on Genomics and Sickle Cell Public Health Interventions: A.W. (principal investigator). Clinical and Genetic Studies of Hereditary Neurological Disorders in Mali: G.L. (principal investigator), L.C., S.D. and O.S. Collaborative African Genomics Network (CAfGEN): G. Anabwani., M. Matshaba (principal investigators), S.W.M., A.K., M.J., G. Mardon (co-principal investigators), B.M., G.R., N.A.H., L.W., S.M. and S.K. H3BioNet: A. Moussa and A.B. 81. Piñero, J. et al. DisGeNET: a comprehensive platform integrating information on human disease-associated genes and variants. Nucleic Acids Res. 45, D833–D839 (2017). 82. Babbi, G. et al. eDGAR: a database of disease–gene associations with annotated relationships among genes. BMC Genomics 18, 554 (2017). relationships among genes. BMC Genomics 18, 554 (2017). 83. Davis, A. P. et al. The Comparative Toxicogenomics Database: update 2019. Nucleic Acids Res. 47, D948–D954 (2019). 83. Davis, A. P. et al. The Comparative Toxicogenomics Database: update 2019. Nucleic Acids Res. 47, D948–D954 (2019). 84. ACMG Board of Directors. ACMG policy statement: updated recommendations regarding analysis and reporting of secondary findings in clinical genome-scale sequencing. Genet. Med. 17, 68–69 (2015). 84. ACMG Board of Directors. ACMG policy statement: updated recommendations regarding analysis and reporting of secondary findings in clinical genome-scale sequencing. Genet. Med. 17, 68–69 (2015). 85. Landrum, M. J. et al. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 46, D1062–D1067 (2018). 85. Landrum, M. J. et al. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 46, D1062–D1067 (2018). Acknowledgements We thank the members of the wider H3Africa Consortium (www.h3africa. org) for their support and input, particularly J. Troyer and A. Duncanson; S. Tishkoff, J. Lupski, J. Belmont and C. Tyler-Smith for comments and feedback on the manuscript; K. Garson, A. Gillum and K. Schulze for their help with figure visualizations and for giving permission for the use of these figures; M. Cherif Rahimy for their assistance with recruitment in Benin and L. Sergeevna Mainzer, G. Rendon and V. Data availability 64. Zerbino, D. R. et al. Ensembl 2018. Nucleic Acids Res. 46, D754–D761 ( WGS data used in this paper are available through the European Genome-phenome Archive (EGA) under study accession number: EGAS00001002976. The data include genomic (BAMs and VCFs) and minimal phenotypic data from appropriately consented individuals. In compliance with current international standards to protect par- ticipant confidentiality, the H3Africa-generated data are available to bona fide researchers within the wider scientific community through a controlled access process. Some of the DNA samples are archived in H3Africa biorepositories as part of the H3Africa Consortium agree- ment. To gain access to data in the EGA or biospecimens in the biore- positories, requests must be submitted to dbac@h3africa.org, or requested through the H3Africa Data and Biospecimen Catalogue (https://catalogue.h3africa.org). Requests are subject to approval by an independent H3Africa Data and Biospecimen Access Committee (DBAC). Novel SNVs identified and reported here will be deposited 65. Amberger, J. S., Bocchini, C. A., Scott, A. F. & Hamosh, A. OMIM.org: leveraging knowledge across phenotype-gene relationships. Nucleic Acids Res. 47, D1038–D1043 (2019). 66. Stelzer, G. et al. The GeneCards suite: from gene data mining to disease genome sequence analyses. Curr. Protoc. Bioinformatics 54, 1.30.31–1.30.33 (2016). sequence analyses. Curr. Protoc. Bioinformatics 54, 1.30.31–1.30.33 (2016). 67. Pybus, M. et al. 1000 Genomes Selection Browser 1.0: a genome browser dedicated to signatures of natural selection in modern humans. Nucleic Acids Res. 42, D903–D909 (2014). 68. Sabeti, P. C. et al. Genome-wide detection and characterization of positive selection in human populations. Nature 449, 913–918 (2007). 69. Pickrell, J. K. et al. Signals of recent positive selection in a worldwide sample of human populations. Genome Res. 19, 826–837 (2009). 70. Danecek, P. et al. The variant call format and VCFtools. Bioinformatics 27, 2156–2158 (2011). 71. GTEx Consortium. The Genotype-Tissue Expression (GTEx) project. Nat. Genet. 45, 580–585 (2013). 72. Cingolani, P. et al. Using Drosophila melanogaster as a model for genotoxic chemical mutational studies with a new program, SnpSift. Front. Genet. 3, 35 (2012). 73. Sherry, S. T. et al. dbSNP: the NCBI database of genetic variation. Nucleic Acids Res. 29, 308–311 (2001). Studies of Hereditary Neurological Disorders in Mali study was funded by the NHGRI/NIH grant U01HG007044. The Collaborative African Genomics Network (CAfGEN) is funded by the National Institute of Allergy and Infectious Diseases (NIAID) of NIH and the NHGRI of the NIH (U54AI110398). Data availability Jongeneel from the HPCBio team at the University of Illinois Urbana-Champaign for the initial processing and variant calling of the high depth H3A-Baylor dataset using the Blue Waters supercomputing centre. WGS in H3Africa cohorts was supported by a grant from the National Human Genome Research Institute, National Institutes of Health (NIH/NHGRI) U54HG003273. The African Collaborative Center for Microbiome and Genomics Research (ACCME) is funded by NIH/ NHGRI grant U54HG006947. The AWI-Gen Collaborative Centre is funded by NIH grant U54HG006938. The Exploring Perspectives on Genomics and Sickle Cell Public Health Interventions was funded by NHGRI/NIH grant U01HG007459. The Clinical and Genetic Competing interests The authors declare no competing interests. Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41586-020- 2859-7. Supplementary information is available for this paper at https://doi.org/10.1038/s41586-020- 2859-7. Correspondence and requests for materials should be addressed to A.A.A., Z.L. or N.A.H. Peer review information Nature thanks Laura Gauthier, Joanna Mountain and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Reprints and permissions information is available at http://www.nature.com/reprints. Reprints and permissions information is available at http://www.nature.com/reprints. Article Extended Data Fig. 1 \| ADMIXTURE clustering analysis of H3A-WGS samples. Existing African datasets from AGVP4, 1000 Genomes project2, SAHGP17 and previously published studies9,14 and a representative European population (CEU) from the 1000 Genomes Project are included panels. K values from 2 to 10 are shown. See Supplementary Tab definitions of abbreviations. Article Extended Data Fig. 1 \| ADMIXTURE clustering analysis of H3A-WGS samples. Existing African datasets from AGVP4, 1000 Genomes project2, SAHGP17 and previously published studies9,14 and a representative European population (CEU) from the 1000 Genomes Project are included as reference panels. K values from 2 to 10 are shown. See Supplementary Table 22 for definitions of abbreviations. Extended Data Fig. 2 \| Characteristics of known and regional selected loci. a, CLR score distributions in known selected genes (significant population-specific outlier scores (that is, with P < 0.01) for the window overlapping the gene are indicated by an asterisk). b, Summary of PBS comparisons. Genes with longer branch lengths in WGR compared to BOT and CAM are circled in blue; longer branch lengths in BOT and CAM to the other two populations are encircled in brown and dark g respectively. c, Overlap between the proportion of KS ancestr score across chromosome 6 in BOT. Extended Data Fig. 2 \| Characteristics of known and regional selected loci. a, CLR score distributions in known selected genes (significant population-specific outlier scores (that is, with P < 0.01) for the window overlapping the gene are indicated by an asterisk). b, Summary of PBS comparisons. Genes with longer branch lengths in WGR compared to BOT and CAM are circled in blue; longer branch lengths in BOT and CAM in comparison to the other two populations are encircled in brown and dark green, respectively. c, Overlap between the proportion of KS ancestry (%) and CLR score across chromosome 6 in BOT. Article ded Data Fig. 3 \| See next page for caption. Extended Data Fig. west African populations). The blue dotted line represents the mean correlation for the same correlations generated using 1,000 permutations of random genes; the s.e.m. for all populations is shown in grey. e, Correlation statistics (adjusted R2) for the putative LOF ratio for genes related to hepatitis C (HCV, n = 190 direct genes, n = 1837 indirect genes), HIV(n = 724 direct genes, n = 1351 indirect genes), influenza in west African countries (CAM, MAL, FNB and BRN), and malaria (n = 484 direct genes, n = 1554 indirect genes) are shown as red dots against the box plot distributions of correlation statistics (adjusted R2) generated using 1,000 permutations of random genes (Supplementary Table 18). Box plots show the median value (centre line), whiskers indicate the limits of the highest (fourth) and lowest (first) quartiles of the data; distribution outliers are shown as dots. Additional information 3 \| See next page for caption. west African populations). The blue dotted line represents the mean correlation for the same correlations generated using 1,000 permutations of random genes; the s.e.m. for all populations is shown in grey. e, Correlation statistics (adjusted R2) for the putative LOF ratio for genes related to hepatitis C (HCV, n = 190 direct genes, n = 1837 indirect genes), HIV(n = 724 direct genes, n = 1351 indirect genes), influenza in west African countries (CAM, MAL, FNB and BRN), and malaria (n = 484 direct genes, n = 1554 indirect genes) are shown as red dots against the box plot distributions of correlation statistics (adjusted R2) generated using 1,000 permutations of random genes (Supplementary Table 18). Box plots show the median value (centre line), whiskers indicate the limits of the highest (fourth) and lowest (first) quartiles of the data; distribution outliers are shown as dots. Extended Data Fig. 3 \| Highly divergent and putative LOF variants. a, EFO traits from the GWAS catalogue reflected by highly divergent SNVs within 50 kb of GWAS hits. From left to right, ribbons illustrate the relative representation of variants across pairwise population comparisons, GWAS ancestry, EFO top label, EFO trait or disease label, and disease or traits mapped to the EFO label. b, Distribution and sharing of common (MAF > 5%) putative LOF variants between two or more populations (coloured bars) and between all populations surveyed (red bars). c, Specific disease classes to which 5% or more genes with putative LOF variants shared between all populations were mapped. d, Correlation (Pearson) between WHO mortality rates for influenza and ratio of putative LOF variants in direct (n = 181) compared with indirect (n = 1842) influenza-associated genes (red solid line, all populations; red dotted line, Article cle Data Fig. 4 \| Distribution of G6PD variants and ClinVar ic variants across H3Africa populations. a, Frequency distribution enic and likely pathogenic variants (n = 287) in H3Africa HC-WGS ns. Disease genes with variants that had an allele frequency > 5% ltiple populations (shown in Fig. 4c) are highlighted. Box plots show nvalue (centre line),whiskersindicatethelimitsofthehighest (fourth) and lowest (first) quartiles of the data; distributi as dots. b, Relative frequencies of 11 G6PD deficiency-ass each population separated by sex. G6PD A− 202A and 376 A-deficiency associated with either rs1050828 (c.202G>A (c.376A>G) (MIM 305900). Extended Data Fig. Corresponding author(s): Adebowale Adeyemo, Zane Lombard, Neil A.Hanchard Last updated by author(s): 2020 - 04 - 03 Reporting Summary Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist. Nature Research wishes to improve the reproducibility of the work that we publish. This form provides struct in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Po Additional information 4 \| Distribution of G6PD variants and ClinVar pathogenic variants across H3Africa populations. a, Frequency distribution of pathogenic and likely pathogenic variants (n = 287) in H3Africa HC-WGS populations. Disease genes with variants that had an allele frequency > 5% across multiple populations (shown in Fig. 4c) are highlighted. Box plots show the median value (centre line), whiskers indicate the limits of the highest (fourth) and lowest (first) quartiles of the data; distribution outliers are shown as dots. b, Relative frequencies of 11 G6PD deficiency-associated alleles within each population separated by sex. G6PD A− 202A and 376G refer to the A-deficiency associated with either rs1050828 (c.202G>A) or rs1050829 (c.376A>G) (MIM 305900). (fourth) and lowest (first) quartiles of the data; distribution outliers are shown as dots. b, Relative frequencies of 11 G6PD deficiency-associated alleles within each population separated by sex. G6PD A− 202A and 376G refer to the A-deficiency associated with either rs1050828 (c.202G>A) or rs1050829 (c.376A>G) (MIM 305900). Extended Data Fig. 4 \| Distribution of G6PD variants and ClinVar pathogenic variants across H3Africa populations. a, Frequency distribution of pathogenic and likely pathogenic variants (n = 287) in H3Africa HC-WGS populations. Disease genes with variants that had an allele frequency > 5% across multiple populations (shown in Fig. 4c) are highlighted. Box plots show the median value (centre line), whiskers indicate the limits of the highest Extended Data Fig. 4 \| Distribution of G6PD variants and ClinVar Extended Data Fig. 4 \| Distribution of G6PD variants and ClinVar pathogenic variants across H3Africa populations. a, Frequency distribution of pathogenic and likely pathogenic variants (n = 287) in H3Africa HC-WGS populations. Disease genes with variants that had an allele frequency > 5% across multiple populations (shown in Fig. 4c) are highlighted. Box plots show the median value (centre line), whiskers indicate the limits of the highest nature research \| reporting summary Statistics For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section. A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one- or two-sided A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one- or two-sided A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted Give P values as exact values whenever suitable. For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted Give P values as exact values whenever suitable. For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes Policy information about availability of data Policy information about availability of data All manuscripts must include a data availability statement. This - Accession codes, unique identifiers, or web links for publicly available datasets cession codes, unique identifiers, or web links for publicly avai - A list of figures that have associated raw data - A description of any restrictions on data availability - A description of any restrictions on data availability WGS data from the H3Africa projects used in this study have been deposited in the Europ WGS data from the H3Africa projects used in this study have been deposited in the European Genome-phenome Archive ( ojects used in this study have been deposited in the European Genome-phenome Archive (EGA) with accession numbers: H3AfricaChipDesign: Study EGAS00001002976 TrypanoGEN1: Data Set EGAD00001004393 TrypanoGEN2: Data Set EGAD00001004220 MALSIC: Data Set EGAD00001004557 AWI-GEN: Data Set EGAD00001004448 CAfGEN: Data Set EGAD00001004533 ELSI: Data Set EGAD00001004316 NEEDI: Data Set EGAD00001004334 ACCME: Data Set EGAD00001004505 H3AfricaChipDesign: Study EGAS00001002976 TrypanoGEN1: Data Set EGAD00001004393 TrypanoGEN2: Data Set EGAD00001004220 MALSIC: Data Set EGAD00001004557 AWI-GEN: Data Set EGAD00001004448 CAfGEN: Data Set EGAD00001004533 ELSI: Data Set EGAD00001004316 NEEDI: Data Set EGAD00001004334 ACCME: Data Set EGAD00001004505 DNA samples are archived in H3Africa biorepositories as part of the H3Africa Consortium agreement. The Data and biospecimens are available on request through the Data and Biospecimen Access Committee of the H3Africa consortium. Novel SNVs identified and reported here will be deposited into the database of Single Nucleotide Polymorphisms (dbSNP). Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection. Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sec he best fit for your research. If you are not sure, read the appropriate sections before making your selection. Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf Life sciences Behavioural & social sciences Ecological, evolutionary & environmen F f f th d t ith ll ti t /d t / ti fl t df For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf Software and code gnomAD 2.0.2 exomeAD 2.0.3 BioMart 2.34.2 eDGAR 10-11-2016 HGVD 2.3 GWAS Catalogue v.2019-10-14 VEP 90 GTEx 7 MarViN r1 https://github.com/Illumina/MarViN EIGENSOFT v.7.2.1 MALDER v.1.0 Beagle 4.1. RFMix_v2 SNPEff 4.3-3 ExAC r2.01 Haplogrep2 2.1.1. AMY-tree 2.0 ClueGO v.2.5.1 ALOFT 1.0 DisGenet 6.0 eDGAR database release 10/11/2016 CTDbase update 2019 SWEED 3.3.1 dbPHSP https://jjwanglab.org/dbphsp t l ith ft th t t l t th h b t t t d ib d i bli h d lit t ft t b d il bl t dit / i nature research \| reporting summary \| reporting summary For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information. Reporting for specific materials, systems and methods We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whethe system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selectin Materials & experimental systems n/a Involved in the study Antibodies Eukaryotic cell lines Palaeontology Animals and other organisms Human research participants Clinical data Methods n/a Involved in the study ChIP-seq Flow cytometry MRI-based neuroimaging n/a Invo Life sciences study design All studies must disclose on these points even when the disclosure is negative. No statistical method was used to determine sample size. Sample size was maximised to include as wide a diversity of African populations and ethnolinguistic groups as possible. For specific analysis methods, subsets of the data (e.g. high depth of coverage data) were used to ensure robust conclusions. Sample size Data exclusions Data exclusions Exclusion criteria were pre-established on the basis of relatedness and sequencing quality. Samples were excluded based on sequencing- quality control metrics generated prior to any downstream analysis; this included poor mapping of reads to the reference genome, missingness thresholds, and familial relatedness. Single nucleotide variants were excluded based on excessive missingness, deviations from nature research \| reporting summary Hard-Weinberg equilibrium, call ambiguity and low minor allele frequencies. In addition, where a genetically homogeneous set of individuals was required for an analysis, possible outliers were identified and removed. Replication Results were not externally replicated. Randomization The experiments were not randomized. Blinding Investigators were not blinded to the allocation during analyses, since this it not an association study. Hard-Weinberg equilibrium, call ambiguity and low minor allele frequencies. In addition, where a genetically homogeneous set of individuals was required for an analysis, possible outliers were identified and removed. Replication Results were not externally replicated. Randomization The experiments were not randomized. Blinding Investigators were not blinded to the allocation during analyses, since this it not an association study. Results were not externally replicated. The experiments were not randomized. Investigators were not blinded to the allocation during analyses, since this it not an association study. Human research participants Policy information about studies involving human research participants Note that full information on the approval of the study protocol must also be provided in the manuscript. Policy information about studies involving human research participants There are no covariate-relevant population characteristics. Individuals were chosen to represent a broad geographical and ethnolinguistic diversity and were sampled from three ongoing genomics consortium projects on the continent - the H3Africa Consortium, the Southern African Human Genome Programme (SAHGP), and the Trypanosomiasis Genomics Network of the H3Africa Consortium (TryopanoGEN). Population characteristics The participants in the study were previously recruited to individual cohorts as part of ongoing genomics projects on the continent: (i) Samples from 519 individuals from 8 projects in the H3Africa consortium were submitted for consideration. Individuals were predominantly recruited as control individuals from the respective studies, with the exception of those from case-only studies in Cameroon (Sickle cell), Botswana (HIV), Mali (families with neurological disease), and Benin (sickle cell). Individuals with high relatedness (PIHAT> 0.18) were removed from the analysis. Individuals comprising the Berom from Nigeria were recruited as controls for a cervical cancer study, and thus are all female. (ii) A total of 300 (control) individuals were recruited from five countries (Uganda, Zambia, Democratic Republic of the Congo, Cameroon and Ivory Coast) participating in the TrypanoGEN project to study the host and parasite genetics and genomics of trypanosomiasis infection across Africa. (iii) 16 (unaffected) individuals were recruited by the Southern African Human Genome Programme. These individuals met inclusion criteria: male, over the age of 18 years, four grandparents who speak the same language as the participant, not known to be related to the other participants in the study, and willing to provide broad informed consent (including consent to share data and DNA for future studies approved by the HREC (Medical)). Two main Bantu-speaking ethnolinguistic groups were included: The Sotho (Sotho-Tswana speakers; n=8) were recruited from in and around the town of Ventersburg in the Free State Province and the Xhosa speakers (Nguni language; n=7) were recruited from the Eastern Cape Province. One individual was a Zulu speaker (Nguni language) from Johannesburg. Ethics approval (country and institutional) was obtained by each PI of the individual studies. We attach further details of each study's approval, which will be included in the online methods and supplementary materials section of the paper (detailed information was not included in initial submission). Note that full information on the approval of the study protocol must also be provided in the manuscript.
https://openalex.org/W3030936982	https://www.biorxiv.org/content/biorxiv/early/2020/05/30/2020.05.28.121301.full.pdf	English	null	The Atypical Fibrin Fibre Network in Rheumatoid Arthritis and its Relation to Autoimmunity, Inflammation and Thrombosis	bioRxiv (Cold Spring Harbor Laboratory)	2,020	cc-by	16,518	Corresponding authors: p g Etheresia Pretorius Department of Physiological Sciences, Faculty of Science, Stellenbosch University, South Africa. resiap@sun.ac.za http://www.resiapretorius.net/ ORCID: 0000-0002-9108-2384 Douglas B Kell Department of Biochemistry and Systems Biology, Institute of Systems, Molecular and Integrative Biology, Faculty of Health and Life Sciences, University of Liverpool, UK. dbk@liv.ac.uk Etheresia Pretorius Department of Physiological Sciences, Faculty of Science, Stellenbosch University, South Africa. resiap@sun.ac.za http://www.resiapretorius.net/ ORCID: 0000-0002-9108-2384 Douglas B Kell Department of Biochemistry and Systems Biology, Institute of Systems, Molecular and Integrative Biology, Faculty of Health and Life Sciences, University of Liverpool, UK. dbk@liv.ac.uk The Atypical Fibrin Fibre Network in Rheumatoid Arthritis and its Relation to Autoimmunity, Inflammation and Thrombosis Johannes A. Bezuidenhout1, Chantelle Venter1, Timothy Roberts1,3,5, Gareth Tarr2, Douglas B. Kell1,3,4 & Etheresia Pretorius1 The Atypical Fibrin Fibre Network in Rheumatoid Arthritis and its Relation to Autoimmunity, Inflammation and Thrombosis Johannes A. Bezuidenhout1, Chantelle Venter1, Timothy Roberts1,3,5, Gareth Tarr2, Douglas B. Kell1,3,4* & Etheresia Pretorius1 1Department of Physiological Sciences, Faculty of Science, Stellenbosch University, Stellenbosch, Private Bag X1 Matieland, 7602, South Africa 1Department of Physiological Sciences, Faculty of Science, Stellenbosch University, Stellenbosch, Private Bag X1 Matieland, 7602, South Africa 2Division of Rheumatology, Department of Medicine, Faculty of Medicine and Health Sciences, Stellenbosch University, PO Box 241 Cape Town, 8000, South Africa 2Division of Rheumatology, Department of Medicine, Faculty of Medicine and Health Sciences, Stellenbosch University, PO Box 241 Cape Town, 8000, South Africa 3 Department of Biochemistry and Systems Biology, Institute of Systems, Molecular and Integrative Biology, Faculty of Health and Life Sciences, University of Liverpool, Crown St, Liverpool L69 7ZB, UK 4The Novo Nordisk Foundation Centre for Biosustainability, Building 220, Chemitorvet 200, Technical University of Denmark, 2800 Kongens Lyngby, Denmark 4The Novo Nordisk Foundation Centre for Biosustainability, Building 220, Chemitorvet 200, Technical University of Denmark, 2800 Kongens Lyngby, Denmark 5University College London Hospital NHS Foundation Trust, 250 Euston Road, London, NW1 2PB. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Abstract Whole blood coagulation was assessed using Thromboelastography. Fibrin clot networks and fiber structure was investigated using Scanning Electron Microscopy. The detection and quantification of citrullination in formed fibrin clots were performed using a fluorescently labeled Citrulline monoclonal antibody with Confocal Microscopy. Results. Concentrations of SAA, CRP and ICAM-1 were significantly elevated in RA patients compared to controls. TEG parameters relating to coagulation initiation (R and K), rate of fibrin cross-linking (α-Angle), and time to reach maximum thrombus generation (TMRTG) were attenuated in RA patients. Parameters relating to clot strength (MA, MRTG, TGG) did not statistically differ between RA and controls. Logistic regression modelling revealed stronger association between acute phase reactants (CRP, SAA) with TEG parameters than endothelial function markers. Microscopic analysis revealed denser networks of thicker fibrin fibers in RA patients compared to controls [median (interquartile range) 214 (170-285) vs 120 (100-144) nm respectively, p<0.0001, Odds ratio=22.7). Detection of multiple citrullinated regions within fibrin clot structures in RA patients, which was less prevalent in control samples (p<0.05, OR=2.2). Conclusion. Patients with active RA display a coagulation profile that is dissimilar to general findings associated with other inflammatory conditions. The alteration of protein structures by autoimmune linked citrullination could play a role in determining the structure of fibrin and the potential of conferring a heightened thrombotic risk in RA patients. Douglas B Kell Department of Biochemistry and Systems Biology, Institute of Systems, Molecular and Integrative Biology, Faculty of Health and Life Sciences, University of Liverpool, UK. 1 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Rheumatoid Arthritis, Coagulation, Fibrinogen, Citrullination, CRP, SAA, ICAM-1, VCAM-1, Cardiovascular Risk Abstract Objective. The risk of cardiovascular events in patients with RA is disproportionately heightened as a result of systemic inflammation. The relative effect of autoimmune- associated citrullination on the structure and thrombotic potential of fibrin(ogen) remains unknown. We therefore compared indices of vascular function, inflammation, coagulation and fibrin clot composition in RA patients with healthy controls and evaluated inter-parameter relationships. Methods. Blood samples were collected from 30 RA patients and 25 age- and gender-matched healthy volunteers. Levels of SAA, CRP, ICAM-1 and VCAM-1 was measured using a sandwich immunoassay. Whole blood coagulation was assessed using Thromboelastography. Fibrin clot networks and fiber structure was investigated using Scanning Electron Microscopy. The detection and quantification of citrullination in formed fibrin clots were performed using a fluorescently labeled Citrulline monoclonal antibody with Confocal Microscopy. Results. Concentrations of SAA, CRP and ICAM-1 were significantly elevated in RA patients compared to controls. TEG parameters relating to coagulation initiation (R and K), rate of fibrin cross-linking (α-Angle), and time to reach maximum thrombus generation (TMRTG) were attenuated in RA patients. Parameters relating to clot strength (MA, MRTG, TGG) did not statistically differ between RA and controls. Logistic regression modelling revealed stronger association between acute phase reactants (CRP, SAA) with TEG parameters than endothelial function markers. Microscopic analysis revealed denser networks of thicker fibrin fibers in RA patients compared to controls [median (interquartile range) 214 (170-285) vs 120 (100-144) nm respectively, p<0.0001, Odds ratio=22.7). Detection of multiple citrullinated regions within fibrin clot structures in RA patients, which was less prevalent in control samples (p<0.05, OR=2.2). Conclusion. Patients with active RA display a coagulation profile that is dissimilar to general findings associated with other inflammatory conditions. The alteration of protein structures by autoimmune linked citrullination could play a role in determining the structure of fibrin and the potential of conferring a heightened thrombotic risk in RA patients. Objective. The risk of cardiovascular events in patients with RA is disproportionately heightened as a result of systemic inflammation. The relative effect of autoimmune- associated citrullination on the structure and thrombotic potential of fibrin(ogen) remains unknown. We therefore compared indices of vascular function, inflammation, coagulation and fibrin clot composition in RA patients with healthy controls and evaluated inter-parameter relationships. Methods. Blood samples were collected from 30 RA patients and 25 age- and gender-matched healthy volunteers. Levels of SAA, CRP, ICAM-1 and VCAM-1 was measured using a sandwich immunoassay. Key words: Rheumatoid Arthritis, Coagulation, Fibrinogen, Citrullination, CRP, SAA, ICAM-1, VCAM-1, Cardiovascular Risk 2 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint The Atypical Fibrin Fibre Network in Rheumatoid Arthritis and its Relation to Autoimmunity, Inflammation and Thrombosis......................................................... 1 Abstract ................................................................................................................. 2 Key words:............................................................................................................. 2 Introduction ............................................................................................................ 4 Materials and methods .......................................................................................... 6 Ethical considerations ............................................................................................ 6 Study population .................................................................................................... 7 Blood sampling ...................................................................................................... 7 Thromboelastography® .......................................................................................... 7 Scanning electron microscopy ............................................................................... 8 Vascular injury panel analysis .............................................................................. 10 Confocal laser scanning microscopy .................................................................... 10 Statistical analysis ............................................................................................... 10 Results ................................................................................................................ 11 Subjects .............................................................................................................. 11 Confirmation of altered inflammatory and vascular function profile in RA ............. 12 Functional coagulation assessment indicates a prothrombotic tendency in RA .... 12 Association of vascular injury biomarkers to thrombotic parameters indices ........ 13 SEM analysis exposes anomalous fibrin network architecture in formed RA clots 15 Confocal analysis of plasma clots reveals a higher number of citrullinated sites in RA fibrin networks compared to controls .............................................................. 17 Discussion ........................................................................................................... 20 Concluding remarks ............................................................................................. 25 References .......................................................................................................... 27 3 3 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Introduction Rheumatoid Arthritis (RA) is a chronic, systemic autoimmune disease characterized by both peripheral joint and extra-articular site inflammation, with an increased predisposition to a higher incidence of cardiovascular disease (CVD) (1, 2). CVD (including stroke and myocardial infarction) is almost 50% more common in RA patients than the general population and is the most frequent cause of early mortality (3). Traditional risk factors for CVD (age, hypertension, obesity, etc.), do not fully account for the elevated occurrence of CVD events, and thus RA (genetics and disease characteristics) has been identified as a strong independent risk factor (4). The interdependence of inflammatory and hemostatic pathways is well established and observable in multiple types of tissue, organs and pathologies (5). Disruption of the tightly regulated homeostatic control of immune and hemostatic systems could result in a rapid progression towards a prothrombotic tendency, a central cause of ischaemic stroke and myocardial infarction (6). This circumstance holds true for RA, with elevated levels of both pro-inflammatory [e.g. C-reactive protein (CRP) (7-11), Tumor necrosis factor alpha (TNFα) (7, 9, 11), Interleukin-6 (IL-6) (7-12), IL-1β (7) and Serum Amyloid A (SAA) (10, 13, 14)] and prothrombotic markers [e.g. D-dimer (8, 9, 11, 15, 16), Fibrinogen (10, 11, 16, 17), Tissue Factor (TF) (15), and von Willebrand factor (vWF) (8, 16, 18)], which is associated with one another (7, 19) and with the risk of future cardiovascular complications (20-23). Key intermediaries of this manifestation are the structural components of formed thrombi. Soluble fibrinogen is cleaved by thrombin in order to form dense matrices of thin fibrous protein known as fibrin (24). Polymerized fibrin networks are essential for wound healing and other occlusive physiological processes (24). However, exposure to inflammatory biomarker stimuli [such as CRP (25), SAA (26), and pro- inflammatory cytokines (27, 28)] can result in the alteration of mechanical and viscoelastic properties of fibrin clots into a prothrombotic phenotype. This phenomenon has previously been observed in RA plasma clots (29, 30). Various immunopathogenic processes related to RA development can exert upstream amplification of the coagulation cascade as well as impairing fibrin clot dissolution (9, 19, 31). Fibrin(ogen) is also a potent pro-inflammatory signaling entity itself, mainly through ligand-receptor interactions with immune cells that further propagates pro- 4 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. Introduction CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Figure 1: Overview of the overlapping processes of inflammation and coagulation in both synovial and vascular compartments. 1. The chronic and systemic nature of the inflammatory response in RA characterizes the disease as an independent risk factor for CVD. 2. The movement of leukocytes, inflammatory cytokines, procoagulant factors and immune complexes are aided by vascular endothelial dysfunction and neovascularization of hyperproliferative joint tissues. 3. The role of fibrin(ogen) is integral to the formation of hyperplastic and destructive synovial tissue (pannus) and vascular thrombosis, while being a prominent self-protein target of aberrant citrullination and autoimmunogenicity in RA. Figure 1: Overview of the overlapping processes of inflammation and coagulation in both synovial and vascular compartments. 1. The chronic and systemic nature of the inflammatory response in RA characterizes the disease as an independent risk factor for CVD. 2. The movement of leukocytes, inflammatory cytokines, procoagulant factors and immune complexes are aided by vascular endothelial dysfunction and neovascularization of hyperproliferative joint tissues. 3. The role of fibrin(ogen) is integral to the formation of hyperplastic and destructive synovial tissue (pannus) and vascular thrombosis, while being a prominent self-protein target of aberrant citrullination and autoimmunogenicity in RA. Introduction It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint inflammatory effects (32-36). The deimination of particular arginine residues in fibrin(ogen), known as citrullination, is a distinctive RA posttranslational modification that alters normal protein structure and function that confers antigenicity to modified proteins (37-43). The functional relationship between citrullination and the presence of a prothrombotic fibrin clot phenotype is still poorly understood. Some studies have shown that citrullination of fibrinogen prevents thrombin digestion and subsequent fibrinogenesis (44-46). However, the experimental conditions upon which these findings are based do not reflect physiological coagulation and is inconsistent with a predominantly hypercoagulable state seen in RA (47). Inflammation-induced fibrin formation is equally present in RA synovial spaces as it is in circulation (Refer to Figure 1). Synovial coagulation is a key step in pannus formation, where fibrin provides the structural scaffold for immune cells that are responsible for synovial membrane disintegration and eventual joint damage (31, 48). Endothelial tissue dysfunction is a key process that facilitates this ubiquitous distribution of aberrant fibrin deposition in both synovia and vasculature. This pathophysiological state is characterized by the expression of cell adhesion molecules [intercellular cell adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1)], pro-inflammatory cytokines and pro-thrombotic markers (49, 50). This allows for the recruitment, translocation and propagation of inflammatory and thrombotic mediators across the synovial barrier (51-53). There is significant overlap in inflammatory pathways responsible for joint damage in RA and hypercoagulation, coupled with the fact that disease severity has been correlated to more adverse cardiovascular complications (21, 54, 55). It is therefore prudent that these processes and their relevant markers be examined systemically in RA, and not isolated to either vascular or synovial compartment. The aim of this study was to examine the extent to which the coagulation profiles and fibrin network architecture of RA patients are influenced by acute phase inflammation, endothelial dysfunction an autoimmune-related protein modification. 5 . Ethical considerations Ethical approval for this study was given by the Health Research Ethics Committee (HREC) of Stellenbosch University (reference number: 6983). This study was carried out in strict adherence to the International Declaration of Helsinki, South African Guidelines for Good Clinical Practice and the South African Medical Research Council (SAMRC) Ethical Guidelines for research. Written consent was obtained from all participants (RA patients and healthy participants) prior to any sample collection. 6 6 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Study population The RA sample group consisted of 30 patients (24 female and 6 male) that visited the Winelands Rheumatology Clinic (Stellenbosch, South Africa) for routine check- ups. All patients fulfilled the 2010 American College of Rheumatism/European League against Rheumatism (ACR/EULAR) classification criteria for RA diagnosis (56). The mean age of RA group was 53.4 years (range 22-75 years) with a mean disease duration of 10.5 years (range 1-39 years). RA participants were excluded from the study if they presented with other severe comorbidities (such as cancer or diabetes), existing cardiovascular disease or taking anticoagulant medication. RA participants were not excluded on the basis of any antirheumatic drug treatment or the use of glucocorticosteroids. The majority of RA patients (87%) were on a schedule of non-biologic disease modifying antirheumatic drugs (DMARDS, such as methotrexate, hydroxychloroquine, sulfasalazine, or leflunomide), while a lower proportion of patients were on biologic DMARDs (60%) and cortisone (14%, 5-10mg dosage). The control group consisted of 30 age- (mean: 53.9 years) and gender- (22 female and 8 male) matched volunteer blood donors. The inclusion criteria for healthy controls were: (i) no history of thrombotic disease or inflammatory conditions (ii) no use of any chronic medication (ii) no use of anticoagulant therapy (iii) non- smokers (iv) females not taking contraceptive medication or hormone replacement therapy (v) females that are not pregnant or lactating. All demographic information is summarized in Table 1. Blood sampling Whole blood (WB) samples were collected in vacutainer tubes using 3.8% sodium citrate as anticoagulant. Blood drawing on all participants was performed by a qualified nurse, or phlebotomist by sterile puncture of the antecubital vein. Blood tubes were incubated at room temperature for a minimum duration of 30 minutes prior to the commencement of any whole blood analysis. In order to obtain platelet poor plasma (PPP), sodium citrated blood tubes were centrifuged at 3000xg for 15 minutes, aliquoted into Eppendorf tubes and stored at -80oC until further analysis. Thromboelastography® Analysis of dynamic coagulation kinetics were performed on RA and control WB by means of Thromboelastograph® (TEG®) 5000 Haemostasis Analyzer System (Haemonetics®, 07-033). In brief, coagulation is initiated by recalcification of 340µL 7 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint WB with 20µL of 0.2mM Calcium chloride (CaCl2) (Haemonetics®, 7003) in a disposable TEG® cup (Haemonetics®, 6211). Various kinetic clotting parameters are determined by assessing the resistance that the forming thrombus provides against the oscillating pin of the instrument (measuring at 37oC). Parameters derived from the thromboelastograph tracing consist of: reaction time (R, time from test start to initial fibrin formation in minutes), kinetics [K, time required to reach an amplitude (clot thickness) of 20mm, in minutes], alpha angle (α, rate of fibrin cross linking indicated by degrees), maximal amplitude (MA, maximum strength of formed clot in millimeters), maximum rate of thrombus generation (MRTG, in dynes.cm-2.s-1), time to maximum rate of thrombus generation (TMRTG, in minutes) and total thrombus generation (TTG, in dynes.cm-2).(57) Scanning electron microscopy CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Confocal laser scanning microscopy Confocal laser scanning microscopy In order to determine the extent of protein deamination in fibrin networks, PPP aliquots of RA samples (n=10) and control samples (n=10) were thawed and fibrin clots prepared (refer to SEM method) on glass microscope slides in a dark room. Samples were fixed with 4% PFA, washed 3x with PBS, and blocked with 5% Goat serum solution (Abcam, ab7481) for 30 minutes. Clots were then stained with a 1:50 dilution Citrulline Monoclonal Antibody (2D3.1) (Thermo Fisher Scientific, Cat #MA5- 27573) and incubated for 1 hour. Following another 3x PBS wash to remove unbound antibodies, samples were then stained with 1:200 dilution Goat Anti-Mouse IgG Secondary antibody conjugated to AlexaFluor 488 (Thermo Fisher Scientific, Cat #A327273) and incubated for 1 hour. Slides were washed 3x with PBS to remove unbound antibody, allowed to dry, and mounted with a glass coverslip. Samples were viewed with a Zeiss LSM 780 Confocal laser scanning microscope (CLSM) with a Plan-Apochromat 63x/1.4 oil DIC M27 objective. AlexaFluor488 was excited with the 488nm laser and emission was detected at 508-570nm. Three representative micrographs per sample were analysed for fluorescent particle distribution using ImageJ (Version 1.52p). Images were calibrated to scale, and a global threshold (27 pixel cut-off) applied to all analysed micrographs. Scanning electron microscopy The ultrastructure of fibrin networks and individual fibrin fibers were examined using scanning electron microscopy (SEM). In summary, clots were prepared from thawed PPP samples of RA patients (n=10) and controls (n=10) by addition of 5µL human thrombin (provided by South African National Blood Service) to 10µL PPP on a glass coverslip and transferred to a 24-well plate. Preparation consisted of washing with 10X Gibco® phosphate-buffered saline (PBS, pH 7.4) (ThermoFisher Scientific, 10010015), chemical fixation with 4% Paraformaldehyde (PFA) (Sigma-Aldrich, P6148) and then 1% Osmium Tetrahydroxide (OsO4) (Sigma-Aldrich, 75632), followed by dehydration with increasing grades of ethanol and 99.9% Hexamethyldisilizane (HMDS) (Sigma-Aldrich, 37921) [for detailed protocols please refer to (57)]. Samples were carbon coated using a Quorom Q150T E carbon coater. Images were captured at an electron high tension (EHT) of 1kV using a high resolution InLens detector of the Zeiss Merlin™ (Gemini II) FE SEM (Carl Zeiss Microscopy, Munich, Germany). Fibrin fiber diameters representative of each respective sample group (RA and control) was determined by means of image analysis software ImageJ (Version 1.52p). Three representative micrographs (78,98µm2 image size, 10 000x magnification) were calibrated to scale and overlaid with a non-destructive grid (2µm2 tile size). Single representative fibrin fibers were measured in 28 tiles per image, producing 84 fiber diameter measurements per sample (method illustrated in figure 2). 8 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Figure 2: Fibrin fiber diameter measurement scheme. SEM micrographs of a prepared PPP clot from a representative healthy control (top) and RA (bottom). Figure 2: Fibrin fiber diameter measurement scheme. SEM micrographs of a prepared PPP clot from a representative healthy control (top) and RA (bottom). Figure 2: Fibrin fiber diameter measurement scheme. SEM micrographs of a prepared PPP clot from a representative healthy control (top) and RA (bottom). 9 9 . Vascular injury panel analysis Plasma concentrations of soluble ICAM-1, VCAM-1, CRP, and SAA were measured by sandwich immunoassay [Meso Scale Discover (MSD) Vascular Injury Panel (V- plex) 2 (human) kits, catalogue number K15198D]. RA (n=30) and control (n=30) PPP samples and reagents were prepared as per manufacturer’s protocol. Samples were run in duplicate and measurements read on an MSD Discovery Workbench 4. Analyte concentrations were calculated from the calibration curve generated by absorbance measurements of manufacturer supplied calibrator standards. Statistical analysis Statistical analysis was performed using R version 4.0. Specifically, univariate logistic regression was performed to determine odds ratios (OR) for experimental variables using the logistic model in the rstanarm package (with default priors). ORs and 95% confidence intervals were extracted in the corresponding unit system (i.e. not z-scaled) for all variables except Fibrin fiber diameter and citrulline particle number shown in Table 4 which are z-scaled to aid interpretation. Tables 2 and 3 10 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint show ORs after adjustment for age and gender with unadjusted analysis identifying the same significance and effect sizes. In addition, results for classical statistical tests are reported as follows. The distribution of sample datasets for each variable experimental was determined using the Shapiro-Wilk test. Accordingly, p-values for each variable comparing RA to healthy controls were calculated using either a Mann- Whitney U test for nonparametric data or a Student t-test for parametric data. Statistical significance was set at p<0.05. One can see close alignment between all these and the OR results. Results Subjects j Demographic information of all study participants is listed in Table 1. The RA sample group closely resembles the general population distribution for age (median: 54 years) and sex (80% female) of the disease (58). The control group of healthy volunteers was closely matched to the RA group with regards to age (median: 50 years) and sex (73% female). The RA sample group was heterogeneous with respect to clinical presentation, with most patients on an anti-rheumatic drug therapy regime. The majority of RA patients also presented with positive titers for anti-cyclic citrullinated peptide (CCP) (77%) and rheumatoid factor (97%) autoantibodies. Table 1: Demographic and clinical characteristics of study participants. Risk Factor RA (n=30) Control (n=30) Age (years) 53.5 (22-75) 50 (28-79) Women [n (%)] 24 (80%) 22 (73.3%) Smokers [n (%)] 3 (10%) - Disease duration (years) 6.5 (1-39) - Hypertension [n (%)] 5 (16.7%) - Hypercholesterolemia [n (%)] 3 (10%) - Disease activity (DAS-28) classification [n (%)] - Remission - Low - Moderate - High 1 (3.3%) 17 (56.7%) 4 (13.3%) 8 (26.7%) - Autoantibody seropositivity [n (%)] - RF-/Anti-CCP- - RF+/Anti-CCP- - RF-/Anti-CCP+ - RF+/Anti-CCP+ 1 (3.3%) 6 (20%) - 23 (76.7%) Conventional DMARD use [n (%)] 26 (86.7%) - Biologic DMARD use [n (%)] 18 (60%) - Table 1: Demographic and clinical characteristics of study participants. 11 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Cortisone use [n (%)] 13 (43%) - Values expressed as median and interquartile range (IQR) unless otherwise stated. Disease activity scores (DAS-28), were categorized in RA patients as remission, low, moderate or high (59). For full demographic information and specified list of disease modifying anti-rheumatic drugs (DMARDs) please refer to full study database in supplementary material. 13 (43%) CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint patients showed significantly altered rates of clot formation compared to healthy controls. This included shortened clot initiation (R; OR=0.675, p=0.0007 and K; OR=0.539, p=0.0196), augmented fibrin cross-linking (α; OR=1.15, p=0.0006) and shortened time to maximal thrombus formation (TMRTG; OR=0.737, p=0.0001). Measures of overall clot strength (MA) and growth (TTG) were attenuated in RA but did not statistically differ from those of controls. Table 3: Thromboelastography® analysis. TEG® Parameter RA (n=30) Control (n=30) Adjusted OR (95% CI) P-value R (min) 8.3 [7.5-9.5] 11.9 [8.3-15.1] 0.675 (0.528-0.827) 0.0007*** K (min) 2.6 [2.2-3.2] 3.6 [2.6-4.5] 0.539 (0.308-0.847) * 0.0196* α (o) 63.2 [57.6-68.6] 54.2 [48.4-60.5] 1.15 (1.07-1.27) * 0.0006*** MA (mm) 55.9 [52.0-61.7] 58.9 [54.5-63.9] 0.960 (0.889-1.03) 0.1646 MRTG (dyn·cm-2·s-1) 5.01 [3.89-6.02] 4.23 [3.26-5.85] 1.03 (0.767-1.40) 0.2745 TMRTG (min) 11.67 [9.73-12.96] 17.43 [12.48-21.13] 0.737 (0.607-0.858) * 0.0001*** TTG (dyn·cm-2) 633 [544.1-775.2] 719.3 [602.1-879.8] 0.997 (0.995-1.00) 0.0899 Viscoelastic coagulation measurements for RA and control are expressed as median [IQR]. Odds ratios for age- and gender-adjusted logistic regression model is listed with the 95% confidence interval, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. Table 3: Thromboelastography® analysis. Viscoelastic coagulation measurements for RA and control are expressed as median [IQR]. Odds ratios for age- and gender-adjusted logistic regression model is listed with the 95% confidence interval, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. Viscoelastic coagulation measurements for RA and control are expressed as median [IQR]. Odds ratios for age- and gender-adjusted logistic regression model is listed with the 95% confidence interval, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. Viscoelastic coagulation measurements for RA and control are expressed as median [IQR]. Odds ratios for age- and gender-adjusted logistic regression model is listed with the 95% confidence interval, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. 13 (43%) Confirmation of altered inflammatory and vascular function profile in RA Circulating concentrations of endothelial function and acute phase markers are shown in Table 2. Previous studies have shown these markers (CRP, SAA, sVCAM- 1 and sICAM-1) to be associated with a prothrombotic state and increased CVD risk in RA(7, 13, 60-62), and was therefore measured in this study to determine the extent to which systemic inflammation influences viscoelastic and structural clot properties. As expected, all markers were elevated in RA compared to controls [CRP (median 4.25 µg/mL vs 1.26, OR=1.29), SAA (4.98 µg/mL vs 1.52, OR=1.29), sICAM-1 (378.82 ng/mL vs 289.70, OR=1.01), sVCAM-1 (359.83 ng/mL vs 325.80, OR=1.00)]. Table 2: Vascular injury panel (V-Plex) analysis. V-Plex Analyte RA (n=30) Control (n=25) Adjusted OR (95% CI) P-value CRP (µg/mL) 4.25 [1.89-8.11] 1.26 [0.43-3.15] 1.29 (1.09-1.62) 0.0011** SAA (µg/mL) 4.98 [2.56-8.84] 1.52 [0.64-2.24] 1.25 (1.08-1.55) * <0.0001**** sICAM-1 (pg/mL) 377.82 [295.69-460.49] 289.7 [225.34-365.56] 1.008 (1.0021-1.014) * 0.0202* sVCAM-1 (pg/mL) 359.83 [283.01-394.20] 325.8 [253.73-422.62] 1.001 (0.995-1.007) 0.9 Analyte concentrations for RA and control are expressed as median [IQR]. Odds ratios for age- and gender-adjusted logistic regression model is listed with the 95% confidence interval, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. Table 2: Vascular injury panel (V-Plex) analysis. Analyte concentrations for RA and control are expressed as median [IQR]. Odds ratios for age- and gender-adjusted logistic regression model is listed with the 95% confidence interval, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. Functional coagulation assessment indicates a prothrombotic tendency in RA Whole blood coagulation parameters as measured by TEG® are listed in Table 3. Limited viscoelastic assessment of coagulation in RA has been performed to date (60, 63, 64), with TEG® not commonly used in rheumatology practice (65-67). RA 12 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . 13 (43%) Denotes statistically significant differences in variables between RA and controls. Association of vascular injury biomarkers to thrombotic parameters indices Significantly altered inflammatory and coagulation indices in RA, as determined by an adjusted logistic regression model, are represented by box-and-whisker plots (Figure 3). Additionally, the distribution of RA and control experimental values are plotted and arranged in a lattice (Figure 4) to exhibit correlation values between intra- and inter-assay variables. CRP shows positive and significant association with SAA (r = 0.6) and sICAM-1 (r = 0.52). Further strong correlation existed between respective endothelial markers (sICAM-1 and sVCAM-1, r=0.62), and between all 13 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint TEG® parameters (R, K, α, TMRTG). CRP and SAA display stronger association with thromboelastographic parameters than the cell adhesion molecules. Figure 3: Box-and-whisker plots for statistically significant TEG® and V-plex parameters. Boxes represent the median and IQR. Whiskers indicate upper (75th percentile + 1.5IQR) and lower (25th percentile -1.5IQR) extremes. Outlier values are indicated by black dots Figure 3: Box-and-whisker plots for statistically significant TEG® and V-plex parameters. Boxes represent the median and IQR. Whiskers indicate upper (75th percentile + 1.5IQR) and lower (25th percentile -1.5IQR) extremes. Outlier values are indicated by black dots 14 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. 13 (43%) ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Figure 4: Lattice of cross-plots and correlation values across statistically significant TEG® and V-Plex measurements. The sample values are indicated by crosses (RA: red, Control: black). Strong intra-analysis correlation exists between CRP with SAA and sICAM-1, between sICAM-1 and VCAM-1, and between all TEG parameters (R, K, Angle, TMRTG). CRP and SAA show moderate correlation with inflammatory and endothelial markers. Figure 4: Lattice of cross-plots and correlation values across statistically significant TEG® and V-Plex measurements. The sample values are indicated by crosses (RA: red, Control: black). Strong intra-analysis correlation exists between CRP with SAA and sICAM-1, between sICAM-1 and VCAM-1, and between all TEG parameters (R, K, Angle, TMRTG). CRP and SAA show moderate correlation with inflammatory and endothelial markers. SEM analysis exposes anomalous fibrin network architecture in formed RA clots SEM analysis exposes anomalous fibrin network architecture in formed RA clots Further investigation into the apparent modification of the clot structure in RA was carried out by means of SEM. Figure 2 illustrates the scheme followed for the measurement of fibrin fiber diameters in PPP clots of RA (n=10) and Control (n=10) samples. Results (refer to Table 4) indicate that fibrin fiber diameters in 15 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint representative areas were significantly increased in RA versus controls (median: 214nm vs 120nm, OR=22.732). Examining the networks qualitatively, it is evident that ex vivo formed clots from RA samples have denser, less porous fibrin networks. Figure 5 B-D, representative of the RA sample group, illustrates the amalgamation of fibrin monomers that contribute to increased fibrin fiber diameter and overall network density. This contrasts sharply with the ultrastructural attributes of Figure 5A (Healthy control sample), which demonstrates thinner protein strands and a more permeable fibrin network. Table 4: Microscopy Image Analysis. Imaging Parameter RA (n=10) Control (n=10) OR (95% CI) P-value Fibrin fiber diameter (nm) 214 (170-285) 120 (100-144) 22.732 (17.085-31.441) <0.0001 *** Fluorescent citrulline particles 3279 (1966-4260) 1095 (411-2937) 2.2268 (1.2238-4.3527) 0.0355 * Table 4: Microscopy Image Analysis. Table 4: Microscopy Image Analysis. Imaging Parameter RA (n=10) Control (n=10) OR (95% CI) P-value Fibrin fiber diameter (nm) 214 (170-285) 120 (100-144) 22.732 (17.085-31.441) <0.0001 **** Fluorescent citrulline particles 3279 (1966-4260) 1095 (411-2937) 2.2268 (1.2238-4.3527) 0.0355 * Evaluation of plasma clots using image analysis software (ImageJ) based techniques for representative RA (n=10) and Healthy control (n=10) subjects. Fibrin fiber diameters were determined from SEM micrographs, while fluorescent particle analysis was calculated from CLSM micrographs. Values are expressed as median [interquartile range]. SEM analysis exposes anomalous fibrin network architecture in formed RA clots Scaled odds ratios for the logistic regression model are listed with their 95% confidence intervals, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. p ( ) ( ) ( ) Evaluation of plasma clots using image analysis software (ImageJ) based techniques for representative RA (n=10) and Healthy control (n=10) subjects. Fibrin fiber diameters were determined from SEM micrographs, while fluorescent particle analysis was calculated from CLSM micrographs. Values are expressed as median [interquartile range]. Scaled odds ratios for the logistic regression model are listed with their 95% confidence intervals, while level of statistical significance was set at p<0.05 for calculated Mann-Whitney test values. Denotes statistically significant differences in variables between RA and controls. 16 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Figure 5: Scanning electron micrographs of the fibrin network ultrastructure. Representative micrographs of the fibrin network in healthy controls (A) and RA patients (B-D). The altered clot ultrastructure in RA, consisting of less permeable networks of thicker fibrin fibers, represents a prothrombotic phenotype. A B C D B A B D C D C Figure 5: Scanning electron micrographs of the fibrin network ultrastructure. Representative micrographs of the fibrin network in healthy controls (A) and RA patients (B-D). The altered clot ultrastructure in RA, consisting of less permeable networks of thicker fibrin fibers, represents a prothrombotic phenotype. SEM analysis exposes anomalous fibrin network architecture in formed RA clots Confocal analysis of plasma clots reveals a higher number of citrullinated sites in RA fibrin networks compared to controls Additionall e probed hether a toimm ne related modifications of coag lation Confocal analysis of plasma clots reveals a higher number of citrullinated sites in RA fibrin networks compared to controls Addi i ll b d h h i l d difi i f l i Additionally, we probed whether autoimmune-related modifications of coagulation proteins are distinctive to RA patients compared to healthy individuals. Fibrin and fibrinogen are well known extracellular targets for peptidylarginine deiminase (PAD) enzymes (37, 68), with citrullinated isoforms eliciting the generation of anti-CCP antibodies in large proportions of RA patients (40, 41). Citrullinated fibrin deposits are common synovial compartment manifestations, but this has not been investigated in plasma. The determination of the net-effect of coagulation factor citrullination on hemostatic outcome also remains elusive. To investigate the presence and extent of potential citrullination in RA (n=10) and Control (n=10) PPP thrombi, fluorescence analysis using a Citrulline-identifying monoclonal antibody with CLSM was performed (Figure 6). Acquired image data (Figure 7) suggests a 17 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint relatively higher proportion of citrulline particles in RA fibrin networks versus controls (OR=2.2268). relatively higher proportion of citrulline particles in RA fibrin networks versus controls (OR=2.2268). Figure 6: Confocal microscopy of PPP clots stained with a Citrulline monoclonal antibody. Confocal micrographs of representative control and RA samples. Each row represents identical areas captured of fibrin clot preparations using transmitted light to illustrate fibrin network topography (left-hand column) and a green channel (middle column) to identify fluorescent particles corresponding to citrulline residues. Fluorescent images are also provided in binary (right-hand column) to better illustrate differences seen between RA and control samples with respect to citrulline- coupled fluorescence. Particle analysis confirmed the observable presence of enhanced fluorescent signal in RA samples (n=10) versus healthy controls (n=10). Figure 6: Confocal microscopy of PPP clots stained with a Citrulline monoclona antibody. Confocal micrographs of representative control and RA samples. Eac row represents identical areas captured of fibrin clot preparations using transmitte light to illustrate fibrin network topography (left-hand column) and a green channe (middle column) to identify fluorescent particles corresponding to citrulline residues Fluorescent images are also provided in binary (right-hand column) to bette illustrate differences seen between RA and control samples with respect to citrulline Figure 6: Confocal microscopy of PPP clots stained with a Citrulline monoclonal antibody. Confocal micrographs of representative control and RA samples. Each row represents identical areas captured of fibrin clot preparations using transmitted light to illustrate fibrin network topography (left-hand column) and a green channel (middle column) to identify fluorescent particles corresponding to citrulline residues. Fluorescent images are also provided in binary (right-hand column) to better illustrate differences seen between RA and control samples with respect to citrulline- coupled fluorescence. Particle analysis confirmed the observable presence of enhanced fluorescent signal in RA samples (n=10) versus healthy controls (n=10). 18 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Figure 7: Box-and-whisker plots for micrographs analysis. relatively higher proportion of citrulline particles in RA fibrin networks versus controls (OR=2.2268). Outlier values are indicated by black dots. In comparison to healthy individuals, RA subjects exhibited significantly increased fibrin fiber diameters (median: 214nm vs 120nm, OR=22.7) and detected fluorescent areas (median: 3279 vs 1095, OR=2.23) of fibrin citrullination. Figure 7: Box-and-whisker plots for micrographs analysis. Outlier values are indicated by black dots. In comparison to healthy individuals, RA subjects exhibited significantly increased fibrin fiber diameters (median: 214nm vs 120nm, OR=22.7) d d t t d fl t ( di 3279 1095 OR 2 23) f fib i Figure 7: Box-and-whisker plots for micrographs analysis. Outlier values are indicated by black dots. In comparison to healthy individuals, RA subjects exhibited significantly increased fibrin fiber diameters (median: 214nm vs 120nm, OR=22.7) and detected fluorescent areas (median: 3279 vs 1095, OR=2.23) of fibrin citrullination. 19 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Discussion It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint enzymes (92, 95) by local tissue. SAA showed slightly weaker association with sICAM-1 compared with CRP in our studies (Figure 4). Levels of soluble cell adhesion molecules (sICAM-1 and sVCAM-1) are the most accessible form in which to determine endothelial activity. They are also strong predictive biomarkers for CVD in RA (61, 62, 97). Increased levels of CAMs indicate endothelial dysfunction that facilitates pro-inflammatory and prothrombotic conditions (7). ICAM-1 is a prominent receptor for fibrinogen (98), with interaction fortifying endothelial adhesion and migration of leukocytes (99-101), endothelial-platelet adhesion (102-104), and causes vasoconstriction (105, 106). ICAM-1 signaling pathways can also promote endothelial tissue factor expression (107). The role of VCAM-1 in directly promoting coagulation is not as well understood, but does contribute to atherosclerotic plaque formation as a result of adhering to PBMCs (108, 109) sICAM-1 and sVCAM-1 concentrations were elevated in RA (Table 2) but were not as strongly associated with coagulation function as CRP and SAA (Figure 4). This belies the fact that CAMs present on cell surfaces play more facilitative roles in thrombotic diseases rather than instigating them. After confirming the presence of a systemic inflammatory state in RA patients, we investigated the possible repercussions thereof on the coagulation profile of study participants. Thromboelastography® is a hemostatic function test that measures the rate, strength, and stability of clot formation through the viscoelastic changes induced by fibrin polymerization. Despite its widespread clinical use, especially as point-of-care instruments in surgical settings, TEG® is not often utilized in Rheumatology practice. The advantage that TEG® provides over conventional coagulation tests [such as prothrombin time (PT), activated partial thromboplastin time (aPTT) and D-dimer] is that it measures global hemostatic function and outcome rather than single time points or pathways in the coagulation cascade (110). Türk et al. Discussion There is a need to bridge translational gaps between RA immunopathogenesis and systemic vascular and hemostatic irregularities. The development of RA auto- immunogenicity precedes the onset of joint disease (69, 70). The link between RA autoimmune patterns and its possible role in exacerbating thrombosis is still poorly understood. Crosstalk between immune and hemostatic systems with the endothelium represents a critical interface in which both arthritic and cardiovascular pathologies are initiated and propagated. We therefore analyzed a panel of biomarkers that are representative of this dynamic milieu, is associated with RA disease severity and CVD. Levels of both acute phase reactants (CRP and SAA) were significantly elevated in RA patients (Table 2) and showed a strong association (Figure 4). This was expected as acute phase reactant concentrations rise dramatically under acute inflammatory states, with both CRP and SAA shown to reliably predict disease severity and CVD risk in RA (13, 14, 22, 71, 72). CRP can bind to surface receptors on monocytes, endothelial cells and platelets thereby altering their function to propagate hypercoagulable conditions (73, 74). The ability of CRP and SAA to induce TF expression has been demonstrated in various cell types (13, 14, 22, 71, 72, 75-81) and by in vivo studies (82, 83). Additionally, acute phase reactants can also suppress fibrinolysis by promoting expression of plasminogen activator inbibitor- 1 (PAI-1) (82) and inhibiting TF pathway inhibitor (TFPI) expression (81, 83). CRP is also able to activate endothelial cells and cause the expression of cell adhesion molecules (ICAM-1 and VCAM-1) (84, 85). It should however be noted that cellular effects of CRP have been disputed as being primarily caused by bacterial contaminants in CRP preparations rather than the protein itself (86). SAA is the precursor to amyloid A (AA) protein, which form insoluble fibrillar depositions in major organs as a result of long-term inflammation (87). RA has been prominently implicated as a pre-existing condition for the development of potentially fatal AA amyloidosis (88, 89). Known for being primarily hepatically synthesized, synovial tissue cells (90-93) and chondrocytes (94) can be articular sources of SAA production. Within the synovium SAA promotes pro-arthritic processes through the expression of cell adhesion molecules (95, 96), cytokines (96), and matrix degrading 20 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. Discussion (2018) is the only recent study that has assessed thrombotic tendency in RA patient with thromboelastographic assessment (60) Using a modified version of TEG®, known as rotational thromboelastometry (ROTEM), they found that RA disease markers (CRP, DAS-28) correlated strongly with ROTEM parameters. Our findings show that coagulation initiation was amplified in RA patients with shortened velocity parameters of clot formation (R, K, α, TMRTG) (Table 3) and these indices were moderately associated with levels of SAA and CRP (Figure 3). 21 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Parameters relating to clot strength (MA, TTG) were attenuated in the RA sample group, but did not statistically differ from healthy controls (Table 3). Thus, although the blood clots form rapidly it leads to a weak clot. The mechanics of clot formation as measured by TEG® have been related to the risk of adverse ischaemic events (111). Excessive hepatic production of fibrinogen is highly prevalent in RA (10). Increased plasma fibrinogen concentration is a strong contributing factor to hypercoagulation (112). Fibrin(ogen) is susceptible to structural and functional modifications by certain inflammatory molecules, including CRP (25), SAA (26), and certain bacterial virulence factors (113-115). Fibrin(ogen) is also prone to post- translational modification that relates to the generation of auto-immunogenicity in RA – the relevance of this process was investigated and is discussed below. Evaluating fibrin gel matrices visually can reveal much about thrombotic potential under inflammatory conditions. Denser fibrin fiber networks are accompanied with increased resistance to fibrinolysis and is associated with the risk for thrombotic events (reviewed by Undas and Ariëns, 2011) (116). These structural properties can be viewed and functionally assessed with a high degree of resolution using SEM (57) Our analysis revealed denser fibrin networks in RA prepared ex vivo PPP clots compared to controls (Figures 2 and 5). Discussion This is consistent with a prothrombotic phenotype observed in previous studies that have inspected the fibrin network in RA (29, 30). Furthermore, we determined a relative measure of fibrin fiber diameter using an image analysis software-based technique (Figure 2). The average diameter of fibrin fibers was larger in RA clots compared to controls (Figure 7). This can be clearly seen in the presented micrographs, were there is an apparent amalgamation of single fibrin fibers in RA preparations (Figures 2 and 5). The appearance of very thin fibrin strands (approximately 100nm in diameter) was not readily detectable in the RA fibrin networks. Some studies have indicated that thin fibrin fibers have higher tensile strength than thicker fibers, concluding that dense networks consisting of predominantly thin fibers are more resistant to degradation (117, 118). Fibrin networks of this nature in RA were observed by Vrancic et al. (2019) (119). However, study by Buclay et al. (2015) concluded that thicker fibers are more resistant to plasmin degradation than thinner fibers, owing to their ability to elongate during lysis (120). It is apparent that our investigation into the structural properties of fibrin networks in RA and its relation to hemostatic function has a rather deceptive 22 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint appearance. Our group has previously shown that fibrin networks in multiple inflammatory conditions (113, 121-123) and in plasma exposed to inflammatory stimuli (26, 114, 124) stain positive for amyloid-specific dyes. This constitutes the appearance of misfolded, protein aggregates with enriched β-sheet content that confers a higher degree of insolubility (125). It may therefore be possible that the fibrin network in RA may undergo a similar transition. This, coupled with impairment of fibrinolytic pathways, presents a possible explanation for high thrombotic risk seen in RA patients. Discussion Distinct protein modifications related to the generation of autoimmunity in RA could present an additional complication when attempting to expound underlying mechanisms responsible for excessive thrombotic risk. Citrullination is a post- translational modification in which positively charged arginine are deiminated by peptidylarginine deiminase (PAD) enzymes to form neutrally charged citrulline (126). PADs are usually localized to intracellular environments and requires calcium for catalysis (127). PADs become active under inflammatory and apoptotic conditions where enzymes are liberated to extracellular spaces and exposed to sufficient calcium concentrations for catalytic function (126). Two isoforms of PAD (PAD2 and PAD4) are primarily responsible for generating citrullinated antigens that incite the generation of anti-citrullinated protein antibodies (ACPAs), a hallmark serological feature of RA (128). Fibrinogen and fibrin are prominent substrates for PAD2 and PAD4 and autoantibodies targeting citrullinated fibrin(ogen) have been identified (40, 43, 68, 129-131). The pathogenicity of citrullinated fibrin(ogen) immune complexes have been demonstrated both in vitro(33) and in vivo(132, 133). Citrullinated fibrin deposits are also common manifestations within synovial cavities, where it contributes to self-perpetuating inflammatory processes (48, 134). Our findings provide novel evidence for the citrullination of fibrin within vasculature which is more prominent in RA plasma compared to control plasma (Figures 6 and 7). Previously the presence of citrullinated fibrinogen could only be detected in RA synovial fluid (41). Later research by Zhao et al. (2008) confirmed the presence of citrullinated fibrinogen containing immune complexes in RA plasma (131). The insolubility of fibrin may increase the likelihood of it being citrullinated in circulation. As we could not stain specifically for citrullinated residues in fibrin only, detected fluorescence could also have originated from citrullinated histone derived from neutrophil 23 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint extracellular traps (135). Nevertheless, binding of ACPAs to fibrin could render it less degradable, by decreasing available binding surface to plasmin (136). Funding We thank the Medical Research Council of South Africa (MRC) (Self-Initiated Research Program: A0X331) for supporting this collaboration. Raw data sharing All raw data is available at: All raw data is available at: https://1drv.ms/u/s!AgoCOmY3bkKHiqIaVlNkRzHB8cUo6Q?e=FusgPo https://1drv.ms/u/s!AgoCOmY3bkKHiqIaVlNkRzHB8cUo6Q?e=FusgPo Discussion CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Consent for publication All authors approved submission of the paper. DECLARATIONS Competing interests The authors declare that they have no competing interests. Concluding remarks Inflammatory and thrombotic processes are highly pertinent to the development of joint disease and cardiovascular complications. There is a need to study changes that occur within synovial environments in unison with simultaneously occurring changes within circulatory tracts. Further investigation into overlapping processes that are crucially involved in the concurrent development of both RA and CVD could reveal improved global disease markers and novel targets for therapeutic intervention. The formation and structure of fibrin clots in RA shows an atypical pattern compared to conventional observations of hypercoagulation under inflammatory conditions. We propose determining if citrullination causes a structural and functional shift in the nature of fibrin to represent an amyloid-like state. This protein modification could potentially contribute to the formation of aberrant fibrin clots in RA patients that confer a higher degree of thrombotic risk. Discussion There remains conjecture as to the effect of citrullination on hemostatic outcome. Citrullination of proteins results in structural unfolding (42) and loss of function (137), which increases its antigenic shelf-life. It has been demonstrated that citrullinated fibrinogen is resistant to thrombin digestion, as preferential epitopes for PADs overlap with thrombin binding sites (45, 46, 138). Despite this, fibrinogenesis in RA is by no means impaired, as evidenced by this study and others. It is plausible that high levels of fibrinogen (10) and thrombin activity (17, 139) in RA has much stronger influence on the fate of fibrinogen than PAD enzymes. There is also evidence that upstream coagulation factors and fibrinolytic components are susceptible to citrullination (140, 141). It is therefore difficult to predict a hemostatic endpoint based on overall citrullination and the effect of citrullination on thrombosis cannot be postulated on singular reactions. The implications that citrullination could have on fibrin, being the end-product of coagulation and a major determinant of thrombotic risk, remains intriguing and should be further investigated. Future investigations would be to determine the effect of citrullination on the thrombotic potential of formed thrombi. Our research has previously implicated the role of amyloid resembling structures in fibrin clots that confer a prothrombotic phenotype. There is evidence to suggest that the processes of citrullination and amyloidogenesis may overlap. Citrullination has been shown to affect the aggregation and oligomerization of β-amyloid proteins (142) and that citrullination of human myelin oligodendrocyte glycoprotein (MOG) can lead to amyloid-like behavior shift that has pathogenic implications for multiple sclerosis (143). This study did present some limitations and challenges. The relatively small sample population complicates the determination of correlative associations between inflammatory and hemostatic indices. The determination of true overall fibrin diameter was also not possible with current techniques – statistical analysis revealed a measurement accuracy of 82%. The detection of citrullination in fibrin networks using fluorescent techniques were unspecific to fibrin(ogen) in this study. However, the current study included this analysis only as a preliminary probe into determining if citrullination of clots was a discerning factor between RA and non-RA individuals. 24 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: oRxiv preprint . FIGURE AND LEGENDS Figure 8: Overview of the overlapping processes of inflammation and coagulation in both synovial and vascular compartments. 1. The chronic and systemic nature of the inflammatory response in RA characterizes the disease as an independent risk factor for CVD. 2. The movement of leukocytes, inflammatory cytokines, procoagulant factors and immune complexes are aided by vascular endothelial dysfunction and neovascularization of hyperproliferative joint tissues. 3. The role of fibrin(ogen) is integral to the formation of hyperplastic and destructive synovial tissue (pannus) and vascular thrombosis, while being a prominent self-protein target of aberrant citrullination and autoimmunogenicity in RA. 25 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: xiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint References 1. Avina-Zubieta JA, Choi HK, Sadatsafavi M, Etminan M, Esdaile JM, Lacaille D. Risk of cardiovascular mortality in patients with rheumatoid arthritis: a meta-analysis of observational studies. Arthritis and rheumatism. 2008;59(12):1690-7. 1. Avina-Zubieta JA, Choi HK, Sadatsafavi M, Etminan M, Esdaile JM, Lacaille D. Risk of cardiovascular mortality in patients with rheumatoid arthritis: a meta-analysis of observational studies. Arthritis and rheumatism. 2008;59(12):1690-7. 2. Avina-Zubieta JA, Thomas J, Sadatsafavi M, Lehman AJ, Lacaille D. Risk of incident cardiovascular events in patients with rheumatoid arthritis: a meta-analysis of observational studies. Ann Rheum Dis. 2012;71(9):1524-9. 3. Lopez-Mejias R, Castaneda S, Gonzalez-Juanatey C, Corrales A, Ferraz-Amaro I, Genre F, et al. Cardiovascular risk assessment in patients with rheumatoid arthritis: The relevance of clinical, genetic and serological markers. Autoimmunity reviews. 2016;15(11):1013-30. 4. del Rincon ID, Williams K, Stern MP, Freeman GL, Escalante A. High incidence of cardiovascular events in a rheumatoid arthritis cohort not explained by traditional cardiac risk factors. Arthritis and rheumatism. 2001;44(12):2737-45. ( ) 5. Foley JH, Conway EM. Cross Talk Pathways Between Coagulation and Inflammation. Circulation research. 2016;118(9):1392-408. 6. Maino A, Rosendaal FR, Algra A, Peyvandi F, Siegerink B. Hypercoagulability Is a Stronger Risk Factor for Ischaemic Stroke than for Myocardial Infarction: A Systematic Review. PloS one. 2015;10(8):e0133523-e. 7. Dessein PH, Joffe BI, Singh S. Biomarkers of endothelial dysfunction, cardiovascular risk factors and atherosclerosis in rheumatoid arthritis. Arthritis research & therapy. 2005;7(3):R634. 8. Wallberg-Jonsson S, Cvetkovic JT, Sundqvist KG, Lefvert AK, Rantapaa-Dahlqvist S. Activation of the immune system and inflammatory activity in relation to markers of atherothrombotic disease and atherosclerosis in rheumatoid arthritis. The Journal of rheumatology. 2002;29(5):875-82. gy ( ) 9. Ingegnoli F, Fantini F, Favalli EG, Soldi A, Griffini S, Galbiati V, et al. Inflammatory and prothrombotic biomarkers in patients with rheumatoid arthritis: effects of tumor necrosis factor-α blockade. Journal of autoimmunity. 2008;31(2):175-9. y ( ) 10. Rooney T, Scherzer R, Shigenaga JK, Graf J, Imboden JB, Grunfeld C. Levels of plasma fibrinogen are elevated in well-controlled rheumatoid arthritis. Rheumatology (Oxford, England). 2011;50(8):1458-65. ( g ) ( ) 11. Bunescu A, Seideman P, Lenkei R, Levin K, Egberg N. Enhanced Fcgamma receptor I, alphaMbeta2 integrin receptor expression by monocytes and neutrophils in rheumatoid arthritis: interaction with platelets. The Journal of rheumatology. 2004;31(12):2347-55. p gy 12. Berckmans RJ, Nieuwland R, Kraan MC, Schaap MC, Pots D, Smeets TJ, et al. Figure 9: Fibrin fiber diameter measurement scheme. SEM micrographs of a prepared PPP clot from a representative healthy control (top) and RA (bottom). CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint Figure 9: Fibrin fiber diameter measurement scheme. SEM micrographs of a prepared PPP clot from a representative healthy control (top) and RA (bottom). Figure 9: Fibrin fiber diameter measurement scheme. SEM micrographs of a prepared PPP clot from a representative healthy control (top) and RA (bottom). Figure 10: Box-and-whisker plots for statistically significant TEG® and V-plex parameters. Boxes represent the median and IQR. Whiskers indicate upper (75th percentile + 1.5IQR) and lower (25th percentile -1.5IQR) extremes. Outlier values are indicated by black dots Figure 11: Lattice of cross-plots and correlation values across statistically significant TEG® and V-Plex measurements. The sample values are indicated by crosses (RA: red, Control: black). Strong intra-analysis correlation exists between CRP with SAA and sICAM-1, between sICAM-1 and VCAM-1, and between all TEG parameters (R, K, Angle, TMRTG). CRP and SAA show moderate correlation with inflammatory and endothelial markers. Figure 12: Scanning electron micrographs of the fibrin network ultrastructure. Representative micrographs of the fibrin network in healthy controls (A) and RA patients (B- D). The altered clot ultrastructure in RA, consisting of less permeable networks of thicker fibrin fibers, represents a prothrombotic phenotype. Figure 13: Confocal microscopy of PPP clots stained with a Citrulline monoclonal antibody. Confocal micrographs of representative control and RA samples. Each row represents identical areas captured of fibrin clot preparations using transmitted light to illustrate fibrin network topography (left-hand column) and a green channel (middle column) to identify fluorescent particles corresponding to citrulline residues. Fluorescent images are also provided in binary (right-hand column) to better illustrate differences seen between RA and control samples with respect to citrulline-coupled fluorescence. Particle analysis confirmed the observable presence of enhanced fluorescent signal in RA samples (n=10) versus healthy controls (n=10). Figure 14: Box-and-whisker plots for micrographs analysis. Outlier values are indicated by black dots. In comparison to healthy individuals, RA subjects exhibited significantly increased fibrin fiber diameters (median: 214nm vs 120nm, OR=22.7) and detected fluorescent areas (median: 3279 vs 1095, OR=2.23) of fibrin citrullination. Table 5: Demographic and clinical characteristics of study participants. Table 6: Vascular injury panel (V-Plex) analysis. Table 6: Vascular injury panel (V-Plex) analysis. Table 7: Thromboelastography® analysis. Table 7: Thromboelastography® analysis. Table 8: Microscopy Image Analysis. Table 8: Microscopy Image Analysis. 26 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint . . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint References Synovial microparticles from arthritic patients modulate chemokine and cytokine release by synoviocytes. Arthritis Res Ther. 2005;7(3):R536-44. y y ; ( ) 13. Targonska-Stepniak B, Majdan M. Serum amyloid A as a marker of persistent inflammation and an indicator of cardiovascular and renal involvement in patients with rheumatoid arthritis. Mediators of inflammation. 2014;2014:793628. 14. Shen C, Sun XG, Liu N, Mu Y, Hong CC, Wei W, et al. Increased serum amyloid A and its association with autoantibodies, acute phase reactants and disease activity in patients with rheumatoid arthritis. Molecular medicine reports. 2015;11(2):1528-34. 15. So AK, Varisco PA, Kemkes-Matthes B, Herkenne-Morard C, Chobaz-Peclat V, Gerster JC, et al. Arthritis is linked to local and systemic activation of coagulation and fibrinolysis pathways. Journal of thrombosis and haemostasis : JTH. 2003;1(12):2510-5. 16. McEntegart A, Capell HA, Creran D, Rumley A, Woodward M, Lowe GD. Cardiovascular risk factors, including thrombotic variables, in a population with rheumatoid arthritis. Rheumatology (Oxford, England). 2001;40(6):640-4. 17. Undas A, Gissel M, Kwasny-Krochin B, Gluszko P, Mann KG, Brummel-Ziedins KE. Thrombin generation in rheumatoid arthritis: dependence on plasma factor composition. Thrombosis and haemostasis. 2010;104(2):224-30. 27 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 18. Daza L, Aguirre M, Jimenez M, Herrera R, Bollain JJ. Common carotid intima-media thickness and von Willebrand factor serum levels in rheumatoid arthritis female patients without cardiovascular risk factors. Clin Rheumatol. 2007;26(4):533-7. 19. Beinsberger J, Heemskerk JW, Cosemans JM, editors. Chronic arthritis and cardiovascular disease: altered blood parameters give rise to a prothrombotic propensity. Seminars in arthritis and rheumatism; 2014: Elsevier. 20. Södergren A, Karp K, Bengtsson C, Möller B, Rantapää-Dahlqvist S, Wållberg- Jonsson S. Biomarkers associated with cardiovascular disease in patients with early rheumatoid arthritis. PloS one. 2019;14(8). ( ) 21. Maradit-Kremers H, Nicola PJ, Crowson CS, Ballman KV, Gabriel SE. Cardiovascular death in rheumatoid arthritis: a population-based study. Arthritis and rheumatism. 2005;52(3):722-32. References ( ) 22. Graf J, Scherzer R, Grunfeld C, Imboden J. Levels of C-reactive protein associated with high and very high cardiovascular risk are prevalent in patients with rheumatoid arthritis. PLoS One. 2009;4(7):e6242. 23. del Rincon I, Freeman GL, Haas RW, O'Leary DH, Escalante A. Relative contribution of cardiovascular risk factors and rheumatoid arthritis clinical manifestations to atherosclerosis. Arthritis and rheumatism. 2005;52(11):3413-23. ( ) 24. Litvinov RI, Weisel JW. What Is the Biological and Clinical Relevance of Fibrin? Seminars in thrombosis and hemostasis. 2016;42(4):333-43. 25. Salonen EM, Vartio T, Hedman K, Vaheri A. Binding of fibronectin by the acute phase reactant C-reactive protein. The Journal of biological chemistry. 1984;259(3):1496- 501. 26. Page MJ, Thomson GJ, Nunes JM, Engelbrecht A-M, Nell TA, de Villiers WJ, et al. Serum amyloid A binds to fibrin (ogen), promoting fibrin amyloid formation. Scientific reports. 2019;9(1):1-14. ; ( ) 27. Bester J, Pretorius E. Effects of IL-1β, IL-6 and IL-8 on erythrocytes, platelets and clot viscoelasticity. Scientific reports. 2016;6(1):1-10. 28. Page MJ, Bester J, Pretorius E. The inflammatory effects of TNF-α and complement component 3 on coagulation. Scientific reports. 2018;8(1):1-9. p g p ; ( ) 29. Kwasny-Krochin B, Gluszko P, Undas A. Unfavorably altered fibrin clot properties in patients with active rheumatoid arthritis. Thromb Res. 2010;126(1):e11-6. 30. Pretorius E, Oberholzer HM, van der Spuy WJ, Swanepoel AC, Soma P. Scanning electron microscopy of fibrin networks in rheumatoid arthritis: a qualitative analysis. Rheumatology international. 2012;32(6):1611-5. 31. Busso N, Hamilton JA. Extravascular coagulation and the plasminogen activator/plasmin system in rheumatoid arthritis. Arthritis and rheumatism. 2002;46(9):2268- 79. 32. Smiley ST, King JA, Hancock WW. Fibrinogen stimulates macrophage chemokine secretion through toll-like receptor 4. Journal of immunology (Baltimore, Md : 1950). 2001;167(5):2887-94. ; ( ) 33. Sokolove J, Zhao X, Chandra PE, Robinson WH. Immune complexes containing citrullinated fibrinogen costimulate macrophages via Toll-like receptor 4 and Fcgamma receptor. Arthritis and rheumatism. 2011;63(1):53-62. p ( ) 34. Rubel C, Gomez S, Fernandez GC, Isturiz MA, Caamano J, Palermo MS. Fibrinogen-CD11b/CD18 interaction activates the NF-kappa B pathway and delays apoptosis in human neutrophils. European journal of immunology. 2003;33(5):1429-38. p p p p j gy 35. Perez RL, Roman J. Fibrin enhances the expression of IL-1 beta by human peripheral blood mononuclear cells. Implications in pulmonary inflammation. Journal of immunology (Baltimore, Md : 1950). 1995;154(4):1879-87. gy ( ) ( ) 36. Perez RL, Ritzenthaler JD, Roman J. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint References Transcriptional regulation of the interleukin- 1beta promoter via fibrinogen engagement of the CD18 integrin receptor. American journal of respiratory cell and molecular biology. 1999;20(5):1059-66. 28 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 37. Blachere NE, Parveen S, Frank MO, Dill BD, Molina H, Orange DE. High-Titer Rheumatoid Arthritis Antibodies Preferentially Bind Fibrinogen Citrullinated by Peptidylarginine Deiminase 4. Arthritis & rheumatology (Hoboken, NJ). 2017;69(5):986-95. 38. Raijmakers R, van Beers JJ, El-Azzouny M, Visser NF, Bozic B, Pruijn GJ, et al. Elevated levels of fibrinogen-derived endogenous citrullinated peptides in synovial fluid of rheumatoid arthritis patients. Arthritis Res Ther. 2012;14(3):R114. 39. Zhao X, Okeke NL, Sharpe O, Batliwalla FM, Lee AT, Ho PP, et al. Circulating immune complexes contain citrullinated fibrinogen in rheumatoid arthritis. Arthritis Res Ther. 2008;10(4):R94. ( ) 40. Hill JA, Al-Bishri J, Gladman DD, Cairns E, Bell DA. Serum autoantibodies that bind citrullinated fibrinogen are frequently found in patients with rheumatoid arthritis. The Journal of rheumatology. 2006;33(11):2115-9. gy ( ) 41. Takizawa Y, Suzuki A, Sawada T, Ohsaka M, Inoue T, Yamada R, et al. Citrullinated fibrinogen detected as a soluble citrullinated autoantigen in rheumatoid arthritis synovial fluids. Ann Rheum Dis. 2006;65(8):1013-20. ( ) 42. Tarcsa E, Marekov LN, Mei G, Melino G, Lee SC, Steinert PM. Protein unfolding by peptidylarginine deiminase. Substrate specificity and structural relationships of the natural substrates trichohyalin and filaggrin. The Journal of biological chemistry. 1996;271(48):30709-16. 43. Masson-Bessiere C, Sebbag M, Girbal-Neuhauser E, Nogueira L, Vincent C, Senshu T, et al. The major synovial targets of the rheumatoid arthritis-specific antifilaggrin autoantibodies are deiminated forms of the alpha- and beta-chains of fibrin. Journal of immunology (Baltimore, Md : 1950). 2001;166(6):4177-84. gy ( ) ( ) 44. Damiana T, Damgaard D, Sidelmann JJ, Nielsen CH, de Maat MP, Münster A-MB, et al. Citrullination of fibrinogen by peptidylarginine deiminase 2 impairs fibrin clot structure. References It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 56. Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham CO, 3rd, et al. 2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Ann Rheum Dis. 2010;69(9):1580-8. 56. g gy p eague Against Rheumatism collaborative initiative. Ann Rheum Dis. 2010;69(9):1580-8 57. Pretorius E, Swanepoel AC, DeVilliers S, Bester J. Blood clot parameters: Thromboelastography and scanning electron microscopy in research and clinical practice. Thromb Res. 2017;154:59-63. 58. Alamanos Y, Drosos AA. Epidemiology of adult rheumatoid arthritis. Autoimmunity reviews. 2005;4(3):130-6. 59. Aletaha D, Ward MM, Machold KP, Nell VP, Stamm T, Smolen JS. Remission and active disease in rheumatoid arthritis: defining criteria for disease activity states. Arthritis and rheumatism. 2005;52(9):2625-36. ( ) 60. Turk SM, Cansu DU, Teke HU, Kasifoglu T, Meltem Akay O, Bilgin M, et al. Can we predict thrombotic tendency in rheumatoid arthritis? A thromboelastographic analysis (with ROTEM). Clin Rheumatol. 2018;37(9):2341-9. ) ( ) 61. Wallberg-Jonsson S, Ohman M, Rantapaa-Dahlqvist S. Which factors are related to the presence of atherosclerosis in rheumatoid arthritis? Scandinavian journal of rheumatology. 2004;33(6):373-9. gy 62. Jonsson SW, Backman C, Johnson O, Karp K, Lundstrom E, Sundqvist KG, et al. Increased prevalence of atherosclerosis in patients with medium term rheumatoid arthritis. The Journal of rheumatology. 2001;28(12):2597-602. gy ( ) 63. Motohashi Y, Nakazawa T. [Blood coagulation, fibrinolysis and various values of Lansbury's activity index in rheumatoid arthritis patients]. Arerugi = [Allergy]. 1992;41(2 Pt 1):148-55. ) 64. Motohashi Y, Nakazawa T, Kojima A. [Analysis of blood coagulation and fibrinolysis in rheumatoid arthritis using the urokinase activated thromboelastography]. Ryumachi [Rheumatism]. 1992;32(2):123-30. 65. Gong H, Shi C, Zhou Z, Teng J, Sun Y, Yang C, et al. Evaluating hypercoagulability in new-onset systemic lupus erythematosus patients using thromboelastography. Journal of clinical laboratory analysis. 2019:e23157. y y 66. Collins KS, Balasubramaniam K, Viswanathan G, Natasari A, Tarn J, Lendrem D, et al. Assessment of blood clot formation and platelet receptor function ex vivo in patients with primary Sjogren's syndrome. BMJ open. 2013;3(5). p y j g y p ( ) 67. Yasar Bilge NS, Akay OM, Kasifoglu T, Kus G, Korkmaz C. The role of hemostatic mechanisms in the development of thrombosis in Behcet's disease: an analysis by modified rotation thromboelastogram (ROTEM). Clin Rheumatol. 2013;32(12):1815-8. 68. References Clinica Chimica Acta. 2020;501:6-11. 45. Okumura N, Haneishi A, Terasawa F. Citrullinated fibrinogen shows defects in FPA and FPB release and fibrin polymerization catalyzed by thrombin. Clinica chimica acta; international journal of clinical chemistry. 2009;401(1-2):119-23. j y ; ( ) 46. Nakayama-Hamada M, Suzuki A, Furukawa H, Yamada R, Yamamoto K. Citrullinated fibrinogen inhibits thrombin-catalysed fibrin polymerization. Journal of biochemistry. 2008;144(3):393-8. y ( ) 47. Bisoendial RJ, Levi M, Tak PP, Stroes ES. The prothrombotic state in rheumatoid arthritis: an additive risk factor for adverse cardiovascular events. Seminars in thrombosis and hemostasis. 2010;36(4):452-7. 48. Sanchez-Pernaute O, Largo R, Calvo E, Alvarez-Soria MA, Egido J, Herrero- Beaumont G. A fibrin based model for rheumatoid synovitis. Ann Rheum Dis. 2003;62(12):1135-8. ( ) 49. Murdaca G, Colombo BM, Cagnati P, Gulli R, Spanò F, Puppo F. Endothelial dysfunction in rheumatic autoimmune diseases. Atherosclerosis. 2012;224(2):309-17. 50. Hoppe B, Dörner T. Coagulation and the fibrin network in rheumatic disease beyond haemostasis. Nature Reviews Rheumatology. 2012;8(12):738-46. 50. Hoppe B, Dörner T. Coagulation and the fibrin network in rheumatic disease: a role beyond haemostasis. Nature Reviews Rheumatology. 2012;8(12):738-46. 51. Elshabrawy HA, Chen Z, Volin MV, Ravella S, Virupannavar S, Shahrara S. The pathogenic role of angiogenesis in rheumatoid arthritis. Angiogenesis. 2015;18(4):433-48. 51. Elshabrawy HA, Chen Z, Volin MV, Ravella S, Virupannavar S, Shahrara pathogenic role of angiogenesis in rheumatoid arthritis. Angiogenesis. 2015;18(4):433 52. Veale DJ, Maple C. Cell adhesion molecules in rheumatoid arthritis. Drugs & aging. 1996;9(2):87-92. 53. Paleolog EM. Angiogenesis in rheumatoid arthritis. Arthritis research. 2002;4 Suppl 3:S81-90. 54. Myasoedova E, Chandran A, Ilhan B, Major BT, Michet CJ, Matteson EL, et al. The role of rheumatoid arthritis (RA) flare and cumulative burden of RA severity in the risk of cardiovascular disease. Ann Rheum Dis. 2016;75(3):560-5. ( ) 55. Book C, Saxne T, Jacobsson LT. Prediction of mortality in rheumatoid arthritis based on disease activity markers. The Journal of rheumatology. 2005;32(3):430-4. 29 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. References Damgaard D, Bawadekar M, Senolt L, Stensballe A, Shelef MA, Nielsen CH. Relative efficiencies of peptidylarginine deiminase 2 and 4 in generating target sites for anti- citrullinated protein antibodies in fibrinogen, alpha-enolase and histone H3. PLoS One. 2018;13(8):e0203214. ( ) 69. Rantapaa-Dahlqvist S, de Jong BA, Berglin E, Hallmans G, Wadell G, Stenlund H, et al. Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development of rheumatoid arthritis. Arthritis and rheumatism. 2003;48(10):2741-9. 70. Nielen MM, van Schaardenburg D, Reesink HW, Van de Stadt RJ, van der HorstBruinsma IE, de Koning MH, et al. Specific autoantibodies precede the symptoms of rheumatoid arthritis: a study of serial measurements in blood donors. Arthritis & Rheumatism: Official Journal of the American College of Rheumatology. 2004;50(2):380-6. g gy ( ) 71. Dessein PH, Joffe BI, Stanwix AE. High sensitivity C-reactive protein as a disease activity marker in rheumatoid arthritis. The Journal of rheumatology. 2004;31(6):1095-7. 72. Zhang J, Chen L, Delzell E, Muntner P, Hillegass WB, Safford MM, et al. The association between inflammatory markers, serum lipids and the risk of cardiovascular events in patients with rheumatoid arthritis. Ann Rheum Dis. 2014;73(7):1301-8. p ; ( ) Fay WP. Linking inflammation and thrombosis: Role of C-reactive protein. World of cardiology. 2010;2(11):365. j gy ( ) 74. Badimon L, Pena E, Arderiu G, Padro T, Slevin M, Vilahur G, et al. C-Reactive Protein in Atherothrombosis and Angiogenesis. Front Immunol. 2018;9:430. 30 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 75. Cermak J, Key NS, Bach RR, Balla J, Jacob HS, Vercellotti GM. C-reactive protein induces human peripheral blood monocytes to synthesize tissue factor. Blood. 1993;82(2):513-20. ; ( ) 76. Cirillo P, Golino P, Calabro P, Cali G, Ragni M, De Rosa S, et al. C-reactive protein induces tissue factor expression and promotes smooth muscle and endothelial cell proliferation. Cardiovascular research. 2005;68(1):47-55. References p ( ) 77. Cai H, Song C, Endoh I, Goyette J, Jessup W, Freedman SB, et al. Serum amyloid A induces monocyte tissue factor. Journal of immunology (Baltimore, Md : 1950). 2007;178(3):1852-60. 78. Cai X, Freedman SB, Witting PK. Serum amyloid A stimulates cultured endothelial cells to migrate and proliferate: inhibition by the multikinase inhibitor BIBF1120. Clinical and experimental pharmacology & physiology. 2013;40(9):662-70. 79. Song C, Shen Y, Yamen E, Hsu K, Yan W, Witting PK, et al. Serum amyloid A may potentiate prothrombotic and proinflammatory events in acute coronary syndromes. Atherosclerosis. 2009;202(2):596-604. ( ) 80. Witting PK, Song C, Hsu K, Hua S, Parry SN, Aran R, et al. The acute-phase protein serum amyloid A induces endothelial dysfunction that is inhibited by high-density lipoprotein. Free radical biology & medicine. 2011;51(7):1390-8. gy ( ) 81. Zhao Y, Zhou S, Heng CK. Impact of serum amyloid A on tissue factor and tissue factor pathway inhibitor expression and activity in endothelial cells. Arteriosclerosis, thrombosis, and vascular biology. 2007;27(7):1645-50. gy ( ) 82. Ji Y, Fish PM, Strawn TL, Lohman AW, Wu J, Szalai AJ, et al. C-reactive protein induces expression of tissue factor and plasminogen activator inhibitor-1 and promotes fibrin accumulation in vein grafts. Journal of thrombosis and haemostasis : JTH. 2014;12(10):1667-77. ( ) 83. Wu J, Stevenson MJ, Brown JM, Grunz EA, Strawn TL, Fay WP. C-reactive protein enhances tissue factor expression by vascular smooth muscle cells: mechanisms and in vivo significance. Arteriosclerosis, thrombosis, and vascular biology. 2008;28(4):698-704. g gy ( ) 84. Pasceri V, Willerson JT, Yeh ET. Direct proinflammatory effect of C-reactive protein on human endothelial cells. Circulation. 2000;102(18):2165-8. 85. Verma S, Li SH, Badiwala MV, Weisel RD, Fedak PW, Li RK, et al. Endothelin antagonism and interleukin-6 inhibition attenuate the proatherogenic effects of C-reactive protein. Circulation. 2002;105(16):1890-6. p ( ) 86. van den Berg CW, Taylor KE. Letter in response to Bisoendial et al: "activation of inflammation and coagulation after infusion of C-reactive protein in humans". Circulation research. 2005;97(1):e2. 87. Papa R, Lachmann HJ. Secondary, AA, Amyloidosis. Rheumatic diseases clinics of North America. 2018;44(4):585-603. 88. Koivuniemi R, Paimela L, Suomalainen R, Tornroth T, Leirisalo-Repo M. Amyloidosis is frequently undetected in patients with rheumatoid arthritis. Amyloid : the international journal of experimental and clinical investigation : the official journal of the International Society of Amyloidosis. 2008;15(4):262-8. y y ( ) 89. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whi this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint References Koivuniemi R, Paimela L, Suomalainen R, Leirisalo-Repo M. Amyloidosis as a cause of death in patients with rheumatoid arthritis. Clinical and experimental rheumatology. 2008;26(3):408-13. 90. Tanaka F, Migita K, Kawabe Y, Aoyagi T, Ida H, Kawakami A, et al. Interleukin-18 induces serum amyloid A (SAA) protein production from rheumatoid synovial fibroblasts. Life sciences. 2004;74(13):1671-9. ; ( ) 91. O'Hara R, Murphy EP, Whitehead AS, FitzGerald O, Bresnihan B. Acute-phase serum amyloid A production by rheumatoid arthritis synovial tissue. Arthritis research. 2000;2(2):142-4. 92. O'Hara R, Murphy EP, Whitehead AS, FitzGerald O, Bresnihan B. Local expression of the serum amyloid A and formyl peptide receptor-like 1 genes in synovial tissue is associated with matrix metalloproteinase production in patients with inflammatory arthritis. Arthritis and rheumatism. 2004;50(6):1788-99. 31 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 93. Kumon Y, Suehiro T, Hashimoto K, Nakatani K, Sipe JD. Local expression of acute phase serum amyloid A mRNA in rheumatoid arthritis synovial tissue and cells. The Journal of rheumatology. 1999;26(4):785-90. 93. gy ( ) 94. Momohara S, Okamoto H, Yamanaka H. Chondrocyte of rheumatoid arthritis serve as a source of intra-articular acute-phase serum amyloid A protein. Clinica chimica acta; international journal of clinical chemistry. 2008;398(1-2):155-6. 95. Mullan RH, Bresnihan B, Golden-Mason L, Markham T, O'Hara R, FitzGerald O, et al. Acute-phase serum amyloid A stimulation of angiogenesis, leukocyte recruitment, and matrix degradation in rheumatoid arthritis through an NF-kappaB-dependent signal transduction pathway. Arthritis and rheumatism. 2006;54(1):105-14. 96. Connolly M, Rooney PR, McGarry T, Maratha AX, McCormick J, Miggin SM, et al. References Acute serum amyloid A is an endogenous TLR2 ligand that mediates inflammatory and angiogenic mechanisms. Ann Rheum Dis. 2016;75(7):1392-8. g g ( ) 97. Dessein PH, Joffe BI, Singh S. Biomarkers of endothelial dysfunction, cardiovascular risk factors and atherosclerosis in rheumatoid arthritis. Arthritis Res Ther. 2005;7(3):R634- 43. 98. Tsakadze NL, Zhao Z, D'Souza SE. Interactions of intercellular adhesion molecule-1 with fibrinogen. Trends in cardiovascular medicine. 2002;12(3):101-8. 99. Sriramarao P, Languino LR, Altieri DC. Fibrinogen mediates leukocyte-endothelium bridging in vivo at low shear forces. Blood. 1996;88(9):3416-23. 100. Sans E, Delachanal E, Duperray A. Analysis of the roles of ICAM-1 in neutrophil transmigration using a reconstituted mammalian cell expression model: implication of ICAM- 1 cytoplasmic domain and Rho-dependent signaling pathway. Journal of immunology (Baltimore, Md : 1950). 2001;166(1):544-51. ( ) ( ) 101. Languino LR, Plescia J, Duperray A, Brian AA, Plow EF, Geltosky JE, et al. Fibrinogen mediates leukocyte adhesion to vascular endothelium through an ICAM-1- dependent pathway. Cell. 1993;73(7):1423-34. 102. Bombeli T, Schwartz BR, Harlan JM. Adhesion of activated platelets to endothelial cells: evidence for a GPIIbIIIa-dependent bridging mechanism and novel roles for endothelial intercellular adhesion molecule 1 (ICAM-1), alphavbeta3 integrin, and GPIbalpha. The Journal of experimental medicine. 1998;187(3):329-39. p ( ) 103. Massberg S, Enders G, Matos FC, Tomic LI, Leiderer R, Eisenmenger S, et al. Fibrinogen deposition at the postischemic vessel wall promotes platelet adhesion during ischemia-reperfusion in vivo. Blood. 1999;94(11):3829-38. p ; ( ) 104. Khandoga A, Biberthaler P, Enders G, Axmann S, Hutter J, Messmer K, et al. Platelet adhesion mediated by fibrinogen-intercelllular adhesion molecule-1 binding induces tissue injury in the postischemic liver in vivo. Transplantation. 2002;74(5):681-8. 105. Lominadze D, Tsakadze N, Sen U, Falcone JC, D'Souza SE. Fibrinogen and fragment D-induced vascular constriction. American journal of physiology Heart and circulatory physiology. 2005;288(3):H1257-64. y p y gy ( ) 106. Sen U, Tyagi N, Patibandla PK, Dean WL, Tyagi SC, Roberts AM, et al. Fibrinogen- induced endothelin-1 production from endothelial cells. American journal of physiology Cell physiology. 2009;296(4):C840-7. p y gy 107. Schmid E, Muller TH, Budzinski RM, Binder K, Pfizenmaier K. Signaling by E-selectin and ICAM-1 induces endothelial tissue factor production via autocrine secretion of platelet- activating factor and tumor necrosis factor alpha. Journal of interferon & cytokine research : the official journal of the International Society for Interferon and Cytokine Research. 1995;15(9):819-25. 108. References Cybulsky MI, Iiyama K, Li H, Zhu S, Chen M, Iiyama M, et al. A major role for VCAM- 1, but not ICAM-1, in early atherosclerosis. The Journal of clinical investigation. 2001;107(10):1255-62. 109. Elices MJ, Osborn L, Takada Y, Crouse C, Luhowskyj S, Hemler ME, et al. VCAM-1 on activated endothelium interacts with the leukocyte integrin VLA-4 at a site distinct from the VLA-4/fibronectin binding site. Cell. 1990;60(4):577-84. 32 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 111. Yao X, Dong Q, Song Y, Wang Y, Deng Y, Li Y. Thrombelastography maximal clot strength could predict one-year functional outcome in patients with ischemic stroke. Cerebrovascular diseases (Basel, Switzerland). 2014;38(3):182-90. ( , ) ; ( ) 112. Harr JN, Moore EE, Ghasabyan A, Chin TL, Sauaia A, Banerjee A, et al. Functional fibrinogen assay indicates that fibrinogen is critical in correcting abnormal clot strength following trauma. Shock (Augusta, Ga). 2013;39(1):45-9. 113. de Waal GM, Engelbrecht L, Davis T, de Villiers WJS, Kell DB, Pretorius E. Correlative Light-Electron Microscopy detects lipopolysaccharide and its association with fibrin fibres in Parkinson's Disease, Alzheimer's Disease and Type 2 Diabetes Mellitus. Sci Rep. 2018;8(1):16798. p ( ) 114. Pretorius E, Page MJ, Hendricks L, Nkosi NB, Benson SR, Kell DB. Both lipopolysaccharide and lipoteichoic acids potently induce anomalous fibrin amyloid formation: assessment with novel Amytracker stains. Journal of the Royal Society, Interface. 2018;15(139). 115. Adams B, Nunes JM, Page MJ, Roberts T, Carr J, Nell TA, et al. Parkinson's Disease: A Systemic Inflammatory Disease Accompanied by Bacterial Inflammagens. Front Aging Neurosci. 2019;11:210-. 116. Undas A, Ariens RA. Fibrin clot structure and function: a role in the pathophysiology of arterial and venous thromboembolic diseases. Arteriosclerosis, thrombosis, and vascular biology. 2011;31(12):e88-99. gy ; ( ) 117. Li W, Sigley J, Pieters M, Helms CC, Nagaswami C, Weisel JW, et al. References CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: bioRxiv preprint 129. Robinson WH, Sokolove J. Citrullination of fibrinogen: generation of neoepitopes and enhancement of immunostimulatory properties. Arthritis Research & Therapy. 2012;14(Suppl 1):O30-O. ) 130. Blachère NE, Parveen S, Frank MO, Dill BD, Molina H, Orange DE. High-Titer Rheumatoid Arthritis Antibodies Preferentially Bind Fibrinogen Citrullinated by Peptidylarginine Deiminase 4. Arthritis & rheumatology (Hoboken, NJ). 2017;69(5):986-95. p y g gy ( ) ( ) 131. Zhao X, Okeke NL, Sharpe O, Batliwalla FM, Lee AT, Ho PP, et al. Circulating immune complexes contain citrullinated fibrinogen in rheumatoid arthritis. Arthritis research & therapy. 2008;10(4):R94-R. py 132. Yue D, Brintnell W, Mannik LA, Christie DA, Haeryfar SM, Madrenas J, et al. CTLA- 4Ig blocks the development and progression of citrullinated fibrinogen-induced arthritis in DR4-transgenic mice. Arthritis and rheumatism. 2010;62(10):2941-52. g ( ) 133. Cordova KN, Willis VC, Haskins K, Holers VM. A citrullinated fibrinogen-specific T cell line enhances autoimmune arthritis in a mouse model of rheumatoid arthritis. Journal of immunology (Baltimore, Md : 1950). 2013;190(4):1457-65. gy ( ) ( ) 134. Sanchez-Pernaute O, Filkova M, Gabucio A, Klein M, Maciejewska-Rodrigues H, Ospelt C, et al. Citrullination enhances the pro-inflammatory response to fibrin in rheumatoid arthritis synovial fibroblasts. Ann Rheum Dis. 2013;72(8):1400-6. y ( ) 135. Varju I, Kolev K. Networks that stop the flow: A fresh look at fibrin and neutrophil extracellular traps. Thromb Res. 2019;182:1-11. p 136. Sebbag M, Moinard N, Auger I, Clavel C, Arnaud J, Nogueira L, et al. Epitopes of human fibrin recognized by the rheumatoid arthritis-specific autoantibodies to citrullinated proteins. European journal of immunology. 2006;36(8):2250-63. 137. Travers TS, Harlow L, Rosas IO, Gochuico BR, Mikuls TR, Bhattacharya SK, et al. Extensive Citrullination Promotes Immunogenicity of HSP90 through Protein Unfolding and Exposure of Cryptic Epitopes. Journal of immunology (Baltimore, Md : 1950). 2016;197(5):1926-36. 138. Damiana T, Damgaard D, Sidelmann JJ, Nielsen CH, de Maat MPM, Munster AB, et al. Citrullination of fibrinogen by peptidylarginine deiminase 2 impairs fibrin clot structure. Clinica chimica acta; international journal of clinical chemistry. 2020;501:6-11. ; j y ; 139. Kern A, Balog A, Dulic S, Barabas E, Kiszelak M, Vasarhelyi B. References Fibrin Fiber Stiffness Is Strongly Affected by Fiber Diameter, but Not by Fibrinogen Glycation. Biophysical journal. 2016;110(6):1400-10. p y j ( ) 118. Undas A. Fibrin clot properties and their modulation in thrombotic disorders. Thrombosis and haemostasis. 2014;112(1):32-42. 119. Vranic A, Pruner I, Veselinovic M, Soutari N, Petkovic A, Jakovljevic V, et al. Assessment of hemostatic disturbances in women with established rheumatoid arthritis. Clin Rheumatol. 2019;38(11):3005-14. ; ( ) 120. Bucay I, O'Brien ET, 3rd, Wulfe SD, Superfine R, Wolberg AS, Falvo MR, et al. Physical determinants of fibrinolysis in single fibrin fibers. PLoS One. 2015;10(2):e0116350. 121. de Villiers S, Bester J, Kell DB, Pretorius E. A Possible Role of Amyloidogenic Blood Clotting in the Evolving Haemodynamics of Female Migraine-With-Aura: Results From a Pilot Study. Front Neurol. 2019;10:1262-. y 122. Pretorius E, Bester J, Page MJ, Kell DB. The Potential of LPS-Binding Protein to Reverse Amyloid Formation in Plasma Fibrin of Individuals With Alzheimer-Type Dementia. Front Aging Neurosci. 2018;10:257. g g ; 123. Pretorius E, Page MJ, Engelbrecht L, Ellis GC, Kell DB. Substantial fibrin amyloidogenesis in type 2 diabetes assessed using amyloid-selective fluorescent stains. Cardiovasc Diabetol. 2017;16(1):141. ( ) 124. Pretorius E, Mbotwe S, Bester J, Robinson CJ, Kell DB. Acute induction of anomalous and amyloidogenic blood clotting by molecular amplification of highly substoichiometric levels of bacterial lipopolysaccharide. Journal of the Royal Society, Interface. 2016;13(122):20160539. 125. Rambaran RN, Serpell LC. Amyloid fibrils: abnormal protein assembly. Prion. 2008;2(3):112-7. 126. Gyorgy B, Toth E, Tarcsa E, Falus A, Buzas EI. Citrullination: a posttranslational modification in health and disease. Int J Biochem Cell Biol. 2006;38(10):1662-77. 127. Fujisaki M, Sugawara K. Properties of peptidylarginine deiminase from the epidermis of newborn rats. The Journal of Biochemistry. 1981;89(1):257-63. 128. Foulquier C, Sebbag M, Clavel C, Chapuy-Regaud S, Al Badine R, Mechin MC, et al. Peptidyl arginine deiminase type 2 (PAD-2) and PAD-4 but not PAD-1, PAD-3, and PAD-6 are expressed in rheumatoid arthritis synovium in close association with tissue inflammation. Arthritis and rheumatism. 2007;56(11):3541-53. 33 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted May 30, 2020. ; https://doi.org/10.1101/2020.05.28.121301 doi: Rxiv preprint . References Alterations of the thrombin generation profile in rheumatoid arthritis. J Thromb Thrombolysis. 2016;41(3):359- 64. 140. Chang X, Yamada R, Sawada T, Suzuki A, Kochi Y, Yamamoto K. The inhibition of antithrombin by peptidylarginine deiminase 4 may contribute to pathogenesis of rheumatoid arthritis. Rheumatology (Oxford, England). 2005;44(3):293-8. gy ( g ) ( ) 141. Tilvawala R, Nguyen SH, Maurais AJ, Nemmara VV, Nagar M, Salinger AJ, et al. The Rheumatoid Arthritis-Associated Citrullinome. Cell chemical biology. 2018;25(6):691-704.e6. 142. Osaki D, Hiramatsu H. Citrullination and deamidation affect aggregation properties of amyloid beta-proteins. Amyloid : the international journal of experimental and clinical investigation : the official journal of the International Society of Amyloidosis. 2016;23(4):234- 41. 143. Araman C, van Gent ME, Meeuwenoord NJ, Heijmans N, Marqvorsen MHS, Doelman W, et al. Amyloid-like Behavior of Site-Specifically Citrullinated Myelin Oligodendrocyte Protein (MOG) Peptide Fragments inside EBV-Infected B-Cells Influences Their Cytotoxicity and Autoimmunogenicity. Biochemistry. 2019;58(6):763-75. 34
https://openalex.org/W2971362240	https://ecommons.aku.edu/cgi/viewcontent.cgi?article=1824&context=pakistan_fhs_mc_surg_surg	English	null	Incidence of Intraocular Lens Exchange after Cataract Surgery	Scientific reports	2,019	cc-by	6,764	Authors Authors Maram E. A. Abdalla Elsayed, Khabir Ahmad, Abdulelah A. Al-Abdullah, Rizwan Malik, Rajiv Khandekar, Hernan Martinez-Osorio, Marco Mura, and Patrik Schatz Authors Authors Maram E. A. Abdalla Elsayed, Khabir Ahmad, Abdulelah A. Al-Abdullah, Rizwan Malik, Rajiv Khandekar, Hernan Martinez-Osorio, Marco Mura, and Patrik Schatz eCommons@AKU eCommons@AKU Department of Surgery Department of Surgery 9-9-2019 Incidence of intraocular lens exchange after cataract surgery Incidence of intraocular lens exchange after cataract surgery Maram E. A. Abdalla Elsayed Khabir Ahmad Abdulelah A. Al-Abdullah Rizwan Malik Rajiv Khandekar See next page for additional authors Follow this and additional works at: https://ecommons.aku.edu/pakistan_fhs_mc_surg_surg Part of the Ophthalmology Commons, and the Surgery Commons eCommons@AKU eCommons@AKU Department of Surgery Department of Surgery Department of Surgery Maram E. A. Abdalla Elsayed Khabir Ahmad Abdulelah A. Al-Abdullah Rizwan Malik Rajiv Khandekar See next page for additional authors ow this and additional works at: https://ecommons.aku.edu/pakistan_fhs_mc_surg_surg Part of the Ophthalmology Commons, and the Surgery Commons OPEN Received: 12 April 2019 Accepted: 16 August 2019 Published: xx xx xxxx Intraocular lens (IOL) exchange after cataract surgery is unusual but may be associated with suboptimal visual outcome. The incidence of IOL exchange has not been consistently estimated. Such information is invaluable when counseling patients prior to cataract surgery. We examined the incidence of, and indications and risk factors for, IOL exchange after cataract surgery. We also assessed visual outcome of eyes that had an IOL exchange. A cohort design was used to estimate the incidence of IOL exchange and a case-control design to identify factors associated with it. All phacoemulsification surgeries with IOL (n = 17415 eyes) during 2010–2017 and those that had a subsequent IOL removal or replacement during the same time period were identified (n = 34 eyes). The incidence of IOL exchange was 2 per 1000 surgeries (95% confidence interval [CI] 1 to 3) over 8 years. Eyes that underwent subsequent IOL removal or replacement were compared with eyes that had cataract surgery only (n = 47) across demographic and clinical characteristics. In a binary logistic regression analysis, two factors were significantly associated with IOL exchange/removal: an adverse event during cataract surgery (adjusted odds ratio [aOR] 19.45; 95% CI 4.89–77.30, P < 0.001) and a pre-existing ocular comorbidity (aOR 10.70; 95% CI 1.69–67.63, P = 0.021). The effect of gender was marginally significant (P = 0.077). Eyes that underwent IOL exchange or explantation were nearly two and a half times more likely to have a final best-corrected visual acuity of <20/60 compared to those that had cataract surgery alone (adjusted RR 2.60 95% CI, 1.13–6.02; P = 0.025). Age-related cataract remains the leading cause of blindness worldwide, despite the fact that cataract surgery is one of the most frequently performed and cost effective surgical interventions to date1–5. The outcomes of cataract surgery are generally very good. However, potentially vision-threatening complications may occur in a small pro- portion of patients. Most of these complications (estimated incidence) have been extensively studied including endophthalmitis (0.05–0.15%)6, retinal detachment (0.39%)7, and expulsive hemorrhage (0.05–0.1%)8. Seemingly less severe and often seen as a late complication, the need for intraocular lens (IOL) exchange after primary cata- ract surgery is unusual and may lead to suboptimal visual outcome9–15. However the incidence of this surgery has not been consistently estimated previously. Authors Authors www.nature.com/scientificreports OPEN Here, we studied the incidence of IOL exchange among patients that had primary surgery at a single center over a period of 8 years. We also examined the risk factors for, and visual outcomes after intraocular lens (IOL) removal/exchange after cataract surgery. Results id Characteristic IOL exchange/explantation group n = 34 Control group n = 47 Freq % Freq % Age (years)* 58.32 ± 13.51 58.26 ± 16.41 Gender Male 20 58.8 22 46.8 Female 14 41.2 25 53.2 Eye Right 23 67.6 26 55.3 Left 11 32.4 21 44.7 Pre-existing systemic comorbidity† Yes 24 70.6 36 76.6 No 10 29.4 11 23.4 Pre-existing ocular comorbidity‡ Yes 7 20.6 2 4.3 No 27 79.4 45 95.7 Axial length (millimeter)* 25.10 ± 2.14 24.10 ± 1.68 Adverse event during cataract surgery§ Yes 17 50 4 8.5 No 17 50 43 91.5 Type of surgery Phacoemulsification 20 58.8 43 91.5 Phaco + vitrectomy/other procedure 14 41.2 4 8.5 IOL position In the bag 25 73.5 44 93.6 Anterior chamber/sulcus 9 26.5 3 6.4 Table 1 Characteristics of cases and controls Data are presented as means and Table 1. Characteristics of cases and controls. Data are presented as means and SD. †Frequent systemic comorbidities included diabetes and hypertension ‡Pre-existing ocular comorbidities included glaucoma Table 1. Characteristics of cases and controls. Data are presented as means and SD. †Frequent systemic comorbidities included diabetes and hypertension. ‡Pre-existing ocular comorbidities included glaucoma, high myopia, previous intraocular surgery, uveitis, pseudoexfoliation and Fuchs’ dystrophy. §Frequent adverse events during cataract surgery included posterior capsular tear with or without vitreous prolapse, or subluxated intraocular lens. Table 1. Characteristics of cases and controls. Data are presented as means and SD. †Frequent systemic comorbidities included diabetes and hypertension. ‡Pre-existing ocular comorbidities included glaucoma, high myopia, previous intraocular surgery, uveitis, pseudoexfoliation and Fuchs’ dystrophy. §Frequent adverse events during cataract surgery included posterior capsular tear with or without vitreous prolapse, or subluxated intraocular lens. that underwent IOL exchange/ex-plantation were almost comparable to controls in terms of most characteristics (i.e., age, pre-existing systemic comorbidities, and operated eye), except that eyes with IOL exchange had a greater mean axial length and a higher burden of pre-existing ocular comorbidities such as glaucoma (3 eyes), high myo- pia (2 eyes), and uveitis, pseudoexfoliation and Fuchs’ dystrophy and previous intraocular surgery (1 eye each) compared to control eyes. Results id that underwent IOL exchange/ex-plantation were almost comparable to controls in terms of most characteristics (i.e., age, pre-existing systemic comorbidities, and operated eye), except that eyes with IOL exchange had a greater mean axial length and a higher burden of pre-existing ocular comorbidities such as glaucoma (3 eyes), high myo- pia (2 eyes), and uveitis, pseudoexfoliation and Fuchs’ dystrophy and previous intraocular surgery (1 eye each) compared to control eyes. Factors associated with of IOL exchange/explantation. In the multivariable binary logistic regres- sion analysis (Table 2), an adverse event during cataract surgery and a pre-existing ocular comorbidity were significantly associated with the outcome. The effect of gender was marginally significant and that of axial length was not statistically significant. Briefly, the risk of IOL exchange/explantation was higher among eyes with an eventful cataract surgery (adjusted odds ratio [aOR] 19.45; 95% CI 4.89–77.30, P < 0.001); among those with a pre-existing ocular comorbidity (aOR 10.70; 95% CI 1.69–67.63, P = 0.012), among males (aOR 2.94; 95% CI 0.89–9.73, P = 0.077), and among eyes with an axial length of ≥26 mm (aOR 2.92; 95% CI 0.75 11.28 P = 0.121), compared with their reference groups. Indication for IOL exchange and location of re-implanted IOL. The most frequent indications for IOL exchange (Table 3) were: In-the-bag subluxation in 8 eyes (23.5%), out-of-the bag subluxation and refractive error in 7 each (20.6% each) and uveitis in 6 (17.6%). Of 7 eyes with refractive error, 4 eyes had undergone a pre- vious myopic LASIK (laser in situ keratomileusis) procedure. Of the 3 remaining eyes, one had preoperative axial length measure using ultrasound instead of optical biometry, because of mature cataract. The remaining 2 eyes had received the wrong IOL power without any clear reason. For all the 7 eyes that were given wrong IOL power, the final IOL position (after IOL exchange) was in the bag. it Of the explanted IOLs, 25 were one-piece acrylic hydrophobic, 2 were three-piece acrylic hydrophobic and 7 were one-piece polymethyl methacrylate. The most frequent locations of the re-implanted IOLs were: sulcus in 9 eyes (26.5%), in-the-bag in 8 (23.5%), iris fixated and scleral fixated in 5 each (14.7% each) and anterior chamber in 4 (11.8%). Anterior vitrectomy and pars plana vitrectomy (PPV) were needed at the time of IOL exchange in 5 and 6 eyes, respectively, where PPV alone was done for those that had posterior IOL dislocation. Results id Incidence of IOL exchange. During 2010–2017, a total of 17415 eyes underwent phacoemulsification with IOL implantation in King Khaled Eye Specialist Hospital (KKESH). Of these, we initially identified 45 eyes of 45 patients who had IOL removal or exchange in the same period. Eleven eyes (of 11 patients) were excluded because these had phakic IOL removal with cataract surgery. Thus 34 eyes (of 34 patients) that underwent subsequent IOL removal or replacement during same period remained, equaling an overall cumulative incidence of IOL exchange/explantation of 2 per 1000 surgeries (95% confidence interval [CI] 1 to 3). Characteristics of eyes/individuals with IOL exchange/explantation and their controls. The emographic and clinical information for cases and controls is summarized in Table 1. Briefly, individuals/eyes 1Jeddah Eye Hospital, Jeddah, Saudi Arabia. 2Research Department, King Khaled Eye Specialist Hospital, Riyadh, Saudi Arabia. 3Vitreoretinal Division, King Khaled Eye Specialist Hospital, Riyadh, Saudi Arabia. 4Glaucoma Division, King Khaled Eye Specialist Hospital, Riyadh, Saudi Arabia. 5Anterior Segment Division, King Khaled Eye Specialist Hospital, Riyadh, Saudi Arabia. 6Ophthalmology Department, University of Illinois, Chicago, USA. 7Department of Ophthalmology, Clinical Sciences, Skane University Hospital, Lund University, Lund, Sweden. Correspondence and requests for materials should be addressed to P.S. (email: pschatz@kkesh.med.sa) Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 www.nature.com/scientificreports/ Characteristic IOL exchange/explantation group n = 34 Control group n = 47 Freq % Freq % Age (years)* 58.32 ± 13.51 58.26 ± 16.41 Gender Male 20 58.8 22 46.8 Female 14 41.2 25 53.2 Eye Right 23 67.6 26 55.3 Left 11 32.4 21 44.7 Pre-existing systemic comorbidity† Yes 24 70.6 36 76.6 No 10 29.4 11 23.4 Pre-existing ocular comorbidity‡ Yes 7 20.6 2 4.3 No 27 79.4 45 95.7 Axial length (millimeter)* 25.10 ± 2.14 24.10 ± 1.68 Adverse event during cataract surgery§ Yes 17 50 4 8.5 No 17 50 43 91.5 Type of surgery Phacoemulsification 20 58.8 43 91.5 Phaco + vitrectomy/other procedure 14 41.2 4 8.5 IOL position In the bag 25 73.5 44 93.6 Anterior chamber/sulcus 9 26.5 3 6.4 Table 1. Characteristics of cases and controls. Data are presented as means and SD. †Frequent systemic comorbidities included diabetes and hypertension. ‡Pre-existing ocular comorbidities included glaucoma, high myopia, previous intraocular surgery, uveitis, pseudoexfoliation and Fuchs’ dystrophy. §Frequent adverse events during cataract surgery included posterior capsular tear with or without vitreous prolapse, or subluxated intraocular lens. Results id ‡Frequent adverse events during cataract surgery included posterior capsular tear with or without vitreous prolapse, or subluxated intraocular lens. Table 2. Bivariate and multivariable binary logistic regression analysis of factors associated with IOL exchange/ explantation. Frequent systemic comorbidities included diabetes and hypertension. †Pre-existing ocular comorbidities included glaucoma, high myopia, previous intraocular surgery uveitis, pseudoexfoliation and Fuchs’ dystrophy. ‡Frequent adverse events during cataract surgery included posterior capsular tear with or without vitreous prolapse, or subluxated intraocular lens. Visual outcome of IOL exchange/explantation. A higher proportion of eyes that underwent IOL exchange or explantation had suboptimal vision (best corrected visual acuity [BCVA] <20/60) than the control eyes at the most recent visit (Table 4). Eyes that underwent IOL exchange or explantation were three times (crude relative risk [RR] 3.11 [95% CI, 1.35–7.15]; P = 0.008) more likely to have a BCVA of <20/60 than those that had cataract surgery alone. This relative risk remained high and statistically significant even after adjustment for base- line BCVA (adjusted RR 2.60 [95% CI, 1.13–6.02]; P = 0.025). Results id Another 2 eyes needed subsequent PPV for reasons unrelated to the IOL exchange; 1 of these for diabetes tractional detachment and vitreous hemorrhage and 1 for panuveitis. Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 www.nature.com/scientificreports/ Characteristic Crude OR (95% CI) P Adjusted OR (95% CI) P Age, years ≤59 1.28 (0.53–3.10) 0.586 ≥60 1.0 Gender Male 1.62 (0.67–3.96) 0.287 2.94 (0.89–9.73) 0.077 Female 1.0 Eye Right 1.69 (0.67–4.24) 0.264 Left 1.0 Pre-existing systemic comorbidity* Yes 0.73 (0.27–1.99) 0.543 No 1.0 Pre-existing ocular comorbidity† Yes 5.83 (1.13–30.14) 0.035 10.70 (1.69–67.63) 0.012 No 1.0 Axial length, mm ≤25 1.0 ≥26 2.38 (0.80–7.09) 0.150 2.92 (0.75–11.28) 0.121 Adverse event during cataract surgery‡ Yes 10.75 (3.16–36.61) <0.001 19.45 (4.89–77.30) <0.001 No 1.0 1.0 Type of surgery Phacoemulsification 1.0 Phaco + vitrectomy/other procedure 7.53 (2.20–25.78) 0.001 IOL position In the Bag 1.0 Anterior chamber/sulcus 5.28 (1.31–21.32) 0.019 Table 2. Bivariate and multivariable binary logistic regression analysis of factors associated with IOL exchange/ explantation. Frequent systemic comorbidities included diabetes and hypertension. †Pre-existing ocular comorbidities included glaucoma, high myopia, previous intraocular surgery uveitis, pseudoexfoliation and Fuchs’ dystrophy. ‡Frequent adverse events during cataract surgery included posterior capsular tear with or without vitreous prolapse, or subluxated intraocular lens. Characteristic Crude OR (95% CI) P Adjusted OR (95% CI) P Age, years ≤59 1.28 (0.53–3.10) 0.586 ≥60 1.0 Gender Male 1.62 (0.67–3.96) 0.287 2.94 (0.89–9.73) 0.077 Female 1.0 Eye Right 1.69 (0.67–4.24) 0.264 Left 1.0 Pre-existing systemic comorbidity Yes 0.73 (0.27–1.99) 0.543 No 1.0 Pre-existing ocular comorbidity† Yes 5.83 (1.13–30.14) 0.035 10.70 (1.69–67.63) 0.012 No 1.0 Axial length, mm ≤25 1.0 ≥26 2.38 (0.80–7.09) 0.150 2.92 (0.75–11.28) 0.121 Adverse event during cataract surgery‡ Yes 10.75 (3.16–36.61) <0.001 19.45 (4.89–77.30) <0.001 No 1.0 1.0 Type of surgery Phacoemulsification 1.0 Phaco + vitrectomy/other procedure 7.53 (2.20–25.78) 0.001 IOL position In the Bag 1.0 Anterior chamber/sulcus 5.28 (1.31–21.32) 0.019 Table 2. Bivariate and multivariable binary logistic regression analysis of factors associated with IO explantation. Frequent systemic comorbidities included diabetes and hypertension. †Pre-existing o comorbidities included glaucoma, high myopia, previous intraocular surgery uveitis, pseudoexfolia Fuchs’ dystrophy. ‡Frequent adverse events during cataract surgery included posterior capsular tear without vitreous prolapse, or subluxated intraocular lens. Table 2. Bivariate and multivariable binary logistic regression analysis of factors associated with IOL exchange/ explantation. Frequent systemic comorbidities included diabetes and hypertension. †Pre-existing ocular comorbidities included glaucoma, high myopia, previous intraocular surgery uveitis, pseudoexfoliation and Fuchs’ dystrophy. Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 www.nature.com/scientificreports/ estimated the annual incidence of late in-the-bag IOL dislocation over a 21-year period, reviewing 123 patients (eyes) and an equal number of controls, and data from the Swedish National Cataract Registry and life tables from official statistics on the Swedish population19. The annual incidence varied between 0 and 0.08%. The cumulative risk was 0.09% and 0.55% over 5 and 10 years, respectively. Risk factors included PXF, zonular dehiscence, pseu- dophakodonesis and increased axial length19.h These studies have reported slightly higher incidence estimates compared to our study, which may be related to the fact that most patients in these studies needed only IOL repositioning17–20 as opposed to IOL exchange, whereas our study focused specifically on IOL removal/exchange. On the other hand such differences may, in part, be due to differences in surgical approach (which is partly determined by surgeon’s preference and which we could not control for in our retrospective study) to manage eyes with IOL dislocations. y g y Jin et al. 2007 focused on IOL exchange due to postoperative refractive error20. Incorrect corneal power esti- mation, followed by error in measurement of axial length and inserting a wrong IOL were the most common reasons for IOL exchange20. In this study, we identified 7 eyes that needed IOL exchange due to refractive error (Table 3). Of these, 4 had incorrect corneal power estimation related to previous myopic LASIK procedures, lead- ing to a low postoperative corneal power. This, in turn, tends to lead to an underestimation of the effective lens position (ELP) and therefore to an underestimation of the correct IOL power for emmetropia. Of the 3 remaining eyes, one had preoperative axial length measure using ultrasound instead of optical biometry, because of mature cataract. Ultrasound tends to lead to an underestimation of the correct axial length and subsequent overestima- tion of the correct IOL power, due to compression of the globe during the ultrasound examination.h The indications for IOL removal/exchange have been examined in several previous studies. Davies et al. found that the indications IOL exchange were IOL dislocation to the posterior pole, dissatisfaction with multifocal IOL and uveitis9. In a study by Chan et al., the most important indication for in the bag IOL exchange were IOL decentration, caused by complicated surgeries, and high myopia11. The most common indication for sulcus IOL exchange was glaucoma-hyphema-uveitis syndrome, especially for the one-piece IOL11. Jones et al. www.nature.com/scientificreports/ By contrast, of 47 eyes in the control group, 6(12.8%) and 2 (4.3%) had a BCVA <20/60–20/200 and <20/200, respectively. Table 4. Best corrected visual acuity at most recent follow up in eyes that had an IOL exchange/explantation after cataract surgery, and in control eyes that had cataract surgery only. Adjusted for baseline BCVA using Cox’s regression. P values for crude and adjusted RR = 0.008 and 0.025. Of the 34 eyes in the IOL exchange/ explantation group, 11(32.4%) and 7(20.6%) had a BCVA <20/60–20/200 and <20/200, respectively. By contrast, of 47 eyes in the control group, 6(12.8%) and 2 (4.3%) had a BCVA <20/60–20/200 and <20/200, respectively. Table 4. Best corrected visual acuity at most recent follow up in eyes that had an IOL exchange/explantation after cataract surgery, and in control eyes that had cataract surgery only. Adjusted for baseline BCVA using Cox’s regression. P values for crude and adjusted RR = 0.008 and 0.025. *Of the 34 eyes in the IOL exchange/ explantation group, 11(32.4%) and 7(20.6%) had a BCVA <20/60–20/200 and <20/200, respectively. By contrast, of 47 eyes in the control group, 6(12.8%) and 2 (4.3%) had a BCVA <20/60–20/200 and <20/200, respectively. Table 4. Best corrected visual acuity at most recent follow up in eyes that had an IOL exchange/explantation after cataract surgery, and in control eyes that had cataract surgery only. Adjusted for baseline BCVA using Cox’s regression. P values for crude and adjusted RR = 0.008 and 0.025. *Of the 34 eyes in the IOL exchange/ explantation group, 11(32.4%) and 7(20.6%) had a BCVA <20/60–20/200 and <20/200, respectively. By contrast, of 47 eyes in the control group, 6(12.8%) and 2 (4.3%) had a BCVA <20/60–20/200 and <20/200, respectively. for dislocated IOL, resulting in a 10 year incidence of 0.6% for this complication18. Kinga Dabrowska-Kloda et al. estimated the annual incidence of late in-the-bag IOL dislocation over a 21-year period, reviewing 123 patients (eyes) and an equal number of controls, and data from the Swedish National Cataract Registry and life tables from official statistics on the Swedish population19. The annual incidence varied between 0 and 0.08%. The cumulative risk was 0.09% and 0.55% over 5 and 10 years, respectively. Risk factors included PXF, zonular dehiscence, pseu- dophakodonesis and increased axial length19. for dislocated IOL, resulting in a 10 year incidence of 0.6% for this complication18. Kinga Dabrowska-Kloda et al. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Characteristic Number % Indication In the bag subluxation 8 23.5 Out of the bag subluxation 7 20.6 Refractive error 7 20.6 Uveitis 6 17.6 Opacification of intraocular lens 3 8.8 Corneal decompensation 2 5.9 Trauma 1 2.9 Location of re-implanted IOL Sulcus 9 26.5 In the bag 8 23.5 Iris fixated 5 14.7 Scleral fixated 5 14.7 Anterior chamber 4 11.8 Aphakia 3 8.8 Table 3. Indication for IOL exchange and location of re-implanted IOL after cataract surgery (n = 34 eyes of 34 patients). Table 3. Indication for IOL exchange and location of re-implanted IOL after cataract surgery (n = 34 eyes of 34 patients). Table 3. Indication for IOL exchange and location of re-implanted IOL after cataract surgery (n = 34 eyes of patients). Best corrected visual acuity at the last follow up Relative risk (95 CI %) for BCVA <20/60 Group** 20/20–20/60 <20/60–20/200 Freq (%) Freq (%) Crude Adjusted IOL exchange/ explantation(n = 34 eyes) 16 (47.1) 18 (52.9) 3.11 (1.35–7.15) 2.60 (1.13–6.02) Control (n = 47 eyes) 39(83.0) 8 (17.0) 1.0 1.0 Table 4. Best corrected visual acuity at most recent follow up in eyes that had an IOL exchange/explantation after cataract surgery, and in control eyes that had cataract surgery only. Adjusted for baseline BCVA using Cox’s regression. P values for crude and adjusted RR = 0.008 and 0.025. Of the 34 eyes in the IOL exchange/ explantation group, 11(32.4%) and 7(20.6%) had a BCVA <20/60–20/200 and <20/200, respectively. By contrast, of 47 eyes in the control group, 6(12.8%) and 2 (4.3%) had a BCVA <20/60–20/200 and <20/200, respectively. Best corrected visual acuity at the last follow up Relative risk (95 CI %) for BCVA <20/60 Group** 20/20–20/60 <20/60–20/200 Freq (%) Freq (%) Crude Adjusted IOL exchange/ explantation(n = 34 eyes) 16 (47.1) 18 (52.9) 3.11 (1.35–7.15) 2.60 (1.13–6.02) Control (n = 47 eyes) 39(83.0) 8 (17.0) 1.0 1.0 Table 4. Best corrected visual acuity at most recent follow up in eyes that had an IOL exchange/explantation after cataract surgery, and in control eyes that had cataract surgery only. Adjusted for baseline BCVA using Cox’s regression. P values for crude and adjusted RR = 0.008 and 0.025. *Of the 34 eyes in the IOL exchange/ explantation group, 11(32.4%) and 7(20.6%) had a BCVA <20/60–20/200 and <20/200, respectively. Discussion Several retrospective studies have looked at the risk factors for and visual outcomes after IOL removal/ exchange10–15. However, to the best of our knowledge this is the first study to report the incidence of IOL removal/ exchange (2 per 1000 cataract surgeries (95% confidence interval [CI] 1 to 3, over 8 years, corresponding to an average annual incidence of 2.5 per 10,000). This information is invaluable for counseling patients prior to cata- ract surgery about the risk of an IOL exchange and its sequelae.i g y g q IOL removal/exchange incidence estimates that exist in the literature are all based on one specific subgroup of these patients; namely those that had late IOL dislocations16–20. Furthermore, most of the patients exam- ined in those studies only needed IOL repositioning as opposed the exchange or replacement17–20. Focusing on late IOL dislocation, Clark et al. reported the 5-year cumulative incidence of IOL dislocation to be 0.3% in a population-based study, with an increase in the rate of this complication during a study period of 22 years16. This figure is somewhat higher compared to our study (0.2% over 8 years). However, it is notable that Clark et al. did not mention the degree of dislocation and whether IOL exchange was required or not16.i g g q Jakobsson et al. characterized the risk factors of late IOL dislocation in a sample of 84 eyes, finding an over- representation of pseudoexfoliation (PXF), glaucoma and cataract surgeries complicated by zonular dehiscence17. They used data from the Swedish National Cataract Register between 2004 and 2006 to estimate the incidence of IOL repositioning surgery, by dividing the number of IOL repositioning surgeries by the total number of pseu- dophakic individuals during the same period, arriving at an annual incidence of 0.05%, which is slightly higher than the average annual incidence described in our study (0.025%)17. However only 7% of their patients required IOL exchange or removal and the remaining patients had scleral suturing of the dislocated IOL17. Monestam et al. followed 810 patients for 10 years after cataract surgery and identified 5 patients that required further surgery Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 www.nature.com/scientificreports/ www.nature.com/scientificreports/ In their study, the main indication was subluxation with IOL dislocation, especially associated with pseudoex- foliation syndrome12. Incorrect IOL power calculation, calcification and corneal endothelial decompensation were also found12. In our study, the most common indication for IOL exchange/removal was also IOL dislocation (Table 3). ( ) Jakobsson et al. 2013 examined the outcomes after surgical repositioning of IOLs after late dislocation and found that 94% could be repositioned using scleral sutures. Pseudoexfoliation was detected in 57% of the eyes, which was also reflected in the present study, where ocular comorbidity including pseudexfoliation, was an inde- pendent risk factor for IOL exchange (Table 2). They found that 13 eyes (out of 91) had additional surgical pro- cedures, and 3 of those had postoperative retinal detachments. More than half of the eyes obtained visual acuities more than 20/40 and 23% of the eyes had a decrease in the visual acuity following the surgery21. In our study, eyes that underwent IOL exchange were nearly two and a half times more likely to have a final visual acuity of <20/60 compared to the control eyes (Table 4). p y ( ) Kamiya et al. found that the most common complaints for multifocal IOL removal were waxy vision, glare and halos, blurred vision (at far, at near, or at intermediate distances), and dysphotopsia,. The most common reasons for multifocal IOL removal were decreased contrast sensitivity, followed by photic phenomenon, unknown ori- gin including neuroadaptation failure, incorrect IOL power, preoperative excessive expectation, IOL dislocation/ decentration, and anisometropia. The mean ± SD axial length was 25.13 ± 1.83 mm in their study22. Myopia was also overrepresented in the IOL removal/exchange group in our study, and 4 of the patients had undergone pre- vious LASIK for high myopia (>6D), leading to difficulties in estimating the central corneal power. However, in our final multivariable model, axial length, when used as a categorical variable, did not turn out to be significantly associated with IOL exchange, although it was significant when used as a continuous variable (data not shown). This is possibly because of a relatively limited sample size in our study (Table 2). h p y y p y ( ) In our study, the presence of an intraoperative adverse event and a pre-existing ocular co-morbidity (such as pseudexfoliations) were significant predictors of IOL exchange. This finding generally concur with existing literature and is plausible. Methods Th This retrospective study was undertaken at the King Khaled Eye Specialist Hospital, Riyadh, a tertiary eye care hospital. Ethics approval was obtained from the Institutional Review Board of KKESH. Informed consent for intraocular surgery was obtained from all patients. The research followed the tenets of the Declaration of Helsinki. A cohort design was employed to estimate the incidence of IOL exchange and a case-control design was used to identify factors associated with it.hi The main outcome was the cumulative incidence of IOL exchange which was defined as the number of eyes that underwent IOL exchange/explantation, divided by the total number of eyes that underwent cataract surgery with IOL during the study period23. To achieve this objective, all eyes that underwent phacoemulsification cata- ract surgery with IOL implantation during 2010–2017 and had a subsequent IOL removal or replacement during the same time period were included in the study. Eyes that underwent phakic IOL removal with phacoemulsifi- cation and IOL were excluded. Diagnostic coding was used to identify the total number of cataract surgeries with IOL and those that had a subsequent exchange/explantation. For the risk-factor component of the study, all eyes that underwent IOL exchange (n = 34) were compared with a subset of eyes that had cataract surgery alone (n = 47) during the same period. Specifically, a control eye/ patient for a given case eye/patient was selected from the complete list of entries coded as “phacoemulsification with IOL”, where the first entry fulfilling the criterion of age matching the patients age ±5 years, was selected as a control. A structured proforma (Supplemental File 1) was used to collect data from medical records on age, gender, pre-operative characteristics, indication for IOL exchange/explantation, the type of IOL that was removed, oper- ative and postoperative complications, and visual outcomes. All case and control eyes had baseline preoperative IOL calculations using Zeiss IOL Master 500 (Carl Zeiss Meditec, AG Jena, Germany), except where hazy media (such as mature cataract) precluded the use of optical biometry. In such cases, applanation ultrasound biometry (Cinescan S or Aviso, Quantel Medical Cedex - France) was used to estimate the axial length. The Holladay 2 formula was routinely used as the standard IOL calculation formula. However other formulas could be requested in addition, at the discretion of the surgeon. Emmetropia or mild myopia was the refractive target for all patients and controls. www.nature.com/scientificreports/ p Limitations of this study include its retrospective design and the inability to control or adjust for variabil- ity regarding surgeons preferred technique of IOL exchange or choice of IOL exchange versus IOL reposition. The figure of two IOL exchanges/removals per 1000 cataract surgeries may be an underestimate given that the follow-up time after cataract surgery was at most 8 years in our study. For the calculation of cumulative incidence of an event, the specified time interval should generally be the same for all members of a group. However, it may vary in certain situations23. A sufficiently large multi-center prospective study is warranted to more accurately estimate the true incidence of IOL exchange/removal and other rare complications of cataract surgery and to examine their risk factors and visual outcomes. We conclude that IOL exchange is rare yet may affect the final visual outcome adversely. The most frequent indication was IOL dislocation. Two factors were significantly associated with IOL exchange/explantation: an adverse event during cataract surgery, and a pre-existing ocular comorbidity. Eyes that underwent IOL exchange or explantation were nearly two and a half time times more likely to have a final visual acuity of <20/60 compared to those that had cataract surgery alone. www.nature.com/scientificreports/ found that IOL dislocation, refractive error, patient dissatisfaction and implant opacification were the main indications for IOL exchange10. Neuhann et al. presented results of 105 eyes from 100 patients undergoing IOL exchange/explant. Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 Methods Th Visual acuity was measured in Snellen format. Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 www.nature.com/scientificreports/ Statistical analysis. All data were entered in a Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) database by a trained data entry operator and analyzed using SPSS for Windows version 20.0 (IBM SPSS, Inc., Chicago, IL, USA). For the present analysis, pre-existing ocular comorbities were grouped into binary cate- gories with “Yes” indicating the presence of the condition and “No” otherwise, as were the pre-existing systemic comorbidities, adverse event during cataract surgery, type of surgery, and IOL position. Age was categorized as ≤59 years and ≥60 years, and axial length as ≤25 mm and ≥26 mm. Counts and percentages were used to describe categorical data and means (with SD) were computed to describe continuous variables. g ( ) p A binary logistic regression analysis was performed to identify factors associated with IOL exchange/explanta- tion. Odds ratios (OR) with 95% confidence intervals (CI) were calculated to indicate an association. Nine varia- bles were examined in the univariate analysis. Of these, adverse event during cataract surgery, pre-existing ocular comorbidity, axial length, type of surgery and IOL position were significantly associated with IOL exchange/ explantation. Both type of surgery and IOL position were not considered for the multivariable logistic regression models as they were strongly associated with an adverse event during cataract, as indicated by a high phi coeffi- cients of 0.63 and 0.71, respectively. The final model included adverse event during cataract surgery, pre-existing ocular comorbidity, and axial length was adjusted for gender due to slight gender imbalance between case and control eyes. y Next, the best-corrected visual acuity (BCVA) at the last follow up was compared between eyes with and without subsequent IOL exchange/explantation. Using cox regression analysis, relative risk (RR) with 95% CIs were calculated to determine the association between IOL exchange/explantation and BCVA <20/60 at the last follow up. For all the analyses, 2-sided p-values were calculated. A p-value < 0.05 was considered significant. Supplemental File 1: data collection form and Supplemental File 2: Excel datasheet are appended. References 1. Lundström, M. & Wendel, E. Duration of self assessment benefit of cataract extraction: a long term study. Br J Ophthalmol. 89 1017–1120 (2005).f 2. Busbee, B. G., Brown, M. M., Brown, G. C. & Sharma, S. Incremental cost-effectiveness of initial cataract surgery. Ophthalmology 2002 109, 606–613 (2002). 3. Erie, J. C. Rising Cataract Surgery Rates: Demand and Supply. Ophthalmology 121, 2–4 (2014). 4. Haddad, N. M., Sun, J. K., Abujaber, S., Schlossman, D. K. & Silva, P. S. Cataract surgery and its complications in diabetic patients Semin Ophthalmol 29, 329–337 (2014). 4. Haddad, N. M., Sun, J. K., Abujaber, S., Schlossman, D. K. & Silva, P. S. Cataract surgery and its complications in diabetic patients. Semin Ophthalmol 29, 329–337 (2014). 5. Nagy, Z. Z. et al. Complications of femtosecond laser-assisted cataract surgery. J Cataract Refract Surg. 40, 20–28 (2014). h l d f d h h l ft d d 6. Creuzot-Garcher, C. et al. Incidence of Acute Postoperative Endophthalmitis after Cataract Surgery: A Nationwide Study i from 2005 to 2014. Ophthalmology 123, 1414–1420 (2016).ht p gy ( ) 7. Olsen, T. & Jeppesen, P. The incidence of retinal detachment after cataract surgery. Open Ophthalmol J. 6, 79–82 (2012). ht 8. Payne, J. W., Kameen, A. J., Jensen, A. D. & Christy, N. E. Expulsive hemorrhage: its incidence in cataract surgery and a report o bilateral cases. Trans Am Ophthalmol Soc. 83, 181–204 (1985). 9. Davies, E. C. & Pineda, R. Intraocular lens exchange surgery at a tertiary referral center: Indications, complications, and visual outcomes. J Cataract Refract Surg. 42, 1262–1267 (2016). f g 0. Jones, J. J., Jones, Y. J. & Jin, G. J. C. Indications and outcomes of intraocular lens exchange during a recent 5-year period. Am Ophthalmol. 157, 154–162 (2014).ff 11. Chan, T. C., Lok, J. K., Jhanji, V. & Wong, V. W. Intraocular lens removal in Chinese patients: different patterns and different responses. Int Ophthalmol. 35, 679–684 (2015). 12. Neuhann, I., Fleischer, F. & Neuhann, T. Reasons for exchange and removal of intraocular lenses. Klin Monbl Augenheilkd. 229, 794–797 (2012).t 3. Oltulu, R. et al. Intraocular lens removal or exchange: indications, postoperative interventions, and outcomes. Arq Bras Oftalmol. 78 154–157 (2015).f 4. Narang, P., Steinert, R., Little, B. & Agarwal, A. Intraocular lens scaffold to facilitate intraocular lens exchange. J Cataract Refrac Surg. 40, 1403–1407 (2014). g 15. Tassignon, M. J., Bartholomeeusen, E., Rozema, J. References J., Jongenelen, S. & Mathysen, D. G. Feasibility of multifocal intra-ocular lens exchange and conversion to the bag-in-the-lens implantation. Acta Ophthalmol. 92, 265–269 (2014). g g p 6. Clark, A., Morlet, N., Ng, J. Q., Preen, D. B. & Semmens, J. B. Whole population trends in complications of cataract surgery over 22 years in Western Australia. Ophthalmology 118, 1055–1061 (2011). years in Western Australia. Ophthalmology 118, 1055–1061 (2011 y p gy 7. Jakobsson, G. et al. Late dislocation of in-the-bag and out-of-the bag intraocular lenses: ocular and surgical characteristics and time to lens repositioning. J Cataract Refract Surg. 36, 1637–1644 (2010). . Jakobsson, G. et al. Late dislocation of in-the-bag and out-of-the b to lens repositioning. J Cataract Refract Surg. 36, 1637–1644 (201 p g f g 8. Mönestam, E. I. Incidence of dislocation of intraocular lenses and pseudophakodonesis 10 years after cataract surgery Ophthalmology 116, 2315–2320 (2009). 18. Mönestam, E. I. Incidence of dislocation of intraocular lenses and pseudophakodonesis 10 years after cataract surgery. Ophthalmology 116, 2315–2320 (2009). 19. Dabrowska-Kloda, K., Kloda, T., Boudiaf, S., Jakobsson, G. & Stenevi, U. Incidence and risk factors of late in-the-bag intraocular lens p p y g y Ophthalmology 116, 2315–2320 (2009). 19. Dabrowska-Kloda, K., Kloda, T., Boudiaf, S., Jakobsson, G. & Stenevi, U. Incidence and risk factors of late in-the-bag intraocular lens dislocation evaluation of 140 eyes between 1992 and 2012 J Cataract Refract Surg 41 1376 1382 (2015) p gy 9. Dabrowska-Kloda, K., Kloda, T., Boudiaf, S., Jakobsson, G. & Stenevi, U. Incidence and risk factors of late in-the-bag intraocular len dislocation: evaluation of 140 eyes between 1992 and 2012. J Cataract Refract Surg. 41, 1376–1382 (2015). y f g 20. Jin, G. J., Crandall, A. S. & Jones, J. J. Intraocular lens exchange due to incorrect lens power. Ophthalmology 114, 417–424 (20 g p p gy 21. Jakobsson, G., Zetterberg, M., Sundelin, K. & Stenevi, U. Surgical repositioning of intraocular lenses after late dislocation: complications, effect on intraocular pressure, and visual outcomes. J Cataract Refract Surg. 39, 1879–1885 (2013). f 22. Kamiya, K. et al. Survey Working Group of the Japanese Society of Cataract and Refractive Surgery. Multifocal intraocular lens removal: a case series of 50 eyes. Am J Ophthalmol. 158, 215–220 (2014). y p 3. Last, J. M. Cumulative incidence. A Dictionary of Epidemiology. International Epidemiological Association. New York: Oxford University Press (2001). Author Contributions Design of the study (M.A.E., K.A., A.A.A., R.K. and P.S.); Conduct of the study (M.A.E., K.A., A.A.A., R.M., R.K., H.M.O., M.M. and P.S.); Collection, management, analysis, and interpretation of the data (M.A.E., K.A., A.A.A., R.M., R.K., H.M.O., M.M. and P.S.); Preparation, review and final approval of the manuscript (M.A.E., K.A., A.A.A., R.M., R.K., H.M.O., M.M. and P.S.). Design of the study (M.A.E., K.A., A.A.A., R.K. and P.S.); Conduct of the study (M.A.E., K.A., A.A.A., R.M., R.K. H.M.O., M.M. and P.S.); Collection, management, analysis, and interpretation of the data (M.A.E., K.A., A.A.A. R.M., R.K., H.M.O., M.M. and P.S.); Preparation, review and final approval of the manuscript (M.A.E., K.A. Additional Information 1. Lundström, M. & Wendel, E. Duration of self assessment benefit of cataract extraction: a long term study. Br J Ophthalmo 1017–1120 (2005).f Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-49030-2. Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-49030-2. Competing Interests: The authors declare no competing interests. Competing Interests: The authors declare no competing interests. Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2 www.nature.com/scientificreports/ Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. fi Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. Additional Information The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2019 Scientific Reports \| (2019) 9:12877 \| https://doi.org/10.1038/s41598-019-49030-2
https://openalex.org/W2891486861	https://www.nature.com/articles/s41598-018-31881-w.pdf	English	null	Spatial separation of catches in highly mixed fisheries	Scientific reports	2,018	cc-by	11,881	Mixed fisheries and the EU landing obligationfi i g g Recent efforts to reduce exploitation rates in commercial fisheries have begun the process of rebuilding depleted fish populations1. Improved management of fisheries has the potential to increase population sizes and allow increased sustainable catches, yet fisheries catch globally remains stagnant2. In light of a projected increase in demand for fish protein3 there is an important role for well managed fisheries in supporting future food security4 necessitating that fisheries are managed efficiently to maximise productivity. gi gfi y p y A particular challenge in realising increased catches from rebuilt populations is maximising yields from mixed fisheries5–7. In mixed fisheries, the predominant type of fishery worldwide, several fish species are caught together in the same net or fishing operation (known as a “technical interaction”). If managed by individual quotas, and catches do not match available stock quotas, either a vessel must stop fishing when the first quota is reached (the “choke” species) or overexploitation of the weaker species occurs while fishers continue to catch more healthy species and throw back (“discard”) the fish for which they have no quota8. There is, therefore, a pressing need for scientific tools to simplify the complexities of mixed fisheries and help avoid discarding.ii i yi g Sustainability of European fisheries has been hampered by the “mixed fishery problem” for decades with large-scale discarding resulting9,10. Mixed fisheries require specific management approaches to avoid overfishing and a paradigm shift is being introduced under the EU Common Fisheries Policy (CFP) reform of 2012 through two significant management changes. First, by 2019 all fish that are caught are due to be counted against the respective stock quota even if they are discarded; second, by 2020 all fish stocks must be fished at an exploitation rate corresponding to their Maximum Sustainable Yield (MSY)11. These changes are expected to contribute to attainment of the goal of Good Environmental Status (GES) under the European Marine Strategy Framework Directive (MSFD12) and move Europe towards an ecosystem based approach to fisheries management13.l i Conflicts between overall management goals and drivers for individual actors must be overcome to achieve sustainability. Societal objectives for fisheries to achieve MSY across ecosystem components are paralleled by individual fishers goals to maximise utility; whether that be profit, income or the continuance of traditional practices14. Spatial separation of catches in highly mixed fisheries Paul J. Dolder 1,2, James T. Thorson 3 & Cóilín Minto 1 Mixed fisheries are the dominant type of fishery worldwide. Overexploitation in mixed fisheries occurs when catches continue for available quota species while low quota species are discarded. As EU fisheries management moves to count all fish caught against quota (the “landing obligation”), the challenge is to catch available quota within new constraints, else lose productivity. A mechanism for decoupling exploitation of species caught together is spatial targeting, which remains challenging due to complex fishery and population dynamics. How far spatial targeting can go to practically separate species is often unknown and anecdotal. We develop a dimension-reduction framework based on joint dynamic species distribution modelling to understand how spatial community and fishery dynamics interact to determine species and size composition. In application to the highly mixed fisheries of the Celtic Sea, clear common spatial patterns emerge for three distinct assemblages. While distribution varies interannually, the same species are consistently found in higher densities together, with more subtle differences within assemblages, where spatial separation may not be practically possible. We highlight the importance of dimension reduction techniques to focus management discussion on axes of maximal separation and identify spatiotemporal modelling as a scientific necessity to address the challenges of managing mixed fisheries. Received: 27 March 2018 Accepted: 31 July 2018 Published online: 17 September 2018 www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports/ when reaching their first restrictive quota. This introduces a potential significant cost to fishers of under-utilised quota7,15 and provides a strong incentive to mitigate such losses16,17. q p g g To align catch with available quota depends on the ability to exploit target species while avoiding unwanted catch. Methods by which fishers can alter their fishing patterns include switching fishing method (e.g. trawling to netting), changing technical gear characteristics (e.g. introducing escapement panels in nets), or altering the timing and location of fishing activity18,19. For example, otter trawl gears are known to have higher catch rates of roundfish due to the higher headline and wider sweeps, which herd demersal fish into the net; conversely, beam trawls that employ chain mesh to “dig” benthic flatfish species, have higher catch rates for these species20. Fishing location choice also has a significant effect on catch21, something that fishers routinely consider in their decision making based on their own knowledge of good fishing locations.i g g gi g In the past, spatiotemporal management measures (such as time-limited fishery closures) have been applied to reduce unwanted catch with varying degrees of success (e.g.22–25) while move-on rules have also been proposed or implemented to influence catch rates of particular vulnerable species to reduce or eliminate discards (e.g.26–28). However, such measures have generally been targeted at individual species without considering associations and interactions among several species. Highly mixed fisheries are complex with spatial, technological and commu- nity interactions combining. The design of spatiotemporal management measures that aim to allow exploitation of high quota stocks while protecting low quota stocks requires understanding these interactions at a scale mean- ingful to managers and fishers. While fisheries surveys and commercial fishing routinely generate a large amount of geo-referenced information on numbers and weight of fish caught, integrating spatiotemporal information from across multiple sources of fisheries-dependent and independent survey data requires an effective framework to reduce and understand the complexities of the system. p y Here, our goal is to develop a framework for understanding these complexities. Framework for analysing spatiotemporal mixed fisheries interactions Framework for analysing spatiotemporal mixed fisheries interactions We present a framework for analysing how far spatiotemporal avoidance can contribute towards mitigating imbalances in quota in mixed fisheries. Fisheries-independent survey data are used to characterise the spatio- temporal dynamics of key components of a fish community by employing a geostatistical Vector Autoregressive Spatiotemporal model (VAST). Therein, a factor analysis decomposition was used to describe trends in spati- otemporal dynamics of the different species as a function of latent variables29 representing spatial variation (9 factors; termed “average” spatial variation) and spatiotemporal variation (9 factors) for encounter probability and positive catch rates (termed “positive density”) separately30. Resultant factor analyses identify community dynamics and drivers common among 9 species, each analysed separately for juvenile and adult stages. We refer to each combination of species and size class as a “species”, and present results for the 18 species through trans- formation of the loading matrices using PCA rotation. This PCA rotation is used to visualise a reduced num- ber of orthogonal factors representing average spatial variation or spatiotemporal variation while explaining the majority of covariation among catch rates, as well as the association of each species with these maps. We refer to the association of each species with a given factor as its “association with this factor”, and the value of each factor at a given location as its “ ‘coefficient’ at that location”. By describing the species dynamics through underlying spatiotemporal factors we can take account of how the factors contribute to affect catches of the species in mixed fisheries. Gaussian Markov Random Fields (GMRFs) capture spatial and temporal dependence within and among species for both encounter probability and positive density31. VAST is set in a mixed modelling framework which allows estimation of fixed effects to account for systematic differences driving encounter and catches, such as differences in sampling efficiency (catchability), while random effects capture the spatiotemporal dynamics of the fish community. www.nature.com/scientificreports/ We do so by (1) implementing a spatiotemporal dimension reduction method that estimates the correlation in catches for multiple species at each fishing location, (2) using the results to draw inference on the fishery-community dynamics, (3) creating a framework to identify common trends among species, and (4) describing the potential for and limitation of spa- tial measures to mitigate unwanted catches in highly mixed fisheries. Mixed fisheries and the EU landing obligationfi Under the new policy, unless fishers can avoid catch of unwanted species they will have to stop fishing 1Marine and Freshwater Research Centre, Galway-Mayo Institute of Technology (GMIT), Dublin Road, Galway, H91 T8NW, Ireland. 2Centre for Environment, Fisheries and Aquaculture Science (Cefas), Pakefield Road, Lowestoft, Suffolk, NR33 0HT, United Kingdom. 3Fisheries Resource Analysis and Monitoring Division, Northwest Fisheries Science Center, National Marine Fisheries Service, NOAA, 2725 Montlake Blvd E, Seattle, Washington, 98112, USA. Correspondence and requests for materials should be addressed to P.J.D. (email: paul.dolder@gmit.ie) Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 1 www.nature.com/scientificreports/ Spatial distributions indicate three species assemblages. A spatial Spatial distributions indicate three species assemblages. A spatial dynamic factor analysis was used to decompose the dominant spatial patterns driving differences in average spatial variation. The first three factors (after PCA rotation) account for 83.7% of the between species variance in the probability of encountering a spe- cies (the “average encounter probability”) and 69% of the explained variance in catch rates on encounter (“aver- age positive density”). A clear spatial pattern can been seen both for average encounter probability and average positive density, with a positive coefficient value associated with the first factor in the inshore north easterly part of the Celtic Sea into the Bristol Channel and Western English Channel, moving to a negative coefficient value offshore in the south-westerly waters (Fig. 1). The species loadings show plaice, sole and whiting to be positively associated with the first factor for average encounter probability while the other species are negatively associated. For average positive density, positive associations are also found for haddock and juvenile cod (weakly positive), indicative of a more inshore distribution for these species. p On the second spatial factor for average encounter probability a north/south split can be seen at approxi- mately 49°N while positive density is more driven by a positive coefficient in the deeper westerly waters as well as some inshore areas. Species loadings for the second factor indicate there are positive associations for juvenile white-bellied anglerfish, juvenile hake, juvenile megrim, plaice and juvenile whiting with average positive density, which may reflect two different spatial distributions in the more offshore and in the inshore areas (Fig. 1).fi lff On the third factor, there is a positive coefficient for the easterly waters for encounter probability and negative coefficient with the westerly waters. This splits the roundfish species (cod, haddock and whiting, that all have a positive association with the third factor for average encounter probability) from the rest of the species (that have a negative association). Positive density is driven by a north/south split (Fig. 1), with positive coefficient values in the northerly areas. Juvenile anglerfishes (white- and black- bellied), cod, juvenile haddock, hake, adult plaice and whiting are also positively associated with the third factor towards the north while adult anglerfishes, adult had- dock, megrim, juvenile plaice and sole have negative loadings reflecting their more southerly distribution (Fig. 1). Dynamics of Celtic Sea fisherieshi yi The highly mixed demersal fisheries of the Celtic Sea are used as a case study. The Celtic Sea is a temperate sea where fisheries are spatially and temporally complex; mixed fisheries are undertaken by several nations using different gear types21,32. Close to 150 species have been identified in the commercial catches of the Celtic Sea, with approximately 30 species dominating the catch33.i pp y p g Our spatiotemporal model is parametrised using catch data from seven fisheries-independent surveys under- taken in the Celtic Sea over the period 1990–2015 (Table S1) and include nine of the main commercial spe- cies: Atlantic cod (Gadus morhua), Atlantic haddock (Melanogrammus aeglefinus), Atlantic whiting (Merlangius merlangus), European Hake (Merluccius merluccius), white-bellied anglerfish (Lophius piscatorius), black-bellied anglerfish (Lophius budegassa), megrim (Lepidorhombus whiffiagonis), European plaice (Pleuronectes platessa) and common sole (Solea solea). These species comprise over 60% of landings by towed fishing gears for the area (average 2011–201534). Each species was separated into juvenile and adult size classes based on their legal mini- mum conservation reference size (Table S2).hf The data were analysed to understand how the different associations among species (combination of species and size class) form distinct assemblages with common drivers of spatiotemporal distributions, and how these affect catch compositions for fishers operating in mixed fisheries. We consider how these have changed over time, and the implications for mixed fisheries in managing catches of quota species under the EU landing obligation. 2 Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w www.nature.com/scientificreports/ Figure 1. Factor values for the first three factors for (A) Average encounter probability and (B) Average positive density for the species (outer figures) and spatially (inner figures). Red: positive association to the factor, Blue: negative association. Figure 1. Factor values for the first three factors for (A) Average encounter probability and (B) Average positive density for the species (outer figures) and spatially (inner figures). Red: positive association to the factor, Blue: negative association. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w Results Using relatively few factors in a spatial dynamic factor analysis the Celtic Sea demersal fish community can be partitioned into three species assemblages (roundfish, flatfish and deeper water species). Within these assem- blages there are common trends in spatiotemporal distributions in encounter probability and positive density, which can be partitioned into time invariant (“average effect”) spatial trends and time variant (“spatiotemporal”) trends. We show through presentation of factor coefficients that time invariant trends may be linked to physical characteristics of the system including depth and predominant substrate type, while species loadings on to time varying spatial trends show changes in distribution of species over time to be similar within an assemblage. We demonstrate how this information can be used to help inform spatial targeting and avoidance of the different assemblages. More nuanced differences in spatiotemporal distributions exist within an assemblage presenting a greater challenge to spatially separate catches. Yet we show how this information may be utilised by managers and fishers to better match catch to quota in highly mixed fisheries through changes in gear and locations fished. Spatial distributions indicate three species assemblages. A spatial 1), the interannual differences in factor coefficients show less structure (Figs S5 and S6). These interannual differences are important as they reflect the ability of fishers to predict where they can target or avoid species from one year to the next, without which it may be difficult to balance catches with available quota and avoid unwanted catch.fi yfi q Spatiotemporal factor coefficients for encounter probability and positive density did not show the same spatial pattern driving species distributions from year to year, but when the first two factor loadings are plotted clear relationships in species association with spatiotemporal factor coefficients identify the three different assem- blages (Fig. 2). The same factors appear to drive spatiotemporal (interannual changes in) distributions of megrim, anglerfish species and hake (the deeper water species, forming an assemblage negatively associated with the sec- ond axes of Fig. 2) and the roundfish and flatfish (two assemblages more positively associated with the second axes of Fig. 2A). For spatiotemporal positive density (Fig. 2B) cod, haddock and whiting (the roundfish species) are separated from plaice, sole (the flatfish) and the deeper water assemblage. As such, it can be predicted that higher catches of a species within a assemblage (e.g. cod in roundfish) would be expected when catching another species within that assemblage (e.g. whiting in roundfish). This suggests that one or more common environmen- tal drivers are influencing the distributions of the assemblages, and that driver differentially affects the different assemblages. Temperature is often included as a covariate in species distribution models, but was found not to contribute to the variance in the first factor coefficients (Fig. S6, no correlations found for either spatiotemporal encounter probability or positive density) and so was not included as a covariate in the final model. Covariance in spatiotemporal abundance within species assemblages. To gain greater insight into the community dynamics we considered how species covary in space and time through correlations among species. Pearson correlation coefficients for the modelled average spatial encounter probability (Fig. 3A) show clear strong associations between adult and juvenile size classes for all species (>0.75 for all species except hake, 0.56). Spatial distributions indicate three species assemblages. A spatial l While this exploratory factor analysis models unobserved drivers of distribution, we considered what might be driving the differences seen in the spatial factor coefficients and species loadings. The first factor was highly cor- related with log(depth) for both average encounter probability coefficients (−0.85, CI = −0.88 to −0.81; Fig. S1) and average positive density coefficients (−0.71, CI = −0.77 to −0.65; Fig. S2). A random forest classification tree Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 3 www.nature.com/scientificreports/ Figure 2. Position of each species on the first two axes from the factor analysis for (A) spatiotemporal encounter probability and (B) spatiotemporal positive density. Fish images from The Fisherman/Shutterstock. com and Richard Griffin/Shutterstock.com. Figure 2. Position of each species on the first two axes from the factor analysis for (A) spatiotemporal encounter probability and (B) spatiotemporal positive density. Fish images from The Fisherman/Shutterstock. com and Richard Griffin/Shutterstock.com. assigned 80% of the variance in the first factor for average encounter probability to depth and predominant sub- strate type, with the majority (86%) of the variance explained by depth. The variance explained by these variables dropped to 25% on the second factor with a more even split between depth and substrate, while explaining 60% of the variance on the third factor. For average positive density, the variables explained less of the variance with 62%, 35%, and 31% for each of the factors, respectively. It is clear that depth and to a lesser extent substrate are important variables for describing the main driver of similarities and differences in distributions and abundances for the different species. The first factor corre- lates strongly with these variables, despite them not explicitly being incorporated in the model. While depth and substrate were incorporated as covariates in an alternative model formulation (see Methods), they were not found to improve predictions as the random fields adequately captured the influence of these variables on spatial variation in abundance. The utility of these variables as predictors of species distributions has been identified in other marine species distribution models35. The advantage to the approach taken here is that, where such data is unavailable at an appropriate spatial resolution, the spatial factor analysis can adequately characterised the species spatial dynamics. Species assemblages show similar spatiotemporal patterns. While there are clear spatial patterns in the factor coefficients describing differences in average encounter probability and positive density (Fig. Spatial distributions indicate three species assemblages. A spatial Among species, hierarchical clustering identified the same three common species-groups as our visual inspection of factor loadings above, with roundfish (cod, haddock, whiting) closely grouped, with correlations for adult cod with adult haddock and adult whiting of 0.73 and 0.5 respectively, while adult haddock with adult whiting was 0.63 (Fig. 3A). Spatial distributions indicate three species assemblages. A spatial Flatfish (plaice and sole) are also strongly correlated with adult plaice and sole having Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 4 www.nature.com/scientificreports/ ple_adu ple_juv sol_adu sol_juv pisc_adu pisc_juv hke_adu hke_juv meg_juv meg_adu bud_adu bud_juv whg_adu whg_juv cod_adu cod_juv had_adu had_juv ple_adu ple_juv sol_adu sol_juv pisc_adu pisc_juv hke_adu hke_juv meg_juv meg_adu bud_adu bud_juv whg_adu whg_juv cod_adu cod_juv had_adu had_juv (A) Spatial Encounter probability 1 0.88 0.75 0.51 −0.31 −0.06 −0.33 0.2 −0.01 −0.19 −0.27 −0.2 0.48 0.55 −0.09 0.26 0.02 0.35 0.88 1 0.69 0.64 −0.35 −0.2 −0.4 −0.01 −0.12 −0.39 −0.33 −0.38 0.25 0.5 −0.29 0.14 −0.28 0.06 0.75 0.69 1 0.86 −0.28 0.07 −0.37 0.28 −0.01 −0.28 −0.26 −0.21 0.45 0.63 −0.15 0.12 −0.02 0.33 0.51 0.64 0.86 1 −0.4 −0.2 −0.57 −0.02 −0.31 −0.57 −0.45 −0.53 0.15 0.51 −0.22 0.05 −0.36 −0.13 −0.31 −0.35 −0.28 −0.4 1 0.74 0.52 0.25 0.28 0.59 0.72 0.65 −0.02 −0.35 0.34 0.15 0.32 0.23 −0.06 −0.2 0.07 −0.2 0.74 1 0.73 0.68 0.49 0.56 0.7 0.81 0.34 0.06 0.31 0.26 0.41 0.58 −0.33 −0.4 −0.37 −0.57 0.52 0.73 1 0.54 0.61 0.7 0.68 0.81 0.13 −0.16 0.44 0.27 0.54 0.53 0.2 −0.01 0.28 −0.02 0.25 0.68 0.54 1 0.75 0.66 0.64 0.72 0.48 0.45 0.36 0.34 0.53 0.75 −0.01 −0.12 −0.01 −0.31 0.28 0.49 0.61 0.75 1 0.85 0.82 0.83 0 −0.02 0.01 −0.14 0.44 0.63 −0.19 −0.39 −0.28 −0.57 0.59 0.56 0.7 0.66 0.85 1 0.88 0.88 0.07 −0.19 0.41 0.12 0.69 0.63 −0.27 −0.33 −0.26 −0.45 0.72 0.7 0.68 0.64 0.82 0.88 1 0.93 −0.17 −0.29 0.14 −0.09 0.32 0.39 −0.2 −0.38 −0.21 −0.53 0.65 0.81 0.81 0.72 0.83 0.88 0.93 1 0.05 −0.21 0.21 −0.01 0.48 0.59 0.48 0.25 0.45 0.15 −0.02 0.34 0.13 0.48 0 0.07 −0.17 0.05 1 0.81 0.49 0.72 0.63 0.73 0.55 0.5 0.63 0.51 −0.35 0.06 −0.16 0.45 −0.02 −0.19 −0.29 −0.21 0.81 1 0.17 0.53 0.24 0.45 −0.09 −0.29 −0.15 −0.22 0.34 0.31 0.44 0.36 0.01 0.41 0.14 0.21 0.49 0.17 1 0.87 0.72 0.42 0.26 0.14 0.12 0.05 0.15 0.26 0.27 0.34 −0.14 0.12 −0.09 −0.01 0.72 0.53 0.87 1 0.57 0.42 0.02 −0.28 −0.02 −0.36 0.32 0.41 0.54 0.53 0.44 0.69 0.32 0.48 0.63 0.24 0.72 0.57 1 0.85 0.35 0.06 0.33 −0.13 0.23 0.58 0.53 0.75 0.63 0.63 0.39 0.59 0.73 0.45 0.42 0.42 0.85 1 sol_adu sol_juv ple_adu ple_juv had_juv pisc_juv whg_juv hke_juv meg_adu meg_juv bud_juv hke_adu cod_juv pisc_adu bud_adu cod_adu had_adu whg_adu sol_adu sol_juv ple_adu ple_juv had_juv pisc_juv whg_juv hke_juv meg_adu meg_juv bud_juv hke_adu cod_juv pisc_adu bud_adu cod_adu had_adu whg_adu (B) Spatial Density 1 0.62 0.31 0.2 0.24 0.24 0.41 −0.03 −0.13 −0.61 −0.61 −0.47 −0.14 −0.02 0.09 −0.03 0.21 0.23 0.62 1 0.1 0.43 0.04 0.27 0.23 −0.26 0.35 −0.14 −0.41 −0.26 −0.16 0.18 0.34 −0.12 0.38 0.14 0.31 0.1 1 0.64 0.43 0.19 0.5 0.17 −0.33 −0.36 −0.21 −0.35 0.1 −0.38 −0.31 −0.08 0.17 0.05 0.2 0.43 0.64 1 0.25 0.47 0.6 0.03 −0.06 0.07 0.12 −0.34 −0.18 −0.48 −0.23 −0.63 −0.01 0 0.24 0.04 0.43 0.25 1 0.43 0.4 0.08 0.14 −0.07 0.16 −0.21 0.18 −0.14 −0.26 0.1 0.56 0.44 0.24 0.27 0.19 0.47 0.43 1 0.5 −0.03 −0.09 −0.02 0.14 0.22 0.5 0.29 0.1 −0.08 0.46 0.67 0.41 0.23 0.5 0.6 0.4 0.5 1 0.21 −0.37 −0.48 −0.13 −0.45 0.04 −0.35 −0.02 −0.12 0.07 0.47 −0.03 −0.26 0.17 0.03 0.08 −0.03 0.21 1 0.26 0.21 0.46 −0.17 −0.17 −0.35 0.15 −0.04 −0.18 −0.13 −0.13 0.35 −0.33 −0.06 0.14 −0.09 −0.37 0.26 1 0.67 0.47 0.1 −0.23 0.1 0.23 −0.06 0.24 −0.25 −0.61 −0.14 −0.36 0.07 −0.07 −0.02 −0.48 0.21 0.67 1 0.79 0.2 −0.29 −0.13 0.07 −0.42 −0.08 −0.31 −0.61 −0.41 −0.21 0.12 0.16 0.14 −0.13 0.46 0.47 0.79 1 0.31 −0.08 −0.26 −0.28 −0.41 −0.26 −0.28 −0.47 −0.26 −0.35 −0.34 −0.21 0.22 −0.45 −0.17 0.1 0.2 0.31 1 0.74 0.71 −0.13 0.3 0.06 −0.05 −0.14 −0.16 0.1 −0.18 0.18 0.5 0.04 −0.17 −0.23 −0.29 −0.08 0.74 1 0.74 −0.01 0.63 0.47 0.48 −0.02 0.18 −0.38 −0.48 −0.14 0.29 −0.35 −0.35 0.1 −0.13 −0.26 0.71 0.74 1 0.4 0.67 0.55 0.41 0.09 0.34 −0.31 −0.23 −0.26 0.1 −0.02 0.15 0.23 0.07 −0.28 −0.13 −0.01 0.4 1 0.42 0.47 0.49 −0.03 −0.12 −0.08 −0.63 0.1 −0.08 −0.12 −0.04 −0.06 −0.42 −0.41 0.3 0.63 0.67 0.42 1 0.58 0.47 0.21 0.38 0.17 −0.01 0.56 0.46 0.07 −0.18 0.24 −0.08 −0.26 0.06 0.47 0.55 0.47 0.58 1 0.72 0.23 0.14 0.05 0 0.44 0.67 0.47 −0.13 −0.25 −0.31 −0.28 −0.05 0.48 0.41 0.49 0.47 0.72 1 Figure 3. Spatial distributions indicate three species assemblages. A spatial Inter-species correlations for (A) spatial encounter probability over all years and (B) spatial positive density. Species are clustered into three groups based on a hierarchical clustering method with non-significant correlations (the Confidence Interval [±1.96 * SEs] spanned zero) left blank. a coefficient of 0.75. The final group are principally the species found in the deeper waters (hake, megrim and both anglerfish species) with megrim strongly associated with the black-bellied anglerfish species (0.88). Negative relationships were found between plaice and sole, and white-bellied anglerfish (−0.31 and −0.28 for the adult size class), black-bellied anglerfish (−0.27, −0.26 for the adult size class) and hake (−0.33, −0.37) (Fig. 3A) indicating spatial separation in distributions, with the flatfish found more inshore. This underscores the correlations among species seen in associations of each species with factors, with three distinct assemblages being confirmed.fii p p g gi Correlation coefficients for the average positive density (Fig. 3B) show fewer significant positive or negative relationships among species than for encounter probability, but still evident are the strong correlation among the roundfish with higher catches of cod correlated with higher catches of haddock (0.58) and whiting (0.47), as well as the two anglerfish species (0.71 for white-bellied and 0.44 for black-bellied) and hake (0.73). Similarly, plaice and sole are correlated (0.31) and higher catches of one would expect to see higher catches of the other, but also higher catches of some juvenile size classes of roundfish (whiting and haddock) and anglerfish species. Negative correlation of juvenile megrim, anglerfish (budegassa) and hake with adult sole (−0.61, −0.61 and −0.47 respec- tively), plaice (−0.36 and −0.35 for megrim and hake only) indicate high catches of one can predict low catches of the other successfully. y To understand how stable relationships between catches of pairs of species were from one year to the next, we regressed the correlation coefficients for the average spatial correlations between pairs for species x and species y across all years with those of the spatiotemporal population correlations, representing how correlations between species x and species y change from year to year (Fig. S9). The correlations were 0.60 (0.52–0.66) and 0.47 (0.38–0.55) for encounter probability and positive density respectively (Fig. S9a,b). Spatial distributions indicate three species assemblages. A spatial ple_adu ple_juv sol_adu sol_juv pisc_adu pisc_juv hke_adu hke_juv meg_juv meg_adu bud_adu bud_juv whg_adu whg_juv cod_adu cod_juv had_adu had_juv ple_adu ple_juv sol_adu sol_juv pisc_adu pisc_juv hke_adu hke_juv meg_juv meg_adu bud_adu bud_juv whg_adu whg_juv cod_adu cod_juv had_adu had_juv (A) Spatial Encounter probability 1 0.88 0.75 0.51 −0.31 −0.06 −0.33 0.2 −0.01 −0.19 −0.27 −0.2 0.48 0.55 −0.09 0.26 0.02 0.35 0.88 1 0.69 0.64 −0.35 −0.2 −0.4 −0.01 −0.12 −0.39 −0.33 −0.38 0.25 0.5 −0.29 0.14 −0.28 0.06 0.75 0.69 1 0.86 −0.28 0.07 −0.37 0.28 −0.01 −0.28 −0.26 −0.21 0.45 0.63 −0.15 0.12 −0.02 0.33 0.51 0.64 0.86 1 −0.4 −0.2 −0.57 −0.02 −0.31 −0.57 −0.45 −0.53 0.15 0.51 −0.22 0.05 −0.36 −0.13 −0.31 −0.35 −0.28 −0.4 1 0.74 0.52 0.25 0.28 0.59 0.72 0.65 −0.02 −0.35 0.34 0.15 0.32 0.23 −0.06 −0.2 0.07 −0.2 0.74 1 0.73 0.68 0.49 0.56 0.7 0.81 0.34 0.06 0.31 0.26 0.41 0.58 −0.33 −0.4 −0.37 −0.57 0.52 0.73 1 0.54 0.61 0.7 0.68 0.81 0.13 −0.16 0.44 0.27 0.54 0.53 0.2 −0.01 0.28 −0.02 0.25 0.68 0.54 1 0.75 0.66 0.64 0.72 0.48 0.45 0.36 0.34 0.53 0.75 −0.01 −0.12 −0.01 −0.31 0.28 0.49 0.61 0.75 1 0.85 0.82 0.83 0 −0.02 0.01 −0.14 0.44 0.63 −0.19 −0.39 −0.28 −0.57 0.59 0.56 0.7 0.66 0.85 1 0.88 0.88 0.07 −0.19 0.41 0.12 0.69 0.63 −0.27 −0.33 −0.26 −0.45 0.72 0.7 0.68 0.64 0.82 0.88 1 0.93 −0.17 −0.29 0.14 −0.09 0.32 0.39 −0.2 −0.38 −0.21 −0.53 0.65 0.81 0.81 0.72 0.83 0.88 0.93 1 0.05 −0.21 0.21 −0.01 0.48 0.59 0.48 0.25 0.45 0.15 −0.02 0.34 0.13 0.48 0 0.07 −0.17 0.05 1 0.81 0.49 0.72 0.63 0.73 0.55 0.5 0.63 0.51 −0.35 0.06 −0.16 0.45 −0.02 −0.19 −0.29 −0.21 0.81 1 0.17 0.53 0.24 0.45 −0.09 −0.29 −0.15 −0.22 0.34 0.31 0.44 0.36 0.01 0.41 0.14 0.21 0.49 0.17 1 0.87 0.72 0.42 0.26 0.14 0.12 0.05 0.15 0.26 0.27 0.34 −0.14 0.12 −0.09 −0.01 0.72 0.53 0.87 1 0.57 0.42 0.02 −0.28 −0.02 −0.36 0.32 0.41 0.54 0.53 0.44 0.69 0.32 0.48 0.63 0.24 0.72 0.57 1 0.85 0.35 0.06 0.33 −0.13 0.23 0.58 0.53 0.75 0.63 0.63 0.39 0.59 0.73 0.45 0.42 0.42 0.85 1 sol_adu sol_juv ple_adu ple_juv had_juv pisc_juv whg_juv hke_juv meg_adu meg_juv bud_juv hke_adu cod_juv pisc_adu bud_adu cod_adu had_adu whg_adu sol_adu sol_juv ple_adu ple_juv had_juv pisc_juv whg_juv hke_juv meg_adu meg_juv bud_juv hke_adu cod_juv pisc_adu bud_adu cod_adu had_adu whg_adu (B) Spatial Density 1 0.62 0.31 0.2 0.24 0.24 0.41 −0.03 −0.13 −0.61 −0.61 −0.47 −0.14 −0.02 0.09 −0.03 0.21 0.23 0.62 1 0.1 0.43 0.04 0.27 0.23 −0.26 0.35 −0.14 −0.41 −0.26 −0.16 0.18 0.34 −0.12 0.38 0.14 0.31 0.1 1 0.64 0.43 0.19 0.5 0.17 −0.33 −0.36 −0.21 −0.35 0.1 −0.38 −0.31 −0.08 0.17 0.05 0.2 0.43 0.64 1 0.25 0.47 0.6 0.03 −0.06 0.07 0.12 −0.34 −0.18 −0.48 −0.23 −0.63 −0.01 0 0.24 0.04 0.43 0.25 1 0.43 0.4 0.08 0.14 −0.07 0.16 −0.21 0.18 −0.14 −0.26 0.1 0.56 0.44 0.24 0.27 0.19 0.47 0.43 1 0.5 −0.03 −0.09 −0.02 0.14 0.22 0.5 0.29 0.1 −0.08 0.46 0.67 0.41 0.23 0.5 0.6 0.4 0.5 1 0.21 −0.37 −0.48 −0.13 −0.45 0.04 −0.35 −0.02 −0.12 0.07 0.47 −0.03 −0.26 0.17 0.03 0.08 −0.03 0.21 1 0.26 0.21 0.46 −0.17 −0.17 −0.35 0.15 −0.04 −0.18 −0.13 −0.13 0.35 −0.33 −0.06 0.14 −0.09 −0.37 0.26 1 0.67 0.47 0.1 −0.23 0.1 0.23 −0.06 0.24 −0.25 −0.61 −0.14 −0.36 0.07 −0.07 −0.02 −0.48 0.21 0.67 1 0.79 0.2 −0.29 −0.13 0.07 −0.42 −0.08 −0.31 −0.61 −0.41 −0.21 0.12 0.16 0.14 −0.13 0.46 0.47 0.79 1 0.31 −0.08 −0.26 −0.28 −0.41 −0.26 −0.28 −0.47 −0.26 −0.35 −0.34 −0.21 0.22 −0.45 −0.17 0.1 0.2 0.31 1 0.74 0.71 −0.13 0.3 0.06 −0.05 −0.14 −0.16 0.1 −0.18 0.18 0.5 0.04 −0.17 −0.23 −0.29 −0.08 0.74 1 0.74 −0.01 0.63 0.47 0.48 −0.02 0.18 −0.38 −0.48 −0.14 0.29 −0.35 −0.35 0.1 −0.13 −0.26 0.71 0.74 1 0.4 0.67 0.55 0.41 0.09 0.34 −0.31 −0.23 −0.26 0.1 −0.02 0.15 0.23 0.07 −0.28 −0.13 −0.01 0.4 1 0.42 0.47 0.49 −0.03 −0.12 −0.08 −0.63 0.1 −0.08 −0.12 −0.04 −0.06 −0.42 −0.41 0.3 0.63 0.67 0.42 1 0.58 0.47 0.21 0.38 0.17 −0.01 0.56 0.46 0.07 −0.18 0.24 −0.08 −0.26 0.06 0.47 0.55 0.47 0.58 1 0.72 0.23 0.14 0.05 0 0.44 0.67 0.47 −0.13 −0.25 −0.31 −0.28 −0.05 0.48 0.41 0.49 0.47 0.72 1 Figure 3. Spatial distributions indicate three species assemblages. A spatial Inter-species correlations for (A) spatial encounter probability over all years and (B) spatial positive density. Species are clustered into three groups based on a hierarchical clustering method with non-significant correlations (the Confidence Interval [±1.96 * SEs] spanned zero) left blank. Spatial distributions indicate three species assemblages. A spatial These indicate generally predictable rela- tionships between species from one year to the next and suggests that a positive or negative correlation between two species is likely to persist from one year to the next, and that species are consistently correlated in hauls. However, the regressions between the spatial correlations and the spatiotemporal correlations shows high var- iance (R2 = 0.36 and 0.22 respectively), indicating that the scale of these relationships do change from one year to the next. This unpredictability would have implications for the fishery if, for example, catches of an unwanted species increased when caught with a target species above a level expected in the fishery potentially leading to challenges for fishers when trying to balance catch with quotas in mixed fisheries. It can be seen in the spatial fac- tor maps that there are subtle differences in patterns in spatial factor coefficients from one year to the next (Figs S4 and S5), indicating changes may be driven by temporally changing environmental factors and species behaviour. Potential to separate catches within assemblages under the landing obligation. The analysis shows the interdependence within three assemblages of roundfish, flatfish and deeper water species, where catch- ing one species within the group indicates a high probability of catching the other species. This has important implications for how spatial avoidance can be used to support implementation of the EU’s landing obligation. If production from mixed fisheries is to be maximised, decoupling catches of species between and within the groups will be key. For example, asking where the maximal separation in the densities of two coupled species is likely to occur? To address this requirement, we map the difference in spatial distribution within a species-group for each pair of species for a single year (2015; Fig. 4). p p g y g Cod had a more north-westerly distribution than haddock and a more westerly distributed than whiting roughly delineated by the 7°W line (Fig. 4A). Whiting appeared particularly concentrated in an area between 51 and 52°N and 5 and 7°W, which can be seen by comparing the whiting distribution with both cod (Fig. 4B) and haddock Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 5 www.nature.com/scientificreports/ Figure 4. Differences in the standardised spatial density for pairs of species and expected catch rates for two different gears at three different locations in 2015. Spatial distributions indicate three species assemblages. A spatial A, B and C in sub-figure (H) correspond to the spatial locations illustrated in sub-figure (G). Figure 4. Differences in the standardised spatial density for pairs of species and expected catch rates for two different gears at three different locations in 2015. A, B and C in sub-figure (H) correspond to the spatia locations illustrated in sub-figure (G). (Fig. 4C). For the deeper water species hake are more densely distributed in two locations around 10 W and 48 N and 12 W and 50 N compared to the anglerfish species (anglerfishes have been presented together as they are jointly managed under a single quota) and megrim, which were more widely spatially distributed (Fig. 4D,E). Megrim has a fairly even density across the modelled area as indicated by the large amount of white space in Fig. 4E. For angler- fishes and megrim (Fig. 4F), anglerfishes have a more easterly distribution than megrim. For the flatfish species plaice and sole (Fig. 4G), plaice appear to be more densely distributed along the coastal areas of Ireland and Britain, while sole are more densely distributed in the Southern part of the English Channel along the coast of France.if Predicted catch distribution from a “typical” otter trawl gear and beam trawl fishing at three different locations highlights the differences fishing gear makes on catches (Fig. 4H). Both gear selectivity and location fished have important effects on the catch composition; in the inshore area (location “A”) plaice and sole are the two main species in the catch reflecting their distribution and abundance, though the otter trawl gear catches a greater proportion of plaice to sole than the beam trawl. The area between Britain and Ireland (location “B”) has a greater contribution of whiting, haddock, cod, hake and anglerfishes in the catch with the otter trawl catching a greater proportion of the roundfish, haddock, whiting and cod while the beam trawl catches more anglerfishes and megrim. The offshore area has a higher contribution of megrim, anglerfishes and hake with the otter trawl catch- ing a greater share of hake and the beam trawl a greater proportion of megrim. Megrim dominates the catch for both gears in location “C”, reflecting its relative abundance in the area irrespective of the gear deployed. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w Discussion Our study is framed by the problem of addressing the scientific challenges of implementing the landing obliga- tion for mixed fisheries. In application to the Celtic Sea, we have identified spatial separation of three distinct assemblages (roundfish, flatfish and deeper water species) while showing that only subtle differences exist in dis- tributions within assemblages. The differences in catch compositions between gears at the same location (Fig. 4H) show that changing fishing methods affects catch, yet that differences in catches between locations are likely to be more important. For example, beam trawls fishing at the inshore locations (e.g. location “A” in Fig. 4) are likely to predominately catch plaice and sole, yet switching to the offshore locations (e.g. location “C”) would likely yield greater catches of megrim and anglerfishes. Such changes in spatial fishing patterns are likely to play an important role in supporting implementation of the landing obligation. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 6 www.nature.com/scientificreports/ More challenging is within-group spatial separation due to significant overlap in spatial distributions for the species, driven by common environmental factors. Subtle changes may yield some benefit in changing catch composition, yet the outcome is likely to be much more difficult to predict. For example, subtle differences in the distribution of cod, haddock and whiting can be seen in Fig. 4A–C, showing spatial separation of catches is much more challenging and likely to require support from other measures such as changes to the selectivity char- acteristics of gear36. For example we identified a spatial overlap of flatfish with juvenile roundfish in our species correlations (Fig. 3); reducing catches of incidental bycatch on the main target fishing grounds will likely require adaptations to fishing gear to address bycatch without significant economic impacts on the fishery.f pi g g y gi pi y A role that science could play in supporting effectiveness of spatiotemporal avoidance would be to provide probabilistic advice on hotspots for species occurrence and high species density, which can inform fishing deci- sions. Previous modelling studies have shown how spatiotemporal models could improve predictions of high ratios of bycatch to target species37–39, and geostatistical models are well suited to this as they incorporate spatial dependency while providing for probabilities to be drawn from posterior distributions of the parameter estimates. Discussion We posit that such advice on “hot spots” as a supportive measure to incentivise avoidance of areas of high bycatch risk could be enhanced by integrating data obtained directly from commercial fishing vessels rapidly while modelling densities at small time scales (e.g., weekly). Short-term forecasts of distribution could inform fishing choices while also capturing seasonal differences in distributions, akin to weather forecasting. Advice informed by a model including a seasonal or real-time component could inform optimal policies for time-area closures, move-on rules or even as informal information to be utilised by fishers directly without the need for costly contin- uous data collection on environmental parameters, but by using the “vessels-as-laboratories” approach. p y g pp An important question for the implementation of the EU’s landing obligation is how far spatial avoidance can go to achieving catch balancing in fisheries. Our model captures differences between location fished for two gear types and their broad scale effect on catch composition, information crucial for managers in implementing the landing obligation. It is likely, however, that this analysis reflects a lower bound on the utility of spatial avoidance as fine-scale behavioural decisions such as time-of-day, gear configuration and location choices can also be used to affect catch40,41. Results of empirical studies undertaken elsewhere5,6 suggest limits to the effectiveness of spatial avoidance in situ. For example, differences in ability to change catch composition have been observed for different fleets; in the North Sea targeting ability was found to differ between otter and beam trawlers as well as between vessels of different sizes42. The particular socioeconomic circumstances for individual vessels is therefore impor- tant to take account when considering the effectiveness of spatial targetting and avoidance.ii gf p g g Under the landing obligation the balance of risk-reward for trip level fishing decisions about where to fish may change. For example, are fishers likely to fish in “safe” areas where its known there are lower catches of the target species but also decreased risk of encountering bycatch? How do decisions about level of risk affect the likelihood of overshooting available quota and potential profit and losses for individual trips? Set in this context, the param- eter estimates could be used to simulate from a distribution of catches in the fishery at different locations and therefore inform on the possibility of extreme catch events and potential consequences for overshooting quotas. Discussion Alternatively, where fisheries data is available with factors such as weather, quota uptake and previous catches these could be included as covariates in the model to help identify causes for high bycatch events. This informa- tion may be of interest in identifying optimum strategies, or used in future work to model closure risks for fish- eries operating in different locations and conditions given quota constraints. Such analyses on risk and decision making are likely to hinge on micro-level decisions by fishers and would be a useful compliment to broader scale considerations such as those detailed here. Our framework allows for a quantitative understanding of the broad scale global production set available to fishers43 and thus the extent to which they can alter catch compositions while operating in a mixed fishery. Simulations of spatial effort allocation scenarios based on the production sets derived from the model estimates could be used as inputs to fisher behavioural models to allow for the identification of lower bounds of optimum spatial harvest strategies. Modelling different spatial strategies at the individual or fishery level would provide managers with an information base to examine trade-offs in quota setting, thus providing a scientific basis to assessing the ability of technical measures to meet the goal of maximising catches in mixed fisheries within single stock quota constraints7. Additionally, the correlations among species could provide information on fisheries at risk of capturing protected, endangered or threatened species such as elasmobranchs, and allow identification of areas where there are high ratios of protected to target species.i g p g p Complex environmental, fishery and community drivers of distribution for groups of species highlights the scale of the challenge in separating catches within the assemblages using spatial management measures. This has important implications for management of mixed fisheries under the EU landing obligation. Our analysis identi- fies where it may be easier to separate catches of species (among groups) and where it is more challenging (within groups). We propose that the dimension-reduction framework presented in Figs 1–4 provides a viable route to reducing the complexity of highly mixed fisheries. This can allow informed management discussion over more traditional anecdotal knowledge of single-species distribution in space and time. Model structure. h The Delta-Gamma formulation is then: where p(si, ci, ti) is the predictor for encounter probability for observation i, at location s for species c and time t and r(si, ci, ti) is similarly the predictor for the positive density. β(ci, ti) is the intercept, ω(si, ci) the spatial varia- tion at location s for factor f, with Lω(ci, f) the loading matrix for spatial covariation among species. ε(si, ci, ti) is the linear predictor for spatiotemporal variation, with Lε(ci, f) the loading matrix for spatiotemporal covariance among species and δ(ci, vi) the contribution of catchability covariates for the linear predictor with Qc v,i i the catch- ability covariates for species c and vessel v; * can be either p for probability of encounter or r for positive density. The Delta-Gamma formulation is then: λ λ = = − = \| > = ⋅ ⋅ − Γ − Pr C p Pr C c c p c exp c ( 0) 1 ( 0) ( ) (4) k k k 1 (4) for the probability p of a non-zero catch C given a gamma distribution for for the positive catch with a rate param- eter λ and shape parameter k. Spatiotemporal variation. The spatiotemporal variation is modelled using Gaussian Markov Random Fields (GMRF) where observations are correlated in space through a Matérn covariance function with the parameters estimated within the model. Here, the correlation decays smoothly over space the further from the location and includes geometric anisotropy to reflect that correlations may decline in one direction faster than another (e.g. moving offshore)31. The best fit estimated an anisotropic covariance where the correlations were stronger in a north-east - south-west direction, extending approximately 97 km and 140 km before correlations for encounter probability and positive density reduced to <10%, respectively (Fig. S10). Incorporating the spatiotemporal cor- relations among species provides more efficient use of the data as inference can be made about poorly sampled locations from the covariance structure. Model structure. h Spatiotemporal encounter probability and positive catch rates are modelled sepa- rately with spatiotemporal encounter probability modelled using a logit-link linear predictor; ∑ ∑ ∑ β ω ε δ = + + + ω ε = = = ω ε logit p s c t c t L c f s f L c f s f t v Q c v [ ( , , )] ( , ) ( , ) ( , ) ( , ) ( , , ) ( ) ( , ) (2) i i i p i i f n i p i f n i p i i v n p p i i 1 1 1 v ∑ ∑ β ω ε = + + ω ε = = ω ε logit p s c t c t L c f s f L c f s f t [ ( , , )] ( , ) ( , ) ( , ) ( , ) ( , , ) i i i p i i f n i p i f n i p i i 1 1 ∑δ + = v Q c v ( ) ( , ) v n p p i i 1 v (2) and positive catch rates modelling using a gamma- distribution30. Model structure. h Model structure. VAST (software in the R statistical programming language can be found here: www.github. com/james-thorson/VAST) implements a delta-generalised linear mixed modelling (GLMM) framework that takes account of spatiotemporal correlations among species through implementation of a spatial dynamic factor analysis (SDFA). Spatial variation is captured through a Gaussian Markov Random Field, while we model random variation among species and years. Covariates affecting catchability (to account for differences between fishing surveys) and density (to account for environmental preferences) can be incorporated for predictions of presence and positive density. The following briefly summarises the key methods implemented in the VAST framework. For full details see Thorson et al.44. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 7 www.nature.com/scientificreports/ SDFA. A spatial dynamic factor analysis incorporates advances in joint dynamic species distribution models44 to take account of associations among species by modelling response variables as a multivariate process. This is achieved through implementing a factor analysis decomposition where common latent trends are estimated so that the number of common trends is less than the number of species modelled. The factor coefficients are then associated through loadings for each factor that return a positive or negative association of one or more species with any location. Log-density of any species is then be described as a linear combination of factors and loadings: ∑ ∑ θ ψ γ χ = + = = s t L s t s t ( , ) ( , ) ( , ) (1) c j n c j j k n k c k 1 , 1 , j k (1) where θc(s, t) represents log-density for species c at site s at time t, ψj is the coefficient for factor j, Lc,j the loading matrix representing association of species c with factor j and γk,cχk(s, t) the linear effect of covariates at each site and time45.h The factor analysis can summarize community dynamics and identify which species and life-stages have sim- ilar spatiotemporal patterns. This allows inference regarding species distributions and abundance of poorly sam- pled species through association with other species, and also provides estimates of spatiotemporal correlations among species45. Estimation of abundances. Model structure. h ∑ ∑ β ω ε = + + ω ε = = ω ε log r s c t c t L c f s f L c f s f t [ ( , , )] ( , ) ( , ) ( , ) ( , ) ( , , ) i i i r i i f n i r i f n i r i i 1 1 ∑ ∑ ∑ β ω ε δ = + + + ω ε = = = ω ε log r s c t c t L c f s f L c f s f t v Q c v [ ( , , )] ( , ) ( , ) ( , ) ( , ) ( , , ) ( ) ( , ) (3) i i i r i i f n i r i f n i r i i v n r r i i 1 1 1 v ∑ ∑ ∑ β ω ε δ = + + + ω ε = = ω ε log r s c t c t L c f s f L c f s f t v Q c v [ ( , , )] ( , ) ( , ) ( , ) ( , ) ( , , ) ( ) ( , ) i i i r i i f n i r i f n i r i i n r r i i 1 1 v ∑δ + = v Q c v ( ) ( , ) (3) v r r i i 1 v (3) where p(si, ci, ti) is the predictor for encounter probability for observation i, at location s for species c and time t and r(si, ci, ti) is similarly the predictor for the positive density. β(ci, ti) is the intercept, ω(si, ci) the spatial varia- tion at location s for factor f, with Lω(ci, f) the loading matrix for spatial covariation among species. ε(si, ci, ti) is the linear predictor for spatiotemporal variation, with Lε(ci, f) the loading matrix for spatiotemporal covariance among species and δ(ci, vi) the contribution of catchability covariates for the linear predictor with Qc v,i i the catch- ability covariates for species c and vessel v; * can be either p for probability of encounter or r for positive density. Model structure. h This was implemented using Template Model Builder (TMB47) with com- putation supported by use of the Irish Centre for High End Computing (ICHEC; http://www.ichec.ie) facility. Data. The model integrates data from seven fisheries-independent surveys taking account of correlations among species spatiotemporal distributions and abundances to predict spatial density estimates consistent with the resolution of the data.hi The model was fitted to nine species separated into adult and juvenile size classes (Table S2) to seven survey series (Table S1) in the Celtic Sea bounded by 48°N to 52°N latitude and 12°W to 2°W longitude (Fig. S8) for the years 1990–2015 inclusive.h The following steps were undertaken for data processing: (i) data for survey stations and catches were down- loaded from ICES Datras (www.ices.dk/marine-data/data-portals/Pages/DATRAS.aspx) or obtained directly from the Cefas Fishing Survey System (FSS); (ii) data were checked and any tows with missing or erroneously recorded station information (e.g. tow duration or distance infeasible) removed; (iii) swept area for each of the survey tows was estimated based on fitting a GAM to gear variables so that Doorspread = s(Depth) + DoorWt + WarpLength + WarpDiameter + SweepLength and a gear specific correction factor taken from the literature48; (iii) fish lengths were converted to biomass (Kg) through estimating a von bertalanffy length weight relationship, = ⋅ Wt a Lb, fit to sampled length and weight of fish obtained in the EVHOE survey and aggregated within size classes (adult and juvenile). Details on the downloading and processing of the data are available in Rmarkdown format (code and steps combined) as supplementary material.hi p pp y The final dataset comprised of estimates of catches (including zeros) for each station and species and estimated wept area for the tow. Model setup. The spatial domain was set up to include 250 knots representing the Gaussian Random Fields. Model structure. h A probability distribution for spatiotemporal variation in both encounter probability and positive catch rate was specified, ε(s, p, t), with a three-dimensional multivariate normal distribution so that: ∼ ⊗ ε ⁎ ⁎ ⁎ vec t MVN E R V [ ( )] (0, ) ∼ ⊗ ε ⁎ ⁎ ⁎ vec t MVN E R V [ ( )] (0, ) (5) (5) Here, vec[E(t)] is the stacked columns of the matrices describing ε⁎ s p t ( , , ) at every location, species and time, R* is a correlation matrix for encounter probability or positive catch rates among locations and V* a covariance matrix for encounter probability or positive catch rate among species (modelled within the factor analysis). ⊗ represents the Kronecker product so that the correlation among any location and species can be computed44. 8 Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w www.nature.com/scientificreports/ Incorporating covariates. Survey catchability (the relative efficiency of a gear catching a species) was estimated as a fixed effect in the model, δs(v), to account for differences in spatial fishing patterns and gear characteristics, which affect encounter and capture probability of the sampling gear46. Parameter estimates (Fig. S11) showed clear differential effects of surveys using otter trawl gears (more effective for round fish species) and beam trawl gears (more effective for flatfish species). Incorporating covariates. Survey catchability (the relative efficiency of a gear catching a species) was estimated as a fixed effect in the model, δs(v), to account for differences in spatial fishing patterns and gear characteristics, which affect encounter and capture probability of the sampling gear46. Parameter estimates (Fig. S11) showed clear differential effects of surveys using otter trawl gears (more effective for round fish species) and beam trawl gears (more effective for flatfish species). gfli p No fixed covariates for habitat quality or other predictors of encounter probability or positive density were included. While incorporation may improve the spatial predictive performance44, it was not found to be the case here based on model selection with Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). Parameter estimation. Parameter estimation was undertaken through Laplace approximation of the marginal likelihood for fixed effects while integrating the joint likelihood (which includes the probability of the random effects) with respect to random effects. Data Availabilityi y Data used to fit the model is available via the ICES Datras data portal (http://www.ices.dk/marine-data/data-por- tals/Pages/DATRAS.aspx) for two surveys and on request to the corresponding author for the remaining five surveys. Model structure. h The model was configured to estimate nine factors each to describe the spatial and spatiotemporal encounter probability and positive density parameters, with a logit-link for the linear predictor for encounter probability and log-link for the linear predictor for positive density, with an assumed gamma distribution.hif g p p y g Three candidate models were identified, (i) a base model where the vessel interaction was a random effect, (ii) the base but where the vessel x species effect was estimated as a fixed covariate, (iii) with vessel × species effect estimated, but with the addition of estimating fixed density covariates for both predominant habitat type at a knot and depth. AIC and BIC model selection favoured the second model (Table S3). The final model included estimating 1,674 fixed parameters and predicting 129,276 random effect values. Model validation. Q-Q plots show good fit between the derived estimates and the data for positive catch rates and between the predicted and observed encounter probability (S12, S13). Further, model outputs are consistent with stock-level trends abundances over time from international assessments (S14), yet also provide detailed insight into species co-occurrence and the strength of associations in space and time. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 1. Worm, B. et al. Rebuilding Global. Fisheries. Science 325, 578–585, https://doi.org/10.1126/science.1173146 (2009).hi 2. FAO. The state of world fisheries and aquaculture. Food and Agriculture Oraganization of the United Nations, 218, 92-5-105177-1 978-92-5-106675-1 (2014). References g p g 2. FAO. The state of world fisheries and aquaculture. Food and Agriculture Oraganization of the United Nations, 218, 92-5-105177-1, 978-92-5-106675-1 (2014). é é l b f h d l d d d h d 2. FAO. The state of world fisheries and aquaculture. Food and Agriculture Oraganization of the United Nations, 218, 92-5-105177-1, 978-92-5-106675-1 (2014). Bé é C l C b f F h d A l F d S d P R d A h C E d 3. Béné, C. et al. Contribution of Fisheries and Aquaculture to Food Security and Poverty Reduction: Assessing the Current Evidence World Development 79, 177–196, https://doi.org/10.1016/j.worlddev.2015.11.007 (2016).i p p g j ( ) 4. Mcclanahan, T., Allison, E. H. & Cinner, J. E. Managing fisheries for human and food security. Fish and Fisheries 16, 78–103, https:// doi.org/10.1111/faf.12045 (2015). p p g j 4. Mcclanahan, T., Allison, E. H. & Cinner, J. E. Managing fisheries for human and food security. Fish and Fisheries 16, 78–103, https:/ doi.org/10.1111/faf.12045 (2015).i p p g j 4. Mcclanahan, T., Allison, E. H. & Cinner, J. E. Managing fisheries for human and food security. Fish and Fisheries 16, 78–103, https:// doi.org/10.1111/faf.12045 (2015).i 5. Branch, T. & Hilborn, R. Matching catches to quotas in a multispecies trawl fishery: targeting and avoidance behavior under individual transferable quotas. Canadian Journal of Fisheries and Aquatic Sciences 65, 1435–1446, http://article.pubs.nrc-cnrc.gc.ca/ ppv/RPViewDoc?issn=1205-7533{&}volume=65{&}issue=7{&}startPage=1435{&}ab=y, https://doi.org/10.1139/F08-065 (2008). 5. Branch, T. & Hilborn, R. Matching catches to quotas in a multispecies trawl fishery: targeting and avoidance behavior under individual transferable quotas. Canadian Journal of Fisheries and Aquatic Sciences 65, 1435–1446, http://article.pubs.nrc-cnrc.gc.ca/ ppv/RPViewDoc?issn=1205-7533{&}volume=65{&}issue=7{&}startPage=1435{&}ab=y, https://doi.org/10.1139/F08-065 (2008). 6. Kuriyama, P. T., Branch, T. A., Bellman, M. A. & Rutherford, K. Catch shares have not led to catch-quota balancing in two North Am ri n m lti p i tr l fi h ri M i P li 71 60 70 http //d i r /10 1016/j m rp l 2016 05 010 (2016) 5. Branch, T. & Hilborn, R. Matching catches to quotas in a multispecies trawl fishery: targeting and avoidance behavior under individual transferable quotas. Canadian Journal of Fisheries and Aquatic Sciences 65, 1435–1446, http://article.pubs.nrc-cnrc.gc.ca/ ppv/RPViewDoc?issn=1205-7533{&}volume=65{&}issue=7{&}startPage=1435{&}ab=y, https://doi.org/10.1139/F08-065 (2008). 6. Kuriyama, P. T., Branch, T. A., Bellman, M. A. & Rutherford, K. Catch shares have not led to catch-quota balancing in two North American multispecies trawl fisheries. Marine Policy 71, 60–70, https://doi.org/10.1016/j.marpol.2016.05.010 (2016). www.nature.com/scientificreports/ Empirical move-on rules to inform fishing strategies: A New England case study. Fish and Fisheries 15, 359–375, https://doi.org/10.1111/faf.12019 (2014).h p g 29. Thorson, J. T. et al. Spatial factor analysis: A new tool for estimating joint species distributions and correlations in species r Methods in Ecology and Evolution 6, 627–637, https://doi.org/10.1111/2041-210X.12359 (2015).h 30. Thorson, J. T., Shelton, A. O., Ward, E. J. & Skaug, H. J. Geostatistical delta-generalized linear mixed models improve precisio estimated abundance indices for West Coast groundfishes. ICES Journal of Marine Science 72, 1297–1310, https://doi.org/10.1 icesjms/fsu243 (2015).h j ( ) 31. Thorson, J. T. & Ward, E. J. Accounting for space-time interactions in index standardization models. Fisheries Research 147, 426–433, https://doi.org/10.1016/j.fishres.2013.03.012 (2013). p g ji 32. Ellis, J. R., Rogers, S. I. & Freeman, S. M. Demersal Assemblages in the Irish Sea, St George’s Channel and Bristol Channel. Estuarine, Coastal and Shelf Science 51, 299–315, https://www.sciencedirect.com/science/article/pii/S0272771400906772, https://doi. org/10.1006/ecss.2000.0677 (2000).iii g 3. Mateo, M., Pawlowski, L. & Robert, M. Highly mixed fisheries: fine-scale spatial patterns in retained catches of French fisheries in the Celtic Sea. ICES Journal of Marine Science: Journal du Conseil fsw129, https://doi.org/10.1093/icesjms/fsw129 (2016).i f g j 4. STECF. EU’s Scientific, Technical and Economic Committee on Fisheries (STECF): Fisheries Dependent Information Database https://stecf.jrc.ec.europa.eu/dd/effort/graphs-annex (2017). p j pf g p ( ) 35. Robinson, L. M. et al. Pushing the limits in marine species distribution modelling: Lessons from the land present challenges and opportunities. Global Ecology and Biogeography 20, 789–802, https://doi.org/10.1111/j.1466-8238.2010.00636.x (2011). gy g g p y g j 6. Santos, J. et al. Reducing flatfish bycatch in roundfish fisheries. Fisheries Research 184, 64–73, https://doi.org/10.1016/j fishres.2015.08.025 (2016). i 7. Ward, E. J. et al. Using spatiotemporal species distribution models to identify temporally evolving hotspots of species co-occurrence Ecological Applications 25, 2198–2209, https://doi.org/10.1890/15-0051.1 (2015).i g pp p g 38. Cosandey-Godin, A., Krainski, E. T., Worm, B. & Flemming, J. M. Applying Bayesian spatiotemporal models to fisheries bycatch in the Canadian Arctic. Canadian Journal of Fisheries and Aquatic Sciences 72, 186–197, https://doi.org/10.1139/cjfas-2014-0159 (2015). 39. Breivik, O. N., Storvik, G. & Nedreaas, K. Latent Gaussian models to decide on spatial closures for bycatch management in the Barents Sea shrimp fishery. Canadian Journal of Fisheries and Aquatic Sciences 73, 1271–1280, https://doi.org/10.1139/ cjfas-2015-0322 (2016). j 40. Abbott, J. K., Haynie, A. C. & Reimer, M. N. www.nature.com/scientificreports/ C. & Van Putten, I. E. Human behaviour: The key source of uncertainty in fish management. Fish and Fisheries 12, 2–17, https://doi.org/10.1111/j.1467-2979.2010.00371.x (2011).hl g p g j 19. Van Putten, I. E. et al. Theories and behavioural drivers underlying fleet dynamics models. Fish and Fisheries 13, 216–235, https:// doi.org/10.1111/j.1467-2979.2011.00430.x (2012). g j 20. Fraser, H. M., Greenstreet, S. P. R., Fryer, R. J. & Piet, G. J. Mapping spatial variation in demersal fish species diversity and composition in the North Sea: Accounting for species- and size-related catchability in survey trawls. ICES Journal of Marine Science 65, 531–538, https://doi.org/10.1093/icesjms/fsn036 (2008). 21. Gerritsen, H. D., Lordan, C., Minto, C. & Kraak, S. B. M. Spatial patterns in the retained catch composition of Irish demersal otter trawlers: High-resolution fisheries data as a management tool. Fisheries Research 129–130, 127–136, https://doi.org/10.1016/j. fishres.2012.06.019 (2012). i 22. Needle, C. L. & Catarino, R. Evaluating the effect of real-time closures on cod targeting. ICES Journal of Marine Science 68, 1647–1655, https://doi.org/10.1093/icesjms/fsr092 (2011).i p g j 3. Holmes, S. J. et al. Using fishery-dependent data to inform the development and operation of a co-management initiative to reduce cod mortality and cut discards. ICES Journal of Marine Science 68, 1679–1688, https://doi.org/10.1093/icesjms/fsr101 (2011). l d f h f h l b l h ( ) Study for the Revision of the plaice box - a Final Report. Tech. Rep 24. Beare, D. J. et al. Study for the Revision of the plaice box - a Fin 24. Beare, D. J. et al. Study for the Revision of the plaice box - a Final Report. Tech. Rep (2010). y p 25. Dinmore, T. A., Duplisea, D. E., Rackham, B. D., Maxwell, D. L. & Jennings, S. Impact of a large-scale area closure on patterns of fishing disturbance and the consequences for benthic communities. ICES Journal of Marine Science 60, 371–380, https://doi. org/10.1016/S1054-3139(03)00010-9 (2003). 26. Gardner, B., Sullivan, P. J., Morreale, S. J. & Epperly, S. P. Spatial and temporal statistical analysis of bycatch data: patterns of sea turtle bycatch in the North Atlantic. Canadian Journal of Fisheries and Aquatic Sciences 65, 2461–2470, https://doi.org/10.1139/F08-152 (2008).ii 7. Dunn, D. C., Boustany, A. M. & Halpin, P. N. Spatio-temporal management of fisheries to reduce by-catch and increase fishing selectivity. Fish and Fisheries 12, 110–119, https://doi.org/10.1111/j.1467-2979.2010.00388.x (2011).i y p g j ( ) 28. Dunn, D. C. et al. www.nature.com/scientificreports/ www.nature.com/scientificreports/ 0. Uhlmann, S. S. et al. Discarded fish in European waters: General patterns and contrasts. ICES Journal of Marine Science 71 1235–1245, https://doi.org/10.1093/icesjms/fst030 (2014). 10. Uhlmann, S. S. et al. Discarded fish in European waters: General patterns and contrasts. ICES Journal of Marine Science 71, 1235–1245, https://doi.org/10.1093/icesjms/fst030 (2014). 11 E C i i R l ti (EU) N 1380/2013 f th E P li t d f th C il f 11 D b 2013 th p g j 1. European Commission. Regulation (EU) No. 1380/2013 of the European Parliament and of the Council of 11 December 2013 on the Common Fisheries Policy, amending Council Regulations (EC) No. 1954/2003 and (EC) No. 1224/2009 and repealing Counci Regulations (EC) No 2371/2002 and (EC) (2013).fi g ( ) ( ) ( ) 12. European Parliament. Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009. Official Journal of the European Union 140, 16–62, https://doi.org/10.3000/17252555.L_2009.140.eng, 534 (2009).hi p p g g ( ) 13. Garcia, S. M., Zerbi, A., C, A., Do Chi, T. & Lasserre, G. The ecosystem approach to fisheries. FAO Fisheries Techn http // f /d p/006/Y4773E/ 4773 05 ht l http //d i /10 1079/9781845934149 0000 (2003) p p g g ( ) arcia, S. M., Zerbi, A., C, A., Do Chi, T. & Lasserre, G. The ecosystem approach to fisheries. FAO Fisheries Technical Paper 443, 71, p p g g 13. Garcia, S. M., Zerbi, A., C, A., Do Chi, T. & Lasserre, G. The ecosystem approach to fisheries. FAO Fisheries T http://www.fao.org/docrep/006/Y4773E/y4773e05.html, https://doi.org/10.1079/9781845934149.0000 (2003 hi http://www.fao.org/docrep/006/Y4773E/y4773e05.html, https://doi.org/10.1079/9781845934149.0000 (2003). g p y p g Fishermen Rational? A Fishing Expedition. Marine Resource Econfih . Holland, D. S. Are Fishermen Rational? A Fishing Expedition. M , g p , ( ) 15. Hoff, A. et al. Economic effort management in multispecies fisheries: The FcubEcon model. ICES Journal of Marine Science 67, 1802–1810, https://doi.org/10.1093/icesjms/fsq076 (2010).i 16. Condie, H. M., Grant, A. & Catchpole, T. L. Incentivising selective fishing under a policy to ban discards; lessons from European and global fisheries. Marine Policy 45, 287–292, https://linkinghub.elsevier.com/retrieve/pii/S0308597X1300198X, https://doi. org/10.1016/j.marpol.2013.09.001 (2014).i g j p 17. Condie, H. M., Grant, A. & Catchpole, T. L. Does banning discards in an otter trawler fishery create incentives for more selective fishing? Fisheries Research 148, 137–146, https://linkinghub.elsevier.com/retrieve/pii/S016578361300221X, https://doi. org/10.1016/j.fishres.2013.09.011 (2013). g ji ( ) 18. Fulton, E. A., Smith, A. D., Smith, D. References pi y , , p g j p ( ) 7. Ulrich, C. et al. Achieving maximum sustainable yield in mixed fisheries: A management approach for the North Sea demersal fisheries. ICES Journal of Marine Science 74, 566–575, https://doi.org/10.1093/icesjms/fsw126 (2017). fisheries. ICES Journal of Marine Science 74, 566 575, https://doi.org/10.1093/icesjms/fsw126 (2017). 8. Batsleer, J., Hamon, K. G., Overzee, H. M. J., Rijnsdorp, A. D. & Poos, J. J. High-grading and over-quota discarding in mixed fisheries. Reviews in Fish Biology and Fisheries 25, 715–736, https://doi.org/10.1007/s11160-015-9403-0 (2015).h i J f , , p g j ( ) 8. Batsleer, J., Hamon, K. G., Overzee, H. M. J., Rijnsdorp, A. D. & Poos, J. J. High-grading and over-quota discarding in mixed fisheries. Reviews in Fish Biology and Fisheries 25, 715–736, https://doi.org/10.1007/s11160-015-9403-0 (2015). 9. Borges, L. The evolution of a discard policy in Europe. Fish and Fisheries 534–540, https://doi.org/10.1111/faf.12062 (2015). i J f p g j ( ) 8. Batsleer, J., Hamon, K. G., Overzee, H. M. J., Rijnsdorp, A. D. & Poos, J. J. High-grading and over-quota discarding in mixe fisheries. Reviews in Fish Biology and Fisheries 25, 715–736, https://doi.org/10.1007/s11160-015-9403-0 (2015).h Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 9 www.nature.com/scientificreports/ Piet, G. J., Van Hal, R. & Greenstreet, S. P. R. Modelling the direct impact of bottom trawling on the North Sea fish community to derive estimates of fishing mortality for non-target fish species. ICES Journal of Marine Science 66, 1985–1998, https://doi. org/10.1093/icesjms/fsp162 (2009). 48. Piet, G. J., Van Hal, R. & Greenstreet, S. P. R. Modelling the direct impact of bottom trawling on the North Sea fish community to derive estimates of fishing mortality for non-target fish species. ICES Journal of Marine Science 66, 1985–1998, https://doi. org/10.1093/icesjms/fsp162 (2009). Acknowledgements g Paul J. Dolder acknowledges funding support from the MARES joint doctoral research programme (MARES_14_15) and Cefas seedcorn (DP227AC). The authors gratefully acknowledge the use of desk space in the Branch lab at the University of Washington, computing facilities at the Irish Centre for High-End Computing (ICHEC; http://www.ichec.ie) and the hard-work of many scientists and crew in collecting and storing data through the scientific surveys used in this study without which it would not have been possible. The manuscript benefited from discussions with D. Stokes, C. Lordan, C. Moore and H. Gerritsen (Marine Institute, Ireland), L. Readdy, C. Darby, I. Holmes, S. Shaw and T. Earl (Cefas) and comments from J. Hastie, M. McClure and R. Fonner. Author Contributions P.J.D., C.M. and J.T.T. designed the study. P.J.D. conducted the analysis. All authors contributed to writing the manuscript. www.nature.com/scientificreports/ Hidden Flexibility: Institutions, Incentives, and the Margins of Selectivity in Fis Land Economics 91, 169–195, https://doi.org/10.3368/le.91.1.169 (2015).hf 41. Thorson, J. T. & Kristensen, K. Implementing a generic method for bias correction in statistical models using random effects, with spatial and population dynamics examples. Fisheries Research 175, 66–74, https://doi.org/10.1016/j.fishres.2015.11.016 (2016).i p p p y p p g ji 2. Pascoe, S., Koundouri, P. & Bjørndal, T. Estimating targeting ability in multi-species fisheries: a primal multi-output distance function approach. Land Economics, https://doi.org/10.3368/le.83.3.382 (2007). 43. Reimer, M. N., Abbott, J. K. & Wilen, J. E. Fisheries Production: Management Institutions, Spatial Choice, and the Quest for Policy Invariance. Marine Resource Economics 32, 143–168, https://doi.org/10.1086/690678 (2017).ht 44. Thorson, J. T. & Barnett, L. A. K. Comparing estimates of abundance trends and distribution shifts using single- and multispecies models of fishes and biogenic habitat. ICES Journal of Marine Science: Journal du Conseil fsw193, https://doi.org/10.1093/icesjms/ fsw193 (2017). Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 10 www.nature.com/scientificreports/ 45. Thorson, J. T. et al. Joint dynamic species distribution models: a tool for community ordination and spatio-temporal monito Global Ecology and Biogeography 25, 1144–1158, https://doi.org/10.1111/geb.12464 (2016).hh 45. Thorson, J. T. et al. Joint dynamic species distribution models: a tool for community ordination and spatio-temporal monitoring. Global Ecology and Biogeography 25, 1144–1158, https://doi.org/10.1111/geb.12464 (2016). Th l Th f l d l f h h f d d d l 45. Thorson, J. T. et al. Joint dynamic species distribution models: a tool for community ordination and spatio temporal monitoring. Global Ecology and Biogeography 25, 1144–1158, https://doi.org/10.1111/geb.12464 (2016). 46. Thorson, J. T. et al. The importance of spatial models for estimating the strength of density dependence. Ecology 96, 1202–1212, https://doi.org/10.1890/14-0739.1.sm (2015). 6. Thorson, J. T. et al. The importance of spatial models for estimating the strength of density dependence. Ecology 96, 1202–1212 https://doi.org/10.1890/14-0739.1.sm (2015).f 46. Thorson, J. T. et al. The importance of spatial models for estimating the strength of density dependence. Ecology 96, 1202– https://doi.org/10.1890/14-0739.1.sm (2015).f p g ( ) 47. Kristensen, K., Nielsen, A., Berg, C. W., Skaug, H. & Bell, B. TMB: Automatic Differentiation and Laplace Approximation. Journal of Statistical Software 70, 1–21, https://arxiv.org/abs/1509.00660, https://doi.org/10.18637/jss.v070.i05, 1509.00660 (2016).i p g ( ) 47. Kristensen, K., Nielsen, A., Berg, C. W., Skaug, H. & Bell, B. TMB: Automatic Differentiation and Laplace Approximation. Journal of Statistical Software 70, 1–21, https://arxiv.org/abs/1509.00660, https://doi.org/10.18637/jss.v070.i05, 1509.00660 (2016).i f ft p g p g j 48. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w Additional Information upplementary information accompanies this paper at https://doi.org/10.1038/s41598-018-31881-w. Competing Interests: The authors declare no competing interests. Competing Interests: The authors declare no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Scientific RePortS \| (2018) 8:13886 \| DOI:10.1038/s41598-018-31881-w 11
https://openalex.org/W4384570913	http://www.clausiuspress.com/assets/default/article/2023/07/17/article_1689580791.pdf	English	null	Analysis of Xi'an City Image and Micro-Film Narrative	Art and performance letters	2,023	cc-by	2,928	Keywords: Xi'an city image, Micro-film storytelling, Narrative Abstract: It is important to share the city's story in order to improve the city's image and its ability to communicate. This paper draws on theories and previous research regarding city image and urban narrative. Based on theories and previous research on city image and urban narrative, this paper examines the five dimensions of “people, things, objects, places and situations” in a five-dimensional narrative framework. The five-dimensional narrative framework is used to analyze the micro-images of Xi'an that “tell the story of Xi'an” and to compile a five-dimensional narrative corpus of Xi'an's urban image. 2. Literature Review DOI: 10.23977/artpl.2023.040608 ISSN 2523-5877 Vol. 4 Num. 6 DOI: 10.23977/artpl.2023.040608 ISSN 2523-5877 Vol. 4 Num. 6 DOI: 10.23977/artpl.2023.040608 ISSN 2523-5877 Vol. 4 Num. 6 Art and Performance Letters (2023) Clausius Scientific Press, Canada 1. Introduction Due to the increasing acceleration of economic globalization in the 21st century, China has seen a significant increase in its urbanization process. According to studies, the rate of urbanisation has increased from 36.2% to 51.3% in the period from 2000 to 2011. In developed countries, it is predicted that by 2050 the urbanization rate will reach 86%, while China's urbanisation rate is expected to exceed 70% [1]. In addition to its natural resources, a city's human connotations, living standards, and history are reflected in its image, which can have a significant economic and social impact. Various media channels have been used to convey the image of a city, which is essential to raising its profile and attracting foreign investment and tourists. It was in 2013, at China's National Propaganda Conference, The strategic concept of “telling the Chinese story and spreading the Chinese voice” was proposed. This concept was quickly identified as a national strategy. It was immediately identified as a national strategy, and for cities, following the national strategy and telling a good city story is a means of strengthening their image and improving their discourse skills. The focus of this study is on understanding the city and its branding from the perspective of storytelling, based on Zhan Zhuo's theory of the 'city as story'[1]. In this study, urban cinema and television are analyzed in order to understand what dimensions Xi'an's city story is comprised of, what stories are representative of these dimensions, and what kind of spirit and culture they convey. Analysis of Xi'an City Image and Micro-Film Narrative Wang Jing Xi’an Peihua University, Xi’an, Shannxi, China 2.1 Narrative, Story and Urban Storytelling It was in 1969 when the French structuralist writer Kevin Lynch (1969) coined the term 'Narratology', which marked the formal emergence of Narratology as a discipline[2]. Narrative Narratology is a structuralist-influenced theory for studying narrative, with stories, narrative 42 discourses, and narrative acts as its objects. Storytelling, narrative discourse, and narrative behavior are all aspects of narrative. The field of narrative theory is often divided into two stages of development: classical Narratology and post-classical Narratology. Narrative works are viewed as closed symbolic systems in classical narratology, in which the author seeks to clarify the relationships between the different elements of the narrative text as a purely textual study in a closed environment by examining their structure, patterns, laws, and other relationships. Compared to structuralist linguistics, classical narratology borrows a whole set of linguistic terms and research categories from structural linguistics, as well as a whole range of linguistic terms and research perspectives. In addition, the paradigm of narrative linguistics is similar to that of structuralist linguistics in that it involves the study of non-contextual grammar in a closed, independent, and self-sufficient manner. In contrast to classical narratives, which emphasize the story (text, discourse) itself, a key characteristic of post-classical narratives is that the meaning of the things themselves does not determine the meaning of the finished narrative, which is usually a reconstructed system of meaning that is constructed by the author. Thus, post-classical narratology does not primarily concern itself with the contents of a story or its presentation, but rather with its specific manner in which it is presented. Each work can be regarded as a concrete application and embodiment of narrative rules, and the meaning of a story text is developed through the exchange and reception of narrative texts [3]. Zhang Nan (Zhongnan University) first proposed the 'urban storytelling theory'[4]. According to the urban story theory, the essence of urban stories lies in the integration of three aspects: the integration of people and the natural environment, the integration of people and history and culture, and the integration of people's interaction activities. The city unites people's collective memory at the level of time and space, and it is the integration of events with the spatial environment that makes people truly aware of their place in the environment, enhances their sense of identity, and brings them closer to the urban environment. 2.1 Narrative, Story and Urban Storytelling Professor Zhang Nan and the Cambridge Centre for the Study of Narrative Environments in the UK have jointly researched and proposed a classification of urban narrative elements into four categories: natural elements, artificial elements, human elements and event elements [4]. The four elements on urban narrative space go beyond the five material level elements proposed by Lynch (1969) in Urban Imagery, and achieve the unification of material form and ideology [2]. The theoretical implications of urban storytelling theory allow for a deeper integration of the disciplines of spatial narrative and urban design. The theory of urban storytelling takes people, their activities, history and culture, and the urban environment as the subjects of the narrative, while the urban space, in turn, becomes the stage for the narrative. In order to analyze the spatial structure and characteristics of urban narratives, it is necessary to incorporate a number of factors, including human activities, history, and culture, as well as the urban environment. 3. The Analysis of Xi'an's Stories from Five Perspectives This is an indication of the “micro” era of image creation and expression in the new media context, and as new forms of image expression and dissemination, they are generally referred to as “micro-images”. Since Xi'an is considered to be one of the four most famous ancient capitals in China, its image has been superficial and it has been referred to as an “ancient capital” for a long time. The purpose of this study is to examine the narrative discourse of 100 video works relating to the city of Xi'an, and the ways in which they convey the story of Xi'an in five dimensions: people, events, objects, scenes and situations. 2.2 The Five Dimensions of Urban Storytelling In his description of urban imagery, Kevin Lynch identifies five components: the path, the edge, the district, the node, and the landmark. It is a behavioural approach to urban imagery, in which the city is essentially understood as a physical form, without adequate consideration of human initiative and activism (1999) [2]. According to Professor Zhang Nan of Central South University, the 'Urban Story Theory' attempts to elevate the simple materiality of a city to the level of an urban event, in order to realize the pursuit of urban meaning [4]. As part of his theory, he proposes the psychological mechanism of “object/field/event” urban storytelling theory, which integrates people with their natural environment, people with history and culture, and their interaction activities to 43 build a perception of city life. In the initial semantic division of urban narrative, the spatial elements are divided into the following categories: natural, human, and artificial, and events. As part of his examination of Lynch's Five Elements of Urban Imagery, Zhan Zhuo (2018) incorporates a critical examination of Professor Zhang Nan's urban storytelling theory [1]. The purpose of this chapter is to provide a blueprint for research on the psychology of environmental memory, which includes identifying the five dimensions of the story city construct: people, events, things, places, and situations. Essentially, a story city is an organic combination of people, events, buildings, spatial places, contexts, and time, and reading stories is reconstructing the urban ecosystem, which is a combination of people, events, objects, actions, and contexts. People are the soul of the city, events are its memory, objects are its body, fields represent its actions, and context is its ethos. In summary, the above-mentioned scholars' perspectives have been synthesized, and an analysis has been conducted of the cultural branding of Xi'an's urban stories in five dimensions: people, things, objects, fields and realms. Internet technology and mobile smart terminals have profoundly changed the mode of mass communication as a result of the development of new media technologies. Through its fragmentation, interactive nature, and personalisation, it has enabled new forms of communication. Microfilms, microdocumentaries, and short videos have become a new form of visual expression associated with new media, with microfilms, microdocumentaries, and short videos becoming new forms of visual expression [3]. 3.1 Dimensional Analysis of People A key feature of this study is the use of Zhan Zhuo's (2018) research approach, in which urban citizens are divided into three types: citizens of the city, foreign migrants, and foreign tourists, representing city stories as places to live, work, and visit [1]. Twenty-one percent of the micro-images tell two or more Xi'an stories; the stories of cities as tourist destinations are the most popular among the 100 urban micro-images, comprising 87 percent of all micro-images. More than half (62%) of micro-images depict cities as places of living and 41%) depict cities as places of employment. In terms of micro-images, most people are interested in “Datang City of Night”, “Datang Furong Garden”, “Xi'an Bell and Drum Tower”, “Xi'an Ming City Wall”, “Xi'an West City”, and “Xi'an West City”. Most of the entries, such as “Datang West City,” were made by tourists who used Xi'an as a place to visit to tell their stories of the city. A city full of historic heritage, as well as a city filled with modern technology and vitality, Xi'an is viewed by tourists as a city full of historical heritage. According to their accounts, Xi'an is a city with a unique charm and appeal that is a combination of a historic heritage and a contemporary feel. 44 3.2 Dimensional Analysis of Events 3.2 Dimensional Analysis of Events Based on Zhan Zhuo's (2018) categorisation of the narrative dimensional analysis of pairs of events in urban storytelling, this study examines five dimensions: cultural events, economic events, sports events, political events, and comprehensive events [1]. There is a connection between the deep historical and cultural heritage of the Xi'an city story and content with a certain degree of interest in the narrative (See Table 1). 3.3 The Physical Dimension The memory of a city is comprised of several important elements, including ancient ruins, traditional buildings, historic districts, as well as folk art and city life. History and memory can be found in every inch of the city, in every piece of land, in every building, on every street, in every corner of every area [1]. Urban memory consists of two parts - the material and the immaterial - and it also reflects the cultural values of a city. The Big Wild Goose Pagoda, Xingqing Palace, the Tang Furong Garden and the Ming City Wall are the most frequently seen physical images in Xi'an's micro-video works, and they are also the physical materials that carry the city's memory. The Big Wild Goose Pagoda appears as a symbol of Xi'an in many of the city's promotional films. Unfortunately, however, in the Xi'an micro-images, especially in the promotional films, the stories of these objects are not told directly, but rather they are used as the backdrop for the films. 3.1 Dimensional Analysis of People Table 1: The Table of Events Analysis Type of city event Events Cultural events East Asia Cultural Capital; Silk Road International Film Festival,; China-US Tourism Year; Xi'an Silk Road International Art Festival Economic events China International Building Intelligence Summit; The 3rd International Cooperation Summit; The 3rd Silk Road International Cooperation Summit; The 16th China New Energy International Expo and Summit; The China (Xi'an) International Forestry Expo and Forestry Industry Summit Sports events: Xi'an International Marathon; Xi'an City Wall International Marathon; The 14th City Games; Xi'an World Women's Chess Masters Summit;China National Snooker Team Championship; China Taekwondo International Open Political Events Eurasian Economic Forum; Xi'an Economic; Trade and Cultural Tour for Global Ambassadors in China General events The Datang City of Night was selected as a national model pedestrian street; Xi'an was selected as a “double centre” city. Among the most frequently mentioned events is the Silk Road International Film Festival. There is no doubt that the Silk Road International Film Festival is the longest-running and most successful international film festival in Xi'an. “Cultural events are an integral part of a city's identity, reflecting its culture and image as a whole. “As a historical and ancient capital and a cultural city, Xi'an has a large number of cultural events depicted in its micro-images, which is associated with its heritage, and is also in sync with its rapid economic growth and cultural co-prosperity in recent years. Acknowledgments This work was supported by the Shaanxi Provincial Research Project on Major Theoretical and Practical Issues in Philosophy and Social Sciences (2022ND0413). This work was supported by the Shaanxi Provincial Research Project on Major Theoretical and Practical Issues in Philosophy and Social Sciences (2022ND0413). 3.5 Dimensional Analysis of the Realm As a result of this, the “realm” reflects the mood, temperament, and image of a city as a whole. This is reflected in the city's promotional films as a representation of the city's overall appearance and position. As presented in Xi'an's micro-video works, the description of the realm is both reminiscent of a classical painting from the Tang Dynasty and a modern urban scene dominated by tall buildings. 4. Conclusion In order to stay within the time and resource constraints of this study, only one city, Xi'an, was selected as the subject of this study. It appears that the five-dimensional construct of the story city used is more relevant for the study of Xi'an, but further research and verification are required to determine whether it is generalization to other cities. Furthermore, this paper includes 100 micro-images that represent a variety of genres, including city propaganda films that introduce the city's politics, economy, culture, history, and humanity, as well as short city tourism videos that introduce the city's attractions. In addition, it is worth exploring whether there is any difference in the narrative discourse used by the different types of micro-images in their narratives about urban life. 3.4 Dimensional Analysis of the Field The urban storytelling site also serves as a platform for storytelling in the city of Xi'an. There are several successful examples of this, but White Deer Plain Film and TV City has been one of the most successful. In 2016, the Shaanxi Travel Group completed the development of the film and television city, which covered an area of 1,050 mu and cost RMB 600 million (Li J, 2018)[5]. As part of the adaptation of the novel “White Deer Plain” into a film and television series, the city was 45 constructed in accordance with the description of the space and scenes described in the novel and was reconstructed in a realistic manner, including the White Deer Village as well as Zishui, which became the primary location for filming and television productions. It has now become one of the most well-known scenes in Xi'an and has been recognized as a key cultural project in Shaanxi Province due to the success of a series of cultural IP products such as the White Deer Plain film and television series. In order to incorporate film and television creation, cultural leisure and rural tourism into a scenic spot themed on folk culture, the White Deer Plain Film and TV City focuses on architecture, patriarchy, village layout, and folk culture of the Xi'an area, including music, food and housing. Bailu Yuan Film and Television City, which has been long anticipated, is also expected to become an important venue for presenting the story of Xi'an in the future. [1] Zhan Z. A five-dimensional narrative strategy for urban communication: the example of Wuhan city Wuhan City Image Film as an Example (Master dissertation) [D]. Huazhong University of Science and Technology, 2018. [2] Kevin L. The Image of the City, Huaxia Publishing House, 1969, pp. 106. [3] Dai G. Q. The influence of Western narrative theory on Chinese literary criticism, Jiangxi Social Sciences, 2009, 15(03), pp. 30-36. [4] Zhang N. Place Narrative: A New Model for the Construction of Urban Cultural Connotation and Characteristics [J]. Journal of Shanghai Jiaotong University (Philosophy and Social Science Edition), 1999, 20(03), pp: 68-76. [5] Li J. Exploring micro-images and city image communication, Center for the Performing Arts, 2018, 16(02), 59-60. [4] Zhang N. Place Narrative: A New Model for the Construction of Urban Cultural Connotation and Characteristics [J]. Journal of Shanghai Jiaotong University (Philosophy and Social Science Edition), 1999, 20(03), pp: 68-76. [5] Li J. Exploring micro-images and city image communication, Center for the Performing Arts, 2018, 16(02), 59-60. [1] Zhan Z. A five-dimensional narrative strategy for urban communication: the example of Wuhan city Wuhan City Image Film as an Example (Master dissertation) [D]. Huazhong University of Science and Technology, 2018. [2] Kevin L. The Image of the City, Huaxia Publishing House, 1969, pp. 106. p ( ) [ ] z g y f evin L. The Image of the City, Huaxia Publishing House, 1969, pp. 106. ai G. Q. The influence of Western narrative theory on Chinese literary criticism, Jiangxi Social Scienc ), pp. 30-36. References 46
https://openalex.org/W2608868760	https://europepmc.org/articles/pmc5407791?pdf=render	English	null	Characterization of physiological defects in adult SIRT6-/- mice	PloS one	2,017	cc-by	9,347	RESEARCH ARTICLE a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Characterization of physiological defects in adult SIRT6-/- mice Victoria Peshti1☯, Alexey Obolensky2☯, Liat Nahum1☯, Yariv Kanfi1, Moran Rathaus1, Maytal Avraham1, Simon Tinman1, Fredrick W. Alt3, Eyal Banin2, Haim Y. Cohen1* 1 The Mina & Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel, 2 Department of Ophthalmology, Hadassah-Hebrew University Medical Center, Jerusalem, Israel, 3 Harvard University Medical School, Boston, Massachusetts, United States of America ☯These authors contributed equally to this work. * Haim.Cohen@biu.ac.il a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Editor: Barbara Fam, University of Melbourne, AUSTRALIA Received: July 24, 2016 Accepted: April 9, 2017 Published: April 27, 2017 Received: July 24, 2016 Accepted: April 9, 2017 Published: April 27, 2017 Copyright: © 2017 Peshti et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Abstract The NAD+-dependent SIRT6 deacetylase was shown to be a major regulator of lifespan and healthspan. Mice deficient for SIRT6 develop a premature aging phenotype and meta- bolic defects, and die before four weeks of age. Thus, the effect of SIRT6 deficiency in adult mice is unknown. Here we show that SIRT6-/- mice in mixed 129/SvJ/BALB/c background reach adulthood, allowing examination of SIRT6-related metabolic and developmental phe- notypes in adult mice. In this mixed background, at 200 days of age, more than 80% of the female knock-out mice were alive whereas only 10% of male knock-out mice survived. In comparison to their wild-type littermates, SIRT6 deficient mice have reduced body weight, increased glucose uptake and exhibit an age-dependent progressive impairment of retinal function accompanied by thinning of retinal layers. Together, these results demonstrate a role for SIRT6 in metabolism and age-related ocular changes in adult mice and suggest a gender specific regulation of lifespan by SIRT6. OPEN ACCESS Citation: Peshti V, Obolensky A, Nahum L, Kanfi Y, Rathaus M, Avraham M, et al. (2017) Characterization of physiological defects in adult SIRT6-/- mice. PLoS ONE 12(4): e0176371. https:// doi.org/10.1371/journal.pone.0176371 Editor: Barbara Fam, University of Melbourne, AUSTRALIA Characterization of physiological defects in adult SIRT6-/- mice subcutaneous fat, lordokyphosis, and eventually death at about 4 weeks [10]. Serum IGF-1 and glucose levels are both severely reduced in SIRT6-/- mice at the age of 24 days [10]. In addition, SIRT6 deficiency caused major retinal transmission defects concomitant to changes in expres- sion of glycolytic genes and elevated levels of apoptosis in inner retina cells [11]. An additional study that examined the effect of SIRT6 deficiency in a mixed genetic background showed a similar set of systemic phenotypes including the fact that the majority of these mice do not exceed five weeks of age [12]. SIRT6 is required for glucose homeostasis and maintenance of normal IGF-1 levels. Indeed, SIRT6 was found to regulate the expression of glycolytic genes, including the glucose transporter GLUT1 [13]. In the absence of SIRT6, GLUT1 levels increase, resulting in increased glucose uptake and glycolysis activation. SIRT6 also functions at chromatin to attenuate NF-κB signaling via interaction with the NF-κB RELA subunit, and deacetylates H3K9 at NF-κB target gene promoters [7]. collection and analysis, decision to publish, or preparation of the manuscript. collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Competing interests: The authors have declared that no competing interests exist. SIRT6 is an important key player in different DNA repair pathways and it is crucial for genomic stability [8, 10, 14–16]. In vitro, SIRT6 depletion leads to telomere dysfunction with end-to-end chromosomal fusions and premature cellular senescence [14, 17]. Moreover, SIRT6 is required for the stable association of telomeric chromatin with WRN, a protein that is mutated in the premature aging disease Werner syndrome [18]. These findings suggest that SIRT6 is important in telomere maintenance. Recently SIRT6 was shown to be regulated by Lamin A, another protein whose mutant form, Progerin, causes premature aging disease in humans [19]. Taken together, this data suggests that SIRT6 dysfunction may play an important role in human premature aging. SIRT6 ability to control lifespan and healthspan is linked to a number of age-related dis- eases. We found transgenic mice over-expressing exogenous SIRT6 (MOSES) to be protected against the physiological damages of diet-induced obesity [20] including accumulation of tri- glycerides and LDL cholesterol in the serum. Introduction The sirtuins are highly conserved NAD+-dependent deacetylases, homologues of the yeast Sir2 protein. They were shown to regulate lifespan in variety of organisms including mammals. In yeast, nematode and flies, an additional copy of SIR2 increased lifespan [1–3]. In mammals there are seven sirtuins, SIRT1 to SIRT7; among them SIRT1 is the most investigated. Mice over-expressing SIRT1 showed improved healthspan, including lower levels of DNA damage, better glucose tolerance and fewer spontaneous carcinomas and sarcomas. Whole body over- expression of SIRT1 did not extend lifespan [4], but brain specific overexpression in mice did increase lifespan [5]. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: This study was supported by the Israeli Academy of Sciences, I-Core, Binational US-Israel Science Foundation, Israel Cancer Association, Koret Foundation, Israel Cancer Research Fund, the Israel Ministry of Health, Lir’ot association, TEVA NNE program, EFSD, D- Cure, the ERC: European Research Council and the Yedidut Research Fund. The funders had no role in study design, data In the last few years SIRT6 was established as the main mammalian sirtuin which regulates lifespan. SIRT6 is predominantly a nuclear sirtuin which, besides deacetylation [6, 7], also catalyzes ADP-ribosylation [8] and deacylation of long chain fatty acyls [9]. SIRT6-deficient 129/SvJ mice are small, express severe metabolic defects, and by 2–3 weeks of age develop abnormalities usually associated with aging. These include profound lymphopenia, loss of PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 1 / 15 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Metabolic measurements Body composition was analyzed by Dual-Energy X-ray Absorptiometry (DEXA) using a Lunar PIXImus densitometer. For glucose tolerance testing (GTT), mice were subjected to overnight fasting and injected intraperitoneally (IP) with 2g of glucose per kg of body weight. Blood glu- cose was measured prior to glucose administration and 15, 30, 60, 90, and 120 min after using an Ascensia Elite glucose meter (Bayer). For the insulin tolerance test (ITT), mice were fasted for 6h, injected IP with 0.75 U insulin (Eli Lilly) per kg of body weight, and blood glucose levels were measured before and 15, 30, 60, 90 and 120 min after injection. HOMA of insulin resis- tance (HOMA-IR) was calculated using the equation: fasting serum insulin (μU/ml) × fasting plasma glucose (mmol/l)/22.5. Serum insulin was measured by ELISA (Crystal Chem). Materials and methods Mice and MEFs The ethic committee of Bar Ilan University (Animal Care and Use Committee (IACUC) approved this study. Mice were kept under specific pathogen free and 12 h light ⁄ dark condi- tions. SIRT6+/- 129/SvJ mice were mated with BALB/c mice (Jackson). Breeding of SIRT6 het- erozygote F1 generation resulted in SIRT6+/+, SIRT6+/- and SIRT6-/- 129/SvJ/BALB/c mice. The previously published findings that 129/SvJ SIRT6 deficient mice die up to four weeks prematurely was taken into consideration and addressed in the approved protocol by the above-mentioned committee. Thus, the mixed background mice who survived significantly longer were monitored daily. The specific criteria to performing euthanasia were disease, severe injury and apathy. There were no unexpected deaths due to apparent diseases or inju- ries. At the time of the examination, animals that died without euthanasia did not exhibit any f d f b b d h d h dd The ethic committee of Bar Ilan University (Animal Care and Use Committee (IACUC) approved this study. Mice were kept under specific pathogen free and 12 h light ⁄ dark condi- tions. SIRT6+/- 129/SvJ mice were mated with BALB/c mice (Jackson). Breeding of SIRT6 het- erozygote F1 generation resulted in SIRT6+/+, SIRT6+/- and SIRT6-/- 129/SvJ/BALB/c mice. The previously published findings that 129/SvJ SIRT6 deficient mice die up to four weeks prematurely was taken into consideration and addressed in the approved protocol by the above-mentioned committee. Thus, the mixed background mice who survived significantly longer were monitored daily. The specific criteria to performing euthanasia were disease, severe injury and apathy. There were no unexpected deaths due to apparent diseases or inju- ries. At the time of the examination, animals that died without euthanasia did not exhibit any signs of distress or signs of being moribund prior to their death. In addition, no apparent cause of death was found in these mice. General microbial examinations were performed routinely by the animal research facility staff. For GTT and ITT, Esracain cream (Lidocaine 5%, Rafa, Israel) was used as topical analge- sic. Before ERG procedures mice were anesthetized by intraperitoneal injections of a mixture of ketamine (Bedford Laboratories, Bedford, OH) and xylazine (VMD, Arendonk, Belgium), with doses suitable to their body weight. After completion of ERG recordings, while still anes- thetized, mice were sacrificed by cervical dislocation. Mice that were not examined by ERG recordings and did not die naturally were put to death by CO2. Materials and methods Mice and MEFs Mouse Embryo Fibroblasts (MEFs) were isolated 13 days after observation of vaginal plugs. Cells were plated into DMEM media supplemented with 10% fetal bovine serum (FBS). In addition, SIRT6 regulates healthy liver via the repression of SREBP1/2, two transcription factors that control triglycerides and cholesterol synthesis and via repressing miR122, the most abundant microRNA in the liver [21], [22]. Indeed, MOSES males have a significantly longer lifespan than their wild-type (WT) litter- mates. Moreover, MOSES mice have reduced IGF-1 levels in sera and IGF-1 signaling was sig- nificantly reduced in white adipose tissues (WAT) [23]. In addition, SIRT6 has a pivotal role in carcinogenesis [24]. It was found to be a tumor suppressor by controlling Myc transcrip- tional activity [24, 25]. SIRT1 is another sirtuin whose deficiency also causes a severe phenotype. Interestingly, the effect of SIRT1 deficiency is strain dependent, especially as relates to mice survival. Whereas SIRT1-deficient mice on 129/SvJ-CP background were smaller and invariably died within 1 month after birth, mice on 129/CD1 mixed background more often survived to adulthood. In 129/SvEv/C57BL/6 background, about 90% of mutant mice died at perinatal or early postnatal stages, and the remaining died about 3 months after birth [26, 27]. In another study only 1% of the 129/SvEv/FVB background and 9.3% on the 129/SvEv/FVB/Black Swiss background survived to adulthood [28]. These observations suggest the reported phenotypes of 129/SvJ SIRT6-/- mice might be strain dependent as well. Due to the very short lifespan of the various SIRT6-/- animal models which usually do not exceed five weeks of age, our knowledge regard- ing the role of SIRT6 in adult and aging mice is still limited. Indeed, all of the reported pheno- types for SIRT6 deficiency are based on experiments that were done in mice younger than two months old. Moreover, given that the characterization of SIRT6 effects was done in young mice it is not known if any of the reported phenotypes are gender specific. Therefore, two criti- cal pieces are missing in this puzzle: what would be the effect of SIRT6 deficiency in adult mice and whether this effect is gender specific. Here we show that SIRT6-/- mice in a mixed 129/SvJ/ PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 2 / 15 Characterization of physiological defects in adult SIRT6-/- mice BALB/c background display increased survival that allows us to explore new SIRT6-related metabolic and developmental phenotypes in adult, aging mice. Electroretinography (ERG) Full field ERG was performed in anesthetized animals following overnight dark adaptation using a Ganzfeld dome and a computerized system (Espion E2, Diagnosys LLC, Littleton, MA). Pupils were dilated with 1% tropicamide and 2.5% phenylephrine and local anesthetic drops (benoxinate HCl, 0.4%; all ocular drops from Fisher Pharmaceuticals, Tel-Aviv, Israel), were administered prior to placing gold-wire active electrodes on the central cornea. A refer- ence electrode was placed on the tongue and a needle ground electrode was placed intramus- cularly in the hip area. Dark-adapted rod and mixed cone-rod as well as light-adapted 16Hz cone flicker responses to a series of white flashes of increasing intensities (0.00008–9.6 cd•s/ m2) were recorded. All ERG responses were filtered at 0.3–500Hz, and signal averaging was applied. Quantitative real time PCR Total RNA was extracted using RNeasy kit according to the manufacturer’s protocol. cDNA was generated using SuperScript Vilo (Invitrogen). Quantitative real-time PCR was performed in triplicate using Absolute blue SYBR Green (Thermo) in a Chromo4 instrument cycler (Bio- Rad). Ct values were normalized to 18S. Primer sequences were as follows: Glut1, 5'-TCAACGAGCATCTTCGAGAAGGCA-3' and 5'-TCGTCCAGCTCGCTCTACAACAAA-3'; 18S, 5'-AGGAATTGACGGAAGGGCAC-3' and 5'-GTGCAGCCCCGGACATCTAAG-3'. Primer sequences were as follows: Glut1, 5'-TCAACGAGCATCTTCGAGAAGGCA-3' and 5'-TCGTCCAGCTCGCTCTACAACAAA-3'; 18S, 5'-AGGAATTGACGGAAGGGCAC-3' and 3 / 15 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Characterization of physiological defects in adult SIRT6-/- mice Histology Tissues were fixed in 4% buffered formaldehyde, embedded in paraffin, sectioned and stained with hematoxylin and eosin. The histology analysis of skin sections was done by Patholab, Israel. Ocular histology and measurements of outer nuclear layer (ONL) thickness Eyes were enucleated, fixed in Davidson solution, embedded in paraplast, and serially cut in 4 μm sections through the center of the optic nerve. For descriptive histology and quantitative analysis, sections were stained with hematoxylin and eosin. All observations and photography were performed using an Olympus BX41 microscope equipped with a DP70 digital camera. Image processing and quantification were performed using Adobe Photoshop CS3 and Image- Pro Plus 6.0 software. Thickness of the ONL was measured as previously described [29]. Briefly, ONL thickness was measured three times in the central part of each hemiretina and results from the corre- sponding retinal areas were averaged. Statistical analysis Results are expressed as mean ± SEM. P-values were calculated by unpaired Student’s t-tests or one-way ANOVA using SPSS Statistics 17.0 software. 1Within 15 females and 26 males, only one male died before 30 days of age. https://doi.org/10.1371/journal.pone.0176371.t001 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Results SIRT6+/- 129/SvJ mice [10] were mated with BALB/c mice to examine the effect of SIRT6 defi- ciency in this hybrid background. SIRT6 deficiency was validated in MEFs, liver and muscle (S1 Fig). Breeding of the resulting SIRT6 heterozygote F1 generation obtained SIRT6+/+, SIRT6+/- and SIRT6-/- 129/SvJ/BALB/c mice. In 182 newborns, genotype segregation did not show the expected Mendelian ratio of 1:2:1 (Table 1), with a higher prevalence of WT mice Table 1. Genotypes of progeny from crosses of F1 SIRT6+/−mice. Stage Percent (and number) of offspring WT HET KO E13.5 30.4% (14) 50% (23) 19.6% (9) Born 34% (62) 43.4% (79) 22.5% (411) 1Within 15 females and 26 males, only one male died before 30 days of age. https://doi org/10 1371/journal pone 0176371 t001 Table 1. Genotypes of progeny from crosses of F1 SIRT6+/−mice. PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 4 / 15 Characterization of physiological defects in adult SIRT6-/- mice (χ2 test analysis, p< 0.05). Interestingly, the genotype ratio in embryos at day 13.5 (E13.5) was as expected from mendelian genetics, suggesting that the change in the ratio observed in born mice might be caused by selective and increased prenatal mortality of SIRT6+/- and SIRT6-/- embryos between E13.5 and birth (Table 1). Therefore, in the 129/SvJ/BALB/c background, partial or complete absence of SIRT6 had a significant effect on prenatal survival. The survival of male and female 129/SvJ/BALB/c mice was compared between SIRT6 KO and WT littermates. In both genders, death of KO mice began to occur within 30 days. Yet, whereas the median survival time of SIRT6 KO male mice was 124 days and 90% of them died before 200 days of age, more than 75% of the female KO mice survived over 300 days of age (Fig 1). These findings show a gender specific effect of SIRT6 on mouse survival. Next, the effect of SIRT6 deficiency on metabolism in this hybrid background was exam- ined. In comparison to their WT littermates, the body weight of SIRT6 KO mice was signifi- cantly lower in both genders (Fig 2A), as has been seen in a previous study [10]. Differently, both male and female KO mice did not show significant change in total body fat in comparison to their WT littermates (Fig 2B), only a trend. Such a change was seen in another mixed genetic background [12]. Results In both genders, the weight of SIRT6 KO mice was significantly lower as compared with age-matched WT animals. (B) KO mice showed a trend towards lower percentage of total body fat in comparison to their WT littermates. The percentage of total body fat was measured at 6–8 months of age. Results are presented as mean ± SEM. n = 3–14 per group. https://doi.org/10.1371/journal.pone.0176371.g002 in Fig 3D, the KO mice secreted similar levels of insulin upon glucose stimulation in compari- Fig 2. Effect of SIRT6 deficiency on body weight and adiposity. (A) Average body weight of WT and KO male (left panel) and female (right panel) mice. In both genders, the weight of SIRT6 KO mice was significantly lower as compared with age-matched WT animals. (B) KO mice showed a trend towards lower percentage of total body fat in comparison to their WT littermates. The percentage of total body fat was measured at 6–8 months of age. Results are presented as mean ± SEM. n = 3–14 per group. https://doi.org/10.1371/journal.pone.0176371.g002 Fig 2. Effect of SIRT6 deficiency on body weight and adiposity. (A) Average body weight of WT and KO male (left panel) and female (right panel) mice. In both genders, the weight of SIRT6 KO mice was significantly lower as compared with age-matched WT animals. (B) KO mice showed a trend towards lower percentage of total body fat in comparison to their WT littermates. The percentage of total body fat was measured at 6–8 months of age. Results are presented as mean ± SEM. n = 3–14 per group. https://doi org/10 1371/journal pone 0176371 g002 https://doi.org/10.1371/journal.pone.0176371.g002 https://doi.org/10.1371/journal.pone.0176371.g002 in Fig 3D, the KO mice secreted similar levels of insulin upon glucose stimulation in compari- son to their WT littermates. Basal insulin level is known to be significantly low in SIRT6 defi- cient mice [12, 13]. As seen in Fig 3D, basal insulin levels are significantly lower only in KO male mice whereas in females there was no significant change, although both male and female KO mice presented higher glucose uptake compared to their WT littermates. Altogether these findings suggest that in this mixed background the increased glucose uptake in SIRT6 KO mice is insulin independent. GLUT1 is an insulin independent glucose transporter. Thus, the levels of GLUT1 in these mice were measured. First, GLUT1 RNA levels were measured in mouse embryonic fibroblasts (MEF’s), liver and muscle of WT and SIRT6 KO mice. Results Skin sections stained with hematoxylin and eosin (H&E) revealed loss of subcutaneous fat in KO male mice (S2 Fig), similar to changes in subcutaneous fat that have been previously reported [10]. The effect of SIRT6 deficiency on glucose homeostasis was examined initially through a glu- cose tolerance test (GTT) in mice fed with chow diet. Significantly, increased glucose uptake was found in both male and female KO mice during the 2 hours after glucose injection as com- pared to WT littermates (Fig 3A and 3B). The calculated homeostasis model assessment (HOMA) measure of insulin resistance was lower in KO mice as compared with WT animals (0.41±0.07 vs. 0.85±0.13, P = 0.033 for males and 0.32±0.22 vs. 0.82±0.12, P = 0.07 in females). The observed increased glucose uptake in KO mice could result from increased insulin sensi- tivity, an increase in insulin secretion or from insulin independent increased glucose uptake. In order to examine whether there is a change in insulin sensitivity, an insulin tolerance test (ITT) was performed. No significant difference was found between the experimental groups (Fig 3C). In order to explore the possibility that the KO mice secrete more insulin in response to glucose stimulus, we measured insulin blood levels after glucose stimulation (GSIS). As seen Fig 1. Gender specific effect of SIRT6 deficiency on mouse survival. Kaplan—Meier survival curves of WT, HET and KO male (left panel) and female (right panel) mice in 129/SvJ/BALB/c background. 90% of KO male mice died before 200 days of age and more than 75% of KO female mice survived over 300 days of age. Males, n = 23–39 per group; Females, n = 14–40 per group. https://doi.org/10.1371/journal.pone.0176371.g001 Fig 1. Gender specific effect of SIRT6 deficiency on mouse survival. Kaplan—Meier survival curves of WT, HET and KO male (left panel) and female (right panel) mice in 129/SvJ/BALB/c background. 90% of KO male mice died before 200 days of age and more than 75% of KO female mice survived over 300 days of age. Males, n = 23–39 per group; Females, n = 14–40 per group. https://doi.org/10.1371/journal.pone.0176371.g001 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 5 / 15 Characterization of physiological defects in adult SIRT6-/- mice Fig 2. Effect of SIRT6 deficiency on body weight and adiposity. (A) Average body weight of WT and KO male (left panel) and female (right panel) mice. Characterization of physiological defects in adult SIRT6-/- mice Fig 3. Increased glucose uptake in SIRT6 deficient mice. Glucose tolerance test (GTT) in WT and KO male (A) and female (B) mice. Area under curve (AUC) is shown on the right for each GTT. Both male and female KO mice show increased glucose uptake in comparison to WT mice. (C) Insulin tolerance test in male (left panel) and female (right panel) mice showed no significant difference between genotypes. (D) Insulin blood levels after glucose stimulation as determined by ELISA. All assays were done at 3–4 months old. Results are presented as mean ± SEM. n = 5–14 per group. * P < 0.05, ** P < 0.01 Fig 3. Increased glucose uptake in SIRT6 deficient mice. Glucose tolerance test (GTT) in WT and KO male (A) and female (B) mice. Area under curve (AUC) is shown on the right for each GTT. Both male and female KO mice show increased glucose uptake in comparison to WT mice. (C) Insulin tolerance test in male (left panel) and female (right panel) mice showed no significant difference between genotypes. (D) Insulin blood levels after glucose stimulation as determined by ELISA. All assays were done at 3–4 months old. Results are presented as mean ± SEM. n = 5–14 per group. * P < 0.05, ** P < 0.01 https://doi.org/10.1371/journal.pone.0176371.g003 of GLUT1 in the membrane of SIRT6 KO MEFs in this background was found (S3 Fig). How- ever, in muscle, expression levels of GLUT1 in SIRT6 KO 129/SvJ/BALA/c male mice (Fig 4A) was twofold higher than in WT mice (P < 0.05). As the muscle is a major organ of glucose con- sumption, this could explain the higher glucose uptake in these mice. Interestingly, these find- ings are in concordance with a previous report showing similar GLUT1-dependent increased glucose uptake in SIRT6 -/- 129/SvJ mice [13] and unlike SIRT6 -/- 129/Black Swiss/FVB mice in which the higher glucose uptake is controlled by insulin [12]. In SIRT6 deficient 129/SvJ mice, IGF-1 levels were decreased to a level similar to mice with liver specific IGF-1 deficiency [10]. Thus, we measured IGF-1 levels in the sera of SIRT6 KO and SIRT6+/+ 129/SvJ/BALB/c mice. In both male and female KO mice, no significant decrease in IGF-1 levels was found (Fig 4B). PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Results SIRT6 protein deficiency in these KO tissues and cells, was validated by western blot analyses (S1 Fig). In comparison to their WT lit- termates, no significant difference in the RNA expression levels of GLUT1 was found in liver and MEFs of SIRT6 KO 129/SvJ/BALB/c mice (Fig 4A). In addition, no difference in the levels PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 6 / 15 At this time point no differences in outer nuclear layer (ONL, photoreceptor nuclei layer) thickness were found between WT and SIRT6 KO as assessed by quantitative histological techniques (Fig 7A and 7C). However, by six months of age, scotopic a-wave responses at higher intensities showed a trend towards lower amplitudes (Fig 6C) and b-wave amplitudes were significantly reduced in SIRT6 KO animals as compared with WT and HET mice (Fig 6D). Impairment of scotopic retinal function per- sisted through 10 months of age (Fig 6E and 6F). Light-adapted cone responses were also markedly affected at six and ten months of age, being 2.5-3-fold lower in SIRT6 KO mice (Fig 6G and 6H). This was accompanied by thinning of the retina and its layers (Fig 7A and 7B, S4A and S4B Fig). Measurements of ONL thickness correlated with the functional ERG showed preserved a-wave amplitudes in KO as compared to WT and HET eyes (Fig 6A), while a trend towards lower dark-adapted b-wave ERG amplitudes was observed, attaining statistical significance at high stimulus intensities (Fig 6B). At this time point no differences in outer nuclear layer (ONL, photoreceptor nuclei layer) thickness were found between WT and SIRT6 KO as assessed by quantitative histological techniques (Fig 7A and 7C). However, by six months of age, scotopic a-wave responses at higher intensities showed a trend towards lower amplitudes (Fig 6C) and b-wave amplitudes were significantly reduced in SIRT6 KO animals as compared with WT and HET mice (Fig 6D). Impairment of scotopic retinal function per- sisted through 10 months of age (Fig 6E and 6F). Light-adapted cone responses were also markedly affected at six and ten months of age, being 2.5-3-fold lower in SIRT6 KO mice (Fig 6G and 6H). This was accompanied by thinning of the retina and its layers (Fig 7A and 7B, S4A and S4B Fig). Measurements of ONL thickness correlated with the functional ERG Fig 5. Corneal changes in SIRT6 deficient mice over time. (A-B) Photographs and histological sections of corneas in 10-month-old WT and KO mice. Corneas of normal WT mice are fully transparent, with no ingrowth of blood vessels (A, left panel). In contrast, from 6 months of age, corneal scarring and vascularization were often present in KO mice eyes (A, right panel). Corneal injury which we observed in adult KO mice prompted us to study the eye in this model more closely. At one month of age, SIRT6 KO mice had normal corneal appearance (not shown). However, by 5–6 months, severe corneal injury was often observed in SIRT6 KO mice including corneal scarring, vascularization, signs of stromal edema and inflammation (Fig 5A right panel and Fig 5B lower histological section). Assessment of retinal function by full field electroretinography (ERG) showed differences in amplitudes of dark- and light- adapted responses between WT and KO mice (Fig 6). At one month of age, SIRT6 KO mice PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 7 / 15 Characterization of physiological defects in adult SIRT6-/- mice Fig 4. Effect of SIRT6 deficiency on GLUT1 expression and serum IGF-1 levels. (A) Relative expression of GLUT1 in MEF’s, liver and muscle of WT and KO male mice as determined by quantitative real-time PCR. (B) Serum IGF-1 levels of WT and KO male and female mice. All measurements were done at 6–8 months of age. Results are presented as mean ± SEM. n = 5–14 per group. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g004 Fig 4. Effect of SIRT6 deficiency on GLUT1 expression and serum IGF-1 levels. (A) Relative expression of GLUT1 in MEF’s, liver and muscle of WT and KO male mice as determined by quantitative real-time PCR. (B) Serum IGF-1 levels of WT and KO male and female mice. All measurements were done at 6–8 months of age. Results are presented as mean ± SEM. n = 5–14 per group. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g004 Fig 4. Effect of SIRT6 deficiency on GLUT1 expression and serum IGF-1 levels. (A) Relative expression of GLUT1 in MEF’s, liver and muscle of WT and KO male mice as determined by quantitative real-time PCR. (B) Serum IGF-1 levels of WT and KO male and female mice. All measurements were done at 6–8 months of age. Results are presented as mean ± SEM. n = 5–14 per group. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g004 https://doi.org/10.1371/journal.pone.0176371.g004 showed preserved a-wave amplitudes in KO as compared to WT and HET eyes (Fig 6A), while a trend towards lower dark-adapted b-wave ERG amplitudes was observed, attaining statistical significance at high stimulus intensities (Fig 6B). Histological section in KO mice (B, lower panel) shows corneal thickening, stromal edema, inflammation and scarring that are not present in WT mice of the same age (B, upper panel). B: Paraffin embedded sections stained with hematoxylin and eosin; C- Cornea; L-Lens; original magnification x10; Scale bar = 200μm. Fig 5. Corneal changes in SIRT6 deficient mice over time. (A-B) Photographs and histological sections of corneas in 10-month-old WT and KO mice. Corneas of normal WT mice are fully transparent, with no ingrowth of blood vessels (A, left panel). In contrast, from 6 months of age, corneal scarring and vascularization were often present in KO mice eyes (A, right panel). Histological section in KO mice (B, lower panel) shows corneal thickening, stromal edema, inflammation and scarring that are not present in WT mice of the same age (B, upper panel). B: Paraffin embedded sections stained with hematoxylin and eosin; C- Cornea; L-Lens; original magnification x10; Scale bar = 200μm. https://doi.org/10.1371/journal.pone.0176371.g005 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 8 / 15 Characterization of physiological defects in adult SIRT6-/- mice Fig 6. Retinal function in SIRT6 deficient mice. Scotopic (A-F) and photopic (G, H) full-field electroretinography (ERG) recordings may suggest accelerated aging in KO eyes. At one month of age, dark- adapted ERG responses to a series of increasing white flash intensities showed preserved a-wave amplitudes in KO as compared to WT and HET eyes (A), while a trend towards lower amplitudes was observed in the scotopic b-wave amplitudes, attaining statistical significance at 3 high intensities (B). At the age of 6 months, scotopic a- wave responses at higher intensities showed a trend towards lower amplitudes (C), and b-wave amplitudes were markedly reduced (D). At 10 months, both a- and b-wave amplitudes in KO mice were markedly reduced as compared to WT and HET mice (E, F). Light adapted 16Hz flicker responses were also markedly reduced in KO mice at 6 and 10 months (G, H). ERG responses were similar in WT and HET mice at all ages. Results are presented as mean ± SEM; n = 6–9 for WT, n = 7–8 for HET, and n = 4–5 for KO mice. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g006 Characterization of physiological defects in adult SIRT6 mice Fig 6. Retinal function in SIRT6 deficient mice. Scotopic (A-F) and photopic (G, H) full-field Fig 6. Retinal function in SIRT6 deficient mice. Characterization of physiological defects in adult SIRT6-/- mice f d h f h k f l d d h l f Fig 7. Retinal structural changes well correlated with the ERG findings. (A) Representative retinal sections stained with hematoxylin and eosin demonstrate marked decrease of retinal thickness in KO eyes by the age of 6 months with further thinning by 10 months. Original magnification x20; Scale bar = 100μm. (B) Throughout the experiment, total retinal thickness was reduced in KO eyes as compared with WT littermates. (C) Thickness of the outer nuclear layer (ONL, containing the photoreceptor nuclei) did not differ between WT and KO mice at the age of 1 month, but by the age of 6 and 10 months, the ONL was significantly thinner in animals lacking SIRT6 activity. Results are presented as mean ± SEM; n = 5–7 for WT, n = 4–5 for KO mice. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g007 Fig 7. Retinal structural changes well correlated with the ERG findings. (A) Representative retinal sections stained with hematoxylin and eosin demonstrate marked decrease of retinal thickness in KO eyes by the age of 6 months with further thinning by 10 months. Original magnification x20; Scale bar = 100μm. (B) Throughout the experiment, total retinal thickness was reduced in KO eyes as compared with WT littermates. (C) Thickness of the outer nuclear layer (ONL, containing the photoreceptor nuclei) did not differ between WT and KO mice at the age of 1 month, but by the age of 6 and 10 months, the ONL was significantly thinner in animals lacking SIRT6 activity. Results are presented as mean ± SEM; n = 5–7 for WT, n = 4–5 for KO mice. * P < 0.05. Fig 7. Retinal structural changes well correlated with the ERG findings. (A) Representative retinal sections stained with hematoxylin and eosin demonstrate marked decrease of retinal thickness in KO eyes by the age of 6 months with further thinning by 10 months. Original magnification x20; Scale bar = 100μm. (B) Throughout the experiment, total retinal thickness was reduced in KO eyes as compared with WT littermates. (C) Thickness of the outer nuclear layer (ONL, containing the photoreceptor nuclei) did not differ between WT and KO mice at the age of 1 month, but by the age of 6 and 10 months, the ONL was significantly thinner in animals lacking SIRT6 activity. findings: at 6 months of age, ONL thickness was significantly reduced in the central retina of SIRT6 KO as compared with WT mice with further thinning by 10 months (Fig 7C). findings: at 6 months of age, ONL thickness was significantly reduced in the central retina of SIRT6 KO as compared with WT mice with further thinning by 10 months (Fig 7C). Results are presented as mean ± SEM; n = 5–7 for WT, n = 4–5 for KO mice. * P < 0.05. htt //d i /10 1371/j l 0176371 007 https://doi.org/10.1371/journal.pone.0176371.g007 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Scotopic (A-F) and photopic (G, H) full-field electroretinography (ERG) recordings may suggest accelerated aging in KO eyes. At one month of age, dark- adapted ERG responses to a series of increasing white flash intensities showed preserved a-wave amplitudes in KO as compared to WT and HET eyes (A), while a trend towards lower amplitudes was observed in the scotopic b-wave amplitudes, attaining statistical significance at 3 high intensities (B). At the age of 6 months, scotopic a- wave responses at higher intensities showed a trend towards lower amplitudes (C), and b-wave amplitudes were markedly reduced (D). At 10 months, both a- and b-wave amplitudes in KO mice were markedly reduced as compared to WT and HET mice (E, F). Light adapted 16Hz flicker responses were also markedly reduced in KO mice at 6 and 10 months (G, H). ERG responses were similar in WT and HET mice at all ages. Results are presented as mean ± SEM; n = 6–9 for WT, n = 7–8 for HET, and n = 4–5 for KO mice. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g006 Fig 6. Retinal function in SIRT6 deficient mice. Scotopic (A-F) and photopic (G, H) full-field electroretinography (ERG) recordings may suggest accelerated aging in KO eyes. At one month of age, dark- adapted ERG responses to a series of increasing white flash intensities showed preserved a-wave amplitudes in KO as compared to WT and HET eyes (A), while a trend towards lower amplitudes was observed in the scotopic b-wave amplitudes, attaining statistical significance at 3 high intensities (B). At the age of 6 months, scotopic a- wave responses at higher intensities showed a trend towards lower amplitudes (C), and b-wave amplitudes were markedly reduced (D). At 10 months, both a- and b-wave amplitudes in KO mice were markedly reduced as compared to WT and HET mice (E, F). Light adapted 16Hz flicker responses were also markedly reduced in KO mice at 6 and 10 months (G, H). ERG responses were similar in WT and HET mice at all ages. Results are presented as mean ± SEM; n = 6–9 for WT, n = 7–8 for HET, and n = 4–5 for KO mice. * P < 0.05. https://doi.org/10.1371/journal.pone.0176371.g006 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 9 / 15 Discussion SIRT6 deficient 129/SvJ mice display a broad-range of phenotypes and die within 4 weeks after birth [10]. Therefore, our knowledge regarding the effect of SIRT6 in adult mice is lacking. To address this, and to examine whether SIRT6 phenotypes are strain and/or age specific, we gen- erated SIRT6 KO 129/SvJ/BALB/c mice. These KO mice on a mixed background survive to adulthood, thus allowing the opportunity to explore new SIRT6 related metabolic and develop- mental phenotypes unique to adult mice. In comparison to their WT littermates, SIRT6 KO mice had a gender-dependent decreased lifespan, significantly lower body weight, and faster glucose uptake. In addition, ocular changes were noted, including degradation of retinal func- tion and structure that may reflect accelerated aging. Different mouse models of SIRT1 deficiency showed a strain-dependent set of phenotypes [26, 27]. Similarly, deletion of Sir2 in yeast also results in different phenotypes in different backgrounds [3, 30]. In contrast to previously published SIRT6 -/- mouse models, our mixed background animals have a biased mendelian ratio between WT, SIRT6+/- and SIRT6-/-. The PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 10 / 15 Characterization of physiological defects in adult SIRT6-/- mice SIRT6 deficient mice are not hypoglycemic, do not show lower IGF-1 levels, survive to adult- hood and have gender specific effects on lifespan. Moreover, in contrast to 129/Black Swiss/ FVB background animals, our mice have normal insulin-dependent glucose uptake. This strain-dependent variability may stem from the central role of deacetylases that regulate the transcription of many genes, and therefore the observed phenotypes should be considered to be a combinatorial outcome. SIRT6 as a global histone H3 K9 and H3K56 deacetylase nicely fits into such a model. Interestingly, the newborn genotype segregation was not a Mendelian ratio of 1:2:1 (Table 1). Assuming that no death is expected in WT newborn, the number of HET newborn (73) was only 59% of the expected (124 HET newborn). Likewise, based on the WT new- borns, the number of SIRT6 KO newborn is also only 66% of expected. This change in geno- types segregation was significant (P<0.05) whereas similar analysis on the genotype ratio of E13.5 embryos did not show a significant change in the ratio (Table 1). Therefore, SIRT6 deficiency or haplodeficiency results in significantly higher prenatal death rates starting after day E13.5. Discussion A major advantage of the current mouse model over previously published models is that it allows examining the role of SIRT6 deficiency in adult mice. 129/SvJ mice deficient for SIRT6 die prematurely and it is therefore impossible to follow the role of SIRT6 in mice older than 4 weeks in this background. Another published model of SIRT6 deficiency in mixed background results in higher survival rates of KO mice, but in this study, the median lifespan was also about 4 weeks of age with a set of physiological defects similar to those observed in the 129/SvJ background [10, 12]. Due to the young age of the examined animals in both of these SIRT6 KO models, it is not clear if SIRT6-related pathologies are developmental or indeed represen- tative of age-related degenerative diseases. Therefore, our knowledge regarding the role of SIRT6 in adult mice is very limited. Here, we were able to identify SIRT6 effects in mature mice up to the age of 10 months: the metabolic effects, including faster glucose uptake; the male-specific shortening of lifespan and an age-dependent impairment of retinal function accompanied by retinal thinning. Importantly, Xiao and his colleagues also reported some phenotypes for adult SIRT6 mice. However, in that report, only few mice survived for longer than five weeks and therefore these mice in fact represent the exceptions of their experimental conditions. With regard to the metabolic effects of SIRT6 deficiency, the higher glucose uptake which appears in SIRT6 KO mice is a constitutive phenotype that was present in the two mouse mod- els examined so far [12, 13]. Similar to these studies, 129/SvJ/BALB/c males and females exam- ined in the present study have higher glucose uptake in KO mice [13]. In our SIRT6 KO model, as also reported in the 129/SvJ SIRT6 KO mice, an increase in peripheral levels of GLUT1 as measured in muscle seems to be the main cause of this phenotype. ITT and GSIS results suggest that this higher glucose uptake is insulin independent, differing from the study of Xiao et al. that examined SIRT6 deficiency in another mixed background in which insulin dependent hypoglycemia was reported [12]. Interestingly, in contrast to a previous study, no difference was found in the levels of GLUT1 in SIRT6 KO vs SIRT6 WT MEF’s isolated from 129/SvJ/BALB/c mice. PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Characterization of physiological defects in adult SIRT6-/- mice well-known marker of aging and was found in various mouse models that manifest premature aging. Apart from the metabolic abnormalities identified in the SIRT6 KO mice, interesting changes also occurred in the eyes of affected mice including corneal injury and degradation of retinal function and structure over time. In many ways, the retinal changes resemble those seen during aging in normal mice, but appear at much earlier ages. Silberman et al [11] recently reported that young SIRT6 KO mice in 129/SvJ background have attenuated ERG amplitudes, but retinal structure was preserved at this age. In this study animals only up to 1 month of age were examined at a single time point, suggesting that the retinal dysfunction observed is not necessarily degenerative. As can be seen in Fig 6, in WT and HET animals ERG amplitudes decrease between the ages of 6 and 10 months but the reduction is much more pronounced in the 129/SvJ/BALB/c Sirt6 KO mice. At the age of 6 and 10 months, dark-adapted ERG responses to higher stimulus intensities were 40–50% lower in SIRT6 KO mice than in WT. This level of attenuation, when due to aging, was reported to occur in nor- mal B6D2F1/J mice only at the age of 2.5 years as compared to 4 months [31]. In another study, Li et al. found that in C57BL/6 mice amplitudes of the dark-adapted ERG b-wave decreased by about 15% at the age of 6 months and by about 32% by the age of 12 months as compared with 2 month old animals [32]. The decrease noted in SIRT6 KO by 6 and 10 months exceeds these levels, but interestingly, both the functional and the structural degrada- tion are much less severe than those seen in genetically-determined retinal degenerations such as in rd10 mice [33]. Our results may suggest a role for SIRT6 in maintenance and function of retinal cells over time. SIRT6 was shown to regulate genome stability and resistance to oxidative injury caused by H2O2 [10]. Therefore, one can predict that in the absence of SIRT6 there will be increased damage to DNA as well as oxidative injury caused by reactive oxygen species. Oxidative injury has been shown to play an important role in retinal degeneration including age-related retinal disease [34–36], and it is possible that this is one of the mechanisms leading to the accelerated retinal degradation observed in the SIRT6 KO mice. Interestingly, other sirtuins, and especially Sirt1, have been shown to play a role in genetically determined and age-related retinal degener- ation in rodent models [37–39] and activation of Sirt1 by oral supplementation of a SIRT1 acti- vator, resveratrol, was shown to attenuate light-induced retinal injury and age-related retinal dysfunction in mice and rats [39, 40]. Taken together, these findings demonstrate again that SIRT6 may play a role in age-related metabolic and ocular changes, and raises the possibility that this protein may serve as a novel therapeutic target. Discussion The male-specific effect of SIRT6 deficiency on lifespan observed in the present study may correlate with the observation that SIRT6 overexpression in mixed background increased lifespan exclusively in male mice. Lack of SIRT6 similarly affected the body weight, insulin levels after glucose stimulus and GTT in males and females. Therefore the reason that SIRT6 deficient male mice have a shorter lifespan than females in 129/SvJ/BALB/c background remains unclear for now. In addition, 129/SvJ/BALB/c KO male mice showed a decrease in subcutaneous fat similarly to 129SvJ background KO mice. Reduced subcutaneous fat is a PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 11 / 15 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Conclusions The primary goal of this study was to examine the effects of SIRT6 deficiency in a mixed genetic background in order to abolish the potential strain specific dependent biased phenotypes and to allow long term follow-up into adulthood. Indeed, using this new model, we were able to show the effect of SIRT6 in adult mice and solve one of the main limitations of previous SIRT6 KO models, namely that of premature death at one month of age. The present study reveals that part of the severe metabolic phenotypes caused by SIRT6 defi- ciency are not strain specific, such as the higher glucose uptake in mice. However, some characteristics such as shortened lifespan are much less severe in this mixed background and, importantly, gender specific. The long-term evaluation of retinal function and structure afforded by this model suggests that absence of Sirt6 may accelerate retinal dysfunction and ageing. 12 / 15 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Characterization of physiological defects in adult SIRT6-/- mice Supporting information Supporting information S1 Fig. SIRT6 protein level. Western blot analysis for SIRT6 protein expression in WT and KO mice. (PDF) S2 Fig. Loss of subcutaneous fat in SIRT6 deficient mice. Hematoxylin and eosin staining of skin sections from WT and KO male mice, showing an acute loss of subcutaneous fat in the absence of SIRT6. (PDF) S3 Fig. GLUT1 protein levels. Flow cytometry analysis of GLUT1 protein levels in cell mem- brane of WT HET and KO cell MEFs. (PDF) S4 Fig. Thinning of the Ganglion Cell Layer (GCL) in SIRT6 deficient mice with age. Thickness of the GCL was significantly lower in KO mice as compared with WT littermates throughout the experiment. However, measurement of linear density of ganglion cell nuclei did not show a difference between WT and KO mice: number of nuclei decreased in both experimental groups over time. (PDF) S1 Fig. SIRT6 protein level. Western blot analysis for SIRT6 protein expression in WT and KO mice. (PDF) S1 Fig. SIRT6 protein level. Western blot analysis for SIRT6 protein expression in WT and KO mice. (PDF) S2 Fig. Loss of subcutaneous fat in SIRT6 deficient mice. Hematoxylin and eosin staining of skin sections from WT and KO male mice, showing an acute loss of subcutaneous fat in the absence of SIRT6. (PDF) S3 Fig. GLUT1 protein levels. Flow cytometry analysis of GLUT1 protein levels in cell mem- brane of WT HET and KO cell MEFs. (PDF) S3 Fig. GLUT1 protein levels. Flow cytometry analysis of GLUT1 protein levels in cell mem- brane of WT HET and KO cell MEFs. (PDF) S3 Fig. GLUT1 protein levels. Flow cytometry analysis of GLUT1 protein levels in cell mem- brane of WT HET and KO cell MEFs. (PDF) S4 Fig. Thinning of the Ganglion Cell Layer (GCL) in SIRT6 deficient mice with age. Thickness of the GCL was significantly lower in KO mice as compared with WT littermates throughout the experiment. However, measurement of linear density of ganglion cell nuclei did not show a difference between WT and KO mice: number of nuclei decreased in both experimental groups over time. Acknowledgments We thank the members of the Cohen lab, in particular Shoshana Naiman, for their helpful comments on the manuscript. This study was supported by the Israeli Academy of Sciences, I-Core, Binational US-Israel Science Foundation, Israel Cancer Association, Koret Founda- tion, Israel Cancer Research Fund, the Israel Ministry of Health, Lir’ot association, TEVA NNE program, EFSD, D- Cure, the ERC: European Research Council and the Yedidut Research Fund. Conceptualization: HYC. Funding acquisition: HYC. Investigation: LN VP AO. Methodology: HYC EB. PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 References 1. Tissenbaum HA, Guarente L. Increased dosage of a sir-2 gene extends lifespan in Caenorhabditis ele- gans. Nature. 2001; 410(6825):227–30. https://doi.org/10.1038/35065638 PMID: 11242085 2. Rogina B, Helfand SL. Sir2 mediates longevity in the fly through a pathway related to calorie restriction. Proc Natl Acad Sci U S A. 2004; 101(45):15998–6003. https://doi.org/10.1073/pnas.0404184101 PMID: 15520384 3. Kaeberlein M, Kirkland KT, Fields S, Kennedy BK. Sir2-independent life span extension by calorie restriction in yeast. PLoS Biol. 2004; 2(9):E296. Epub 2004/08/26. https://doi.org/10.1371/journal.pbio. 0020296 PMID: 15328540 4. Herranz D, Muñoz-Martin M, Cañamero M, Mulero F, Martinez-Pastor B, Fernandez-Capetillo O, et al. Sirt1 improves healthy ageing and protects from metabolic syndrome-associated cancer. Nat Commun. 2010; 1:3. https://doi.org/10.1038/ncomms1001 PMID: 20975665 5. Satoh A, Brace CS, Rensing N, Cliften P, Wozniak DF, Herzog ED, et al. Sirt1 extends life span and delays aging in mice through the regulation of Nk2 homeobox 1 in the DMH and LH. Cell Metab. 2013; 18(3):416–30. https://doi.org/10.1016/j.cmet.2013.07.013 PMID: 24011076 6. Gil R, Barth S, Kanfi Y, Cohen HY. SIRT6 exhibits nucleosome-dependent deacetylase activity. Nucleic Acids Res. 2013; 41(18):8537–45. https://doi.org/10.1093/nar/gkt642 PMID: 23892288 7. Kawahara TL, Michishita E, Adler AS, Damian M, Berber E, Lin M, et al. SIRT6 links histone H3 lysine 9 deacetylation to NF-kappaB-dependent gene expression and organismal life span. Cell. 2009; 136 (1):62–74. Epub 2009/01/13. https://doi.org/10.1016/j.cell.2008.10.052 PMID: 19135889 8. Mao Z, Hine C, Tian X, Van Meter M, Au M, Vaidya A, et al. SIRT6 promotes DNA repair under stress by activating PARP1. Science. 2011; 332(6036):1443–6. https://doi.org/10.1126/science.1202723 PMID: 21680843 9. Jiang H, Khan S, Wang Y, Charron G, He B, Sebastian C, et al. SIRT6 regulates TNF-α secretion through hydrolysis of long-chain fatty acyl lysine. Nature. 2013; 496(7443):110–3. https://doi.org/10. 1038/nature12038 PMID: 23552949 10. Mostoslavsky R, Chua KF, Lombard DB, Pang WW, Fischer MR, Gellon L, et al. Genomic instability and aging-like phenotype in the absence of mammalian SIRT6. Cell. 2006; 124(2):315–29. Epub 2006/ 01/28. https://doi.org/10.1016/j.cell.2005.11.044 PMID: 16439206 11. Silberman DM, Ross K, Sande PH, Kubota S, Ramaswamy S, Apte RS, et al. SIRT6 is required for nor- mal retinal function. PLoS One. 2014; 9(6):e98831. https://doi.org/10.1371/journal.pone.0098831 PMID: 24896097 12. Xiao C, Kim HS, Lahusen T, Wang RH, Xu X, Gavrilova O, et al. SIRT6 deficiency results in severe hypoglycemia by enhancing both basal and insulin-stimulated glucose uptake in mice. J Biol Chem. 2010; 285(47):36776–84. https://doi.org/10.1074/jbc.M110.168039 PMID: 20847051 13. Zhong L, D’Urso A, Toiber D, Sebastian C, Henry RE, Vadysirisack DD, et al. Project administration: EB HYC. Project administration: EB HYC. Resources: FWA. Supervision: EB HYC. Supervision: EB HYC. Validation: LN VP AO YK MR MA ST HYC EB. Validation: LN VP AO YK MR MA ST HYC EB. Visualization: LN VP AO YK. Writing – original draft: HYC EB. Writing – review & editing: LN VP AO YK EB HYC. Writing – review & editing: LN VP AO YK EB HYC. 13 / 15 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 Characterization of physiological defects in adult SIRT6-/- mice PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 References The histone deacetylase Sirt6 regulates glucose homeostasis via Hif1alpha. Cell. 2010; 140(2):280–93. Epub 2010/02/10. https://doi.org/10.1016/j.cell.2009.12.041 PMID: 20141841 14. Gertler AA, Cohen HY. SIRT6, a protein with many faces. Biogerontology. 2013; 14(6):629–39. https:// doi.org/10.1007/s10522-013-9478-8 PMID: 24213807 15. Kaidi A, Weinert BT, Choudhary C, Jackson SP. Human SIRT6 promotes DNA end resection through CtIP deacetylation. Science. 2010; 329(5997):1348–53. https://doi.org/10.1126/science.1192049 PMID: 20829486 16. Atsumi Y, Minakawa Y, Ono M, Dobashi S, Shinohe K, Shinohara A, et al. ATM and SIRT6/SNF2H Mediate Transient H2AX Stabilization When DSBs Form by Blocking HUWE1 to Allow Efficient γH2AX Foci Formation. Cell Rep. 2015; 13(12):2728–40. https://doi.org/10.1016/j.celrep.2015.11.054 PMID: 26711340 17. Tennen RI, Chua KF. Chromatin regulation and genome maintenance by mammalian SIRT6. Trends Biochem Sci. 2010; 36(1):39–46. Epub 2010/08/24. https://doi.org/10.1016/j.tibs.2010.07.009 PMID: 20729089 18. Michishita E, McCord RA, Berber E, Kioi M, Padilla-Nash H, Damian M, et al. SIRT6 is a histone H3 lysine 9 deacetylase that modulates telomeric chromatin. Nature. 2008; 452(7186):492–6. Epub 2008/ 03/14. https://doi.org/10.1038/nature06736 PMID: 18337721 19. Ghosh S, Liu B, Wang Y, Hao Q, Zhou Z. Lamin A Is an Endogenous SIRT6 Activator and Promotes SIRT6-Mediated DNA Repair. Cell Rep. 2015; 13(7):1396–406. https://doi.org/10.1016/j.celrep.2015. 10.006 PMID: 26549451 20. Kanfi Y, Peshti V, Gil R, Naiman S, Nahum L, Levin E, et al. SIRT6 protects against pathological dam- age caused by diet-induced obesity. Aging Cell. 2010; 9(2):162–73. Epub 2010/01/06. https://doi.org/ 10.1111/j.1474-9726.2009.00544.x PMID: 20047575 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 14 / 15 Characterization of physiological defects in adult SIRT6-/- mice 21. Elhanati S, Kanfi Y, Varvak A, Roichman A, Carmel-Gross I, Barth S, et al. Multiple regulatory layers of SREBP1/2 by SIRT6. Cell Rep. 2013; 4(5):905–12. https://doi.org/10.1016/j.celrep.2013.08.006 PMID: 24012758 22. Elhanati S, Ben-Hamo R, Kanfi Y, Varvak A, Glazz R, Lerrer B, et al. Reciprocal Regulation between SIRT6 and miR-122 Controls Liver Metabolism and Predicts Hepatocarcinoma Prognosis. Cell Rep. 2016; 14(2):234–42. https://doi.org/10.1016/j.celrep.2015.12.023 PMID: 26748705 23. Kanfi Y, Naiman S, Amir G, Peshti V, Zinman G, Nahum L, et al. The sirtuin SIRT6 regulates lifespan in male mice. Nature. 2012; 483(7388):218–21. https://doi.org/10.1038/nature10815 PMID: 22367546 24. Lerrer B, Gertler AA, Cohen HY. The complex role of SIRT6 in carcinogenesis. Carcinogenesis. 2016; 37(2):108–18. https://doi.org/10.1093/carcin/bgv167 PMID: 26717993 25. Sebastia´n C, Zwaans BM, Silberman DM, Gymrek M, Goren A, Zhong L, et al. The histone deacetylase SIRT6 is a tumor suppressor that controls cancer metabolism. Cell. 2012; 151(6):1185–99. https://doi. org/10.1016/j.cell.2012.10.047 PMID: 23217706 26. References McBurney MW, Yang X, Jardine K, Hixon M, Boekelheide K, Webb JR, et al. The mammalian SIR2al- pha protein has a role in embryogenesis and gametogenesis. Mol Cell Biol. 2003; 23(1):38–54. Epub 2002/12/17. https://doi.org/10.1128/MCB.23.1.38-54.2003 PMID: 12482959 27. Cheng HL, Mostoslavsky R, Saito S, Manis JP, Gu Y, Patel P, et al. Developmental defects and p53 hyperacetylation in Sir2 homolog (SIRT1)-deficient mice. Proc Natl Acad Sci U S A. 2003; 100 (19):10794–9. Epub 2003/09/10. https://doi.org/10.1073/pnas.1934713100 PMID: 12960381 28. Wang RH, Sengupta K, Li C, Kim HS, Cao L, Xiao C, et al. Impaired DNA damage response, genome instability, and tumorigenesis in SIRT1 mutant mice. Cancer Cell. 2008; 14(4):312–23. Epub 2008/10/ 07. https://doi.org/10.1016/j.ccr.2008.09.001 PMID: 18835033 29. Galbinur T, Obolensky A, Berenshtein E, Vinokur V, Chowers I, Chevion M, et al. Effect of para-amino- benzoic acid on the course of retinal degeneration in the rd10 mouse. J Ocul Pharmacol Ther. 2009; 25 (6):475–82. Epub 2009/12/24. https://doi.org/10.1089/jop.2009.0020 PMID: 20028256 30. Lin SJ, Defossez PA, Guarente L. Requirement of NAD and SIR2 for life-span extension by calorie restriction in Saccharomyces cerevisiae. Science. 2000; 289(5487):2126–8. Epub 2000/09/23. PMID: 11000115 31. Kolesnikov AV, Fan J, Crouch RK, Kefalov VJ. Age-related deterioration of rod vision in mice. J Neu- rosci. 2010; 30(33):11222–31. Epub 2010/08/20. https://doi.org/10.1523/JNEUROSCI.4239-09.2010 PMID: 20720130 32. Li C, Cheng M, Yang H, Peachey NS, Naash MI. Age-related changes in the mouse outer retina. Optom Vis Sci. 2001; 78(6):425–30. Epub 2001/07/11. PMID: 11444632 33. Obolensky A, Berenshtein E, Lederman M, Bulvik B, Alper-Pinus R, Yaul R, et al. Zinc-desferrioxamine attenuates retinal degeneration in the rd10 mouse model of retinitis pigmentosa. Free Radic Biol Med. 2011; 51(8):1482–91. Epub 2011/08/10. https://doi.org/10.1016/j.freeradbiomed.2011.07.014 PMID: 21824515 34. Hollyfield JG. Age-related macular degeneration: the molecular link between oxidative damage, tissue- specific inflammation and outer retinal disease: the Proctor lecture. Invest Ophthalmol Vis Sci. 2010; 51 (3):1275–81. Epub 2010/02/27. 35. Barot M, Gokulgandhi MR, Mitra AK. Mitochondrial dysfunction in retinal diseases. Curr Eye Res. 2011; 36(12):1069–77. Epub 2011/10/08. https://doi.org/10.3109/02713683.2011.607536 PMID: 21978133 36. Beatty S, Koh H, Phil M, Henson D, Boulton M. The role of oxidative stress in the pathogenesis of age- related macular degeneration. Surv Ophthalmol. 2000; 45(2):115–34. Epub 2000/10/18. PMID: 11033038 37. Jaliffa C, Ameqrane I, Dansault A, Leemput J, Vieira V, Lacassagne E, et al. Sirt1 involvement in rd10 mouse retinal degeneration. Invest Ophthalmol Vis Sci. 2009; 50(8):3562–72. https://doi.org/10.1167/ iovs.08-2817 PMID: 19407027 38. PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017 References Ozawa Y, Kubota S, Narimatsu T, Yuki K, Koto T, Sasaki M, et al. Retinal aging and sirtuins. Ophthalmic Res. 2010; 44(3):199–203. https://doi.org/10.1159/000316484 PMID: 20829644 39. Zeng Y, Yang K. Sirtuin 1 participates in the process of age-related retinal degeneration. Biochem Bio- phys Res Commun. 2015; 468(1–2):167–72. https://doi.org/10.1016/j.bbrc.2015.10.139 PMID: 26522222 40. Kubota S, Kurihara T, Ebinuma M, Kubota M, Yuki K, Sasaki M, et al. Resveratrol prevents light-induced retinal degeneration via suppressing activator protein-1 activation. Am J Pathol. 2010; 177(4):1725–31. https://doi.org/10.2353/ajpath.2010.100098 PMID: 20709795 15 / 15 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176371 April 27, 2017
https://openalex.org/W2184164827	https://he02.tci-thaijo.org/index.php/bkkmedj/article/download/218650/151416	English	null	Update of Inpatient Management of Polytrauma Patients	THE BANGKOK MEDICAL JOURNAL	2,013	cc-by	3,227	Update of Inpatient Management of Polytrauma Patients I npatient care of polytrauma victims requires careful attention to detail, and meticulous treatment of ongoing injuries and complications after the immediate resuscitation and damage control surgery. This discussion provides an update on the management of coagulopathy, respiratory failure, and prevention and treatment of venous thromboembolism in polytrauma victims. SpeciÀ cally, the role of thromboelastography, ratios of blood product transfusions, newer modes of respiratory support including airway pressure release ventilation and veno-venous extracorporeal membrane oxygenation, and the use of mechanical and chemoprophylaxis for prophylaxis and management of venous thromboembolism are addressed. Weinacker A, MD email: AWeinacker@stanfordmed.org I Weinacker A, MD email: AWeinacker@stanfordmed.org Inpatient and intensive care of polytrauma patients requires careful attention to detail and frequent reassessment after transfer from the emergency department or operating room. Immediate concerns include evaluation and management of shock, control of bleeding, support of the respiratory system, prevention of complications, and correction of coagulopathy that may contribute to morbidity and mortality. A complete review of the management of polytrauma patients is beyond the scope of this paper, and a number of good reviews have been written to address this subject.1,2 The goal of this discussion therefore, will be to provide an update on the management of coagulopathy, respiratory failure, and prevention and treatment of venous thromboembolism in polytrauma victims. Ann Weinacker, MD1 Ann Weinacker, MD1 Keywords: inpatient management of polytrauma patients, intensive care of polytrauma patients Initial management of shock in patients who have suffered polytrauma is aimed at damage control resuscitation with trans- fusions of blood products and Á uids as the mainstay of therapy, and should be guided by repeated physical and laboratory exami- nations.2 The goal of damage control resuscitation is to restore systolic blood pressure only to 80-90 mmHg (100 mmHg in patients with traumatic brain injury) to minimize bleeding from relative hypertension.3 In general, intravascular volume should be repleted with blood products rather than isotonic Á uids, and component therapy should be used to correct coagulopathy.1 It is also critical to recognize that ongoing bleeding may be occult. 1 Professor of Medicine, Chief of Staff, Associate Director of the Intensive Care Units,Stanford University Hospital Many patients who are initially stabilized in the emergency department or operating room require delayed deÀ nitive treatment of other known injuries only when their conditions allow. Special Feature Special Feature 104 The Bangkok Medical Journal Vol. 6; September 2013 ISSN 2287-0237 (online)/ 2287-9674 (print) Keywords: inpatient management of polytrauma patients, intensive care of polytrauma patients 1 Professor of Medicine, Chief of Staff, Associate Director of the Intensive Care Units,Stanford University Hospital Coagulopathy In the absence of thromboelastography, treatment of coagulopathy with blood products and clotting factors based on clinical data is the norm. In addition, tranexamic acid has been shown to be associated with decreased mor- tality when given within 3 hours of injury.8 Tranexamic acid is a synthetic lysine derivative that inhibits lysine binding sites on plasminogen, blocking conversion of plas- minogen to plasmin. It also inhibits the proteolytic action of plasmin on À brin clot and platelet receptors and is an effective antiÀ brinolytic. Coagulopathy often contributes to ongoing blood loss and shock from both missed and recognized injuries, and may be extremely difÀ cult to manage. The damage control model of resuscitation relies on empiric ratios of transfusion products and includes goal directed hemo- static resuscitation with serial measurements of gener- ally accepted measures of coagulation. Prothrombin time (PT), partial thromboplastin time (PTT), international normalized ratio, À brinogen levels and platelet counts guide therapy but are grossly inadequate measures of coagulopathy. Thus, the ideal transfusion ratio of red cells to plasma to platelets remains unclear, and is being addressed by the Pragmatic Randomized Optimum Platelet and Plasma Ratio (PROPPR) trial, an ongoing multicenter, prospective, randomized study that will evaluate different blood product ratios to be administered to trauma patients in need of > 10 U of PRBCs in the À rst 24 hours (http://ccctsp.sph.uth.tmc.edu/proppr_trial/Site- Pages/Home.aspx, accessed March 25, 2013). Recombinant activated factor VII is another effective procoagulant therapy. It acts locally at the site of tissue injury by binding to exposed tissue factor at the site of tissue injury. It increases thrombin generation and generates a tight À brin hemostatic plug, thereby decreasing blood loss.9 In spite of this, however, it has not been shown to alter mortality in trauma victims and it is extremely expensive. Likewise, prothrombin complex concentrate containing factors II, VII, IX, and X is also capable of decreasing blood loss and may be useful in trauma,9 but it is also expensive and more studies are needed to determine its role in the treatment of polytrauma victims. Update of Inpatient Management of Polytrauma Patients It is important to avoid pitfalls in the care of trauma victims, including under-estimated bleeding from pelvic fractures, development of compartment syndrome, and missed injuries. Missed injuries are common in polytrauma patients, and have been reported in 1.5-39% of patients. Although most missed injuries are not life threatening, as many as 6.5% of trauma-related deaths are due to undiagnosed injuries. Most missed injuries involve the extremities, and most *Address Correspondence to author: E-mail: AWeinacker@stanfordmed.org Update of Inpatient Management of Polytrauma Patients transfusion of blood products and may improve outcomes in critically injured polytrauma victims.4 missed injuries are associated with an injury severity score 16 and a Glasgow coma scale < 8. Coagulopathy Coagulopathy in polytrauma victims may be caused by a variety of factors, including acidosis, hemodilution, hypothermia, traumatic brain injury and the release of brain thromboplastin, disseminated intravascular coagu- lation, and the “acute traumatic coagulopathy” charac- terized by the activation of the thrombomodulin-Protein C system.4,5 Routine tests of coagulation may be inadequate to accurately assess speciÀ c deÀ ciencies in clotting mech- anisms, and thromboelastography may be useful to guide goal-directed treatment of life-threatening coagulopathy with speciÀ c coagulation factors. Although thromboelas- tography was À rst described in 1948, it is not widely avail- able, but it has recently been recognized as a useful tool to manage bleeding disorders from a variety of causes. 105 The Bangkok Medical Journal Vol. 6; September 2013 ISSN 2287-0237 (online)/ 2287-9674 (print) Respiratory failure The goals of respiratory support are to ensure adequate oxygenation and ventilation and to prevent ventilator associated lung injury. Positive end expiratory pressure should be used with caution as high levels may impair venous return, especially in patients with inadequately restored intravascular volume, and the effects of lung or chest injury must be taken into consideration. Transfusion associated lung injury or circulatory overload may also affect the need for and manner of providing respiratory support. Veno-venous extracorporeal membrane oxygenation (ECMO) has been increasingly used in trauma patients with respiratory failure, although there have been no large-scale trials to substantiate its beneÀ t. Accepted indications for ECMO include severe hypoxemia or respiratory acidosis in spite of maximal conventional support. Relative contraindications include irreversible lung injury, contraindications to anticoagulation, and respiratory failure requiring mechanical ventilation for longer than a week. In contrast to venoarterial ECMO, veno-venous ECMO provides respiratory but not hemodynamic support. In veno-venous ECMO, blood is extracted from the vena cava or right atrium and is returned to the right atrium. Cannulae are usually placed into the right common femoral vein (for drainage) and right internal jugular vein (for infusion), although double lumen cannulae are now available for placement in the internal jugular vein. The advantages of the use of a double lumen internal jugular cannula include the necessity for only a single venipuncture site, and increased mobility of the patient during recovery. A retrospective study showed that 20 of 28 consecutive trauma patients referred for ECMO survived, suggesting a possible role of ECMO in trauma patients with acute respiratory failure. (Cordell-Smith, Injury 2006) Small case reports have also suggested ECMO as a viable treatment option for respiratory failure in polytrauma victims.12,13 More studies are needed, however, before recommendations can be made regarding the use of ECMO in this population. The mainstay of respiratory support is lung protective ventilation with low (6 ml/kg predicted body weight) tidal volumes and limitation of airway plateau pressures ( 30 cm H2O pressure). Respiratory failure Respiratory failure is common in severely injured patients, and results not only from pulmonary contusions or other trauma to the lungs or chest, but may also be caused by the development of the acute respiratory distress syndrome, transfusion related acute lung injury, or transfusion associated circulatory overload (Figure 1). Figure 1: Chest supine reveals bilateral patchy inÀ ltration. Thromboelastography is a functional assay that assesses the viscoelastic properties of clot formation and synthe- sizes information obtained from the PT, PTT, thrombin time, À brinogen level and platelet counts. It then provides information regarding clot formation, clot strength, and À brinolysis and may be used to guide therapy as well as to monitor the progress or resolution of an existing coagu- lopathy after trauma.6 Rapid thromboelastography provides data in real time7 and is particularly useful in the operating room and intensive care units where ongoing life-threatening bleeding may occur. The data generated can then be used to direct which blood products are needed to treat ongoing hemorrhage - platelets, plasma, cryoprecipitate, desmopressin, aminocaproic acid, or speciÀ c clotting factors. Use of thromboelastography has been shown to decrease the Figure 1: Chest supine reveals bilateral patchy inÀ ltration. Weinacker A Figure 2: CT of chest axial section reveals bilateral conÁ uent ground-glass appearance. P-low, and is dependent in part on the duration (T-low) of the reduced pressure. Patients can be ventilated with APRV whether they are ventilating spontaneously or not, and weaning is achieved by incrementally reducing T-high and P-high, and by increasing T-low and P-low. (For tutorials on the use of APRV and other ventilation strategies, see www.ccmtutorials.com/rs/mv.) Purported advantages of APRV include improved alveolar recruit- ment, especially in dependent lung zones, decreased peak airway pressures, and minimal hemodynamic effects. In a small study of trauma patients with acute lung injury, those patients ventilated with APRV required less sedation and neuromuscular blockade, less time on mechanical ventilation, and less time in the intensive care unit.11 In spite of these advantages however, mortality was not improved in this study. APRV must be used with caution in patients with signiÀ cant airÁ ow obstruction, as signiÀ cant pulmonary hyperinÁ ation may occur, potentially resulting in impaired venous return and decreased cardiac function. Figure 2: CT of chest axial section reveals bilateral conÁ uent ground-glass appearance. Respiratory failure In patients with the acute respiratory distress syndrome this remains the single therapy that has been demonstrated to reduce mortality.10 Newer modes of ventilatory support including airway pressure release ven- tilation (APRV), high frequency oscillatory ventilation, and veno-venous extracorporeal membrane oxygenation (ECMO) have been used in these patients with increasing success, and further studies are needed to determine whether these modes of support will ultimately improve outcomes in trauma patients with respiratory failure. Noninvasive modes of ventilation may also be used to avoid intubation and mechanical ventilation in select, less severely injured patients, and may be useful in some spinal cord injured patients. Because of increasing interest in APRV and ECMO, these modalities will be discussed in more detail. Airway pressure release ventilation is a form of inverse ratio ventilation whereby a sustained level of increased positive airway pressure (P-high) is interrupted peri- odically and very brieÁ y to a much lower level (P-low). The sustained positive pressure is intended to improve alveolar recruitment and thus improve oxygenation, which is dependent on both the amount of pressure delivered at P-high, and the amount of time (T-high) spent at P-high. Carbon dioxide elimination is achieved during 106 The Bangkok Medical Journal Vol. 6; September 2013 ISSN 2287-0237 (online)/ 2287-9674 (print) Update of Inpatient Management of Polytrauma Patients Polytrauma victims are at increased risk for PE because of immobilization due to spinal cord injury, long bone or pelvic fractures, and major surgery. Vascular injuries and the increasing use of central venous catheters also contribute to the risk. patients who are not at high risk for major bleeding, or mechanical prophylaxis with intermittent pneumatic compression devices.14 Patients at higher risk (approxi- mately 6%) for thromboembolism should receive a combination of low molecular weight or unfractionated heparin and mechanical prophylaxis. Recommendations for orthopedic surgery patients are similar but also include the possible use of anti-Factor Xia drugs, direct thrombin inhibitors, or other anticoagulant therapy.15 Although the use of vena cava À lters has become more widespread, particularly since the advent of removable À lters, these devices are not recommended as primary prevention. The hemodynamic consequences of PE include right ventricular strain or even failure from vascular obstruc- tion and vasoconstriction. In patients with massive PE, right ventricular failure and circulatory collapse is the primary cause of death. The pulmonary consequences of PE include inequality in the ratio of ventilation to perfusion as blood Á ow is redistributed. Blood Á ow is decreased distal to the embolic obstruction and normal regions of the lungs are subsequently over-perfused. Alveolar dead space increases and mixed venous oxygen saturation decreases. Atelectasis may occur distal to the embolic obstruction as a result of surfactant loss and alveolar hemorrhage, thereby worsening hypoxemia. Diagnosis of symptomatic pulmonary embolism is typically with CT angiogram (Figure 3), although conventional pulmonary angiogram and ventilation/perfusion scans are done preferentially in some centers. In patients with contraindications to radiocontrast agents, a presumptive diagnosis of PE can be made by demonstrating deep venous thrombosis using ultrasound, although there is a high false negative rate using this technique.16 To prevent this complication, prophylaxis must be initiated as soon as possible, ideally with a combination of mechanical and chemoprophylaxis. Ongoing bleeding, severe head trauma, or lower extremity fractures may preclude the use of anticoagulants or sequential compres- sion devices however. Treatment of acute PE or deep venous thrombosis is typically with parenteral anticoagulation, and massive PE resulting in hemodynamic compromise requires thrombo- lytic therapy unless contraindicated by high risk or recent injury.17 Patients with small incidentally-discovered PE may not always require therapy. Update of Inpatient Management of Polytrauma Patients Recommendations from the American College of Chest Physicians for prophylaxis in nonorthopedic surgical patients at moderate (approximately 3%) risk for deep venous thrombosis or PE include low molecular weight heparin or low dose unfractionated heparin in In a prospective study of 90 moderately to severely injured trauma victims with an Injury Severity Score of 9, 22 (24%) were found to have asymptomatic PE detected by CT scanning.18 Of these,17 patients had only minor clot burden (segmental and/or subsegmental clots only) and were not treated for thromboembolism. There were no deaths in the untreated group. Figure 3: CT pulmonary angiogram reveals small pulmonary embolism at descending branches of pulmonary arteries. 107 The Bangkok Medical Journal Vol. 6; September 2013 ISSN 2287-0237 (online)/ 2287-9674 (print) Pulmonary embolism Concomitant with ongoing evaluation and deÀ nitive treatment of injuries in polytrauma victims, avoidance of complications is crucial. Pulmonary embolism (PE) is common and may range in severity from trivial and incidentally discovered to massive and ultimately fatal. Update of Inpatient Management of Polytrauma Patients References 1. Shere-Wolfe RF, Galvagno SM Jr, Grissom TE. Critical care considerations in the management of the trauma patient following initial resuscitation. Scand J Trauma Resusc Emerg Med 2012;20:68. 11. Putensen C, Zech S, Wrigge H, et al. Long-term effects of spontaneous breathing during ventilatory support in patients with acute lung injury. Am J Respir Crit Care Med 2001;164:43-9. 2. Tobin JM, Varon AJ. Review article: update in trauma anesthesiology: perioperative resuscitation management. Anesth Analg 2012;115:1326-33. 12. Firstenberg MS, Nelson K, Abel E, et al. Extracorporeal membrane oxygenation for complex multiorgan system trauma. Case Rep Surg 2012;2012:897184. 3. Duchesne JC, McSwain NE Jr, Cotton BA et al. Damage control resuscitation: the new face of damage control. J Trauma 2010;69:976-90. 13. Madershahian N, Wittwer T, Strauch J, et al. Application of ECMO in multitrauma patients with ARDS as rescue therapy. J Card Surg 2007;22:180-4. 4. Brohi K, Singh J, Heron M, et al. Acute traumatic coagu- lopathy. J Trauma 2003;54:1127-30. 14. Gould MK, Garcia DA, Wren SM, et al. Prevention of VTE in nonorthopedic surgical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:S227-77. 5. MacLeod JB, Lynn M, McKenney MG, et al. Early coagu- lopathy predicts mortality in trauma. J Trauma 2003;55: 39-44. 6. Kashuk JL, Moore EE, Sawyer M, et al. Postinjury coagulopathy management: goal directed resuscitation via POC thrombelastography. Ann Surg 2010;251:604-14. 15. Falck-Ytter Y, Francis CW, Johanson NA, et al. Preven- tion of VTE in orthopedic surgery patients: Antithrom- botic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:S278-325. 7. Cotton BA, Faz G, Hatch QM, et al. Rapid thrombelastog- raphy delivers real-time results that predict transfusion within 1 hour of admission. J Trauma 2011;71:407-14. Clinical Practice Guidelines. Chest 2012;141:S278-325. 16. Bates SM, Jaeschke R, Stevens SM, et al. Diagnosis of DVT: Antithrombotic Therapy and Prevention of Throm- bosis, 9th ed: American College of Chest Physicians Evi dence-Based Clinical Practice Guidelines. Chest 2012; 14:S351-418. 8. CRASH-2 trial collaborators, Shakur H, Roberts I, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010;376:23-32. 17. Kearon C, Akl EA, Comerota AJ, et al. 108 The Bangkok Medical Journal Vol. 6; September 2013 ISSN 2287-0237 (online)/ 2287-9674 (print) Conclusion In patients who have suffered polytrauma, resuscitation continues after initial damage control resuscitation and surgery. Circulatory support is paramount and is typically best accomplished with blood product transfusions rather than crystalloids, but the proper ratio of blood products for transfusion remains unclear. Coagulopathy is common in polytrauma victims and has many causes including acidosis, hemodilution, hypothermia, traumatic brain injury, disseminated intravascular coagulation, and activation of the thrombomodulin-Protein C system. Correction of coagulopathy should be guided by laboratory studies and by ongoing clinical assessment, and may be aided by the use of thromboelastography where available. Respiratory failure is also common, particularly in patients with chest trauma, and is best treated with lung protective ventilation using low tidal volumes and limited airway plateau pressures, but other modes such as APRV Figure 3: CT pulmonary angiogram reveals small pulmonary embolism at descending branches of pulmonary arteries. 107 The Bangkok Medical Journal Vol. 6; September 2013 ISSN 2287-0237 (online)/ 2287-9674 (print) 107 Weinacker A the extent and nature of injuries may limit prophylactic options. Appropriate therapy for PE is typically with systemic anticoagulation, but depends on the clinical situation. or even ECMO may have a role. Pulmonary embolism is a potentially fatal complication of polytrauma, so prevention is paramount. Although prophylaxis is ideally achieved with a combination of mechanical and chemoprophylaxis, References Antithrombotic therapy for VTE disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:S419-94. 9. Fries D. The early use of À brinogen, prothrombin complex concentrate, and recombinant-activated factor VIIa in massive bleeding. Transfusion 2013;53:S91-5. 10. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000;342: 1301-8. 18. Schultz DJ, Brasel KJ, Washington L, et al. Incidence of asymptomatic pulmonary embolism in moderately to severely injured trauma patients. J Trauma 2004;56:727-31.
https://openalex.org/W3153199332	https://bmcpsychiatry.biomedcentral.com/track/pdf/10.1186/s12888-021-03201-6	English	null	Aripiprazole in the real-world treatment for irritability associated with autism spectrum disorder in children and adolescents in Japan: 52-week post-marketing surveillance	BMC psychiatry	2,021	cc-by	9,861	Abstract Background: The purpose of this study was to evaluate the post-marketing safety and effectiveness of aripiprazole in treating irritability in pediatric patients (6–17 years) with autism spectrum disorder (ASD) in actual clinical sites of Japan. Methods: In this post-marketing surveillance, patients were enrolled into the multicenter, prospective, non- interventional, observational study for 52 weeks, and were dosed with aripiprazole (1–15 mg/day) under daily clinical settings in Japan. Results: In 510 patients, the continuation rate of aripiprazole treatment was 84.6% at day 168 (week 24) and 78.1% at day 364 (week 52). Adverse drug reactions (ADRs) occurred in 22.7% of patients (n = 116), and the most common ADRs were somnolence (9.4%), followed by weight increased (3.3%). At week 4, the mean change from baseline in the irritability subscale score for the Aberrant Behavior Checklist Japanese version (ABC-J) was −5.7 ± 6.8 (n = 288). Based on multiple regression analysis, comorbid attention deficit and hyperactivity did not affect the ABC-J irritability subscale score at endpoint. At week 24, the mean change from baseline for the Strengths and Difficulties Questionnaire was −3.3 ± 4.9 (n = 215) for the total difficulties score and 0.6 ± 1.7 (n = 217) for the prosocial behavior subscale score. Results: In 510 patients, the continuation rate of aripiprazole treatment was 84.6% at day 168 (week 24) and 78.1% at day 364 (week 52). Adverse drug reactions (ADRs) occurred in 22.7% of patients (n = 116), and the most common ADRs were somnolence (9.4%), followed by weight increased (3.3%). At week 4, the mean change from baseline in the irritability subscale score for the Aberrant Behavior Checklist Japanese version (ABC-J) was −5.7 ± 6.8 (n = 288). Based on multiple regression analysis, comorbid attention deficit and hyperactivity did not affect the ABC-J irritability subscale score at endpoint. At week 24, the mean change from baseline for the Strengths and Difficulties Questionnaire was −3 3 ± 4 9 (n = 215) for the total difficulties score and 0 6 ± 1 7 (n = 217) for the prosocial p g y yp y irritability subscale score at endpoint. At week 24, the mean change from baseline for the Strengths and Difficulties Questionnaire was −3.3 ± 4.9 (n = 215) for the total difficulties score and 0.6 ± 1.7 (n = 217) for the prosocial behavior subscale score. © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Sugimoto et al. BMC Psychiatry (2021) 21:204 https://doi.org/10.1186/s12888-021-03201-6 Sugimoto et al. BMC Psychiatry (2021) 21:204 https://doi.org/10.1186/s12888-021-03201-6 Open Access Aripiprazole in the real-world treatment for irritability associated with autism spectrum disorder in children and adolescents in Japan: 52-week post-marketing surveillance Aripiprazole in the real-world treatment for irritability associated with autism spectrum disorder in children and adolescents in Japan: 52-week post-marketing surveillance Yuna Sugimoto1, Kayo Yamamura1, Tomoyo Takayama1, Yasuhiko Fukuta1, Kazuo Aoki2, Katsunaka Mikami3 and Akemi Tomoda4 Abstract Conclusions: Aripiprazole was well tolerated and effective in the long-term treatment of irritability associated with ASD in Japanese pediatric patients in the real-world clinical practice. Trial registration: This surveillance was registered with Clinical Trial.gov (no. NCT03179787) on June 7, 2017 (retrospectively registered). Trial registration: This surveillance was registered with Clinical Trial.gov (no. NCT03179787) on June 7, 2017 (retrospectively registered). Keywords: Aripiprazole, Autism spectrum disorder, Children and adolescents, Irritability, Post-marketing surveillance Keywords: Aripiprazole, Autism spectrum disorder, Children and adolescents, Irritability, Post-marketing surveillance © The Author(s). 2021 Open Access This article is licensed under a which permits use, sharing, adaptation, distribution and reproductio appropriate credit to the original author(s) and the source, provide changes were made. The images or other third party material in thi licence, unless indicated otherwise in a credit line to the material. If licence and your intended use is not permitted by statutory regulat permission directly from the copyright holder. To view a copy of th The Creative Commons Public Domain Dedication waiver (http://cre data made available in this article, unless otherwise stated in a cred Correspondence: Sugimoto.Yuna@otsuka.jp 1Pharmacovigilance Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan Full list of author information is available at the end of the article * Correspondence: Sugimoto.Yuna@otsuka.jp p g @ jp 1Pharmacovigilance Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan Full list of author information is available at the end of the article Japan Full list of author information is available at the end of the article Correspondence: Sugimoto.Yuna@otsuka.jp 1Pharmacovigilance Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Methods Patients Patients newly treated with aripiprazole for irritability associated with ASD in children and adolescents (≥6 years and < 18 years) were included in the surveillance. The target sample size was calculated to be 300 patients on the assumption of detecting adverse drug reactions (ADRs) occurring at a frequency of 1% with a confidence of at least 95%. Although there are no approved pharmacologic treat- ments that target core deficits of ASD, associated sec- ondary symptoms such as irritability may be ameliorated by a combination of behavioral and pharmacologic ap- proaches, including the use of atypical antipsychotics [3]. Aripiprazole is an atypical antipsychotic drug devel- oped by Otsuka Pharmaceutical Co., Ltd. that is charac- terized by partial agonism at dopamine D2 receptors and serotonin 5-HT1A receptors and antagonism at 5-HT2A receptors [4, 5]. In the pediatric field, the indications for adolescent schizophrenia (13 to 17 years), pediatric bipo- lar mania (10 to 17 years), irritability associated with aut- istic disorder (6 to 17 years), and Tourette’s disorder (6 to 18 years) in the United States, and schizophrenia in adolescents aged 15 years and older in Europe have been approved [6, 7]. Aripiprazole may have a more favorable side-effect profile than other antipsychotics in child and adolescent patients with mental health disorder [8] be- cause of its unique mechanism of action. Although there are no approved pharmacologic treat- ments that target core deficits of ASD, associated sec- ondary symptoms such as irritability may be ameliorated by a combination of behavioral and pharmacologic ap- proaches, including the use of atypical antipsychotics [3]. Background clinical study. Also, the criteria for patient selection and concomitant medications are strictly specified in ran- domized controlled clinical trials and do not necessarily reflect the actual clinical treatment environment. There- fore, this post-marketing surveillance was conducted to assess the real-world safety and effectiveness of aripipra- zole for children and adolescents with ASD-related irrit- ability in Japan. g Autism spectrum disorder (ASD) is characterized by persistent impairment in reciprocal social communica- tion and social interaction, and restricted, repetitive pat- terns of behavior, interests, or activities. These symptoms are present from early childhood and limit or impair everyday functioning. The impairments in com- munication and social interaction are pervasive and sus- tained [1]. Understandably, these symptoms can have a substantial impact on the individuals and their families. This impact can be further increased by the presence of associated behaviors such as irritability, which may manifest as tantrums, aggressiveness, self-injurious be- haviors, and sudden mood changes, all of which can have a significant impact on education and social devel- opment [2]. Assessment points for this surveillance were selected based on prior clinical trials in Japan [9–11] and the United States [13–16]. © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 12 Sugimoto et al. BMC Psychiatry (2021) 21:204 Sugimoto et al. BMC Psychiatry (2021) 21:204 Page 2 of 12 Page 2 of 12 Study design Thi ill This surveillance was conducted as a multicenter, pro- spective, non-interventional, observational study for 52 weeks (1 year). The registry and case reports were encoded by the physicians through Electronic Data Cap- ture system during the period from April 2017 to Sep- tember 2019. This surveillance was conducted in compliance with the Ministerial Ordinance on Standards for Conducting Post-marketing Surveys and Studies on Drugs; MHLW Ordinance No. 171 issued on December 20, 2004 (GPSP). As the surveillance is a non-interventional study in accordance with GPSP, the need for ethics approval and consent were waived. The surveillance was designed by Otsuka, reviewed by the Japanese Pharmaceutical and Medical Devices Agency (PMDA), and also registered at Clinical Trials.gov (identifier: NCT03179787). The indication for irritability associated with autistic disorder (6 to 17 years) with oral aripiprazole has been approved in the United States since 2009 and in 5 coun- tries thereafter, and the efficacy was also confirmed in clinical trials in children and adolescents with irritability associated with ASD in Japan [9–11]. Based on the trials, the indication for irritability associated with ASD in chil- dren and adolescents has been approved since Septem- ber 2016 in Japan. Aripiprazole was administered orally once daily with a starting dose of 1 mg daily and a maintenance dose of 1–15 mg daily according to the package insert in Japan [17]. The dose could be adjusted according to the sever- ity of patient’s symptoms, however the dose increase per day was to be 3 mg or less and the daily dose was not to exceed 15 mg [17]. The clinical studies that started before 2013 included the patients diagnosed with autistic disorder based on the diagnostic criteria of Diagnostic and Statistical Man- ual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) [12]. Before starting this surveillance, DSM-IV-TR had been updated to DSM-5 [1], and the indication was changed to ASD. Since diagnosis with co- morbid ASD and attention deficit and hyperactivity dis- order (ADHD) was recognized, the scope of applicable patients in this surveillance has been broader than in the Patient demographics Patients’ gender, age, body height, weight, severity of symptoms, duration of illness, comorbidities, medical history, concomitant medications, etc. were recorded be- fore aripiprazole administration. Page 3 of 12 Page 3 of 12 Sugimoto et al. BMC Psychiatry (2021) 21:204 Sugimoto et al. BMC Psychiatry (2021) 21:204 Effectiveness Using caregiver-rated Aberrant Behavior Checklist- Japanese version (ABC-J), the degree of aberrant behav- iors was scored on four levels ranging from 0 ‘not at all a problem’ to 3 ‘the problem is severe in degree’, totaling 58 items. These items were classified into five subscales: irritability (15 items), lethargy/social withdrawal (16 items), stereotypy (7 items), hyperactivity (16 items), and inappropriate speech (4 items), and scores were calcu- lated [18]. The Japanese version of the ICH Medical Dictionary for Regulatory Activities (MedDRA/J version 22.1) was used in coding and classifying adverse events. Adverse events for which a causal relationship to aripiprazole could not be ruled out were considered as ADRs. All statistical analyses were performed using SAS software (version 9.3; SAS Institute Inc.). Statistical analysis l Results were summarized using descriptive statistics. Continuous variables were described as mean ± standard deviation (SD). The LOCF method was used with imput- ation of the latest observed values. An applicable paired t-test or Wilcoxon signed-rank sum test was applied to compare the pre-dose and post-dose effect. The level of statistical significance was set at two-sided 5% and the confidence interval at two-sided 95%. The adverse events (if any occurred), onset date, serious- ness, outcome, date of outcome, causality to aripiprazole, other possible causal factors, and treatment for adverse events during the observation period were recorded. Ser- iousness was determined by the physicians. Body weight and height were also recorded. The shift in the percentile body weight category from baseline to end-point was calculated using predefined body weight percentile categories. The treatment continuation rate was analyzed by Kaplan-Meier method, with the reason for discontinu- ation of aripiprazole censored as “transfer,”, “lost to follow-up,” or “improved symptoms”. Results Physician-rated Clinical Global Impression- Improvement (CGI-I) and Severity of illness (CGI-S) scales quantified the physician’s impression at baseline (CGI-S only), week 4, week 8, week 16, week 24, and week 52, or at the time of discontinuation of aripiprazole administration. The CGI-I scale was scored from 1 ‘very much improved’ to 7 ‘very much worse’, and CGI-S was scored from 1 ‘normal’ to 7 ‘very much severely ill’. Aripiprazole dosing duration by age group was predefined using modified method reported by Hirshkowitz M et al. [21], and the shift in sleep duration from baseline to end-point (last observation carried forward; LOCF) was calculated. The daily dose of aripiprazole, administration period (start date and end date), and reasons for discontinu- ation (if discontinued) were recorded. Patient disposition Five hundred and twenty-eight patients were enrolled at 100 sites across Japan, and 526 case reports were col- lected during the period of April 2017 to September 2019. The safety analysis population consisted of 510 pa- tients. Among excluded patients, 15 patients were lost to follow-up, and one patient was not treated with aripipra- zole. After excluding 21 patients with the protocol devia- tions (including all effectiveness measures were not assessed after treatment: n = 11, all effectiveness mea- sures were assessed over 7 days after last dose: n = 7, the daily dose was > 15 mg: n = 2, all effectiveness were assessed after day 393: n = 1), the effectiveness analysis population consisted of 489 patients (Fig. 1). y y Caregiver-rated Strengths and Difficulties Question- naire for Children (SDQ) was used to score the adapta- tion and mental health conditions on three levels: 0 ‘not true’, 1 ‘somewhat true’, and 2 ‘certainly true’, with a total of 25 items. These items consist of five subscales: conduct problems (5 items), hyperactivity (5 items), emotional problems (5 items), peer problems (5 items), and prosocial behavior (5 items), each yielding scores be- tween 0 and 10. The Total Difficulties Score ranging from 0 to 40 was derived from the sum of the four sub- scale scores, excluding prosocial behavior [19]. Banding of raw scores obtained with the five SDQ subscales and the Total Difficulties Score, the scale properties were assessed as ‘Low Need’, ‘Some Need’, or ‘High Need’, ap- plying recommended banding of raw scores by Matsuishi T et al. [20] both in ≤13 years and in > 13 years old. Patient demographics Baseline patient characteristics are summarized in Table 1. Among 510 patients included in the safety ana- lysis, the mean age of the patient population at baseline was 10.4 ± 3.1 years, and the majority of the patients were males (73.7%) and aged 6 years or older but youn- ger than 13 years (71.8%). Major comorbidities (≥1%) were ADHD in 53.1% (n = 271), sleep disorders in 17.6% (n = 90), learning disorders (LD) in 11.2% (n = 57), and tic disorders in 6.9% (n = 35). There were 24.1% (n = 123) of patients with intellectual disabilities. In addition, the average sleep time duration in the last 4 weeks, which was reported by patients or caregivers and not actigraphy-measured data, was recorded at week 24, week 52, and/or end of study. Recommended sleep Sugimoto et al. BMC Psychiatry (2021) 21:204 Page 4 of 12 Sugimoto et al. BMC Psychiatry Fig. 1 Patient disposition Fig. 1 Patient disposition (32.5 days [1–383 days]), increased appetite (120.5 days [29–179 days]), weight increased (168.0 days [1–320 days]), and obesity (192.0 days [76–365 days]). The most common concomitant medications (drugs for ASD irritability, ≥1%) were antidepressants in 7.1% (n = 36), risperidone in 5.1% (n = 26), and ADHD medi- cations in 1.2% (n = 6). Serious ADRs were epilepsy, partial seizures, and renal impairment (0.2% each). On the physician-rated CGI-S scale, most patients were assessed as being moderately ill (55.3%, n = 282), followed by markedly ill (29.4%, n = 150). Most common ADRs leading to discontinuation with an incidence of > 0.5% were somnolence (2.0%), weight increased (0.8%), and insomnia (0.6%). Safety In 510 patients included in the safety analysis, adverse events occurred in 24.5% (n = 125) and ADRs based on physician-assessed causality occurred in 22.7% (n = 116) (Table 3). The most common ADRs with an incidence of ≥1% were somnolence (9.4%, n = 48), followed by weight increased (3.3%, n = 17), nausea (1.4%, n = 7), in- creased appetite and headache (1.2% each, n = 6), and obesity (1.0%, n = 5). Fig. 2 shows the duration to onset date of ADRs from the first dose, and the summary sta- tistics (median [min-max]) were headache (3.5 days [1– 284 days]), nausea (8.0 days [2–82 days]), somnolence Aripiprazole dosing The most common ADRs (≥0.5%) associated with extrapyramidal symptoms were tic disorders and akathi- sia (0.8% each), both of which were non-serious. The mean duration of aripiprazole treatment was 290.6 ± 151.5 days and the mean daily dose was 2.20 ± 1.85 mg. Weight increase-related ADRs occurred in 4.3% of pa- tients (n = 22), including weight increased (3.3%) and obesity (1.0%), all of which were non-serious. Among these ADRs, one event led to treatment discontinuation (weight increased, 0.6%). A total of 201 patients (39.4% of the safety population) discontinued aripiprazole treatment during the surveil- lance period. Of the 201 patients who discontinued treatment, the most common reason for discontinuation (multiple responses were allowed) was ‘request for dis- continuation from the patient or family’ in 33.8% (n = 68) (Table 2) and the mean duration of treatment prior to discontinuation was 133.6 ± 113.4 days. The mean percentile of weight was 48.36% ± 31.86% at baseline, 52.35% ± 32.86% at end-point (LOCF), and the mean change from baseline to end-point was 3.99% ± 11.65%. Percentile body weight categorical shift data are presented in Table 4. All patients with a weight in the > 90th percentile at end-point (LOCF) were in the > 75th percentile at baseline, and all patients with a weight in the ≤10th percentile at end-point (LOCF) were in the ≤25th percentile at baseline. There were no patients in which the percentiles of weight differed greatly between before and after aripiprazole treatment. The continuation rate of aripiprazole treatment was 84.6% at day 168 (week 24) and 78.1% at day 364 (week 52) from the first dose (see Additional file 1). Effectiveness The mean SDQ prosocial b subscale score was 3.5 ± 2.5 (n = 316) at baselin Table 1 Baseline demographics and clinical characteristics (n = 510) Value, n (%) Gender Male 376 (73.7) Female 134 (26.3) Age, years 6 to 12 years 366 (71.8) 13 to 17 years 144 (28.2) Mean ± SD 10.4 ± 3.1 Median (min to max) 10.0 (6 to 17) Treatment Category Hospitalization 22 (4.3) Outpatient 488 (95.7) Duration of ASD, years < 1 year 216 (42.4) ≥1 to < 2 years 46 (9.0) ≥2 to < 3 years 25 (4.9) ≥3 years 86 (16.9) Unknown 137 (26.9) Mean ± SD (n = 373) 1.70 ± 2.51 Median (min to max) 0.50 (0 to 15.3) Intellectual disability None 387 (75.9) Total 123 (24.1) Mild 68 (13.3) Moderate 30 (5.9) Severe 19 (3.7) Most severe 6 (1.2) Comorbidities None 144 (28.2) Total 366 (71.8) ADHDa 271 (53.1) LDa 57 (11.2) Tic disordersa 35 (6.9) Sleep disordersa 90 (17.6) Concomitants for ASD None 461 (90.4) Total 49 (9.6) Risperidonea 26 (5.1) Antidepressantsa 36 (7.1) ADHD drugsa 6 (1.2) Concomitants for other condition None 257 (50.4) Total 253 (49.6) Methylphenidate hydrochlorideab 86 (16.9) Table 1 Baseline demographics and clinical characteristics (n (Continued) Value Atomoxetine hydrochlorideab 42 (8. Risperidoneab 30 (5. Ramelteonab 17 (3. Sodium valproateab 12 (2. Yokukansanab 12 (2. CGI-S 1: Normal 1 (0.2 2: Minimally ill 1 (0.2 3: Mildly ill 42 (8. 4: Moderately ill 282 (5 5: Makedly ill 150 (2 6: Severely ill 30 (5. 7: Very severely ill 4 (0.8 Mean ± SD 4.3 ± Median (min to max) 4.0 (1 a: Multiple responses were allowed in a patient b: Drugs using in ≥2% of Patiants Gender Male Female Age, years 6 to 12 years 13 to 17 years Mean ± SD Median (min to max) Treatment Category Hospitalization Outpatient Duration of ASD, years < 1 year ≥1 to < 2 years ≥2 to < 3 years ≥3 years Unknown Mean ± SD (n = 373) Median (min to max) Intellectual disability None Total Mild Moderate Severe Most severe Comorbidities None Total ADHDa LDa Tic disordersa Sleep disordersa Concomitants for ASD None Total Risperidonea Antidepressantsa ADHD drugsa Concomitants for other condition None Total Methylphenidate hydrochlorideab Guanfacine hydrochlorideab 6.8 ± 8.3 at end-point (LOCF), with significant reduc- tions in scores compared with baseline at each assess- ment point (p < 0.0001). Effectiveness week 24, 0.8 ± 2.1 (n = 200) at week 52, ristics (n = 510) Value, n (%) 376 (73.7) 134 (26.3) 366 (71.8) 144 (28.2) 10.4 ± 3.1 10.0 (6 to 17) 22 (4.3) 488 (95.7) 216 (42.4) 46 (9.0) 25 (4.9) 86 (16.9) 137 (26.9) 1.70 ± 2.51 0.50 (0 to 15.3) 387 (75.9) 123 (24.1) 68 (13.3) 30 (5.9) 19 (3.7) 6 (1.2) 144 (28.2) 366 (71.8) 271 (53.1) 57 (11.2) 35 (6.9) 90 (17.6) 461 (90.4) 49 (9.6) 26 (5.1) 36 (7.1) 6 (1.2) 257 (50.4) 253 (49.6) 86 (16.9) 43 (8.4) Table 1 Baseline demographics and clinical characte (Continued) Atomoxetine hydrochlorideab Risperidoneab Ramelteonab Sodium valproateab Yokukansanab CGI-S 1: Normal 2: Minimally ill 3: Mildly ill 4: Moderately ill 5: Makedly ill 6: Severely ill 7: Very severely ill Mean ± SD Median (min to max) a: Multiple responses were allowed in a patient b: Drugs using in ≥2% of Patiants 6.8 ± 8.3 at end-point (LOCF), with significant tions in scores compared with baseline at each ment point (p < 0.0001). Patients showed a improvement in all ABC-J subscales at all asse points and end-point (LOCF) compared with b (Fig. 3). Interestingly, based on multiple regressio lysis, comorbid ADHD was not selected as the v affecting the ABC-J irritability subscale score a point (see Additional file 2). From the results of the improvement assessm CGI-I, the percentage of patients scoring 3 ‘mi improved’ or less was 74.3% (n = 306/412) at w 79.4% (n = 342/431) to 85.4% (n = 239/280) after and 77.0% (n = 376/488) at end-point (LOCF), w increasing trend throughout the treatment peri Additional file 3). From the results of the severity assessment by the percentage of patients who were assessed t ‘normal’ to 3 ‘mildly ill’ was 8.8% (n = 43/489) a line, 43.9% (n = 180/410) at week 4, 51.3% (n = 2 at week 8, and 59.7% (n = 292/489) at end-point and tended to increase throughout the treatment (see Fig. 4 and Additional file 3). The mean baseline SDQ Total Difficulties Sco 21.1 ± 5.5 (n = 309); mean changes from baseline 3.3 ± 4.9 (n = 215) at week 24, −4.3 ± 6.1 (n = week 52, and −3.8 ± 5.6 (n = 309) at end-point (p < 0.0001) (Fig. 5). Effectiveness The mean of ABC-J irritability subscale scores at base- line was 19.8 ± 9.5 (n = 396), decreased to 14.7 ± 8.8 (n = 288) at week 4 and to 12.9 ± 8.1 (n = 316) at week 8, remained at 12.7 to 13.1 at week 16 to 52, and was 13.0 ± 9.0 (n = 396) at end-point (LOCF). Mean changes of ABC-J irritability subscale scores from baseline were −5.7 ± 6.8 at 4 weeks, −7.0 to −8.3 after week 8 and − Sugimoto et al. BMC Psychiatry (2021) 21:204 Page 5 of 12 6.8 ± 8.3 at end-point (LOCF), with signi tions in scores compared with baseline at ment point (p < 0.0001). Patients show improvement in all ABC-J subscales at a points and end-point (LOCF) compared (Fig. 3). Interestingly, based on multiple re lysis, comorbid ADHD was not selected a affecting the ABC-J irritability subscale s point (see Additional file 2). From the results of the improvement a CGI-I, the percentage of patients scoring improved’ or less was 74.3% (n = 306/412 79.4% (n = 342/431) to 85.4% (n = 239/280) and 77.0% (n = 376/488) at end-point (LO increasing trend throughout the treatmen Additional file 3). From the results of the severity assessme the percentage of patients who were asse ‘normal’ to 3 ‘mildly ill’ was 8.8% (n = 43/ line, 43.9% (n = 180/410) at week 4, 51.3% at week 8, and 59.7% (n = 292/489) at end- and tended to increase throughout the trea (see Fig. 4 and Additional file 3). The mean baseline SDQ Total Difficulti 21.1 ± 5.5 (n = 309); mean changes from ba 3.3 ± 4.9 (n = 215) at week 24, −4.3 ± 6.1 week 52, and −3.8 ± 5.6 (n = 309) at end-p (p < 0.0001) (Fig. 5). The mean SDQ proso subscale score was 3.5 ± 2.5 (n = 316) at mean changes from baseline were 0.6 ± 1. Effectiveness Patients showed a mean improvement in all ABC-J subscales at all assessment points and end-point (LOCF) compared with baseline (Fig. 3). Interestingly, based on multiple regression ana- lysis, comorbid ADHD was not selected as the variable affecting the ABC-J irritability subscale score at end- point (see Additional file 2). 6.8 ± 8.3 at end-point (LOCF), with significant reduc- tions in scores compared with baseline at each assess- ment point (p < 0.0001). Patients showed a mean improvement in all ABC-J subscales at all assessment points and end-point (LOCF) compared with baseline (Fig. 3). Interestingly, based on multiple regression ana- lysis, comorbid ADHD was not selected as the variable affecting the ABC-J irritability subscale score at end- point (see Additional file 2). 15.3) From the results of the improvement assessment by CGI-I, the percentage of patients scoring 3 ‘minimally improved’ or less was 74.3% (n = 306/412) at week 4, 79.4% (n = 342/431) to 85.4% (n = 239/280) after week 8, and 77.0% (n = 376/488) at end-point (LOCF), with an increasing trend throughout the treatment period (see Additional file 3). From the results of the severity assessment by CGI-S, the percentage of patients who were assessed to be 1 ‘normal’ to 3 ‘mildly ill’ was 8.8% (n = 43/489) at base- line, 43.9% (n = 180/410) at week 4, 51.3% (n = 220/429) at week 8, and 59.7% (n = 292/489) at end-point (LOCF) and tended to increase throughout the treatment period (see Fig. 4 and Additional file 3). The mean baseline SDQ Total Difficulties Score was 21.1 ± 5.5 (n = 309); mean changes from baseline were − 3.3 ± 4.9 (n = 215) at week 24, −4.3 ± 6.1 (n = 200) at week 52, and −3.8 ± 5.6 (n = 309) at end-point (LOCF) (p < 0.0001) (Fig. 5). The mean SDQ prosocial behavior subscale score was 3.5 ± 2.5 (n = 316) at baseline, and mean changes from baseline were 0.6 ± 1.7 (n = 217) at week 24, 0.8 ± 2.1 (n = 200) at week 52, and 0.7 ± 2.0 Sugimoto et al. Effectiveness BMC Psychiatry (2021) 21:204 Page 6 of 12 Table 2 Reason for discontinuation and the timing Reason for discontinuationa Timing of discontin Table 2 Reason for discontinuation and the timing Table 2 Reason for discontinuation and the timing Reason for discontinuationa Timing of discontinuation, n (%) Total Day 1- < 29 Day 29- < 61 Day 61- < 91 Day 91- < 121 Day 121- < 181 Day 181- < 270 Day 270- < 365 Day ≥365 1. Adverse events 43 (21.4) 12 (6.0) 9 (4.5) 4 (2.0) 4 (2.0) 5 (2.5) 7 (3.5) 2 (1.0) 0 (0.0) 2. Worsening of symptoms 14 (7.0) 2 (1.0) 2 (1.0) 1 (0.5) 1 (0.5) 2 (1.0) 3 (1.5) 3 (1.5) 0 (0.0) 3. Improved symptoms 32 (15.9) 4 (2.0) 1 (0.5) 5 (2.5) 4 (2.0) 3 (1.5) 8 (4.0) 6 (3.0) 1 (0.5) 4. Request for discontinuation from patient or family 68 (33.8) 14 (7.0) 13 (6.5) 5 (2.5) 5 (2.5) 7 (3.5) 13 (6.5) 9 (4.5) 2 (1.0) 5.Transfer to another hospital 21 (10.4) 2 (1.0) 4 (2.0) 0 (0.0) 0 (0.0) 7 (3.5) 4 (2.0) 4 (2.0) 0 (0.0) 6. Lost to follow-up 38 (18.9) 7 (3.5) 7 (3.5) 3 (1.5) 4 (2.0) 5 (2.5) 7 (3.5) 5 (2.5) 0 (0.0) 7. Other 15 (7.5) 2 (1.0) 1 (0.5) 0 (0.0) 2 (1.0) 7 (3.5) 2 (1.0) 0 (0.0) 1 (0.5) Totalb 201 (100.0) 37 (18.4) 34 (16.9) 16 (8.0) 17 (8.5) 30 (14.9) 39 (19.4) 24 (11.9) 4 (2.0) a: Multiple responses were allowed in a patient. Concurrent reasons are: 3 and 4 (n = 9), 1 and 4 (n = 8), 1 and 2 (n = 4), 1 and 7 (n = 3), 4 and 7 (n = 2), 2 and 4 (n = 1), 1 and 6 (n = 1), 1, 2, and 4 (n = 1) b: If more than one reason for discontinuation was observed in a patient, the data were summarized as 1 patient were shown in hyperactivity/inattention, emotional symptoms, and prosocial behavior subscales. (n = 316) at end-point (p < 0.0001) (Fig. 5). All SDQ sub- scale scores and Total Difficulties Score were signifi- cantly decreased at all assessment points compared with baseline. The shift in SDQ scale properties from baseline to end-point (LOCF) is shown in Table 5. Effectiveness For conduct problems, the percentage of patients changing from ‘High Need’ to ‘Some Need’ or ‘Low need’ was 38.8% (n = 83/214) and the percentage of patients changing from ‘Some Need’ to ‘Low need’ was 51.1% (n = 24/47). In contrast, the percentage of patients changing from ‘Low Need’ to ‘Some Need’ or ‘High need’ was 23.1% (n = 12/52) and the patients changing from ‘Some Need’ to ‘High need’ was 19.1% (n = 9/47). Also, similar trends The shift in mean sleep time duration from baseline to end-point (LOCF) is shown in Fig. 6. The percentage of patients changing from ‘Too short’ to ‘May be appropri- ate’ or ‘Recommended’ was 55.6% (n = 15/27). In con- trast, the percentage of patients changing from ‘Recommended’ or ‘May be appropriate’ to ‘Too short’ was 1.6% (n = 4/258). Only one patient changed from ‘Too long’ to ‘Recommended’. No patients had a mean sleep time duration of ‘May be appropriate (longer)’ or ‘Too long’ at end-point (LOCF). : MedDRA/J version (22.1) Discussion In particular, baseline severity between the short-term clinical study and this surveillance was differ- ent; CGI-S scores were 4.9 ± 0.1 (mean ± standard error; SE) and 4.3 ± 0.8, and ABC-J irritability subscale scores were 26.9 ± 1.0 (mean ± SE) and 19.7 ± 9.6, respectively [9], aligning with our findings that aripiprazole was ad- ministered to patients with milder symptoms in routine practice than in the clinical study. frequently in the early phase of treatment, and there was no tendency for increased incidence with long-term treatment. Somnolence was observed through the sur- veillance period, and there were no reports of fall or traumatic injury related to somnolence. However, since the event is considered to affect school-aged pediatric patients’ academic performance, monitoring of the events is needed. In contrast, the incidence of treatment-emergent ad- verse events (TEAEs) was 97.7% (n = 84/86) in the Japa- nese clinical study, and the major events were somnolence (32.6%), influenza (29.1%) and weight in- creased (24.2%) [11]. Also, the major TEAEs were weight increased (23.0%) and vomiting (18.8%) in American 52- week clinical study [15]. Although the backgrounds of patients, the frequency of observations, and the dose dif- In 510 patients included in the safety analysis, ADRs occurred in 22.7% (n = 116/510) and the events that oc- curred in ≥1% of patients were somnolence (9.4%), weight increased (3.3%), nausea (1.4%), increased appe- tite and headache (1.2% each), and obesity (1.0%). Discussion Table 3 Adverse events and adverse drug reactions occurring in ≥0.5% of patients Table 3 Adverse events and adverse drug reactions occurring in ≥0.5% of patients Adverse Events n = 510 Adverse Drug Reactions* n = 510 125 (24.5%) 116 (22.7%) Somnolence 48 (9.4%) 48 (9.4%) Weight increased 18 (3.5%) 17 (3.3%) Headache 8 (1.6%) 6 (1.2%) Nausea 7 (1.4%) 7 (1.4%) Increased appetite 6 (1.2%) 6 (1.2%) Obesity 5 (1.0%) 5 (1.0%) Insomnia 4 (0.8%) 3 (0.6%) Tic disorders 4 (0.8%) 4 (0.8%) Akathisia 4 (0.8%) 4 (0.8%) Constipation 4 (0.8%) 4 (0.8%) Malaise 4 (0.8%) 4 (0.8%) Irritability 3 (0.6%) 3 (0.6%) Dizziness 3 (0.6%) 3 (0.6%) Vomiting 3 (0.6%) 3 (0.6%) *: MedDRA/J version (22.1) Atypical antipsychotics are used to treat irritability, one of the behavioral symptoms associated with ASD in pediatric patients, yet there is a need for additional real- world data on the long-term safety and effectiveness of specific agents. In this surveillance, the safety and effect- iveness of aripiprazole in patients (≥6 years to < 18 years old) newly treated with aripiprazole for irritability associ- ated with ASD in children and adolescents were exam- ined in a 52-week observation in daily clinical practice after marketing in Japan. In the 510 patients included in the safety analysis, the treatment continuation rate was 78.1% at week 52, and the most common reason for discontinuation was ‘re- quest for discontinuation from patient or family’ (33.8%). Treatment continuation might be decided by consult- ation between the physician and the patient or family members depending on the patient’s conditions or school environment. The mean daily dose for the entire treatment period in this surveillance was 2.20 ± 1.85 mg, and the mean daily dose of the long-term clinical study was 7.2 ± 4.0 mg [11]. The surveillance cannot be compared with the Sugimoto et al. BMC Psychiatry (2021) 21:204 Page 7 of 12 Fig. 2 First onset date of adverse drug reactions that occurred in ≥1% of patients Fig. 2 First onset date of adverse drug reactions that occurred in ≥1% of patients clinical study because of differences in inclusion/exclu- sion criteria of patients and dose-escalation/reduction methods. Discussion Inappropriate speech: 3.6 ± 3.1 at baseline, 3.0 ± 2.8 at Week 4, 2.7 ± 2.6 at Week 8, 2.5 ± 2.6 at Week 16, 2.5 ± 2.5 at Week 52, 2.4 ± 2.6 at end-point (LOCF, n = 399) Fig. 3 ABC-J score. Irritability: 19.8 ± 9.5 at baseline, 14.7 ± 8.8 at Week 4, 12.9 ± 8.1 at Week 8, 12.7 ± 8.3 at Week 16, 13.1 ± 9.1 at Week 52, 13.0 ± 9.0 at end-point (LOCF, n = 396). Lerthargy / social withdrawal: 13.2 ± 9.6 at baseline, 11.5 ± 8.9 at Week 4, 9.4 ± 8.5 at Week 8, 8.3 ± 7.8 at Week 16, 8.9 ± 8.4 at Week 52, 9.1 ± 8.4 at end point (LOCF, n = 392). Stereotypy: 4.5 ± 4.8 at baseline, 3.4 ± 4.3 at Week 4, 3.0 ± 4.1 at Week 8, 2.8 ± 3.8 at Week 16, 2.7 ± 3.9 at Week 52, 2.7 ± 4.0 at end-point (LOCF, n = 398). Hyperactivity: 20.2 ± 11.6 at baseline, 15.3 ± 9.6 at Week 4, 13.4 ± 9.0 at Week 8, 12.6 ± 8.8 at Week 16, 13.4 ± 10.1 at Week 52, 12.9 ± 9.9 at end-point (LOCF, n = 387). Inappropriate speech: 3.6 ± 3.1 at baseline, 3.0 ± 2.8 at Week 4, 2.7 ± 2.6 at Week 8, 2.5 ± 2.6 at Week 16, 2.5 ± 2.5 at Week 52, 2.4 ± 2.6 at end-point (LOCF, n = 399) study, and −3.2 ± 8.1 at end-point in the long-term clinical study in Japan [9–11]. Also, the mean change from baseline was −6.5 ± 11.1 at end-point in Ameri- can 52-week clinical study [16]. In these clinical stud- ies, enrolled patients with ABC-J irritability subscale score of 18 or higher cannot be directly compared with the surveillance because of differences in patients and treatment characteristics. However, ABC-J irrit- ability subscale score showed that aripiprazole im- proved the symptoms even in patients with relatively mild symptoms who were treated in post-marketing setting. fered from those in the clinical study, somnolence and weight increased were also the most common events in the real-world treatment. For the effect on weight and growth, 0.6% of weight increase-related ADRs (weight increased and obesity) led to treatment discontinuation, and this was considered not to have a major effect on long-term treatment con- tinuation. Discussion Among these, headache and nausea were observed more Table 4 Shift from baseline to end-point (LOCF) in percentile distribution of body weight End Point Percentile Category (LOCF) Baseline Percentile Category, n (%) ≤5 > 5- ≤10 > 10- ≤25 > 25- ≤50 > 50- ≤75 > 75- ≤90 > 90- ≤95 > 95 Total ≤5 22 (73.3) 1 (11.1) 1 (2.3) 0 0 0 0 0 24 (9.6) > 5- ≤10 6 (20.0) 3 (33.3) 5 (11.6) 0 0 0 0 0 14 (5.6) > 10- ≤25 1 (3.3) 3 (33.3) 17 (39.5) 8 (17.4) 0 0 0 0 29 (11.6) > 25- ≤50 1 (3.3) 2 (22.2) 19 (44.2) 19 (41.3) 9 (15.0) 0 0 0 50 (19.9) > 50- ≤75 0 0 1 (2.3) 16 (34.8) 29 (48.3) 3 (10.3) 0 0 49 (19.5) > 75- ≤90 0 0 0 3 (6.5) 22 (36.7) 14 (48.3) 3 (20.0) 1 (5.3) 43 (17.1) > 90- ≤95 0 0 0 0 0 9 (31.0) 6 (40.0) 2 (10.5) 17 (6.8) > 95 0 0 0 0 0 3 (10.3) 6 (40.0) 16 (84.2) 25 (10.0) Total 30 (100.0) 9 (100.0) 43 (100.0) 46 (100.0) 60 (100.0) 29 (100.0) 15 (100.0) 19 (100.0) 251 (100.0) Table 4 Shift from baseline to end-point (LOCF) in percentile distribution of body weight Sugimoto et al. BMC Psychiatry (2021) 21:204 Page 8 of 12 Sugimoto et al. BMC Psychiatry Fig. 3 ABC-J score. Irritability: 19.8 ± 9.5 at baseline, 14.7 ± 8.8 at Week 4, 12.9 ± 8.1 at Week 8, 12.7 ± 8.3 at Week 16, 13.1 ± 9.1 at Week 52, 13.0 ± 9.0 at end-point (LOCF, n = 396). Lerthargy / social withdrawal: 13.2 ± 9.6 at baseline, 11.5 ± 8.9 at Week 4, 9.4 ± 8.5 at Week 8, 8.3 ± 7.8 at Week 16, 8.9 ± 8.4 at Week 52, 9.1 ± 8.4 at end point (LOCF, n = 392). Stereotypy: 4.5 ± 4.8 at baseline, 3.4 ± 4.3 at Week 4, 3.0 ± 4.1 at Week 8, 2.8 ± 3.8 at Week 16, 2.7 ± 3.9 at Week 52, 2.7 ± 4.0 at end-point (LOCF, n = 398). Hyperactivity: 20.2 ± 11.6 at baseline, 15.3 ± 9.6 at Week 4, 13.4 ± 9.0 at Week 8, 12.6 ± 8.8 at Week 16, 13.4 ± 10.1 at Week 52, 12.9 ± 9.9 at end-point (LOCF, n = 387). Discussion The mean change in body weight from base- line increased over time throughout the treatment period, but there were no major changes in weight per- centile category from baseline, suggesting that weight in- crease is due to natural growth and not a crucial safety issue. The percentile and z-score of height and body mass index (BMI) are also suggesting that (see Additional file 4). Interestingly, although patients with comorbid ADHD were excluded in the clinical trials under diagnosis based on DSM-IV-TR, 53.1% of patients in this surveillance had comorbidities of ADHD that did not affect the ABC-J irritability subscale score at end-point (LOCF). ABC-J irritability subscale score was 19.8 ± 9.5 (n = 396) at baseline, and the mean changes from baseline were −5.7 ± 6.8 at week 4 and −6.8 ± 8.3 at end-point, which were significantly decreased (p < 0.0001). The corresponding score was 26.9 ± 1.0 (mean ± SE) at the baseline of the short-term clinical study and 17.6 ± 10.0 at the baseline of the long-term clinical study in Japan. The mean changes from baseline were −11.4 ± 1.3 (mean ± SE) at week 8 in the short-term clinical In addition, overall aberrant behaviors including hyperactivity, stereotypy, inappropriate speech, and leth- argy/social withdrawal were improved concurrently with irritability. SDQ for children has not been assessed in the clinical trials [11], while total 25 items in SDQ are recorded by caregivers in this surveillance. ASD diagnosis has Sugimoto et al. BMC Psychiatry (2021) 21:204 Page 9 of 12 Page 9 of 12 Sugimoto et al. BMC Psychiatry Fig. 4 Patient distribution of CGI-S score at baseline and endpoint Sleep duration has not been measured in prior clinical trials, whereas adequate sleep duration may not be maintained sufficiently due to irritability in the patients with ASD. The average sleep time duration in the last 4 weeks was reported by patients or caregivers and assessed in consideration of the recommended sleep time for each age [21] in this surveillance. Although con- comitant use of hypnotics was not taken into account, the improvement of irritability might lead to proper sleep duration as shown in Fig. 6. The results and inter- pretation of sleep time duration have certain limitations as there are no actigraphy-measured data. Discussion To our knowledge, this is the first study assessing the tolerability and effectiveness of aripiprazole in pediatric patients with ASD associated irritability over the long term, collecting a large amount of data and under actual clinical practice settings, not only for ABC-J or CGI scale but also SDQ scale, even though there are reports of clinical trials including systematic review and meta- analysis [24–28]. The goals of treatment are mainly to maximize an individual’s functional independence and quality of life through development and learning, im- provements in social skills and communication, reduc- tions in disability and comorbidity, and promotion of independence [29]. The results of ABC-J and SDQ assess- ments suggested that aripiprazole improved the symptoms of irritability associated with ASD, which may have a sec- ondarily effect in addressing these treatment goals. been associated with the low scores on the prosocial subscale in UK cohorts [22], and social impairments are core deficits [23]. Our findings suggest that pro- social behavior, which refers to positive interactions with other people, including helping, sharing, cooper- ating, and comforting, may also be improved in school based settings from a social perspective as well as irritability symptoms. The results and interpretation of safety and effectiveness have certain limitations. This surveillance was a prospective study in which the evaluation methods and measurement scales were determined in advance, and it was an observa- tional study in daily clinical practice without a comparison group. In addition, there were variations in the reporting and evaluation and deficiencies in the assessments based on the reports made by the physicians and caregivers. Fig. 5 SDQ score. Total difficulties score: 21.1 ± 5.5 at baseline, 18.0 ± 5.5 at Week 24, 17.0 ± 6.0 at Week 52, 17.3 ± 5.8 at end-point (LOCF). Prosocial behavior subscale: 3.5 ± 2.5 at baseline, 4.2 ± 2.6 at Week 24, 4.3 ± 2.5 at Week 52, 4.3 ± 2.7 at end-point (LOCF) Fig. 5 SDQ score. Total difficulties score: 21.1 ± 5.5 at baseline, 18.0 ± 5.5 at Week 24, 17.0 ± 6.0 at Week 52, 17.3 ± 5.8 at end-point (LOCF). Prosocial behavior subscale: 3.5 ± 2.5 at baseline, 4.2 ± 2.6 at Week 24, 4.3 ± 2.5 at Week 52, 4.3 ± 2.7 at end-point (LOCF) Sugimoto et al. Supplementary Information h l l Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12888-021-03201-6. Additional file 1. Treatment continuation rate (Kaplan-Meier plot). Additional file 2. Factors affecting ABC-J irritability subscale. Additional file 3. Patient distribution of CGI-I and CGI-S score. Additional file 4. Changes from baseline in height, weight and BMI. The online version contains supplementary material available at https://doi. org/10.1186/s12888-021-03201-6. Additional file 1. Treatment continuation rate (Kaplan-Meier plot). Additional file 2. Factors affecting ABC-J irritability subscale. Additional file 3. Patient distribution of CGI-I and CGI-S score. Additional file 4. Changes from baseline in height, weight and BMI. Additional file 1. Treatment continuation rate (Kaplan-Meier plot). Additional file 2. Factors affecting ABC-J irritability subscale. Additional file 3. Patient distribution of CGI-I and CGI-S score. Additional file 4. Changes from baseline in height, weight and BMI. Fig. 6 Shift from baseline to end-point (LOCF) in sleep time duration Discussion BMC Psychiatry (2021) 21:204 Page 10 of 12 Table 5 Shift from baseline to end-point (LOCF) in SDQ scale properties SDQ scale properties at baseline (bandings of raw scores) SDQ scale properties at end-point (LOCF), n (%) Low Need Some Need High Need Total Conduct problems Low Need (0 to 3) 40 (76.9) 9 (17.3) 3 (5.8) 52 (100.0) Some Need (4) 24 (51.1) 14 (29.8) 9 (19.1) 47 (100.0) High Need (5 to 10) 44 (20.6) 39 (18.2) 131 (61.2) 214 (100.0) Hyperactivity/inattention Low Need (0 to 5) 78 (87.6) 5 (5.6) 6 (6.7) 89 (100.0) Some Need (6) 18 (46.2) 13 (33.3) 8 (20.5) 39 (100.0) High Need (7 to 10) 56 (29.8) 27 (14.4) 105 (55.9) 188 (100.0) Emotional symptoms Low Need (0 to 3) 99 (90.0) 4 (3.6) 7 (6.4) 110 (100.0) Some Need (4) 20 (48.8) 9 (22.0) 12 (29.3) 41 (100.0) High Need (5 to 10) 43 (25.7) 18 (10.8) 106 (63.5) 167 (100.0) Peer problems Low Need (0 to 3) 26 (68.4) 5 (13.2) 7 (18.4) 38 (100.0) Some Need (4) 9 (18.4) 25 (51.0) 15 (30.6) 49 (100.0) High Need (5 to 10) 22 (9.7) 26 (11.5) 179 (78.9) 227 (100.0) Total difficulties score Low Need (0 to 12) 20 (74.1) 3 (11.1) 4 (14.8) 27 (100.0) Some Need (13 to 15) 14 (37.8) 16 (43.2) 7 (18.9) 37 (100.0) High Need (16 to 40) 45 (18.4) 36 (14.7) 164 (66.9) 245 (100.0) Prosocial behavior Low Need (6 to 10) 62 (81.6) 8 (10.5) 6 (7.9) 76 (100.0) Some Need (5) 19 (41.3) 18 (39.1) 9 (19.6) 46 (100.0) High Need (0 to 4) 29 (14.9) 24 (12.4) 141 (72.7) 194 (100.0) Table 5 Shift from baseline to end-point (LOCF) in SDQ scale properties Conclusions The results of the 52-week post-marketing surveillance suggest that aripiprazole was well tolerated and effective in the long-term treatment of irritability associated with ASD in Japanese children and adolescents in the real- world clinical practice. Fig. 6 Shift from baseline to end-point (LOCF) in sleep time duration Fig. 6 Shift from baseline to end-point (LOCF) in sleep time duration Abbreviations ABC-J: Aberrant Behavior Checklist Japanese version; ADHD: Attention Deficit and Hyperactivity Disorder; ADR: Adverse Drug Reaction; ASD: Autism Spectrum Disorder; BMI: Body Mass Index; CGI-I: Clinical Global Impression- Improvement; CGI-S: Clinical Global Impression-Severity of illness; DSM-IV- TR: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision; LD: Learning Disorders; LOCF: Last Observation Carried Forward; MedDRA/J: Japanese version of the ICH Medical Dictionary for Regulatory Activities; PMDA: Japanese Pharmaceutical and Medical Devices Agency; SD: Standard Deviation; SDQ: Strengths and Difficulties Questionnaire for Children; SE: Standard Error; TEAE: Treatment-emergent Adverse Event Supplementary Information Th l l Acknowledgements We thank all the physicians in each medical institution for the participation in this surveillance. The authors also acknowledge CAC Croit Corporation for Page 11 of 12 Page 11 of 12 Sugimoto et al. BMC Psychiatry (2021) 21:204 Sugimoto et al. BMC Psychiatry (2021) 21:204 the statistical analysis, and Ms. Yamada (Otsuka Pharmaceutical Co., Ltd.), Mr. Mamiya and Dr. Yasuda (former employees of Otsuka Pharmaceutical Co., Ltd.) for helpful discussions. 5. Stark AD, Jordan S, Allers KA, Bertekap RL, Chen R, Mistry Kannan T, et al. Interaction of the novel antipsychotic aripiprazole with 5-HT1A and 5-HT 2A receptors: functional receptor binding and in vivo electrophysiological studies. Psychopharmacology. 2007;190(3):373–82. https://doi.org/10.1007/ s00213-006-0621-y. 5. Stark AD, Jordan S, Allers KA, Bertekap RL, Chen R, Mistry Kannan T, et al. Interaction of the novel antipsychotic aripiprazole with 5-HT1A and 5-HT 2A receptors: functional receptor binding and in vivo electrophysiological studies. Psychopharmacology. 2007;190(3):373–82. https://doi.org/10.1007/ s00213-006-0621-y. Availability of data and materials y The data that support the findings of this study are available from Clinical Trial.gov (no. NCT03179787). 10. Ichikawa H, Hiratani M, Yasuhara A, et al. An open-label extension study of the safety and efficacy of aripiprazole for irritability in children and adolescents with autistic disorder: an interim report. Jpn J Clin Psychopharmacol. 2016;19:1481–92 (in Japanese). 10. Ichikawa H, Hiratani M, Yasuhara A, et al. An open-label extension study of the safety and efficacy of aripiprazole for irritability in children and adolescents with autistic disorder: an interim report. Jpn J Clin Psychopharmacol. 2016;19:1481–92 (in Japanese). Consent for publication 13. Owen R, Sikich L, Marcus RN, Corey-Lisle P, Manos G, McQuade RD, et al. Aripiprazole in the treatment of irritability in children and adolescents with autistic disorder. Pediatrics. 2009;124(6):1533–40. https://doi.org/10.1542/ peds.2008-3782. Received: 25 January 2021 Accepted: 6 April 2021 19. Goodman R. The strengths and difficulties questionnaire: a research note. J Child Psychol Psychiatry. 1997;38(5):581–6. https://doi.org/10.1111/j.1469-761 0.1997.tb01545.x. 19. Goodman R. The strengths and difficulties questionnaire: a research note. J Child Psychol Psychiatry. 1997;38(5):581–6. https://doi.org/10.1111/j.1469-761 0.1997.tb01545.x. Authors’ contributions 6. Otsuka America Pharmaceutical, Inc. Aripiprazole [package insert]. Rockville; 2020. https://www.accessdata.fda.gov/scripts/cder/safetylabelingchanges/. 6. Otsuka America Pharmaceutical, Inc. Aripiprazole [package insert]. Rockville; 2020. https://www.accessdata.fda.gov/scripts/cder/safetylabelingchanges/. YS: design, data analyses and writing of this manuscript, KY: design, data collection and management, TT: design and supervising, YF: supervising, KA: interpreting results and critical revision, KM and AT: medical advice for the design, interpreting results and critical revision. All authors read and approved the final manuscript. 7. Otsuka Pharmaceutical Netherlands B.V. Aripiprazole [summary of product characteristics]. Amsterdam; 2020. https://www.ema.europa.eu/en/ medicines/human/EPAR/abilify. 7. Otsuka Pharmaceutical Netherlands B.V. Aripiprazole [summary of product characteristics]. Amsterdam; 2020. https://www.ema.europa.eu/en/ medicines/human/EPAR/abilify. 8. Caccia S. Safety and pharmacokinetics of atypical antipsychotics in children and adolescents. Pediatr Drugs. 2013;15(3):217–33. https://doi.org/10.1007/ s40272-013-0024-6. 8. Caccia S. Safety and pharmacokinetics of atypical antipsychotics in children and adolescents. Pediatr Drugs. 2013;15(3):217–33. https://doi.org/10.1007/ s40272-013-0024-6. Author details 1 1Pharmacovigilance Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan. 2Medical Affairs Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan. 3Department of Psychiatry, Tokai University School of Medicine, Kanagawa, Japan. 4Research Center for Child Mental Development, University of Fukui, Fukui, Japan. 1Pharmacovigilance Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan. 2Medical Affairs Department, Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan. 3Department of Psychiatry, Tokai University School of Medicine, 4 h f h ld l l 17. Otsuka Pharmaceutical Co., Ltd. Aripiprazole [package insert]. Tokyo, Japan, Revised September 2020 (in Japanese). Kanagawa, Japan. 4Research Center for Child Mental Development, University of Fukui, Fukui, Japan. 17. Otsuka Pharmaceutical Co., Ltd. Aripiprazole [package insert]. Tokyo, Japan, Revised September 2020 (in Japanese). Kanagawa, Japan. 4Research Center for Child Mental Development, University of Fukui, Fukui, Japan. 18. Ono Y. Factor validity and reliability for the aberrant behavior checklist- Community in a Japanese population with mental retardation. Res Dev Disabil. 1996;17(4):303–9. https://doi.org/10.1016/0891-4222(96)00015-7. 18. Ono Y. Factor validity and reliability for the aberrant behavior checklist- Community in a Japanese population with mental retardation. Res Dev Disabil. 1996;17(4):303–9. https://doi.org/10.1016/0891-4222(96)00015-7. Competing interests 14. Marcus RN, Owen R, Kamen L, Manos G, McQuade RD, Carson WH, et al. A placebo-controlled, fixed-dose study of aripiprazole in children and adolescents with irritability associated with autistic disorder. J Am Acad Child Adolesc Psychiatry. 2009;48(11):1110–9. https://doi.org/10.1097/CHI. 0b013e3181b76658. This study was funded by Otsuka Pharmaceutical Co., Ltd. KM and AT have received consulting fees from Otsuka Pharmaceutical Co., Ltd. based on contract. YS, KY, TT, YF and KA are employees of Otsuka Pharmaceutical Co., Ltd. KM has received: a) research support from Grant-in-Aid for Scientific Re- search (C) (Grant Number 18 K07611), Taisho Pharmaceutical, Otsuka Pharma- ceutical, Shionogi & Co., Japanese Society for Probiotic Science; b) honoraria from Otsuka Pharmaceutical, Shionogi & Co., Shire Japan, Eli Lilly and Co., Meiji Holdings Co., Takeda Pharmaceutical, MIYARISAN PHARMACEUTICAL CO; and c) a consulting fee from Shionogi & Co and Viatris. 15. Marcus RN, Owen R, Manos G, Mankoski R, Kamen L, McQuade RD, et al. Safety and tolerability of aripiprazole for irritability in pediatric patients with autistic disorder: a 52-week, openlabel, multicenter study. J Clin Psychiatry. 2011;72(09):1270–6. https://doi.org/10.4088/JCP.09m05933. 16. Marcus RN, Owen R, Manos G, Mankoski R, Kamen L, McQuade RD, et al. Aripiprazole in the treatment of irritability in pediatric patients (aged 6-17 years) with autistic disorder: results from a 52-week, open-label study. J Child Adolesc Psychopharmacol. 2011;21(3):229–36. https://doi.org/10.1089/ cap.2009.0121. Ethics approval and consent to participate This surveillance was conducted in compliance with the Ministerial Ordinance on Standards for Conducting Post-marketing Surveys and Studies on Drugs; MHLW Ordinance No. 171 issued on December 20, 2004 (GPSP). As the surveillance is a non-interventional study in accordance with GPSP, the need for ethics approval and consent were waived. The surveillance was designed by Otsuka, reviewed by the Japanese Pharmaceutical and Medical Devices Agency (PMDA). 11. Ichikawa H, Hiratani M, Yasuhara A, Tsujii N, Oshimo T, Ono H, et al. An open-label extension long-term study of the safety and efficacy of aripiprazole for irritability in children and adolescents with autistic disorder in Japan. Psychiatry Clin Neurosci. 2018;72(2):84–94. https://doi.org/10.1111/ pcn.12607. 12. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn, text revision edn. Washington, DC: American Psychiatric Association; 2000. 12. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn, text revision edn. Washington, DC: American Psychiatric Association; 2000. Funding This surveillance was funded by Otsuka Pharmaceutical Co., Ltd. 9. Ichikawa H, Mikami K, Okada T, Yamashita Y, Ishizaki Y, Tomoda A, et al. Aripiprazole in the treatment of irritability in children and adolescents with autism spectrum disorder in Japan: a randomized, double-blind, placebo- controlled study. Child Psychiat Hum D. 2017;48(5):796–806. https://doi. org/10.1007/s10578-016-0704-x. 9. Ichikawa H, Mikami K, Okada T, Yamashita Y, Ishizaki Y, Tomoda A, et al. Aripiprazole in the treatment of irritability in children and adolescents with autism spectrum disorder in Japan: a randomized, double-blind, placebo- controlled study. Child Psychiat Hum D. 2017;48(5):796–806. https://doi. org/10.1007/s10578-016-0704-x. References 1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 5th edn. Arlington, VA: American Psychiatric Publishing; 2013. 1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 5th edn. Arlington, VA: American Psychiatric Publishing; 2013. 20. Matsuishi T, Nagano M, Araki Y, Tanaka Y, Iwasaki M, Yamashita Y, et al. Scale properties of the Japanese version of the strengths and difficulties questionnaire (SDQ): a study of infant and school children in community samples. Brain and Development. 2008;30(6):410–5. https://doi.org/10.1016/j. braindev.2007.12.003. 20. Matsuishi T, Nagano M, Araki Y, Tanaka Y, Iwasaki M, Yamashita Y, et al. Scale properties of the Japanese version of the strengths and difficulties questionnaire (SDQ): a study of infant and school children in community samples. Brain and Development. 2008;30(6):410–5. https://doi.org/10.1016/j. braindev.2007.12.003. 2. Volkmar F, Cook EH, Pomeroy J, Realmuto G, Tanguay P. Practice parameters for the assessment and treatment of children, adolescents, and adults with autism and other pervasive developmental disorders. American Academy of Child and Adolescent Psychiatry working group on quality issues. J Am Acad Child Adolesc Psychiatry. 1999;38(12):32S–54S. https://doi.org/10.1016/ S0890-8567(99)80003-3. 21. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health. 2015;1(1):40–3. https://doi. org/10.1016/j.sleh.2014.12.010. 21. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health. 2015;1(1):40–3. https://doi. org/10.1016/j.sleh.2014.12.010. 3. Myers SM, Johnson CP. American Academy of Pediatrics Council on children with disabilities. Management of children with autism spectrum disorders. Pediatrics. 2007;120(5):1162–82. https://doi.org/10.1542/peds.2 007-2362. 22. Russell G, Steer C. Golding J social and demographic factors that influence the diagnosis of autistic spectrum disorders. Soc Psychiatry Psychiatr Epidemiol. 2011;46(12):1283–93. https://doi.org/10.1007/s00127-010-0294-z. 22. Russell G, Steer C. Golding J social and demographic factors that influence the diagnosis of autistic spectrum disorders. Soc Psychiatry Psychiatr Epidemiol. 2011;46(12):1283–93. https://doi.org/10.1007/s00127-010-0294-z. 4. Tadori Y, Kitagawa H, Forbes RA, McQuade RD, Stark A, Kikuchi T. Differences in agonist/antagonist properties at human dopamine D2 receptors between aripiprazole, bifeprunox and SDZ 208-912. Eur J Pharmacol. 2007;574(2-3):103– 11. https://doi.org/10.1016/j.ejphar.2007.07.031. 23. Iizuka C, Yamashita Y, Nagamitsu S, Yamashita T, Araki Y, Ohya T, et al. Comparison of the strengths and difficulties questionnaire (SDQ) scores between children with high-functioning autism spectrum disorder (HFASD) 23. Iizuka C, Yamashita Y, Nagamitsu S, Yamashita T, Araki Y, Ohya T, et al. References Comparison of the strengths and difficulties questionnaire (SDQ) scores between children with high-functioning autism spectrum disorder (HFASD) Page 12 of 12 Sugimoto et al. BMC Psychiatry (2021) 21:204 Sugimoto et al. BMC Psychiatry (2021) 21:204 and attention-deficit/hyperactivity disorder (AD/HD). Brain and Development. 2010;32(8):609–12. https://doi.org/10.1016/j.braindev.2009.09.009. 24. Robb AS, Andersson C, Bellocchio EE, Manos G, Rojas-Fernandez C, Mathew S, et al. Safety and tolerability of aripiprazole in the treatment of irritability associated with autistic disorder in pediatric subjects (6-17 years old):results from a pooled analysis of 2 studies. Prim Care Companion CNS Disord. 2011;13(1):1–9. https://doi.org/10.4088/PCC.10m01008gry. 24. Robb AS, Andersson C, Bellocchio EE, Manos G, Rojas-Fernandez C, Mathew S, et al. Safety and tolerability of aripiprazole in the treatment of irritability associated with autistic disorder in pediatric subjects (6-17 years old):results from a pooled analysis of 2 studies. Prim Care Companion CNS Disord. 2011;13(1):1–9. https://doi.org/10.4088/PCC.10m01008gry. 25. Mankoski R, Stockton G, Manos G, Marler S, McQuade R, Forbes RA, et al. Aripiprazole treatment of irritability associated with autistic disorder and the relationship between prior antipsychotic exposure, adverse events, and weight change. J Child Adolesc Psychopharmacol. 2013;23(8):572–6. https:// doi.org/10.1089/cap.2012.0075. 26. Coustals N, Menard ML, Cohen D. Aripiprazole in children and adolescents. J Child Adolesc Psychopharmacol. 2021;31(1):4–32. https://doi.org/10.1089/ca p.2020.0014. 27. Maneeton N, Maneeton B, Putthisri S, Suttajit S, Likhitsathian S, Srisurapanont M. Aripiprazole in acute treatment of children and adolescents with autism spectrum disorder: a systematic review and metaanalysis. Neuropsychiatr Dis Treat. 2018;14:3063–72. https://doi.org/1 0.2147/NDT.S174622. 28. Hirsch LE, Pringsheim T. Aripiprazole for autism spectrum disorders (ASD). Cochrane Database Syst Rev. 2016;(6). Art. No.: CD009043. https://doi.org/1 0.1002/14651858.CD009043.pub3. 28. Hirsch LE, Pringsheim T. Aripiprazole for autism spectrum disorders (ASD). Cochrane Database Syst Rev. 2016;(6). Art. No.: CD009043. https://doi.org/1 0.1002/14651858.CD009043.pub3. 29. Lai MC, Lombardo MV, Baron-Cohen S. Autism. Lancet. 2014;383(9920):896– 910. https://doi.org/10.1016/S0140-6736(13)61539-1. 29. Lai MC, Lombardo MV, Baron-Cohen S. Autism. Lancet. 2014;383(9920):896– 910. https://doi.org/10.1016/S0140-6736(13)61539-1. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
https://openalex.org/W3104646047	https://ieeexplore.ieee.org/ielx7/6287639/8948470/09252080.pdf	English	null	SecRand: A Secure Distributed Randomness Generation Protocol With High Practicality and Scalability	IEEE access	2,020	cc-by	12,015	Received September 29, 2020, accepted October 27, 2020, date of publication November 9, 2020, date of current version November 19, 2020. Received September 29, 2020, accepted October 27, 2020, date of publication November 9, 2020, date of current version November 19, 2020. Digital Object Identifier 10.1109/ACCESS.2020.3036698 Digital Object Identifier 10.1109/ACCESS.2020.3036698 SecRand: A Secure Distributed Randomness Generation Protocol With High Practicality and Scalability ZHAOZHONG GUO , LIUCHENG SHI, AND MAOZHI XU School of Mathematical Sciences, Peking University, Beijing 100871, China Corresponding author: Maozhi Xu (mzxu@math.pku.edu.cn) This work was supported in part by the National Key Research and Development Program of China under Grant 2017YFB0802000, and in part by the National Natural Science Foundation of China (NSFC) under Grant 61672059. ZHAOZHONG GUO , LIUCHENG SHI, AND MAOZHI XU School of Mathematical Sciences, Peking University, Beijing 100871, China Corresponding author: Maozhi Xu (mzxu@math.pku.edu.cn) Corresponding author: Maozhi Xu (mzxu@math.pku.edu.cn) This work was supported in part by the National Key Research and Development Program of China under Grant 2017YFB0802000, and in part by the National Natural Science Foundation of China (NSFC) under Grant 61672059. ABSTRACT The ever-increasing pervasiveness of decentralized applications, such as blockchain, is creating challenges for sources of randomness, which play an integral part in decentralized settings. Distributed randomness generation (DRG) protocols, aiming at producing high-quality randomness without a central party, have drawn increased attention from academia as well as industry. Previous approaches lacked security proofs, and their dependence on secure messaging channels reduced their practicality. In this work, we ﬁrst formalize the desired properties of a secure DRG protocol and build a security model using these formal deﬁnitions. To the best of our knowledge, this is the ﬁrst work to build a security model for DRG protocols, which can be used as a general framework for security analysis of DRG protocols. We then present SecRand, a secure DRG protocol with high practicality and scalability. We improve upon previous approaches by modifying the secret generation method in the reconstruction phase, which ensures the same scalability but achieves resistance against an adversary’s malicious behavior of withholding its secret shares. We also provide strict proofs under our security model, showing that SecRand achieves the desired properties and is secure enough to be used in decentralized applications. Furthermore, we present a detailed performance evaluation of SecRand by deploying it on a laptop with a Windows 10 environment in the C language. The experimental data showed that SecRand achieved a better performance compared with previous approaches in the presence of corrupted participants, and this performance advantage grew linearly with the number of corrupted participants. is work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Received September 29, 2020, accepted October 27, 2020, date of publication November 9, 2020, date of current version November 19, 2020. A. RELATED WORK In recent years, a substantial amount of research related to distributed randomness generation has been published by researchers in academia as well as industry. Bitcoin is utilized as a source of randomness [8], [13] that relies on block hashes, which is not guaranteed to be unbiased due to the malicious behaviors of miners. Proof-of-Delay [14] addresses this problem by employing a delay function on top of the PoW block hash, but veriﬁcation of the generated random value is slow. Dﬁnity [15] employs a unique threshold signature as a core primitive in its construction, which solves the funda- mental ‘‘last actor’’ problem, where the last participant in the protocol has an advantage in deciding or precomputing the ﬁnal output. However, a fully connected network of authen- ticated and reliable bidirectional point-to-point messaging channels must be established between the participants, which is unrealistic in distributed applications. RandHound [10] and RandHerd [10] are realized based on publicly veriﬁable secret sharing (PVSS) and achieve scalability by sharding. But both of the protocols have failure probabilities. SCRAPE [16] is a succinct but efﬁcient DRG protocol that achieves scalability by relying on an improved PVSS scheme requiring only a lin- ear number of exponentiations. In brief, SCRAPE is divided into three stages, the Commit stage, Reveal stage, and Recov- ery stage. In the Commit stage, each participant chooses a secret, shares it among all the participants by Shamir’s Secret Sharing, and publishes the corresponding commit- ments. In the Reveal stage, each participant is required to reveal its chosen secret. In the Recovery stage, the unrevealed secrets will be reconstructed through Lagrange interpolation by other participants. The ﬁnal output is computed as the sum of these secrets. Unfortunately, SCRAPE’s performance can be degraded by an adversary’s malicious behavior of withholding its secret in the Reveal stage, which results in extra work in the Recovery stage. • We consider the performance, security, and practical- ity of the design and present ‘‘SecRand’’. SecRand is a variant of SCRAPE with the same scalability but achieves better performance by modifying the method of computing the ﬁnal output. Instead of SCRAPE’s process of revealing, reconstructing, and adding up all the secrets, SecRand directly computes the ﬁnal output through Lagrange interpolation. This feature protects SecRand against an adversary’s malicious behavior of withholding its secret share. Thus, SecRand has a higher and more stable performance than SCRAPE. A. RELATED WORK In contrast to SCRAPE’s security analysis in the stand- alone setting, a composable construction of the security analysis is provided under our security model, show- ing that SecRand is secure for use in decentralized applications. Regarding the practicality, participants’ intermediate results generated during the protocol exe- cution process are exchanged through a public bul- letin board [25], verifying and ﬁltering out invalid data. This design makes it possible for thousands of participants to be involved in SecRand, as establishing a fully connected network of secure messaging channels is unrealistic in decentralized applications and hinders scalability. • We propose two constructions of SecRand in this article to adjust to different usage scenarios, which are denoted as SecRand-ZK and SecRand-Pairing. SecRand-ZK uses zero-knowledge for data veriﬁcation and is suit- able for computing-resource-limited scenarios, while SecRand-Pairing uses pairing for data veriﬁcation and is suitable in space-limited scenarios. Considering previous approaches [8]–[16], the availability, unpredictability, bias resistance, guaranteed output delivery, and public veriﬁability are listed as desiderata of a secure DRG protocol. However, there are still no formal deﬁnitions for these properties, and thus, previous approaches lacked rig- orous security proof. This restriction greatly limits their usage in distributed applications. Furthermore, the dependence on secure messaging channels also reduces their practicality. p • We present a detailed performance evaluation of the SecRand protocol by deploying it on a Windows 10 envi- ronment laptop in the C language. We focus on the runtime and space occupation and make a comparison of SecRand’s two different constructions. Experimental data showed that SecRand-ZK had a higher space occu- pation and a lower computing resource occupation for the underlying public bulletin board, whereas SecRand- Pairing was the opposite. This result is consistent with the theoretical analysis. Thus, users may choose the proper construction of SecRand based on the underlying public bulletin board’s computing and storage capac- ity. We also compared the performance of SecRand and SCRAPE. Experimental data showed that SecRand achieved a better performance than SCRAPE, and this performance advantage grew linearly with the number of corrupted participants. • We present a detailed performance evaluation of the SecRand protocol by deploying it on a Windows 10 envi- ronment laptop in the C language. We focus on the runtime and space occupation and make a comparison of SecRand’s two different constructions. I. INTRODUCTION the algorithm to introduce entropy to ensure manipulation- resistance and unpredictability [7]. Centralized generation of randomness requires trust in a single party and is not appli- cable in decentralized settings. Thus, distributed randomness generation (DRG) protocols, which produce randomness in distributed ways, have attracted increased attention with the rapid growth of blockchain technology. A reliable source of randomness, which has been studied since the early days of cryptography, is not only an essential building block in various cryptographic protocols [1], [2] but also has proven to be a vital component in the design of blockchain proposals. Speciﬁcally, randomness is desired and used in blockchain’s consensus layer and application layer, such as Proof-of-Stake (PoS) [3], sharding [4], and smart contracts [5], [6]. We take PoS as an example to show how it is applied in the blockchain area. The core component of PoS is a leader selection algorithm, which decides the identity of the next block proposer. Randomness is used as the input of Security and performance are two major considerations in the design and evaluation of DRG protocols. With regard to security, active adversaries may behave maliciously to inﬂuence the execution of protocols to their advantage. DRG protocols are supposed to resist these malicious behaviors and produce high-quality randomness in terms of the entropy it provides. With respect to performance, DRG protocols are supposed to have low communication and computation The associate editor coordinating the review of this manuscript and approving it for publication was Kashif Saleem . 203917 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability complexity as well as high scalability to support thousands of participants to be involved at the same time. izations, we deﬁne the DRG protocol’s security. These formalizations together with the security deﬁnition con- stitute a security model for DRG protocols, which can be used as a general framework for security analysis. To the best of our knowledge, this is the ﬁrst work to build a security model for DRG protocols, which ﬁlls the research gaps and may provide guidance for DRG design. A. RELATED WORK Experimental data showed that SecRand-ZK had a higher space occu- pation and a lower computing resource occupation for the underlying public bulletin board, whereas SecRand- Pairing was the opposite. This result is consistent with the theoretical analysis. Thus, users may choose the proper construction of SecRand based on the underlying public bulletin board’s computing and storage capac- ity. We also compared the performance of SecRand and SCRAPE. Experimental data showed that SecRand achieved a better performance than SCRAPE, and this performance advantage grew linearly with the number of corrupted participants. B. OUR RESULTS Our contributions are summarized as follows: C. PAPER ORGANIZATION The remainder of the paper is organized as follows. In Sec. II, we present preliminaries, including basic cryptography tools, security assumptions, the public bulletin board, and the adver- sary model. We formalize the properties of a secure DRG protocol and deﬁne security in Sec. III. SecRand’s protocol description and two constructions are outlined in detail in Sec. IV. We give strict proofs showing that SecRand achieves the desired properties and satisﬁes the security deﬁnition in Sec. V. The experimental performance results obtained from the prototype implementation are shown in Sec. VI. Finally, we present conclusions in Sec. VII. f (x) ∈Zq [x] , deg[f (x)] ≤n −t −1}, where the coefﬁcients vi = Qn j=1,j̸=i 1 i−j, for i = 1, 2, . . . , n. j ,j̸ i j We remark that sharing a secret with Shamir’s Secret Shar- ing is equivalent to encoding the secret with a Reed–Solomon code. This fact is ﬁrst used by SCRAPE [16] to verify the validity of secret shares, and we also follow this approach in the veriﬁcation of secret shares. A. CRYPTOGRAPHY TOOLS There is a zero-knowledge proof of knowledge of a value α ∈ Zq such that x = gα and y = hα given g, x, h, y, which is called the discrete log equivalence (DLEQ) proof. A protocol was constructed previously [21] to perform this proof, and it is described as follows: 4) PAIRING Pairing [22] is usually described as a tuple (q, G, GT , e). G and GT are two cyclic groups of a large prime order q, G is the generator of G, and e: G × G →GT is a bilinear map meaning e (a · G, b · G) = e(G, G)ab, where a, b ∈Zq. We remark that Shamir’s Secret Sharing naturally has addi- tional homomorphism, which means that the sum of shares is the share of the sum of corresponding secrets. Therefore, the sum of secrets can be reconstructed through Lagrange interpolation using the summed shares. 1) SHAMIR’S SECRET SHARING (t, n)-Shamir’s Secret Sharing [17]–[19] is used to share a secret among n participants, and no less than t participants are needed to recover the secret. Suppose P shares secret s among n participants, which are denoted as {P1, P2, . . . , Pn}, then P performs the following operations. First, P chooses a random polynomial of degree t −1, f (x) = s + a1x + . . . + at−1xt−1, where a1, . . . , at−1 ∈Fq (q is a prime number) and at−1 ̸= 0. Second, P randomly selects n distinct numbers x1, x2, . . . , xn ∈Fq and computes f (x1), f (x2) , . . . , f (xn). Third, P secretly sends Pi the secret share si = [xi, f (xi)], where i = 1, 2, . . . , n. Finally, any k ≥t participants can work together to reconstruct s by Lagrange interpolation: • Proof Generation: Select w ∈Zq, and compute a1 = gw, a2 = hw. Set e = hash(a1\|\|a2) and compute z = w−αe. Return proof =(z, a1, a2). • Proof Veriﬁcation: Compute e = hash(a1\|\|a2). Check that a1 = gzxe and a2 = hzye. If both of the equations are satisﬁed, return True; else, return False. We use DLEQ(g, x, h, y) to denote the proof generation process and DLEQ_Verify(g, x, h, y, proof ) to denote the proof veriﬁcation process. s = Xk i=1 f (xi) Yk j=1,j̸=i xj xj −xi . 2) REED–SOLOMON CODE In this subsection, we introduce the computational Difﬁe– Hellman (CDH) assumption [23] and decisional bilinear square (DBS) assumption [24], based on which we prove SecRand’s unpredictability property. We use ‘‘PPT adver- sary’’ to represent that the adversary’s computing capabil- ity is bounded by a probabilistic polynomial time Turing machine. The two hardness assumptions are described as follows: In the [n, k, d] Reed–Solomon code [20], C is a linear error correcting code over Zq (q is a prime number) of length n, dimension k, and minimum distance d. Its dual code C⊥is the vector space consisting of all vectors c⊥∈Zn q such that ⟨c, c⊥⟩= 0 for all c in C, where ⟨⟩denotes the operation of an inner product. Based on basic linear algebra theory, we have the following lemma: Lemma: v ∈Zn q C, and c⊥is chosen uniformly at random in C⊥, then the probability that ⟨v, c⊥⟩= 0 is exactly 1 q. • Computational Difﬁe–Hellman Assumption: Let G be a cyclic group generated by G, whose order is a prime q. For a, b ∈Zq∗, given G, a · G, b · G, it is difﬁcult for a PPT adversary to compute (ab) · G. In this article, we use Reed–Solomon code with the follow- ing form: • Decisional Bilinear Square Assumption: (q, G, GT , e) is a description of a bilinear group. For a generator G ∈G, a, b, r ∈Zq∗, given A = a · G, B = b · G, the following probability distributions are indistinguishable for a PPT adversary: D0 = (G, A, B, T0 = e(A, A)b) and D1 = (G, A, B, T1 = e(A, A)r). • Decisional Bilinear Square Assumption: (q, G, GT , e) is a description of a bilinear group. For a generator G ∈G, a, b, r ∈Zq∗, given A = a · G, B = b · G, the following probability distributions are indistinguishable for a PPT adversary: D0 = (G, A, B, T0 = e(A, A)b) and D1 = (G, A, B, T1 = e(A, A)r). C = {(p (1) , p (2) , . . . , p (n)) : p (x) ∈Zq [x] , deg [p (x)] ≤t −1}, C = {(p (1) , p (2) , . . . Our contributions are summarized as follows: Our contributions are summarized as follows: • We formalize the properties of availability, bias resis- tance, unpredictability, guaranteed output delivery, and public veriﬁability, which are regarded as desirable properties of a secure DRG protocol. The formalizations are game-based and applicable for any DRG protocol regardless of its technical details. Using these formal- VOLUME 8, 2020 203918 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability Its dual code is deﬁned as follows: 2) UNPREDICTABILITY We say that a DRG protocol achieves unpredictability if a PPT adversary Apre has a negligible advantage in the following game played against a challenger: • Corrupted participants have three types of malicious behaviors, namely, refusing to participate in the proto- col, sending invalid data during the execution process, and revealing private data and colluding to perform a coordinated attack. • The challenger creates secret keys for honest partici- pants and sends the corresponding public keys to Apre. • Apre creates secret keys for corrupted participants and sends the corresponding public keys to the challenger. • The challenger chooses value x and runs the DRG on behalf of the honest participants with x as the DRG ﬁnal output. It sends Apre all the public information generated in the protocol’s execution, while Apre runs the DRG on behalf of the corrupted participants. 2) REED–SOLOMON CODE , p (n)) : p (x) ∈Zq [x] , deg [p (x)] ≤t −1}, where p (x) ranges over all the polynomials in Zq [x] with a degree of at most t −1. 203919 VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability D. ADVERSARY MODEL In this subsection, we present the adversary model of SecRand. • Ahalt creates secret keys for corrupted participants and sends the corresponding public keys to the challenger. In an (n, t) SecRand, there are n participants, and at least t participants are needed to reconstruct the ﬁnal output. We consider static adversaries that will corrupt a certain number of participants before the protocol execution begins. The corrupted participants may deviate from the protocol in any arbitrary way. The uncorrupted participants are renamed as honest participants. We have the following settings: • The challenger runs the DRG on behalf of the honest participants and sends Ahalt all the public information generated in the protocol’s execution, while Ahalt runs the DRG on behalf of the corrupted participants. • By the end of the DRG, each honest participant obtains an output. The advantage of Ahalt is deﬁned as Pr[opt = ⊥], where opt is the output of any honest participant and ⊥is the output when the DRG protocol halts. • All n participants use a public bulletin board for data interaction. • All n participants use a public bulletin board for data interaction. • The maximum number of corrupted participants is t −1. • Honest participants follow the protocol strictly and will not reveal any private data to anyone else. They make up a majority of all participants, i.e., n ≥2t −1, which is usually referred to as the honest majority assumption. 1) AVAILABILITY We say that a DRG protocol achieves availability if a PPT adversary Ahalt has a negligible advantage in the following game played against a challenger: • The challenger creates secret keys for honest partici- pants and sends the corresponding public keys to Ahalt. C. PUBLIC BULLETIN BOARD output before the end of the protocol. Bias resistance means no participant can bias the ﬁnal output by specifying its input according to other participants’ inputs. Guaranteed output delivery (G.O.D.) means all the honest participants obtain the same output by the end of the protocol. Public veriﬁa- bility means the validity of the protocol’s ﬁnal output can be veriﬁed publicly, and thus, the adversary is not able to have someone accept a fake value as the protocol’s output. We give a short discussion of the necessity and completeness of the above properties. For necessity, unpredictability and bias resistance are two basic requirements for randomness. Availability and G.O.D. ensure that the DRG protocol can always produce a unique output, which is of great importance in distributed systems. Finally, public veriﬁability is required due to the openness of the blockchain, where anyone can read and verify the data stored in the blockchain. For complete- ness, these ﬁve properties are listed as desiderata of a secure DRG protocol by previous approaches, which are enough to build a security model for DRG protocols. To achieve public veriﬁability and scalability, we replace the standard network model with a public bulletin board, of which availability has been assumed in previous work [25]. Participants of SecRand use a public bulletin board as a reliable and stable information exchange channel, as it is publicly accessible and no posted information can ever be changed or erased. Speciﬁcally, the public bulletin board has three functions: • Pi uploads data to the public bulletin board through the function Upload(data). • Pi receives data sent by other participants from public bulletin board through the function Download(data). • The uploaded data are veriﬁed by the public bulletin board through the function Verify(data). We remark that the public bulletin board described above is an enhanced version of previous approaches with a new func- tion Verify(). In fact, such a public bulletin board is readily available in the applications in which we are interested. One instantiation is Ethereum [5], which is a public blockchain with liveness and persistence. Its smart contract mechanism provides a Turing-complete programming language, which can be used to preset the rules for data veriﬁcation according to the requirements of SecRand. We now present formalizations of the properties. A. PROPERTY FORMALIZATION We ﬁrst give a brief description of the properties to be formalized. Availability means the protocol will never halt, even in the presence of participants’ malicious behaviors. Unpredictability means no participant can predict the ﬁnal • Before any honest participant can obtain the ﬁnal output, Apre outputs a predicted value x ′. The advantage of Apre is deﬁned as Pr[x = x′]. VOLUME 8, 2020 203920 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability IV. SECRAND SecRand consists of three phases, namely, Distribution, Veriﬁcation, and Reconstruction. We ﬁrst give an overall description, in which data validity proof generation, data validity proof veriﬁcation, and ﬁnal output computation are abstracted as functions. Two constructions of SecRand are presented, focusing on the abstracted processes. The advantage of Abias is deﬁned as P = \| Pr[opt ∈S] − s/n\|, where opt is the output of any honest participant, s = \|S\|, and n = \|N\|. 3) BIAS RESISTANCE lenger chooses b ∈{0, 1} uniformly at random and sends the veriﬁer all the public information together with xb. ′ We say that a DRG protocol achieves bias resistance if a PPT adversary Abias has a negligible advantage in the following game played against a challenger: • The veriﬁer gives a guess b ′ ∈{0, 1}. • The veriﬁer gives a guess b ′ ∈{0, 1}. The advantage of APV is deﬁned as 1 −Pr[b′ = b]. The advantage of APV is deﬁned as 1 −Pr[b′ = b]. • The challenger creates secret keys for honest partici- pants and sends the corresponding public keys to Abias. • The challenger creates secret keys for honest partici- pants and sends the corresponding public keys to Abias. A. PROTOCOL DESCRIPTION It is assumed that G is a cyclic subgroup of an elliptic curve with order q, of which g and h are two independently chosen generators. C denotes the Reed–Solomon code correspond- ing to the (n, t)-Shamir’s Secret Sharing, and C⊥is its dual code. SecRand is run among n participants that are denoted as P1, P2, . . . , Pn. Each participant Pi has a keypair ski, pki , where the private key ski ∈Zq and the public key pki = ski · g. Private keys are secretly kept, whereas public keys are published and known publicly. • The challenger creates secret keys for honest partic- ipants and sends the corresponding public keys to AG.O.D.. • AG.O.D. creates secret keys for corrupted participants and sends the corresponding public keys to the chal- lenger. • The challenger runs the DRG on behalf of the honest participants and sends AG.O.D all the public information generated in the protocol’s execution, while AG.O.D. runs the DRG on behalf of the corrupted participants. For simplicity, we introduce the following notation. We denote uniform sampling of a random element from a ﬁnite set D by ←R D, while assigning is denoted by ←. We use a dot operation to represent scalar multiplication on an elliptic curve. GenPf () and VerPf () denote the generation and veriﬁcation process of the data validity proof, respectively. The ﬁnal output computation is denoted as FinalOpt(). • By the end of the DRG, each honest participant obtains an output. The advantage of AG.O.D. is deﬁned as Pr[opt1̸=opt2], where opt1 and opt2 are the outputs of any two different honest participants. 4) GUARANTEED OUTPUT DELIVERY We say that a DRG protocol achieves G.O.D. if a PPT adver- sary AG.O.D. has a negligible advantage in the following game played against a challenger: 5) PUBLIC VERIFIABILITY The Distribution phase is divided into three steps. In Step 1, each participant chooses a random number as a secret and conducts (n, t)-Shamir’s Secret Sharing. In Step 2, the secret shares are committed by scalar multiplication of h and encrypted with the receivers’ public keys. In addition, proofs are generated to ensure that encrypted shares are identical to the committed shares. In Step 3, all the commitments, encrypted shares, and corresponding proofs are uploaded to the public bulletin board. The procedure is shown in Algo- rithm 1. We say that a DRG protocol achieves public veriﬁability if a PPT adversary APV has a negligible advantage in the following game played against a challenger and a veriﬁer: • The challenger creates secret keys for honest partici- pants and sends the corresponding public keys to APV . • APV creates secret keys for corrupted participants and sends the corresponding public keys to the challenger. • The challenger runs the DRG on behalf of honest par- ticipants and sends APV all the public information gen- erated in the protocol’s execution, while APV runs the DRG on behalf of the corrupted participants. B. SECURITY DEFINITION Based on the formalizations above, we deﬁne the security of the DRG protocols. • Abias creates secret keys for corrupted participants and sends the corresponding public keys to the challenger. • The challenger randomly chooses a subset S of the out- put space set N. 1) SECURITY FOR DRG PROTOCOL 1) SECURITY FOR DRG PROTOCOL • The challenger runs the DRG on behalf of the honest participants and sends Abias all the public information generated in the protocol’s execution, while Abias runs the DRG on behalf of the corrupted participants. A DRG protocol is secure if and only if (iff) it achieves the properties of availability, bias resistance, unpredictability, G.O.D., and public veriﬁability. • By the end of the DRG, each honest participant obtains an output. Algorithm 1 Distribution public bulletin board. The invalid data will be ﬁltered out and thus do not affect the protocol’s execution. Algorithm 1 Distribution Each participant executes the following procedure (consider Pi as an example): procedure Distribution si ←R Z∗ q for j < t do ai,j ←R Z∗ q end for fi (x) ←si + ai,1x + . . . + ai,t−1xt−1 for j ≤n do si,j ←fi (j) ▷secret shares ci,j ←si,j · h ▷commit to secret shares ˆsi,j ←si,j · pkj ▷encrypt secret shares pf i,j ←GenPf (ci,j, ˆsi,j) ▷generate validity proofs end for Cmti ←(ci,1, ci,2, . . . , ci,n) EncSi ←(ˆsi,1, ˆsi,2, . . . , ˆsi,n) PFi ←(pf i,1, pf i,2, . . . , pf i,n) UPDi ←(Cmti, EncSi, PFi) Upload(UPDi) end procedure Each participant executes the following procedure (consider Pi as an example): Each participant executes the following procedure (consider Pi as an example): procedure Distribution si ←R Z∗ q for j < t do ai,j ←R Z∗ q end for fi (x) ←si + ai,1x + . . . + ai,t−1xt−1 for j ≤n do si,j ←fi (j) ▷secret shares ci,j ←si,j · h ▷commit to secret shares ˆsi,j ←si,j · pkj ▷encrypt secret shares pf i,j ←GenPf (ci,j, ˆsi,j) ▷generate validity proofs end for Cmti ←(ci,1, ci,2, . . . , ci,n) EncSi ←(ˆsi,1, ˆsi,2, . . . , ˆsi,n) PFi ←(pf i,1, pf i,2, . . . , pf i,n) UPDi ←(Cmti, EncSi, PFi) Upload(UPDi) Pi as an example): Pi as an example): procedure Distribution si ←R Z∗ q for j < t do ai,j ←R Z∗ q end for fi (x) ←si + ai,1x + . . . + ai,t−1xt−1 for j ≤n do si,j ←fi (j) ▷secret shares ci,j ←si,j · h ▷commit to secret shares ˆsi,j ←si,j · pkj ▷encrypt secret shares pf i,j ←GenPf (ci,j, ˆsi,j) ▷generate validity proofs end for Algorithm 3 Reconstruction Each participant executes the following procedures (consider Pi as an example): procedure GenerateGSecret for j ≤t do gssj, pf j ←UPGj end for GSecret ←Pt i=1 (gssi · Qt j=1,j̸=i j j−i) ▷group secret end procedure Algorithm 2 Veriﬁcation Public bulletin board veriﬁes the uploaded data as follows (consider UPDi as an example): procedure Veriﬁcation (Cmti, EncSi, PFi) ←UPDi c⊥←R C⊥ ▷select dual code resulti ←⟨Cmti, c⊥⟩ ▷compute inner product if resulti = O then ▷commitment is valid ﬂag ←True for j ≤n do if VerPf (ci,j, ˆsi,jpf i,j) then ▷verify proofs continue else ﬂag ←False end if end for else ﬂag ←False end if return ﬂag end procedure Algorithm 2 Veriﬁcation Public bulletin board veriﬁes the uploaded data as follows (consider UPDi as an example): Algorithm 2 Veriﬁcation Public bulletin board veriﬁes the uploaded data as follows (consider UPDi as an example): procedure Veriﬁcation (Cmti, EncSi, PFi) ←UPDi c⊥←R C⊥ ▷select dual code resulti ←⟨Cmti, c⊥⟩ ▷compute inner product if resulti = O then ▷commitment is valid ﬂag ←True for j ≤n do if VerPf (ci,j, ˆsi,jpf i,j) then ▷verify proofs continue else ﬂag ←False end if end for else ﬂag ←False end if return ﬂag end procedure procedure Veriﬁcation (Cmti, EncSi, PFi) ←UPDi c⊥←R C⊥ ▷select dual code resulti ←⟨Cmti, c⊥⟩ ▷compute inner product if resulti = O then ▷commitment is valid ﬂag ←True for j ≤n do if VerPf (ci,j, ˆsi,jpf i,j) then ▷verify proofs continue else ﬂag ←False end if end for else ﬂag ←False end if return ﬂag end procedure procedure GenerateOpt ς ←FinalOpt(GSecret)▷the ﬁnal output return ς procedure GenerateOpt ς ←FinalOpt(GSecret)▷the ﬁnal output return ς procedure GenerateOpt ς ←FinalOpt(GSecret)▷the ﬁnal output return ς end procedure We remark that UPGi will be veriﬁed by the public bulletin board in a similar way to that used for UPDi to ensure the decrypted shares and group secret shares are correctly computed. 2) VERIFICATION • Suppose x0 is the output of the DRG, and x1 is chosen by APV in the DRG output space. Furthermore, the chal- In the Veriﬁcation phase, the public bulletin board veriﬁes UPDi, which can be decoded as (Cmti, EncSi, PFi). The VOLUME 8, 2020 203921 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability Algorithm 1 Distribution Each participant executes the following procedure (consider Pi as an example): procedure Distribution si ←R Z∗ q for j < t do ai,j ←R Z∗ q end for fi (x) ←si + ai,1x + . . . + ai,t−1xt−1 for j ≤n do si,j ←fi (j) ▷secret shares ci,j ←si,j · h ▷commit to secret shares ˆsi,j ←si,j · pkj ▷encrypt secret shares pf i,j ←GenPf (ci,j, ˆsi,j) ▷generate validity proofs end for Cmti ←(ci,1, ci,2, . . . , ci,n) EncSi ←(ˆsi,1, ˆsi,2, . . . , ˆsi,n) PFi ←(pf i,1, pf i,2, . . . , pf i,n) UPDi ←(Cmti, EncSi, PFi) Upload(UPDi) end procedure Algorithm 1 Distribution 3) RECONSTRUCTION The Reconstruction phase is also divided into three steps. In Step 1, each participant decrypts the secret shares received from other participants and adds up all the decrypted shares to obtain its group secret share, which will be uploaded to the public bulletin board. In Step 2, each participant reconstructs the complete group secret using t valid group secret shares through Lagrange interpolation. In Step 3, each participant takes the complete group secret as the input and computes the ﬁnal output. The procedure is shown in Algorithm 3. Cmti ←(ci,1, ci,2, . . . , ci,n) EncSi ←(ˆsi,1, ˆsi,2, . . . , ˆsi,n) PFi ←(pf i,1, pf i,2, . . . , pf i,n) UPDi ←(Cmti, EncSi, PFi) Upload(UPDi) end procedure Cmti ←(ci,1, ci,2, . . . , ci,n) EncSi ←(ˆsi,1, ˆsi,2, . . . , ˆsi,n) PFi ←(pf i,1, pf i,2, . . . , pf i,n) UPDi ←(Cmti, EncSi, PFi) Cmti ←(ci,1, ci,2, . . . , ci,n) EncSi ←(ˆsi,1, ˆsi,2, . . . , ˆsi,n) PFi ←(pf i,1, pf i,2, . . . , pf i,n) UPDi ←(Cmti, EncSi, PFi) Algorithm 3 Reconstruction Each participant executes the following procedures (consider Pi as an example): Algorithm 3 Reconstruction Each participant executes the following procedures (consider Pi as an example): Algorithm 3 Reconstruction Each participant executes the following procedures (consider Pi as an example): procedure GenerateGss for j ≤k do (Cmtj, EncSj, PFj) ←UPDj (ˆsj,1, ˆsj,2, . . . , ˆsj,n)←EncSj ˜sj,i ←sk−1 i · ˆsj,i ▷decrypt secret share end for ▷˜sj,i = sj,i · g gssi ←˜s1,i + . . . + ˜sk,i ▷group secret share encssi ←ˆs1,i + . . . + ˆsk,i pf i ←GenPf (gssi, encssi) UPGi ←(gssi, pf i) Upload(UPGi) end procedure procedure GenerateGSecret for j ≤t do gssj, pf j ←UPGj end for GSecret ←Pt i=1 (gssi · Qt j=1,j̸=i j j−i) ▷group secret end procedure validity of Cmti is veriﬁed through the Reed-Solomon code, which implies that the (n, t)-Shamir’s Secret Sharing shares committed are correctly computed. The validity of EncSi is veriﬁed by proofs (i.e., PFi), which ensures ci,j and ˆsi,j are scalar multiples of h and pkj with the same si,j. The procedure is shown in Algorithm 2. FinalOpt (GSecret) = H(GSecret). • For t ≤i ≤n, the challenger chooses uniformly random values ski from Zq, computes pki = ski·h, and sends pki to Apre, which can be thought of deﬁning the secret key as ski a. B. CONSTRUCTIONS In this subsection, we give two constructions of SecRand, namely, SecRand-ZK and SecRand-Pairing. They have the same workﬂow as that described in the previous section. Their differences lie in the realizations of GenPf (), VerPf (), and FinalOpt(), which are abstracted as functions in the protocol description. With H as a secure hash function, we present the technical details below. Without loss of generality, we assume the valid uploaded data are {UPD1, UPD2, . . . , UPDk}. We remark that k ≥t is satisﬁed for the (n, t) setting, as the adversary is not able to prevent honest participants from uploading valid data to the VOLUME 8, 2020 203922 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability 2) SECRAND-PAIRING SecRand-Pairing relies on the bilinearity of pairings for data veriﬁcation. No special proofs must be generated. GenPf () and VerPf () are expressed as follows: • Apre runs the Distribution phase on behalf of t −1 corrupted participants. For 1 ≤ i ≤ t −1, Apre randomly selects secret si and performs (n, t)-Shamir’s Secret Sharing. The generated commitments, encrypted shares, and DLEQ proofs are as follows: GenPf ci,j, ˆsi,j = ⊥, VerPf ci,j, ˆsi,jpf i,j = True, if e ci,j, pkj = e h, ˆsi,j False, if e(ci,j, pkj) ̸= e(h, ˆsi,j). As for the ﬁnal output, FinalOpt() is expressed as As for the ﬁnal output, FinalOpt() is expressed as FinalOpt (GSecret) = H(e(g, GSecret)).    Cmt1 = s1,1 · h, . . . , s1,n · h EncS1 = s1,1 · pk1, . . . , s1,n · pkn ZKPF1 = pf 1,1, . . . , pf 1,n , (1)    Cmt2 = s2,1 · h, . . . , s2,n · h EncS2 = s2,1 · pk1, . . . , s2,n · pkn ZKPF2 = pf 2,1, . . . , pf 2,n , (2) · · · · · · · · · · · ·    Cmtt−1 = st−1,1 · h, . . . , st−1,n · h EncSt−1 = st−1,1·pk1, . . . , st−1,n · pkn ZKPFt−1 = pf t−1,1, . . . , pf t−1,n . (3)    Cmt1 = s1,1 · h, . . . , s1,n · h EncS1 = s1,1 · pk1, . . . , s1,n · pkn ZKPF1 = pf 1,1, . . . , pf 1,n , (1)    Cmt2 = s2,1 · h, . . . , s2,n · h EncS2 = s2,1 · pk1, . . . , s2,n · pkn ZKPF2 = pf 2,1, . . . , pf 2,n , (2) · · · · · · · · · · · ·    Cmtt−1 = st−1,1 · h, . . . , st−1,n · h EncSt−1 = st−1,1·pk1, . . . , st−1,n · pkn ZKPFt−1 = pf t−1,1, . . . , pf t−1,n . (3) (1) FinalOpt (GSecret) = H(e(g, GSecret)). V. SECURITY ANALYSIS (2) In this section, we provide strict proofs showing SecRand’s security. Our idea is to ﬁrst prove SecRand achieves the desired properties. We then naturally arrive at the conclu- sion that SecRand is secure according to the deﬁnition in Sec. III-B. (3) Lemma 1: SecRand achieves availability against a PPT adversary. Apre uploads the generated data to the public bul- letin board. Without loss of generality, we assume the uploaded data are valid, for the invalid data will be ﬁltered out in the Veriﬁcation phase. Proof: The only way for Ahalt to halt SecRand is to prevent valid group secret shares from being uploaded to the public bulletin board in the Reconstruction phase. Thus, the protocol will halt due to a lack of sufﬁcient group secret shares to reconstruct the complete group secret. Under our adversary model, Ahalt cannot prevent honest participants from accessing the public bulletin board. Thus, even if all the corrupted participants controlled by Ahalt refuse to par- ticipate in the protocol, there are still no fewer than t valid group secret shares being uploaded in the Reconstruction phase by honest participants that follow the protocol strictly. The honest participants are then able to compute the complete group secret through Lagrange interpolation and obtain the ﬁnal output. For any honest participant’s output opt, we have Pr [opt = ⊥] = 0. Thus, SecRand achieves availability. • The challenger runs the Distribution phase on behalf of n −t + 1 honest participants. Speciﬁcally, for t ≤ i ≤n −1, the challenger randomly selects secret si, performs (n, t)-Shamir’s Secret Sharing, and generates the corresponding commitments, encrypted shares, and DLEQ proofs:    Cmtt = st,1 · h, . . . , st,n · h EncSt = st,1 · pk1, . . . , st,n · pkn ZKPFt = pf t,1, . . . , pf t,n , (4)    Cmtt+1 = st+1,1 · h, . . . , st+1,n · h EncSt+1 = st+1,1 · pk1, . . . , st+1,n · pkn ZKPFt+1 = pf t+1,1, . . . , pf t+1,n , (5) · · · · · · · · · · · ·    Cmtn−1 = sn−1,1 · h, . . . , sn−1,n · h EncSn−1 = (sn−1,1 · pk1, . . . VOLUME 8, 2020 1) SECRAND-ZK Regarding SecRand-ZK, we show that if there exists an adversary Apre that can break the unpredictability property of SecRand-ZK, then there exists an adversary ACDH that can use Apre to break the CDH assumption with the same advan- tage. Without loss of generality, we assume Apre corrupts the ﬁrst t −1 participants. ) SecRand-ZK uses zero-knowledge proof for data veriﬁcation. We use a previously reported DLEQ [21], so GenPf () and VerPf () are expressed as GenPf ci,j, ˆsi,j = DLEQ(h, ci,j, pkj, ˆsi,j), VerPf pf i,j = DLEQ_Verify(h,ci,j, pkj, ˆsi,j, pf i,j). Letting (G, a·G, b·G) be an instance of the CDH problem, ACDH can simulate an unpredictability game as follows: As for the ﬁnal output, it is directly computed from the group secret by a hash function. Thus, FinalOpt() is realized as • The challenger sets g = a · G, h = G. • The challenger sets g = a · G, h = G. • For 1 ≤i ≤t−1, Apre chooses uniformly random values ski from Zq, computes pki = ski ·g, and sends pki to the challenger. FinalOpt (GSecret) = H(GSecret). V. SECURITY ANALYSIS , sn−1,n · pkn) ZKPFn−1 = (pf n−1,1, . . . , pf n−1,n). (6) (4) Lemma 2: SecRand achieves unpredictability against a PPT adversary. Proof: We make a distinction between SecRand-ZK and SecRand-Pairing in this proof due to their differences in data validity veriﬁcation and ﬁnal output computation. We ﬁrst prove SecRand-ZK achieves unpredictability under the CDH assumption, and then we prove SecRand-Pairing’s unpredictability under the DBS assumption. (6) We next focus on constructing participant n’s data. The challenger ﬁrst randomly chooses t −1 elements VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability sn,1, sn,2, . . . ,sn,t−1 as secret shares for corrupted par- ticipants. The challenger then computes sn · h as Thus, the advantage of ADBS is the same as that of Apre, and SecRand achieves unpredictability regardless of the con- struction. sn · h = b · h − Xn−1 i=1 (si · h). (7) (7) Lemma 3: If the output of the hash function is uniformly distributed, SecRand achieves bias resistance against a PPT adversary. For t ≤i ≤n, the challenger generates the commit- ments p(i) · h, where p(x) is the unique polynomial of degree at most t determined by Proof: The attack of Abias can be divided into two distinct cases: p (0) = b − Xn−1 i=1 si, p (i) = sn,i, for 1 ≤i ≤t −1. • Case 1: Abias does not know the preimage of any element in S. In this case, Abias can only participate in the protocol blindly without knowing the effects on the ﬁnal output. Because of the uniform distribution of the hash function’s output, Pr [opt ∈S] = s/n. Thus, the advan- tage of Abias is 0. We remark that the challenger does not know p (0), but it does know p (0)·h = b·h−Pn−1 i=1 (si · h), which can be successfully computed. After having (p (0) · h, p (1) · h, . . . , p (t −1) · h), the challenger can use Lagrange interpolation in the scalar to compute the expression of p(x) · h and obtain (p (t) · h, p (t + 1) · h, . . . , p (n) · h). For t ≤i ≤n, it also creates the encrypted shares as follows: • Case 2: Abias knows the preimage of some element in S. V. SECURITY ANALYSIS By selecting data according to Apre’s outputs, ACDH makes (ab)·G as the group secret. Thus, the advantage of using ACDH to solve the CDH problem is the same as the advantage of Apre, breaking the unpredictability of SecRand-ZK. • Type 1: loghci,j ̸= logpkjˆsi,j, for some j ∈[t, n] and Pj is an honest participant. • Type 1: loghci,j ̸= logpkjˆsi,j, for some j ∈[t, n] and Pj is an honest participant. p p p y y g g Apre’s outputs, ACDH makes (ab)·G as the group secret. Thus, the advantage of using ACDH to solve the CDH problem is the same as the advantage of Apre, breaking the unpredictability of SecRand-ZK. • Type 2: loghci,j = logpkjˆsi,j for 1 ≤j ≤n, but si,1, si,2, . . . , si,n do not constitute valid sharing of some secret with the (n, t)-Shamir’s Secret Sharing scheme. Regarding SecRand-Pairing, we show that if there exists an adversary Apre that can break the unpredictability property of the SecRand-Pairing, then there exists an adversary ADBS that can use Apre to break the DBS assumption with the same advantage. • Type 3: gssi ̸= Pk j=1 ˜smj,i. • Type 3: gssi ̸= Pk j=1 ˜smj,i. For SecRand-ZK, according to SCRAPE [16, Theorem 2], incorrect data of Type 1 and Type 2 are ﬁltered out by the public bulletin board with a probability of at least 1 −ε − 1 q. Type 3 incorrect data are ﬁltered out with probability 1 −ε, where ε is the soundness error of the DLEQ. Thus, a corrupted participant succeeds in making the outputs of two honest participants different with a probability of no more than 2ε + 1 q, which is negligible. Finally, AG.O.D. controlling t −1 corrupted participants has an advantage less than (t −1)(2ε + 1 q) in the G.O.D. game, which is also negligible. Letting (G,aG, bG, T) be an instance of the DBS problem, the task of ADBS is to guess whether T = e(aG, aG)b or T = e(aG, aG)r for a random r ∈Zq. ADBS simulates the unpredictability game between the challenger and Apre following exactly the same steps as those followed by ACDH in the proof of SecRand-ZK’s unpredictability, except the input of the hash function is e(g, g)b = e(aG,aG)b. Before the Reconstruction phase, Apre outputs a prediction x. V. SECURITY ANALYSIS To maximize its advantage, Abias tries to bias the complete group secret toward this preimage. Assuming the probability that the complete group secret equals this preimage is P, we have P < Ppre, where Ppre is the advantage of Apre. This is because once Abias succeeds in biasing the complete group secret to this preimage, it breaks the unpredictability at the same time. In this case, \|Pr [opt∈S] −s/n\| = P < Ppre. From Lemma 2, we know Ppre is negligible, so the advantage of Abias is also negligible. sn,i · pki = p (i) · (ski · h) = ski · (p (i) · h). (8) (8) From all the computed values in Eq. 7 and 8, the chal- lenger generates the corresponding DLEQ proof. Thus, the challenger successfully generates data for participant n as follows:    Cmtn = (c1, . . . , ct−1, ct, . . . , cn) EncSn = (es1, . . . , est−1, est, . . . , esn) ZKPFn = pf n,1, pf n,2, . . . , pf n,n , (9)    Cmtn = (c1, . . . , ct−1, ct, . . . , cn) EncSn = (es1, . . . , est−1, est, . . . , esn) ZKPFn = pf n,1, pf n,2, . . . , pf n,n , (9) (9) Abias has a negligible advantage in both cases, and thus, SecRand achieves bias resistance. Lemma 4: SecRand achieves G.O.D. against a PPT adver- sary. where ci = sn,i · h and esi = sn,i · pki for 1 ≤i ≤t −1, and ci = p(i) · h and esi = ski · (p (i) · h) for t ≤i ≤n. Proof: In the G.O.D. game, the task of AG.O.D. is to make two honest participants obtain different outputs by the end of the protocol. The only possible way to achieve this is sending incorrect data during the Distribution and Reconstruct phases. Basically, a corrupted participant Pi may send three types of incorrect data: • Before the Reconstruction phase, Apre outputs x as its prediction. • Before the Reconstruction phase, Apre outputs x as its prediction. ACDH solves the CDH problem by taking x as (ab) · G. We remark that the ﬁnal output value generated by the protocol equals the sum of the secrets determined by each participant independently. B. PERFORMANCE ANALYSIS OF SECRAND We remark that SecRand achieves the same scalability as SCRAPE due to its technical design, so experiments of the scalability were not specially conducted. Besides, the com- munication complexity of our protocol is O(n) because par- ticipants exchange data through a public bulletin board and do not need to communicate directly with each other. In this subsection, we deploy SecRand-ZK and SecRand- Pairing with the (n, t) setting (a total of n participants and at least t participants are required to reconstruct the ﬁnal output). We set n from 10 to 100, and t = n 2 . In the Distribution and Veriﬁcation phases, we analyzed the exe- cution times of n participants’ data generation and veriﬁ- cation, whereas for the Reconstruction phase, we analyzed the execution times of t participants’ data decryption and veriﬁcation, which was sufﬁcient for computing the ﬁnal output. VI. PERFORMANCE ANALYSIS In this section, we present a detailed performance analysis of SecRand, which is divided into three parts. In the ﬁrst part, we provide implementation details, including the computing environment, underlying cryptographic operations, and their corresponding time costs and space occupation. In the second part, we focus on the runtime and space occupation in each phase as the number of participants increases and compare the differences between using zero-knowledge for data ver- iﬁcation and using pairing for data veriﬁcation. In the third part, a performance comparison is made between SecRand and SCRAPE [16] in the presence of corrupted participants. The space occupation of the underlying cryptographic data structure is shown in Table 2. We used a compressed form to represent elliptic curve points, which used almost half the space of the uncompressed form. The big num- bers used in our protocol are elements of the prime ﬁeld generated by the order of the elliptic curve group. The DLEQ proof contained two elliptic curve points and one big number. Theorem 1: SecRand is a secure DRG protocol. Theorem 1: SecRand is a secure DRG protocol Proof: From Lemmas 1–5, we know SecRand achieves properties of availability, bias resistance, unpredictability, guaranteed output delivery, and public veriﬁability. Thus, it is a secure DRG protocol according to the deﬁnition in Sec. III-B. The time cost of performing the underlying cryptographic operations will inﬂuence the execution time of the proto- col. Therefore, in this subsection, we study the time cost of the cryptographic operations. For each operation, we performed 1000 experiments to obtain the average time cost, and the values are summarized in Table 1. VI. PERFORMANCE ANALYSIS V. SECURITY ANALYSIS We deployed them on a Windows 10 OS laptop with two cores of a 2.50 GHz Intel(R) Core(TM) i7-7660U CPU and 8 GB of RAM. We utilized the GNU Multiple Preci- sion (GMP) Arithmetic Library [26] to perform big num- ber computations and used the Pairing-Based Cryptograph (PBC) library [27] to compute point multiplication of elliptic curves and pairings. The pairing used was the PBC Type A pairing, which is symmetric and constructed on Barreto- Naehrig (BN) curve y2 = x3 + x over a prime ﬁeld of 512 bits. mentations of SecRand-ZK, SecRand-Pairing, and SCRAPE. We deployed them on a Windows 10 OS laptop with two cores of a 2.50 GHz Intel(R) Core(TM) i7-7660U CPU and 8 GB of RAM. We utilized the GNU Multiple Preci- sion (GMP) Arithmetic Library [26] to perform big num- ber computations and used the Pairing-Based Cryptograph (PBC) library [27] to compute point multiplication of elliptic curves and pairings. The pairing used was the PBC Type A pairing, which is symmetric and constructed on Barreto- Naehrig (BN) curve y2 = x3 + x over a prime ﬁeld of 512 bits. V. SECURITY ANALYSIS If the prediction x = T, then ADBS guesses that T = e(aG, aG)b. If the prediction x ̸= T, then ADBS guesses that T = e(aG,aG)r. For SecRand-Pairing, e(ci,j, pkj) = e(h, ˆsi,j) is satisﬁed iff ˆsi,j is valid, and e (hssi, g) = e(h, gssi) is satisﬁed iff gssi = Pk j=1 ˜smj,i. Thus, Type 1 and Type 3 incorrect data will 203924 VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability TABLE 1. Average time cost of cryptographic operations. TABLE 1. Average time cost of cryptographic operations. be ﬁltered out by the public bulletin board with a probability of 1, whereas Type 2 incorrect data will be ﬁltered out with a probability of (1 −1 q). Thus, a corrupted participant succeeds with a probability no more than 1 q, which is negligible. Finally, AG.O.D. has an advantage of less than (t −1)/q in the G.O.D. game, which is also negligible. Lemma 6: SecRand achieves public veriﬁability against a PPT adversary. Proof: From the protocol description, we know that all the data sent during the execution are uploaded to the public bulletin board, including the data needed to compute the ﬁnal output. Furthermore, the validity of the data can be veriﬁed by public information, which can be accessed by both par- ticipants and the veriﬁer. Thus, the participants of SecRand do not have any advantage over the veriﬁer in verifying the validity of the ﬁnal output. Assuming APV succeeds in confusing the veriﬁer with a fake output, the data used in generating this fake output must have passed the veriﬁcation of the public bulletin board. APV may control a corrupted par- ticipant to send these data during the protocol, which causes a fake output to be the output of an honest participant. In this scenario, APV breaks the property of the G.O.D. We then know that breaking public veriﬁability means breaking the G.O.D. at the same time. Thus, the advantage of APV is no more than the advantage of AG.O.D., which was proven to be negligible in Lemma 4. Therefore, the advantage of APV is also negligible. TABLE 2. Space occupation of cryptographic data structure. TABLE 2. Space occupation of cryptographic data structure. mentations of SecRand-ZK, SecRand-Pairing, and SCRAPE. A. IMPLEMENTATION We set n from 10 to 100 and t = n/2 to compare the runtime as well as the space occupation of SecRand-ZK and SecRand-Pairing in the Distribution phase. 2) VERIFICATION PHASE In this phase, the two protocols differ in the method of verify- ing the encrypted shares. SecRand-ZK uses the DLEQ proof to verify the encrypted shares, whereas SecRand-Pairing uses pairing for veriﬁcation. The computation complexities are detailed in Table 5. According to the experimental results in Fig. 2, the SecRand-Pairing had a longer runtime than the SecRand-ZK because the BN-curve pairing was com- putationally expensive. This overhead was around 35% and remained stable as the number of participants increased. In summary, SecRand-ZK had a larger space occupation because of the usage of the DLEQ proof, whereas SecRand- Pairing had a higher computational resource occupation because of the pairing computation. Thus, SecRand-ZK is more suitable for computing-resource-limited scenarios, and SecRand-Pairing is more suitable in space-limited scenarios. A. IMPLEMENTATION To evaluate the performance of our proposed protocol, we have conducted experiments with prototype C imple- 203925 203925 VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability TABLE 5. Computational complexity comparison of SecRand-ZK and SecRand-Pairing in the Verification phase. FIGURE 2. Performance comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. We set n from 10 to 100 and t = n/2 to compare the runtimes of SecRand-ZK and SecRand-Pairing in the Verification phase. TABLE 6. Computation complexity comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. TABLE 7. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. TABLE 5. Computational complexity comparison of SecRand-ZK and SecRand-Pairing in the Verification phase. FIGURE 2. Performance comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. We set n from 10 to 100 and t = n/2 to compare the runtimes of SecRand-ZK and SecRand-Pairing in the Verification phase. TABLE 3. Computational complexity comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. TABLE 4. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. FIGURE 1. Performance comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. We set n from 10 to 100 and t = n/2 to compare the runtime as well as the space occupation of SecRand-ZK and SecRand-Pairing in the Distribution phase. TABLE 5. Computational complexity comparison of SecRand-ZK and SecRand-Pairing in the Verification phase. TABLE 3. Computational complexity comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. TABLE 4. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. TABLE 4. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. TABLE 4. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. FIGURE 2. Performance comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. We set n from 10 to 100 and t = n/2 to compare the runtimes of SecRand-ZK and SecRand-Pairing in the Verification phase. TABLE 6. Computation complexity comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. TABLE 7. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. TABLE 6. Computation complexity comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. TABLE 6. Computation complexity comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. FIGURE 1. Performance comparison of SecRand-ZK and SecRand-Pairing in the Distribution phase. 1) DISTRIBUTION PHASE In the Distribution phase, each participant of SecRand-ZK must generate n DLEQ proofs for data validation, which are not required in SecRand-Pairing. Concrete computation complexity and space occupation are detailed in Tables 3 and 4, which show SecRand-ZK had the additional compu- tation steps of 2n2 BN-curve point multiplications and 150n2 bytes of extra space occupation compared to the SecRand-Pairing. Furthermore, according to experimental results shown in Fig. 1, SecRand-Pairing exhibited a better performance than SecRand-ZK in the Distribution phase, and the performance advantage grew exponentially as the number of participants increased. TABLE 7. Space occupation comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. pairing in SecRand-Pairing. The computation complexities and space occupations are detailed in Tables 6 and 7, respec- tively. Regarding the space occupation, SecRand-ZK had 150n more bytes of space occupation than SecRand-Pairing. Regarding computation complexity, SecRand-ZK had 7n BN- curve point multiplications, whereas SecRand-Pairing had n BN-curve point multiplications and 2n BN-curve pairings. According to the experimental results in Fig. 3, the two pro- tocols had almost the same runtimes in this phase. However, SecRand-Pairing was superior to SecRand-ZK in terms of the space occupation, and this advantage grew linearly as the number of participants increased. C. PERFORMANCE COMPARISON WITH SCRAPE In this subsection, we present a performance comparison of SecRand and SCRAPE [16] from the aspects of runtime and space occupation. We focus on the Reconstruction phase, as the Distribution and Veriﬁcation phases of SecRand are exactly the same as those of SCRAPE. Without loss of gen- erality, we use SecRand-Pairing for comparison. VII. CONCLUSION In this article, we focused on distributed randomness genera- tion (DRG) protocols, which are a key component in decen- tralized applications. We ﬁrst formalized the desired proper- ties of a secure DRG protocol and built a security model using these formal deﬁnitions. To the best of our knowledge, this is the ﬁrst work to build a security model for DRG. We then pre- sented SecRand, a secure distributed randomness generation protocol with high practicality and scalability. SecRand is a variant of SCRAPE, with the same scalability, but it improves on SCRAPE by modifying the method of computing the ﬁnal output, which results in resistance of the adversary’s malicious behavior of withholding its secret share. Further- more, we provided strict proofs under our security model, showing that SecRand achieves the desired properties and is a secure DRG protocol. We implemented SecRand in the C language and presented a detailed performance evaluation. The experimental data showed that SecRand achieved a bet- ter performance than that of SCRAPE in the presence of corrupted participants, and this performance advantage grew linearly with the number of corrupted participants. In SecRand, the ﬁnal output is computed through Lagrange interpolation as long as no fewer than t valid group secret shares are published in the Reconstruction phase. Thus, the execution ﬂow of SecRand is not inﬂuenced by an adver- sary’s malicious behavior of withholding its group secret share. In SCRAPE, participants are required to publish their group secret shares in the Reveal phase. When a participant fails to publish its group secret share, the remaining partici- pants must work together to recompute the unpublished group secret share through Lagrange interpolation. This procedure results in extra consumption of computational resources and space. Thus, the adversary may take advantage of this and lower the performance of SCRAPE by having the corrupted participants behave honestly in the Distribution phase but refusing to reveal their group secret shares in the Reveal phase. Theoretically, with the (n, t) setting (a total of n partic- ipants and at least t participants are needed to reconstruct the ﬁnal output), SecRand should exhibit a better performance than SCRAPE with a probability of 1 −(1 −p)t−1, and the average performance advantage is p(t −1)0, where p is the probability of participants being corrupted, 0 is the consumption of computing resources and space in recomput- ing one unpublished group secret share. The details are in Appendix 1. 3) RECONSTRUCTION PHASE In the Reconstruction phase, each participant of SecRand- ZK must generate a DLEQ proof to prove the validity of the decrypted group secret shares, which are veriﬁed by 203926 VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability FIGURE 3. Performance comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. We set n from 10 to 100 and t = n/2 to compare the runtime as well as the space occupation of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. FIGURE 4. Performance comparison of SecRand-Pairing and SCRAPE in the Reconstruction phase. With n = 20 and t = 11, we set the number of corrupted participants from 0 to 10 to compare the runtimes and space occupations of SecRand-Pairing and SCRAPE in the Reconstruction phase. FIGURE 4. Performance comparison of SecRand-Pairing and SCRAPE in the Reconstruction phase. With n = 20 and t = 11, we set the number of corrupted participants from 0 to 10 to compare the runtimes and space occupations of SecRand-Pairing and SCRAPE in the Reconstruction phase. FIGURE 4. Performance comparison of SecRand-Pairing and SCRAPE in the Reconstruction phase. With n = 20 and t = 11, we set the number of corrupted participants from 0 to 10 to compare the runtimes and space occupations of SecRand-Pairing and SCRAPE in the Reconstruction phase. FIGURE 3. Performance comparison of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. We set n from 10 to 100 and t = n/2 to compare the runtime as well as the space occupation of SecRand-ZK and SecRand-Pairing in the Reconstruction phase. that of SCRAPE in the presence of corrupted participants, and this performance advantage grew linearly as the number of corrupted participants increased. VOLUME 8, 2020 APPENDIX In the (n, t) setting, n participants and at least t participants are needed to compute the ﬁnal output. With our adversary model, up to t −1 participants are vulnerable to corrup- tion with a probability of p. The consumption of computing resources and space in recomputing one unpublished group secret share is denoted as 0. We use Advi to represent the performance advantage when the number of corrupted partic- ipants is i, and the corresponding possibility is pi. The expec- tation of SecRand’s performance advantage over SCRAPE is denoted as EAdv(m), where m is the maximum number of corrupted participants, and m = t −1 in the (n, t) setting. We experiment by setting n = 20 and t = 11, which means up to 10 participants are vulnerable to corruption due to the honest majority assumption. Our experimental results showed that SecRand’s performance advantage over SCRAPE grew linearly with the number of corrupted participants, as shown in Figure 4. In summary, SecRand achieved a more stable performance than that of SCRAPE due to the resistance of the adver- sary’s malicious behavior of withholding its secret share. Furthermore, SecRand exhibited a better performance than The number of corrupted participants satisﬁes a binomial distribution, and thus, we have the following: Advi = i0, 203927 203927 VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability pi = pi(1 −p)m−i. [4] K. Croman, C. Decker, I. Eyal, and A. E. Gencer, ‘‘On scaling decentralized blockchains,’’ in Proc. 20th Annu. Int. Conf. Fin. Cryptograph. Data Secur. (FC), Christ Church, Barbados, Feb. 2016, pp. 106–125, 2016. The expectation of the performance advantage is computed as [5] V. Buterin, ‘‘A next-generation smart contract and decentralized appli- cation platform,’’ Ethereum Found., Zug, Switzerland, White Paper, Jan. 2014. [Online]. Available: https://ethereum.org/en/whitepaper/ as as EAdv (m) = Xm i=1 Advi · pi = Xm i=1 i0Ci mpi(1 −p)m−i. [6] A. Kosba, A. Miller, E. Shi, Z. Wen, and C. Papamanthou, ‘‘Hawk: The blockchain model of cryptography and privacy-preserving smart con- tracts,’’ in Proc. IEEE Symp. Secur. Privacy (SP), San Jose, CA, USA, May 2016, pp. 839–858. Similarly, we have Similarly, we have y pp [7] A. Kiayias, A. Russell, B. David, and R. Oliynykov, ‘‘Ouroboros: A prov- ably secure proof-of-stake blockchain protocol,’’ in Proc. 37th Annu. Int. Cryptol. Conf., Santa Barbara, CA, USA, Aug. 2017, pp. 357–388. [3] I. Bentov, A. Gabizon, and A. Mizrahi, ‘‘Cryptocurrencies without proof of work,’’ in Proc. 20th Annu. Int. Conf. Fin. Cryptograph. Data Securi. (FC), Christ Church, Barbados, Feb. 2016, pp. 142–157. [2] H. Corrigan-Gibbs, W. Mu, D. Boneh, and B. Ford, ‘‘Ensuring high-quality randomness in cryptographic key generation,’’ in Proc. ACM SIGSAC Conf. Comput. Commun. Secur. (CCS), Berlin, Germany, Nov. 2013, pp. 685–696. EAdv (1) = p0, EAdv (1) = p0, [18] B. Chor, S. Goldwasser, S. Micali, and B. Awerbuch, ‘‘Veriﬁable secret sharing and achieving simultaneity in the presence of faults,’’ in Proc. 26th Annu. Symp. Found. Comput. Sci. (SFCS), Portland, OR, USA, 1985, pp. 383–395. we have we have [19] R. L. Rivest, A. Shamir, and Y. Tauman, ‘‘How to leak a secret,’’ in Proc. 7th Int. Conf. Theory Appl. Cryptol. Inf. Secur., Gold Coast, QLD, Australia, Dec. 2001, pp. 552–565. EAdv (m) = pm0. In summary, in the (n, t) setting where m = t−1, the expec- tation of SecRand’s performance advantage over SCRAPE is p(t −1)0. In addition, the probability of no participants being corrupted is (1 −p)t−1, and only in this situation does SCRAPE exhibit a similar performance to that of SecRand. Thus, SecRand exhibits a higher performance than SCRAPE with a probability of 1 −(1 −p)t−1. [20] R. J. McEliece and D. V. Sarwate, ‘‘On sharing secrets and Reed–Solomon codes,’’ Commun. ACM, vol. 24, no. 9, pp. 583–584, Sep. 1981, doi: 10.1145/358746.358762. [21] D. Chaum and T. Pedersen, ‘‘Wallet databases with observers,’’ in Proc. 12th Annu. Int. Cryptol. Conf., Santa Barbara, CA, USA, Aug. 1992, pp. 89–105. [22] S. D. Galbraith, K. G. Paterson, and N. P. Smart, ‘‘Pairing for cryptogra- phers,’’ Int. Assoc. Cryptologic Res. (IACR), Cambridge, U.K., Tech. Rep. 2006/165, 2006. [Online]. Available: https://eprint.iacr.org/2006/165 [23] D. Boneh and M. Franklin, ‘‘Identity-based encryption from the Weil pairing,’’ in Proc. 21st Annu. Int. Cryptol. Conf., Santa Barbara, CA, USA, Aug. 2001, pp. 213–229. [1] C. Blundo, A. De Santis, and U. Vaccaro, ‘‘Randomness in distribution protocols,’’ Inf. Comput., vol. 131, no. 2, pp. 111–139, Dec. 1996, doi: 10.1006/inco.1996.0095. APPENDIX EAdv (m + 1) = Xm+1 i=1 i0Ci m+1pi(1 −p)m+1−i. [8] J. Bonneau, J. Clark, and S. Goldfeder, ‘‘On Bitcoin as a public randomness source,’’ Int. Assoc. Cryptologic Res. (IACR), Cambridge, U.K., Tech. Rep. 2015/1015, 2015. [Online]. Available: https://eprint.iacr.org/2015/1015 We then compute We then compute EAdv (m + 1) −(1 −p) EAdv (m) = Xm i=1 i0 Ci m+1 −Ci m pi (1 −p)m+1−i + (m + 1)0Cm+1 m+1pm+1. [9] S. Azouvi, P. McCorry, and S. Meiklejohn, ‘‘Winning the caucus race: Con- tinuous leader election via public randomness,’’ 2018, arXiv:1801.07965. [Online]. Available: http://arxiv.org/abs/1801.07965 [10] E. Syta, P. Jovanovic, E. K. Kogias, N. Gailly, L. Gasser, I. Khofﬁ, M. J. Fischer, and B. Ford, ‘‘Scalable bias-resistant distributed random- ness,’’ in Proc. IEEE Symp. Secur. Privacy (SP), San Jose, CA, USA, May 2017, pp. 444–460. Since [11] T. Rabin and M. Ben-Or, ‘‘Veriﬁable secret sharing and multiparty pro- tocols with honest majority,’’ in Proc. 21st Annu. ACM Symp. Theory Comput. (STOC), Seattle, WA, USA, 1989, pp. 73–85. Ci m+1 −Ci m = Ci−1 m (m + 1) 0Cm+1 m+1pm+1 = (m + 1) 0Cm mpm+1, [12] P. Schindler, A. Judmayer, N. Stifter, and E. Weippl, ‘‘HydRand: Efﬁcient continuous distributed randomness,’’ Int. Assoc. Cryptologic Res. (IACR), Cambridge, U.K., Cryptol. ePrint Arch. Rep. 2018/319, 2018. [Online]. Available: https://eprint.iacr.org/2018/319 we have we have EAdv (m + 1) −(1 −p) EAdv (m) = p Xm i=0i0Ci mpi (1−p)m−i+ Xm i=0 0Ci mpi (1−p)m−i = p(EAdv (m) + 0). [13] I. Bentov, A. Gabizon, and D. Zuckerman, ‘‘Bitcoin beacon,’’ 2016, arXiv:1605.04559. [Online]. Available: http://arxiv.org/abs/1605.04559 [14] B. Bunz, S. Goldfeder, and J. Bonneau, ‘‘Proofs-of-delay and ran- domness beacons in Ethereum,’’ presented at the IEEE S&B Work- shop, Paris, France, Apr. 2017. [Online]. Available: http://www.jbonneau. com/doc/BGB17-IEEESB-proof_of_delay_ethereum.pdf = p(EAdv (m) + 0). Thus, we know that Thus, we know that [15] T. Hanke, M. Movahedi, and D. Williams, ‘‘DFINITY technology overview series, consensus system,’’ 2018, arXiv:1805.04548. [Online]. Available: http://arxiv.org/abs/1805.04548 EAdv (m + 1) −(1 −p) EAdv (m) = p (EAdv (m) + 0) EAdv (m + 1) = EAdv (m) + p0. [16] I. Cascudo and B. David, ‘‘SCRAPE: Scalable randomness attested by public entities,’’ in Proc. 15th Int. Conf. Appl. Cryptograph. Netw. Secur. (ACNS), Kanazawa, Japan, Jul. 2017, pp. 537–556. Since Since [17] A. Shamir, ‘‘How to share a secret,’’ Commun. ACM, vol. 22, no. 11, pp. 612–613, Nov. 1979, doi: 10.1145/359168.359176. REFERENCES [1] C. Blundo, A. De Santis, and U. Vaccaro, ‘‘Randomness in distribution protocols,’’ Inf. Comput., vol. 131, no. 2, pp. 111–139, Dec. 1996, doi: 10.1006/inco.1996.0095. [24] S. HeidarVand and J. L. Villar, ‘‘Public veriﬁability from pairing in secret sharing schemes,’’ in Proc. 15th Int. Workshop Sel. Area Cryptograph. (SAC), Sackville, NB, CA, Aug. 2008, pp. 294–308. [25] C. Baum, D. Ivan, and C. Orlandi, ‘‘Publicly auditable secure multi- party computation,’’ in Proc. Int. Conf. Secur. Cryptograph. Netw. (SCN), Amalﬁ, Italy, Sep. 2014, pp. 175–196. [26] GNU Multiple Precision Arithmetic Library. Accessed: Aug. 23, 2019. [Online]. Available: https://github.com/aleaxit/gmpy [27] Pairing-Based Cryptography Library. Accessed: Dec. 8, 2019. [Online]. Available: https://crypto.stanford.edu/pbc 203928 VOLUME 8, 2020 Z. Guo et al.: SecRand: A Secure DRG Protocol With High Practicality and Scalability ZHAOZHONG GUO received the B.S. degree in information and computing sciences from Peking University, China, in 2012, where he is currently pursuing the Ph.D. degree in applied mathematics. His current research interests include blockchain technology and public key cryptography and mul- tiparty computation. MAOZHI XU received the B.S. degree in math- ematics from Huaibei Normal University, China, in 1983, the M.S. degree in mathematics from Wuhan University, China, in 1987, and the Ph.D. degree in mathematics from Peking University, China, in 1994. He is currently a Professor with Peking University. LIUCHENG SHI received the B.S. degree in information and computing sciences from Peking University, China, in 2012, where he is cur- rently pursuing the Ph.D. degree in applied math- ematics. His current research interests include blockchain technology and public key cryptogra- phy and applied cryptography. LIUCHENG SHI received the B.S. degree in information and computing sciences from Peking University, China, in 2012, where he is cur- rently pursuing the Ph.D. degree in applied math- ematics. His current research interests include blockchain technology and public key cryptogra- phy and applied cryptography. 203929 VOLUME 8, 2020
https://openalex.org/W4285445470	https://ejournal.upi.edu/index.php/IJPE/article/download/32883/15717	English	null	Exploration of Holistic Character Values on Integrative Thematic Books	Indonesian journal of primary education	2,021	cc-by-sa	3,970	Exploration of Holistic Character Values on Integrative Thematic Books Saroh Kurnia1,Ganjar Muhammad Ganeswara2 , Edi Suresman3 Universitas Pendidikan Indonesia *Corresponding Author: sarohkurnia@upi.edu1 , ganharmg@upi.edu2, edisuresman@upi.edu3 Received 21 March 2021; Revised 21 June 2021; Accepted 30 June 2021 Published 30 June 2021 Received 21 March 2021; Revised 21 June 2021; Accepted 30 June 2021 Published 30 June 2021 Keywords: Exploration, Holistic Character Value, Integrative Thematic Book. Keywords: Exploration, Holistic Character Value, Integrative Thematic Book. Abstract bst act This research is based on the problems related to content materialthat does not emphasize of character education. This research aims to explore the values of the holistic character contained in the integrated thematic book of fourth grade for elementary school and implication in the learning process or student life. This research used qualitative approach with the content analysis objects were several themes in Integrative Thematic learning for fourth grade. And the methods of research used in this research was descriptive. Data collection techniques using documentation and literature study. Based on the exploration results the, character values contained in the integrative thematic book were: [1] The values of holistic character in the textbook are Love of God and all of His Creations, Responsibility, Tolerance, Mutual Cooperation, and Perseverance; Although lacks in the aspects of language, text function, incomplete discourse and the frequency of uneven character values loads, but with the holistic character, can balance cognitive, affective and psychomotor loads in integrative thematic books [2] The implication for learning or student life is that it has not made abstract characters more concrete because holistic characters are more contained in narrative texts compared to practice during learning and in daily life and the use of forms of linguistic exposure must use language that is often found in elementary school children. The results of this study provide several recommendations for further studies, programs, teachers, and researchers that can be taken into consideration integration character education and improving the contents in integrative thematic books. Vol. 5, No. 1 (2021) 63-69 ISSN: 2597-4866 Indonesian Journal of Primary Education Abstrak Penelitian ini dilatarbelakangi permasalahan isi materi yang kurang memuat pendidikan karakter. Tujuan dari penelitian ini yaitu mengeksplorasi nilai-nilai karakter holistik yang terkandung pada buku tematik integratif kelas IV SD/MI serta implikasinya dalam kehidupan atau pembelajaran siswa. Penelitian ini menggunakan pendekatan kualitatif dengan analisis isinya beberapa tema dalam pembelajaran Tematik Integratif Kelas IV SD/MI. Sedangkan metode penelitian yang digunakan dalam penelitian ini adalah deskriptif. Adapun teknik pengumpulan data menggunakan dokumentasi dan studi pustaka. Berdasarkan hasil eksplorasi nilai-nilai karakter yang terkandung dalam buku tematik integratif kelas IV SD/MI ditemukan: [1] Nilai-nilai karakter holistik dalam buku teks yaitu Cinta Tuhan dan segenap ciptaan-Nya, Tanggung Jawab, Toleransi, Gotong Royong dan Pantang Menyerah; meskipun di dalamnya masih memiliki kekurangan dari aspek kebahasaan, fungsi teks, wacana yang tidak lengkap serta frekuensi muatan nilai karakter yang tidak merata. Namun dengan karakter holistik dapat menyeimbangkan muatan kognitif, afektif dan psikomotorik dalam buku tematik integratif [2] Implikasinya terhadap pembelajaran atau kehidupan siswa yaitu belum menjadikan karakter yang abstrak menjadi lebih konkret karena karakter holistik lebih banyak termuat dalam teks naratif di bandingkan dengan praktek saat pembelajaran maupun dalam keseharian dan penggunaan bentuk paparan kebahasaan harus menggunakan bahasa yang sering ditemukan anak sekolah dasar. Hasil penelitian ini memberikan beberapa rekomendasi bagi program studi, guru dan peneliti selanjutnya yang dapat dijadikan pertimbangan dalam upaya pengintegrasian pendidikan karakter dan memperbaiki muatan isi dalam buku tematik integratif. Kata Kunci : Eksplorasi, Nilai Karakter Holistik, Buku Tematik Integratif. Kata Kunci : Eksplorasi, Nilai Karakter Holistik, Buku Tematik Integratif. INTRODUCTION In line with the pace of community development, according to Zainudin (2008) education becomes very dynamic and adapted to existing developments. The education curriculum is not a standard and static benchmark, but is very dynamic and must adapt to existing situations and conditions. Based on this, it is necessary to reform education in order to remain conducive, p INTRODUCTION In line with the pace of community development, according to Zainudin (2008) education becomes very dynamic and adapted to existing developments. The education © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index Saroh Kurnia1, Ganjar Muhammad Ganeswara2 , Edi Suresman3 Exploration of Holistic Character Values in Integrative Thematic Books 64 integrated meaning or meaningful unity in the curriculum structure, although sometimes it is influenced by views between branches of science.As for spaceThe scope of the General Education program should cover six areas of meaning, namely: symbolics, empiricals, esthetics, synnoetics, ethics, and synoptics. programmed and systemic. Because ideally the educational process that takes place in schools can produce students who have noble character, and are not only concerned with abilities in the cognitive field alone or intellectually intelligent but must optimize other aspects, namely affective and psychomotor because with the provision of noble character, children will develop into children. who are good and will become adults in the future have a strong character for the homeland and nation. So that in the 2013 curriculum, teachers are not only tasked with teaching their students to be academically capable, but also must be able to arouse the abilities of their students. In addition, in the 2013 curriculum, character education also gets more attention, because the purpose of education does not only rely on cognitive aspects or becomes temporary memorization that will be lost over time. However, learning must also provide a meaningful experience for students so that there is a change in attitude for the better. (Ruminiati and Khusubakti, 2016) Furthermore, students, especially in elementary schools, are in the early age range where all aspects of intelligence development such as IQ, EQ and SQ grow and develop extraordinary. In general, this level of development still sees everything as a whole (holistic) and is able to understand the relationship between concepts in a simple way. The learning process still depends on concrete objects and experiences experienced directly (Samsudin, 2008). In accordance with the curriculum policy above, learning that is in accordance with the needs and development of students is thematic learning. © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/ind p INTRODUCTION Meanwhile, according to Rusman (2010) that thematic learning is one of the models in integrated learning (integrated instruction) which is a learning system that allows students, both individually and in groups, to actively explore and discover scientific concepts and principles in a holistic, meaningful and meaningful way. authentic. The holistic paradigm can be interpreted as a comprehensive or comprehensive perspective in perceiving reality. A holistic view means looking at aspects of the whole rather than the parts, having a systemic, integrated, complex, dynamic, non- mechanical, and non-linear pattern (Heriyanto, 2003). Therefore, with a holistic view, students will gain complete knowledge from various aspects of scientific disciplines. So that students will be wiser in looking at things because they are equipped with a complete understanding and in the end can optimize good knowledge, attitudes and activities. In addition, there are two types of textbooks in the 2013 curriculum, namely textbooks for students to use and guide books for teachers in the learning process. In the thematic book, each chapter is equipped with a concept map, introduction, part of student activities both experimental and non- experimental or discussion, practice questions, summaries, evaluations, and assignments for students. Meanwhile, the Student Book is activity based, making it possible for students to complete material from various sources (Multayas, et al. 2020). The researchers focused on analyzing holistic character values, namely in the fourth grade SD/MI semester 1 student books which consisted of five themes, including the following: 1) The Beauty of Togetherness, 2) Always Saving Energy, 3) Caring for the In addition, an understanding of holistic character development also requires an integrative program, one of which isin General Education because according to PH Phenix (1964) the basic principles of the General Education program are directed at mastering knowledge and skills, increasing a sense of social responsibility, knowing several other areas of knowledge, the existence of a relationship between one area of knowledge and another, and what is important is quality. a person's understanding of an area of knowledge or the existence of an © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index Saroh Kurnia1, Ganjar Muhammad Ganeswara2 , Edi Suresman3 Exploration of Holistic Character Values in Integrative Thematic Books 65 Environment, 4 ) Various Jobs and 5) My Hero. Environment, 4 ) Various Jobs and 5) My Hero. p INTRODUCTION SD/MI student books from the aspect of character education and the scientific approach. Furthermore, it can also be used as consideration for writing or revising future editions of student books. Furthermore, the researcher conducted an initial study of several student books and found a number of notes that should be commented on in the student books. Some of the material presented does not highlight the content of character education. The presentation technique also tends not to support the procedures in the scientific approach required in the 2013 Curriculum. Old nuances and patterns can still be found in some parts of the book, such as directly presenting the material without giving students the opportunity to observe, ask questions, reason, and try. Some authors directly mention the content of the material (eg definitions, types, examples) without giving students the opportunity to take the steps in the scientific approach. METHOD This study uses a qualitative approach with content analysis. While the content analysis in this study were several themes in the Integrative Thematic learning for Class IV SD/MI. While the type of research used in this research is exploratory, with descriptive research methods. There are several other instruments needed by researchers in collecting data in this study, namely student book analysis guidelines and Student Book Analysis Worksheets, operational verb sheets, holistic character matrices, and value coding as indicators set by the Ministry of Education and Culture to test the feasibility of the book. The aspects of the study are focused on exploring holistic character values in the Thematic Integrative Class IV SD/MI book. As for the validity of the instrument using triangulation theory. Furthermore, the data collection technique uses documentation and literature study. Meanwhile, data analysis used Miles and Huberman data analysis with the following stages: 1) data collection ordata collection stage in the form of holistic character values of integrative thematic textbooks, 2) data reductionorthe process of selecting and grouping data for research purposes, 3) data displayor the data presentation stage, namely the classification and description of the data and 4) the conclusion/verification stage. The matrix to determine the feasibility of an integrative thematic book is listed in the table below: Based on the above background, the researcher tries to explore the holistic character values in thematic books. The purpose of this research is to explore the holistic character values contained in the fourth grade integrative thematic books of SD/MI and their implications in students' lives or learning. Therefore, the presentation of thematic books in elementary schools must be based on aspects of character education and a scientific approach. Both are the main aspects of realizing the goals of developing the 2013 Curriculum and the goals of national education. The study focused on the books for fourth grade elementary school students in semester 1, because grade IV is a transition period from low to high grade which contains more material and themes. This analysis is urgent to be carried out because it can inform various strengths and weaknesses in the MMatrix of Eligibility for Integrative Thematic Books for Class IV SD/MI Semester 1 No kriteria Complete Not complete 1. Developed competencies - 2. Character Developed - 3. METHOD Language Exposure Form 3.1 Word Aspect - MMatrix of Eligibility for Integrative Thematic Books for Class IV SD/MI Semester 1 No kriteria Complete Not complete 1. Developed competencies - 2. Character Developed - 3. Language Exposure Form 3.1 Word Aspect - MMatrix of Eligibility for Integrative Thematic Books for Class IV SD/MI Semeste 66 3.2 Sentence Aspect - 3.3 Paragraph Aspect - 3.4 Aspects of Discourse - 4. Structure of Learning Materials - 5. Text function - 6. General Education Principles - © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index RESULTS AND DISCUSSION characters, namely (1) Love of God and all of His creation which consists of a united character, (2) Responsibility consisting of trustworthy, independent, disciplined characters. , fair and honest, (3) Tolerance which consists of polite character, (4) Gotong royong/cooperation which consists of helpful character, and a good leader, (5) Never gives up/hard work which consists of creative character. The results of this study relate to the content of holistic character values contained in the integrative thematic book for grade IV SD/MI semester 1 which consists of five themes, including: the beauty of togetherness, always saving energy, caring for the environment, various jobs and my heroes. Based on the results of data analysis, there are 15 holistic character values in the thematic book, but they can be classified into five Table 2 Holistic Character Matrix No Character Value Theme 1 Theme 2 Theme 3 Theme 4 Theme 5 1. Love God and All of His Creation - 2. Responsibility 3. Tolerance - - - - 4. Gotong Royong/ Cooperation - - 5. Never Give Up/ Hard Work Table 2 Holistic Character Matrix Table 2 Holistic Character Matrix can optimize good knowledge, attitudes and activities. Based on the analysis of the overall themes and holistic character values that have been explored. So the integrative thematic approach in the 2013 SD/MI curriculum is not only integrated in terms of cross-subject material, it is also integrated in the realm of learning competencies. This means that in every learning, teachers are required to be able to integrate learning gains in three domains (Cognitive, Affective and Psychomotor/Skills) and one way is through thematic books. So that by having holistic character values, students will be wiser in looking at things because they are equipped with a complete understanding and in the end In addition, the results can be obtained based on the principle of general education, namely the existence of an integrated meaning or meaningful unity in the curriculum structure through the language function, namely to obtain meaning from the text, especially in integrative thematic books that are integrated in multi-disciplinary sciences. RESULTS AND DISCUSSION The aim is that students are able to identify each word, sentence and paragraph; the ability to grasp the explicit meaning of each text or to grasp the implicit meaning of some symbols is to enter the comprehension area of the cognitive realm; and awareness to 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index Saroh Kurnia1, Ganjar Muhammad Ganeswara2 , Edi Suresman3 Exploration of Holistic Character Values in Integrative Thematic Books 67 continuously improve competence and increase understanding is in the affective domain. to be forced and not in accordance with the five main learning experiences presented in the 5M activity (observing, asking, reasoning, trying, and communicating). In addition, the findings on the Always Saving Energy theme have less character content than other themes, lack of coherence in the material, too many practice questions but lack of character content, Furthermore, it can be seen that the frequency of character loading in each theme varies, some are found in all themes and some are found in one theme. This is because the characters developed tend to adapt to the title of the theme and the presence of a dominating character. As is the case with the character of God's Love and all of His creation which tends to be used as a support for other characters. Whereas according to Endang and Zubaedi (2020) the goal achieved from the concept of gratitude is to believe in the existence of God through His creation. So the need for character development evenly and well integrated. As is the case with the character of responsibility which has been integrated into all themes in the thematic book by using everyday language (ordinary language) so that it is easy to understand. , Furthermore, regarding the form of linguistic exposure from the word aspect, single words and complex words were found. Meanwhile, in the sentence aspect, there are four types of sentences, namely interrogative sentences, command sentences, news sentences and exclamation sentences. For paragraphs, four types of paragraphs were found including narrative paragraphs, expositions, persuasion and argumentation paragraphs. Then in the discourse aspect, five findings were found. First, there is a tendency to display non-literary texts more than literary texts. Second, the imbalance in the number of sentences in one paragraph presented in the texts available in the book. © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index RESULTS AND DISCUSSION This is indicated by the presentation of the sub-activities that appear 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index Saroh Kurnia1, Ganjar Muhammad Ganeswara2 , Edi Suresman3 Exploration of Holistic Character Values in Integrative Thematic Books 68 general education and character in terms of objectives, materials and learning materials based on language functions and general education principles. This can be seen from the character values already contained in the book, namely Love of God and All of His Creation, Responsibility, Tolerance, Mutual Cooperation/Cooperation and Never Giving Up/Hard work. Even though it still has flaws in itfrom the linguistic aspect, the function of the text, the discourse is not complete but only fragments and the frequency of appearance of character values is uneven and depends on the focus of material achievement. So that with this holistic character, it can balance the cognitive, affective and psychomotor content in integrative thematic books in accordance with educational goals. Furthermore, so that learning is more meaningful and has a good impact on students. The most important thing is the implication of this thematic book so that the material that has been obtained can be applied in learning and in students' daily lives. The results that have been obtained are that they have not made abstract characters more concrete because holistic characters are more contained in narrative texts compared to practice during learning and in everyday life. While the importance of the text that is seen is not just a collection of words or linguistic rules, but the process of selecting linguistic forms to express meaning. So that the use of language exposure forms can be effective depending on students' understanding when reading. Then the holistic character values contained in the Integrative Thematic book have implications for learning or student life, namely that they have not made abstract characters more concrete because holistic characters are more contained in narrative texts compared to practice during learning and in everyday life. While the importance of the text that is seen is not just a collection of words or linguistic rules, but the process of selecting linguistic forms to express meaning. So that the use of language exposure forms can be effective depending on students' understanding when reading. RESULTS AND DISCUSSION Third, it was found that there was an neglect of the inclusion of the author's name and the source of the citation, especially the works quoted from other sources. Fourth, Another problem is the writer's reluctance to seek the writings of experts. Fifth, it is found that there is a tendency for more discourse in the form of what is shown to be incomplete. In addition, the complete discourse was not found, but only fragments whose contents were not easy to understand because of the incompleteness of the discourse. However, it is the opposite of the character of tolerance which is integrated in only one theme in the thematic book. This is because the characters developed tend to adapt to the title of the theme and dominate on theme 1 "The Beauty of Togetherness" so that it is not found in other themes. In contrast to the character of gotong royong/cooperation which is integrated into several themes in thematic books. Even though the importance of the character of cooperation according to Helmawati (2017) because human abilities are limited or different from one another, humans must work together to achieve a goal. In addition, there are also characters who never give up/hard work integrated into all themes in the thematic books. Overall, each character contains empirical, ethical, nondiscursive, synnoetic, symbolic, aesthetic and ordinary language meanings. The advantages of this integrative thematic book include: teaching materials that are relevant to students' lives, integrated subjects make it easier for students to understand in their entirety, the material is presented with interesting pictures, the structure of learning materials is quite complete, there is reflection on learning as improvement material, there are cooperation with parents. Then from the overall data findings, an interpretation is drawn up that the form of linguistic exposure contained in textbooks uses the form of exposure that is often found by elementary school children. Based on the explanation of the text book above, the researcher can explain some of the weaknesses of the thematic book for grade IV SD/MI, including: important findings in organizing the learning materials presented in each sub-activity that appears to pay less attention to the series of activities in the scientific approach. REFERENCES Endang dan Zubaedi. (2020). Pola Pembelajaran 9 Pilar Karakter Pada Anak Usia Dini dan Dimensi- Dimensinya. Jakarta: Prenadamedia Group. Helmawati. (2017). Pendidikan Karakter Sehari-Hari. Bandung: PT. Remaja Rosdakarya. Heriyanto, Husain. (2003). Paradigma Holistik: Dialog Filsafat, Sains, dan Kehidupan Menurut Shadra dan Whitehead. Bandung: Mizan Media Utama. RESULTS AND DISCUSSION In addition, the implications for learning can be done by introducing values and internalizing good values into student behavior in everyday life through a thematic learning process. In addition to making students master the targeted material, it is also designed to make students recognize, care about, and internalize these values so that they become behaviors. Furthermore, according to Mulayahati (2014) one of the determining factors for the success of teachers and students in using textbooks is determined by the quality of textbooks. In measuring the quality of textbooks, important aspects must be considered, namely the suitability of the content of the material with the curriculum, the coherence of the material, the depth and breadth of the material. As for the results of the study, thatIntegrative thematic book for grade IV SD/MI deserves to be one of the studies of general education and character in terms of objectives, materials and learning materials based on language functions and general education principles. Even though it still has flaws in itwhich must be corrected. CONCLUSION Based onresearch results, findings and discussions that have been analyzed in the previous section, it can be concluded thatIntegrative thematic book for grade IV SD/MI deserves to be one of the studies of © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index Saroh Kurnia1, Ganjar Muhammad Ganeswara2 , Edi Suresman3 Exploration of Holistic Character Values in Integrative Thematic Books 69 Mulayahati, Bunga. (2014). Analisis Buku Tematik Terpadu Kurikulum 2013 Kelas IV Sekolah Dasar. Jurnal Pendidikan. Multayas,dkk. (2020). Analisis Nilai Karakter pada Buku Siswa Tematik Sekolah Dasar Berorientasi Pendidikan Karakter. Jurnal Pemikiran dan Pengembangan Sekolah Dasar), Vol. 8 , No 1 Phenix,P.H., (1964), Realms of Meaning, New York: McGraw-Hill Book Company. Ruminiati dan Khusubakti. (2016). Analisis Kesesuaian Isi Buku Tematik Kelas IV Sekolah Dasar dengan Pendidikan Karakter dan Pendekatan Scientific. Jurnal Sekolah Dasar. © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index Rusman. (2010). Model-model pembelajaran. Bandung: PT Raja Grafindo Persada. Rusman. (2010). Model-model pembelajaran. Bandung: PT Raja Grafindo Persada. Samsudin. (2008). Pembelajaran Pendidikan Jasmani Olahraga dan Kesehatan SD/MI. Jakarta: Litera Prenada Media Grup. Zainuddin. (2008). Reformasi Pendidikan: Kritik Kurikulum dan Manajemen Berbasis Sekolah. Yogyakarta: Pustaka Pelajar. © 2021 - Indonesian Journal of Primary Education – Vol .5, No.1 (2021 ) 63-69 http://ejournal.upi.edu/index.php/IJPE/index
https://openalex.org/W2197028160	http://pure-oai.bham.ac.uk/ws/files/30655482/Vaughan_et_al_Symbolism_over_substance_Legal_Ethics.pdf	English	null	Symbolism over substance? Large law firms and corporate social responsibility	Legal ethics	2,015	public-domain	18,145	University of Birmingham Symbolism over Substance? Large Law Firms and Corporate Social Responsibility Vaughan, Steven; Thomas, Linden; Young, Alastair DOI: 10.1080/1460728x.2015.1119530 License: Creative Commons: Attribution (CC BY) Document Version Publisher's PDF, also known as Version of record Citation for published version (Harvard): Vaughan, S, Thomas, L & Young, A 2015, 'Symbolism over Substance? Large Law Firms and Corporate Social Responsibility', Legal Ethics, vol. 18, no. 2, pp. 138-163. https://doi.org/10.1080/1460728x.2015.1119530 University of Birmingham Symbolism over Substance? Large Law Firms and Corporate Social Responsibility Vaughan, Steven; Thomas, Linden; Young, Alastair Citation for published version (Harvard): Vaughan, S, Thomas, L & Young, A 2015, 'Symbolism over Substance? Large Law Firms and Corporate Social Responsibility', Legal Ethics, vol. 18, no. 2, pp. 138-163. https://doi.org/10.1080/1460728x.2015.1119530 Citation for published version (Harvard): Vaughan, S, Thomas, L & Young, A 2015, 'Symbolism over Substance? Large Law Firms and Corporate Social Responsibility', Legal Ethics, vol. 18, no. 2, pp. 138-163. https://doi.org/10.1080/1460728x.2015.1119530 Link to publication on Research at Birmingham portal Download date: 24. Oct. 2024 General rights l li General rights Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes permitted by law. eely distribute the URL that is used to identify this publication. •Users may freely distribute the URL that is used to identify this publication. Users may freely distribute the URL that is used to identify this publication. •Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private study or non-commercial research. study o o co e c a esea c •User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and P •Users may not further distribute the material nor use it for the purposes of commercial gain. Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document. Where a licence is displayed above, please note the terms and conditions of the licence govern your use of th When citing, please reference the published version. ISSN: 1460-728x (Print) 1757-8450 (Online) Journal homepage: http://www.tandfonline.com/loi/rlet20 Date: 15 September 2016, At: 07:08 Take down policy down policy the University of Birmingham exercises care and attention in making items available there are rare occasions when an it ded in error or has been deemed to be commercially or otherwise sensitive. this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access ely and investigate. If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we the work immediately and investigate. Download date: 24. Oct. 2024 Legal Ethics Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=rlet20 KEYWORDS Legal profession; CSR; big law; diversity; pro bono; sustainability CONTACT Steven Vaughan s.vaughan@bham.ac.uk 1Thomas Clarke and Douglas Branson, ‘Corporate Governance – An Emerging Discipline?’ in Thomas Clarke and Douglas Branson (eds), The SAGE Handbook of Corporate Governance (SAGE 2012) 1–37. While the majority of the literature on CSR has been developed over the last ﬁve decades, the concept has nevertheless existed for a good deal longer. For a good account of the origins of CSR, see Harwell Wells, ‘The Birth of Corporate Governance’ (2010) 33(4) Seattle University Law Review 1247. In their review, Aguinis and Glavas found that more than half of the 588 journal articles they read on CSR had been published since 2005 – see Herman Aguinis and Ante Glavas, ‘What We Know and Don’t Know About CSR: A Review and Research Agenda’ (2012) 38(4) Journal of Management 932. © 2015 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/ licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Steven Vaughan, Linden Thomas & Alastair Young To cite this article: Steven Vaughan, Linden Thomas & Alastair Young (2015) Symbolism over substance? Large law firms and corporate social responsibility, Legal Ethics, 18:2, 138-163, DOI: 10.1080/1460728x.2015.1119530 To link to this article: http://dx.doi.org/10.1080/1460728x.2015.1119530 © 2015 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published online: 22 Dec 2015. Submit your article to this journal Article views: 424 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=rlet20 Download by: [University of Birmingham] LEGAL ETHICS, 2015 VOL. 18, NO. 2, 138–163 http://dx.doi.org/10.1080/1460728x.2015.1119530 LEGAL ETHICS, 2015 VOL. 18, NO. 2, 138–163 http://dx doi org/10 1080/1 Symbolism over substance? Large law ﬁrms and corporate social responsibility Steven Vaughan , Linden Thomas and Alastair Young Law School, University of Birmingham, UK © 2015 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. ( ) y , g y p This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecom licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is pro ABSTRACT ABSTRACT At its core, corporate social responsibility (CSR) concerns the impacts of businesses on their surroundings. Despite their signiﬁcant economic and geographic presence (and, as a corollary, their potential signiﬁcant impacts), and despite the varied disciplinary and conceptual lenses used to study CSR, there is very little existing work looking at law ﬁrms and their own CSR policies. This paper ﬁlls part of that gap. In August 2014, we reviewed the websites of the top 100 English law ﬁrms, as ranked by the trade publication The Lawyer. We were interested in public disclosures made by those law ﬁrms on CSR. These were widespread. The majority of the top 100 ﬁrms say something to the wider world about CSR. However, what is said varies signiﬁcantly. This is, perhaps, unsurprising. What is more surprising is that so few ﬁrms explain why they are committed to CSR. Where ﬁrms do make disclosures on CSR, these tend to group around the following three areas: (i) pro bono and community giving; (ii) diversity and inclusion; and (iii) environmental matters. For a number of ﬁrms, little or no distinction is made between pro bono (i.e. the giving of free legal advice) and wider ‘community giving’. We question whether this is the right approach. We were also concerned that, despite there being regulatory intervention by the Legal Services Board as regards the collection and reporting of diversity data by law ﬁrms (and other lawyers), the quality of disclosures (in terms of the amount, nature and breadth of data reported on) varied to such an extent that we were unable to draw any meaningful comparisons or conclusions on diversity in English law ﬁrms. As a phenomenon and a ﬁeld of scholarly enquiry, corporate social responsibility (CSR) has been a site of signiﬁcant interest for the last ﬁve decades.1 CSR is now ubiquitous As a phenomenon and a ﬁeld of scholarly enquiry, corporate social responsibility (CSR) has been a site of signiﬁcant interest for the last ﬁve decades.1 CSR is now ubiquitous CONTACT Steven Vaughan LEGAL ETHICS 139 in the corporate sphere, and a Google search for the term brings back over 16 million results.2 At its core, CSR concerns the impacts of businesses on their surroundings. Over these same ﬁve decades, the practice of law in England and Wales has changed sig- niﬁcantly. ABSTRACT We have moved from a situation in which the legal profession was a relatively homogenous guild-like institution3 to a world of polar extremes in which the largest ﬁrms employ thousands of lawyers in hundreds of ofﬁces across the globe turning over billions of pounds each year, and in which the smallest ﬁrms of sole practitioners in England and Wales struggle to stay aﬂoat in the face of signiﬁcant cuts to legal aid.4 These largest law ﬁrms are powerful, signiﬁcant economic actors. In 2013, the combined annual turnover of just the top ten of the top 100 English and Welsh ﬁrms was over £10 billion.5 This was more than a third of the turnover of the entire legal services sector in England and Wales for the same year.6 In 2013, nine of the top 10 ﬁrms employed more than 2,000 lawyers each; the tenth and largest (DLA Piper) employed over 4,000 lawyers and 8,000 staff worldwide in 89 separate ofﬁces across 30 countries.7 Despite their signiﬁcant econ- omic standing and geographic presence (and, as a corollary, their potential signiﬁcant social, economic and environmental impacts), and despite the varied disciplinary and con- ceptual lenses used to study CSR,8 there is very little existing work looking at law ﬁrms and their own CSR policies.9 This paper ﬁlls part of that gap. In August 2014, we reviewed the websites of the top 100 English law ﬁrms, as ranked by the trade publication The Lawyer.10 We were interested in public disclosures made by those law ﬁrms on CSR, both because of the impacts referred to above, and because of the potential hypocrisy in law ﬁrms advising clients on CSR but not ‘doing’ CSR them- selves. Our intention in this paper is to offer up a foundational, broad overview and cri- tique of those CSR disclosures. We are hampered in a number of areas by the poor quality of publicly available data. Equally, what law ﬁrms say they do about CSR on their websites may not reﬂect what they actually do. As will be seen, we are able to say the most, and in the most depth, about law ﬁrms and pro bono. In further work, we plan to explore each of the separate elements of law ﬁrm CSR discussed below. Our review in this paper allows for three general conclusions to be drawn. 2Search performed on 29 October 2015. 3Emil A Krause, Death of the Guilds: Professions, States, and the Advance of Capitalism, 1930 to the Present (Yale University Press 1996). 4For a recent account of these changes, and a consideration of the future, see Hilary Sommerlad et al, The Futures of Legal Education and the Legal Profession (Hart 2015). 5See <http://www.thelawyer.com/news/uk-200-2013/> 6Law Society, Legal Services Forecasts (Law Society 2014) 3. 7See <http://www.thelawyer.com/news/uk-200-2013/> 8Aguinis and Glavas (n 1) 932. 9As we shall come to discuss, there are (separate) bodies of work on law ﬁrms and pro bono, and on law ﬁrms and diversity, but very little that brings these matters together under the umbrella of law ﬁrms and CSR. 10August 2013 rankings (as the 2014 results had not yet been published at that time). See <http://www.thelawyer.com/ news/uk-200-2013/> We comment in this paper on data available on law ﬁrm websites. As such, we accept that we may not present the actual reality of CSR in law ﬁrms. As set out in the ﬁnal part of this paper, we hope to conduct further empirical work to add depth to our website review. 2Search performed on 29 October 2015. 3 ABSTRACT First, while most law ﬁrms say something about CSR, how they do so is very ad hoc – different ﬁrms report different things about CSR in different ways. We see both a differentiation of approach and competition between ﬁrms, especially among the top 10. As a general trend, the lower the ranking of the law ﬁrm, the less they say about CSR, although there are some notable exceptions to the rule at both ends of the rankings. We would suggest that most ﬁrms are attempting S. VAUGHAN ET AL. 140 (imperfectly) to map on to their businesses a model of CSR which has developed for non- legal services sectors. Because of this, we would also suggest that a number of ﬁrms struggle to articulate why CSR is, or should be, important to them. We question whether current approaches by law ﬁrms to CSR really reﬂect the nuances (and speciﬁc impacts and responsibilities) of law ﬁrms as organisations. As such, we ask whether law ﬁrms have simply bolted on to their own practices approaches to, and forms of, CSR that were pre- viously, and still are, used by their corporate clients. We suggest that recent practice may have been partly symbolic; more about meeting competition and about demand-side pressures (i.e. appealing to clients), than a substantive, altruistic commitment to CSR.11 Our second general ﬁnding is that, for a number of ﬁrms, pro bono (i.e. the giving of free legal advice) and wider ‘community giving’ have become elided. We suggest that this is a mistake. 11For a wider discussion relating to some of these trends, see Ronen Shamir, ‘The Age of Responsibilization: on Market- Embedded Morality’ (2008) 37(1) Economy and Society 1. 12In this paper, we do not engage in the debate over exactly how large ﬁrms should help to alleviate unmet legal need and acknowledge the argument that such support may, in fact, be more effective in the form of direct ﬁnancial giving to third sector legal advice organisations with relevant expertise in the areas of law where there is the greatest need, rather than via the giving of time by lawyers from large ﬁrms. 13We accept that even were there not this confusion between community projects and advice and assistance with legal problems, it would not be possible for corporate lawyers to do more than offer a drop in the ocean to alleviating unmet legal need. 14Hilary Sommerlad, ‘Editorial’ (2014) 17(2) Legal Ethics iii. 15For other examples of this approach, see the 2014 Special Issue (17(2)) of Legal Ethics. 16Andrew Boon and Robert Abbey, ‘Moral Agendas? Pro Bono Publico in Large Law Firms in the United Kingdom’ (1997) 60 (5) Modern Law Review 630. y , ( ) ( ) g 15For other examples of this approach, see the 2014 Special Issue (17(2)) of Legal Ethics. 16 ABSTRACT If CSR is about the impacts a business can have, and about a sense of respon- sibility, then we would argue that law ﬁrms are well placed to discharge such responsibil- ities via a positive impact on the way in which those unable to secure legal advice receive some form of support.12 While community projects (painting schools, building homes, reading with children etc) serve important functions, and express a symbolic commitment to ‘the public good’, such projects ignore and potentially devalue the ability of lawyers in these large ﬁrms to (part) alleviate unmet legal need.13 Thirdly, despite there being regu- latory intervention as regards the collection and reporting of diversity data by law ﬁrms (and other lawyers) in England and Wales, the quality of diversity disclosures (in terms of the amount, nature and breadth of data reported on) varies to such an extent that we have been unable to draw any meaningful comparisons or conclusions. This is both disappointing and frustrating, and we have suggested that regulatory reform on diversity reporting is worthy of consideration by the Legal Services Board, the overarching regulator of legal services in England and Wales. Our paper is not concerned with ‘applied’ legal ethics, nor with issues of moral philos- ophy, but is instead part of the ‘realist approach’ seen in other Legal Ethics scholarship where ‘the starting point in studying legal institutions should be as they actually are’ in order that they can held up to scrutiny.14 As such, this paper engages with important aspects of contemporary professional life in large law ﬁrms.15 We suggest that the approach of large law ﬁrms to CSR is a useful site in which to explore tensions between the conception of professionalism as a monopolistic legal services market control device and professionalism as an ethical commitment to public service and public beneﬁt.16 The remainder of this paper unfolds in ﬁve parts. Part one begins with an overview of why businesses engage in CSR and explores the extent to which these rationales apply to LEGAL ETHICS 141 141 141 large law ﬁrms. This part then charts how the top 100 English law ﬁrms disclose their CSR efforts. Part two concerns pro bono and wider ‘community giving’ initiatives by law ﬁrms. In part three, we discuss equality, diversity and inclusion (ED&I) in the profession and how large law ﬁrms have reported on their ED&I efforts. Part four reviews environmental matters. 19Ashridge Business School, ‘Catalogue of CSR Activities’ (September 2005). See <https://www.ashridge.org.uk/Media- Library/Ashridge/PDFs/Publications/CSRActivities.pdf> 20Oleg V Petrenko, Federico Aime, Jason Ridge and Aaron Hill, ‘Corporate Social Responsibility or CEO Narcissism? CSR Motivations and Organizational Performance’ (2015) Strategic Management Journal [early online view]; Jim Stevens et al ‘Symbolic or Substantive Document? The Inﬂuence of Ethics Codes on Financial Executives’ Decisions’ (2005) 26 Stra- tegic Management Journal 181; Samuel B Graves and Sandra A Waddock, ‘Institutional Owners and Corporate Social Per- formance’ (1994) 37(4) The Academy of Management Journal 1034. 21Elisabet Garriga and Domènec Melé, ‘Corporate Social Responsibility Theories: Mapping the Territory’ (2004) 53(1) Journal of Business Ethics 51. 22See, Clarke and Branson (n 1); and Brian Chefﬁns, ‘The History of Corporate Governance’ in Mike Wright and others (eds), The Oxford Handbook of Corporate Governance (OUP 2013) 46–65. 23Jill Solomon, Corporate Governance and Accountability (4th edn, Wiley 2013) 152. 24Beginning with The Cadbury Report 1992. ABSTRACT The ﬁnal part brings together our discussions in a short conclusion. 17Commission, ‘A Renewed EU Strategy 2011–14 for Corporate Social Responsibility’ (Communication) COM(211) 681 Final. 18Ibid. 19 7Commission, ‘A Renewed EU Strategy 2011–14 for Corporate Social Responsibility’ (Communication) COM(211) 681 8Ibid p 23Jill Solomon, Corporate Governance and Accountability (4th edn, Wiley 2013) 152. 24 h h db Corporate social responsibility There is no one deﬁnition of CSR. The European Commission suggests it should be under- stood as ‘the responsibility of enterprises for their impacts on society’.17 In order for an enterprise to fully meet their corporate social responsibility, enterprises should have in place a process to integrate social, environment, ethical, human rights and consumer concerns into their business operations and core strategy in close collaboration with their stakeholders.18 However, despite this list, the constituent aspects of CSR also elude a common, agreed deﬁnition. Work by the Ashridge Business School identiﬁed (at least) 31 classes of possible CSR activity which businesses should, or could, consider.19 For the purposes of this paper, we explore three main aspects in relation to law ﬁrms (not least because these are the aspects that those ﬁrms themselves most commonly report on): (i) community giving and pro bono; (ii) diversity and inclusion; and (iii) environmental impacts and sustainabil- ity. As we come on to discuss, the relationship between pro bono, CSR and law ﬁrms is long standing, and environmental impacts are a core aspect of how large organisations (including law ﬁrms) respond to CSR pressures, but diversity is perhaps a little different. Its inclusion here, and within CSR more generally, reﬂects a relatively recent acceptance that large organisations have important social roles to play in how they employ, retain and promote different groups from the communities in which they do business. Businesses are said to engage in CSR because of institutional or reputational pressures to be good corporate citizens20 and/or because of the preferences of powerful individuals within those organisations.21 There is now a greater pressure for corporates to disclose data relating to their non-ﬁnancial commitments and impacts.22 Solomon argues, and we would agree, that transparent data disclosure by corporates on CSR matters is impor- tant, ‘to the efﬁcient functioning of corporate governance systems.’23 For some time, such disclosures have also been said to be vital to third parties to ensure the efﬁcient functioning of capital markets.24 In a post-ﬁnancial crisis world, CSR transparency by businesses is S. VAUGHAN ET AL. Corporate social responsibility 142 said to be important because it provides a way in which institutions are able once again to rebuild the trust of stakeholders, which, in turn, is advantageous for market efﬁciency.25 With bodies corporate, there is said to be an agency problem, whereby ownership and control of those corporates is split between shareholders, on the one hand, and directors and ofﬁces on the other.26 To prevent information asymmetry, and the fraudulent and dis- honest behaviour of principals, data should be disclosed in the most timely and transpar- ent manner to stakeholders.27 Agency theory’s success as a means of increasing information disclosure and corporate governance in corporates, however, has been a subject of scepticism.28 There is also literature to suggest that the relationship between transparency and accountability is uncertain,29 and that heightened disclosure can actually lead to negative economic consequences for businesses.30 We would argue that some of these drivers for CSR in corporates apply imperfectly to law ﬁrms. Only two of the 10,394 law ﬁrms in England and Wales are listed on a stock market,31 and recent work suggests that Alternative Business Structures (law ﬁrms which are, among other matters, able to seek outside investment) are almost exclusively funded by their own members and not by third parties.32 As such, there is no agency problem and no external shareholders looking for information or judging law ﬁrms on their performance. Equally, law ﬁrms have not (as yet) been held to public scrutiny or account for their role (whatever that might have been) in the ﬁnancial crisis.33 There is, however, a wealth of other stake- holders (clients, employees, potential lateral hires, law students, regulators, academics and others) who are, or could be, interested in the impacts of law ﬁrms and their corollary approaches to CSR. However, as we will see below, the approach to CSR remains in its formative stages in many parts of the legal market in England and Wales. 26Adam Smith, The Wealth of Nations (ﬁrst published 1776, Penguin 1999); Kathleen M Eisenhardt, ‘Agency Theory: An Assessment and Review’ (1989) 14(1) The Academy of Management Review 57; Michael C Jensen and William H Meckling, ‘Theory of the Firm: Managerial Behaviour, Agency Costs and Ownership Structure’ (1976) 3(4) Journal of Financial Econ- omics 305. 32Sundeep Aulakh and Ian Kilpatrick, ‘Exogenous Jolts, Re-regulation and Change in Professional Organisational Fields: The Case of UK Legal Services’ (Annual Conference on Professional Service Firms, July 2015, Oxford) – copy on ﬁle with the authors. 33 34For an account of this history, see Richard L Abel, The Legal Profession in England and Wales (Blackwell, 1988); and Andrew Boon and Avis Whyte, ‘Charity and Beating Begins at Home: The Aetiology of the New Culture of Pro Bono Publico’ (1999) 2(2) Legal Ethics 169. 31Slater & Gordon and Gateleys. For statistics on the size of the English legal market, see <http://www.sra.org.uk/sra/how- we-work/reports/data/solicitor_ﬁrms.page>. See also <http://www.cnbc.com/2015/05/12/gateley-becomes-ﬁrst-law- ﬁrm-to-join-uk-stock-market.html> 32 29Jonathan Fox, ‘The Uncertain Relationship between Transparency and Accountability’ (2007) 17(4–5) Development in Prac- tice 663. 30 g y 28Benjamin E Hermalin and Michael S Weisbach, ‘Information Disclosure and Corporate Governance’ (2012) 58(1) The Journal of Finance 195. 29 25Mark Bandsuch, Larry Pate and Jeff Thies, ‘Rebuilding Stakeholder Trust in Business: An Examination of Principle-Centered Leadership and Organizational Transparency in Corporate Governance’ (2008) 113(1) Business and Society Review 99. 27Ross L Watts and Jerald L Zimmerman, Positive Accounting Theory (Prentice-Hall Inc. 1986). 28 3On this, see Doreen J McBarnet, ’Financial Engineering or Legal Engineering? Legal Work, Legal Integrity and the Ban Crisis’ in J O’Brien and Iain Macneil (eds), The Future of Financial Regulation (Hart 2009). 4 30Christian Leuz and Robert E Verrecchia, ‘The Economic Consequences of Increased Disclosure’ (2000) 38 J Accounting Research 91. 31 English law ﬁrms and CSR Many law ﬁrms lay claim to a long history of pro bono work, charitable giving and phi- lanthropic activity.34 However, the genesis of coherent and strategic CSR programmes in 32Sundeep Aulakh and Ian Kilpatrick, ‘Exogenous Jolts, Re-regulation and Change in Professional Organisational Fields: The Case of UK Legal Services’ (Annual Conference on Professional Service Firms, July 2015, Oxford) – copy on ﬁle with the authors. 34For an account of this history, see Richard L Abel, The Legal Profession in England and Wales (Blackwell, 1988); and Andrew Boon and Avis Whyte, ‘Charity and Beating Begins at Home: The Aetiology of the New Culture of Pro Bono Publico’ (1999) 2(2) Legal Ethics 169. LEGAL ETHICS 143 UK law ﬁrms, which encompass these activities and more, only began in the mid-2000s.35 Scholarly writing on the subject of CSR in law ﬁrms reﬂects this slow rise to prominence. More has been written about pro bono and there is an extensive literature on diversity and inclusion. In 1998, Wheeler and Wilson wrote about the divergence between law ﬁrms and their corporate clients in terms of pursing a CSR agenda. They argued that whilst business generally had forged stronger links with the community by that time, law ﬁrms had ‘immersed themselves more deeply in a proﬁt-driven agenda.’36 Wheeler and Wilson show that, following the individualistic conservative politics of the 1980s, the 1990s saw a move towards inclusivity and recognition of ‘duties to community’ and that the legal profession was lagging behind its corporate counterparts in responding to this.37 This work sits within a much wider body of commentary on the commercialisation of large law ﬁrm practice.38 As far as we have been able to identify, the paper by Wheeler and Wilson is the only existing example of signiﬁcant academic consideration of law ﬁrm CSR practice.39 In 2007, Economides and O’Leary raised (but did not answer, either empirically or otherwise) the following question: ‘To what extent are law ﬁrms caught up in the wider trend toward ethicisation of modern businesses and corporate social responsibility?’40 More recently, Whelan and Ziv have looked at how outside counsel guidelines (terms of engagement to which sophisticated clients make their external lawyers sign up) are impacting on law ﬁrm policies (in particular in relation to diversity and as regards the conduct of litigation),41 but did not turn the lens on to wider considerations of law ﬁrms and CSR. Boon and Whyte (n 34) discuss these pressures, and other pressures on ﬁrms, to undertake pro bono. 43Ruth Aguilera and others, ‘Putting the S Back in Corporate Social Responsibility: A Multilevel Theory of Social Change in Organizations.’ (2007) 32 Academy of Management Review 836; Petra Christmann and Glen Taylor, ‘Firm Self-Regulation through International Certiﬁable Standards: Determinants of Symbolic versus Substantive Implementation’ (2006) 37 Journal of International Business Studies 863; and Sankar Sen, CB Bhattacharya and Daniel Korschun, ‘The Role of Corpor- ate Social Responsibility in Strengthening Multiple Stakeholder Relationships: A Field Experiment’ (2006) 34 Journal of the Academy of Marketing Science 158. 35Lisa Keller Glanakos, ‘Corporate Social Responsibility Programs in Law Firms’, Practice Innovations Newsletter (January 2011) <https://info.legalsolutions.thomsonreuters.com/signup/newsletters/practice-innovations/2011-jan/article4.aspx> 36Sally Wheeler and Gary Wilson, ‘Corporate Law Firms and the Spirit of Community’ (1998) 49(3) Northern Ireland Legal Quarterly 241. 37Ibid 254 Boon and Whyte (n 34) discuss these pressures, and other pressures on ﬁrms, to undertake pro bono. 35Lisa Keller Glanakos, ‘Corporate Social Responsibility Programs in Law Firms’, Practice Innovations Newsletter (January 2011) <https://info.legalsolutions.thomsonreuters.com/signup/newsletters/practice-innovations/2011-jan/article4.aspx> 36Sally Wheeler and Gary Wilson, ‘Corporate Law Firms and the Spirit of Community’ (1998) 49(3) Northern Ireland Legal Quarterly 241. 37Ibid, 254. 38For an overview, see Hilary Sommerlad, Sonia Harris-Short, Steven Vaughan and Richard Young, The Futures of Legal Edu- cation and the Legal Profession (Hart, 2015). 39In a 2007 paper, Cameron and Taylor-Sands consider the impact of corporate social responsibility on the conduct of cor- porate litigants and propose ‘a long term approach that involves the courts, complemented by corporations as self-reg- ulators developing internal codes of litigation conduct’: Camille Cameron and Michelle Taylor-Sands, ‘Corporate Governments as Model Litigants’ (2007) 10(2) Legal Ethics 154, 175. 40Kim Economides and Majella O’Leary, ‘The Moral of the Story: Toward an Understanding of Ethics in Organisations and Legal Practice’ (2007) 10(1) Legal Ethics 5, 19. 41Christopher J Whelan and Neta Ziv, ‘Law Firm Ethics in the Shadow of Corporate Social Responsibility’ (University of Oxford Legal Research Paper Series, 2012) 63. 42Boon and Whyte (n 34) discuss these pressures, and other pressures on ﬁrms, to undertake pro bono. 43Ruth Aguilera and others, ‘Putting the S Back in Corporate Social Responsibility: A Multilevel Theory of Social Change in Organizations.’ (2007) 32 Academy of Management Review 836; Petra Christmann and Glen Taylor, ‘Firm Self-Regulation through International Certiﬁable Standards: Determinants of Symbolic versus Substantive Implementation’ (2006) 37 Journal of International Business Studies 863; and Sankar Sen, CB Bhattacharya and Daniel Korschun, ‘The Role of Corpor- ate Social Responsibility in Strengthening Multiple Stakeholder Relationships: A Field Experiment’ (2006) 34 Journal of the Academy of Marketing Science 158. , 38For an overview, see Hilary Sommerlad, Sonia Harris-Short, Steven Vaughan and Richard Young, The Futures of Legal Edu- cation and the Legal Profession (Hart, 2015). 39In a 2007 paper, Cameron and Taylor-Sands consider the impact of corporate social responsibility on the conduct of cor- porate litigants and propose ‘a long term approach that involves the courts, complemented by corporations as self-reg- ulators developing internal codes of litigation conduct’: Camille Cameron and Michelle Taylor-Sands, ‘Corporate Governments as Model Litigants’ (2007) 10(2) Legal Ethics 154, 175. 40Kim Economides and Majella O’Leary ‘The Moral of the Story: Toward an Understanding of Ethics in Organisations and stopher J Whelan and Neta Ziv, ‘Law Firm Ethics in the Shadow of Corporate Social Responsibility’ (University o ord Legal Research Paper Series, 2012) 63. 40Kim Economides and Majella O’Leary, ‘The Moral of the Story: Toward an Understanding of Ethics in Organisations and Legal Practice’ (2007) 10(1) Legal Ethics 5, 19. 39In a 2007 paper, Cameron and Taylor-Sands consider the impact of corporate social responsibility on the conduct of cor- porate litigants and propose ‘a long term approach that involves the courts, complemented by corporations as self-reg- ulators developing internal codes of litigation conduct’: Camille Cameron and Michelle Taylor-Sands, ‘Corporate Governments as Model Litigants’ (2007) 10(2) Legal Ethics 154, 175. 40Kim Economides and Majella O’Leary, ‘The Moral of the Story: Toward an Understanding of Ethics in Organisations and Legal Practice’ (2007) 10(1) Legal Ethics 5, 19. 41Christopher J Whelan and Neta Ziv, ‘Law Firm Ethics in the Shadow of Corporate Social Responsibility’ (University of Oxford Legal Research Paper Series, 2012) 63. 42Boon and Whyte (n 34) discuss these pressures, and other pressures on ﬁrms, to undertake pro bono. 43Ruth Aguilera and others, ‘Putting the S Back in Corporate Social Responsibility: A Multilevel Theory of Social Change in , 38For an overview, see Hilary Sommerlad, Sonia Harris-Short, Steven Vaughan and Richard Young, The Futures of Legal Edu- cation and the Legal Profession (Hart, 2015). 39In a 2007 paper, Cameron and Taylor-Sands consider the impact of corporate social responsibility on the conduct of cor- porate litigants and propose ‘a long term approach that involves the courts, complemented by corporations as self-reg- ulators developing internal codes of litigation conduct’: Camille Cameron and Michelle Taylor-Sands, ‘Corporate Governments as Model Litigants’ (2007) 10(2) Legal Ethics 154, 175. 40Kim Economides and Majella O’Leary, ‘The Moral of the Story: Toward an Understanding of Ethics in Organisations and Legal Practice’ (2007) 10(1) Legal Ethics 5, 19. 41Christopher J Whelan and Neta Ziv, ‘Law Firm Ethics in the Shadow of Corporate Social Responsibility’ (University of Oxford Legal Research Paper Series, 2012) 63. 42Boon and Whyte (n 34) discuss these pressures, and other pressures on ﬁrms, to undertake pro bono. 43Ruth Aguilera and others, ‘Putting the S Back in Corporate Social Responsibility: A Multilevel Theory of Social Change in Organizations.’ (2007) 32 Academy of Management Review 836; Petra Christmann and Glen Taylor, ‘Firm Self-Regulation through International Certiﬁable Standards: Determinants of Symbolic versus Substantive Implementation’ (2006) 37 Journal of International Business Studies 863; and Sankar Sen CB Bhattacharya and Daniel Korschun ‘The Role of Corpor 35Lisa Keller Glanakos, ‘Corporate Social Responsibility Programs in Law Firms’, Practice Innovations Newsletter 2011) htt //i f l l l ti th t / i / l tt / ti i ti /2011 j / ti l 44Edward Weeks, ‘Why Firms Should Embrace CSR’, The Lawyer, 4 December 2006 <http://www.thelawyer.com/why-ﬁrms- should-embrace-csr/123341.article>; Lisa Keller Glanakos, ‘Corporate Social Responsibility Programs in Law Firms’, Prac- tice Innovations Newsletter (January 2011). 45Totum, ‘Diversity and CSR: Why Law Firms are Committing to Change’, The Lawyer, 29 May 2014 <http://www.thelawyer. com/brieﬁngs/diversity-and-csr-why-law-ﬁrms-are-committing-to-change/3021149.article> 46See, for example, Andrew Boon, Liz Duff and Michael Shiner, ‘Career Paths and Choices in a Highly Differentiated Pro- fession: The Position of Newly Qualiﬁed Solicitors’ (2001) 64(4) Modern Law Review 563; and Ronit Dinovitzer and Bryant G. Garth, ‘Lawyer Satisfaction in the Process of Structuring Legal Careers’ (2007) 41 Law & Society Review 1. 47These two arguments are often seen (and set against each other) in the context of debates on diversity in law ﬁrms. For an overview, see Savita Kumra, ‘Busy Doing Nothing: An Exploration of the Disconnect Between Gender Equity Issues Faced by Large Law Firms in the United Kingdom and the Diversity Management Initiatives Devised to Address Them’ (2015) 83(5) Fordham Law Review 2277. 48For readers unfamiliar with the rankings of law ﬁrms in England and Wales, we use ‘(#[rank])’ to give a sense of relative positioning. 49<http://www.wearecr.com/reports.php> English law ﬁrms and CSR We would suggest that demand- side pressures from clients must be exerting some (at present unquantiﬁable) pressure on law ﬁrms to engage in and report on CSR matters.42 This would be consistent with existing empirical evidence on why and how stakeholders pressure corporates to engage in CSR.43 UK law ﬁrms, which encompass these activities and more, only began in the mid-2000s.35 UK law ﬁrms, which encompass these activities and more, only began in the mid-2000s.35 The majority of writing on CSR in law ﬁrms has been written by and aimed at legal practitioners. These pieces suggest that the motivations of ﬁrms in undertaking CSR reﬂect the business case (partly demand-side driven, via speciﬁc clauses in tender docu- ments and law ﬁrm panel appointment processes), positive impacts on law recruitment S. VAUGHAN ET AL. 144 and staff satisfaction,44 together with some suggestions that law ﬁrms adopting CSR pol- icies can lead to increased productivity.45 These pieces from practitioners ﬁt in with wider scholarship on how lawyers perceive the extrinsic and intrinsic beneﬁts of their employ- ment.46 In charting how law ﬁrms report their own CSR practices, we hope not only to provide an overview of current market practice, but also to lay the foundations for further studies that engage with law ﬁrms directly in order to better understand not only what they do in terms of their CSR agenda, but also why they adopt their chosen approach. 44Edward Weeks, ‘Why Firms Should Embrace CSR’, The Lawyer, 4 December 2006 <http://www.thelawyer.com/why-ﬁrms- should-embrace-csr/123341.article>; Lisa Keller Glanakos, ‘Corporate Social Responsibility Programs in Law Firms’, Prac- tice Innovations Newsletter (January 2011). 45T ‘Di i d CSR Wh L Fi C i i Ch ’ Th L 29 M 2014 h // h l um, ‘Diversity and CSR: Why Law Firms are Committing to Change’, The Lawyer, 29 May 2014 <http://www.thelawye m/brieﬁngs/diversity-and-csr-why-law-ﬁrms-are-committing-to-change/3021149.article> readers unfamiliar with the rankings of law ﬁrms in England and Wales, we use ‘(#[rank])’ to give a sense of relativ itioning. tp://www wearecr com/reports php> 50<http://www.mishcon.com/about_us/social_impact> 51<http://www.cliffordchance.com/about_us/corporate_responsibility.html> 52<http://www.cliffordchance.com/content/dam/cliffordchance/CR2013/CRR2013.pdf> 53<http://www.freshﬁelds.com/en/united_kingdom/careers/trainees/About-us/Pro-bono-CSR/> 54<http://www.clarkewillmott.com/Corporate-responsibility/> 55<http://www.lawsociety.org.uk/communities/small-ﬁrms-division/advice/corporate-social-responsibility/>] 56See, for example, David A Jones, ‘Does Serving the Community also Serve the Company? Using Organizational Identiﬁ- cation and Social Exchange Theories to Understand Employee Responses to a Volunteerism Programme’ (2010) 83 Journal of Occupational and Organizational Psychology 857; and Jeremey Galbreath, ‘How does Corporate Social Responsibility Beneﬁt Firms?’ (2010) 22(4) European Business Review 41. 57<http://www.pinsentmasons.com/en/about-us/starﬁsh/> – the website has been updated since our 2014 review and this language no longer appears. Our data Law ﬁrm ranking CSR section Pro bono section Equality and diversity section Environment/sustainability section 1–10a 100% 100% 100% 90% 11–25 93% 60% 93% 87% 26–50 84% 56% 72% 64% 51–100 81% 57% 72% 50% aWe have split the ﬁrms into these groupings because they largely reﬂect the size and turnover of the ﬁrms and because, as we will come to see, there is real variation when we compare the top 10 ﬁrms with other ﬁrms. Table 1. Law ﬁrm website sections. ranking CSR section Pro bono section Equality and diversity section Environment/sustainability section our lawyers to effect the change they want to see in the world.’50 On its website, Clifford Chance (#2) states that they believe that they have a ‘duty to run [their] business respon- sibly’,51 (the moral case) and then, in the ﬁrm’s 2013 CSR report we see the business case come to the fore: ‘we must align our CR and over-arching business strategies. Our CR pro- gramme is therefore focused on those areas that we believe are of greatest relevance to our principal stakeholders and where we have the greatest impact.’52 A number of ﬁrms also linked CSR to the well-being and satisfaction of their employees: Not only is helping others the right thing to do, but those involved also speak highly of the personal satisfaction they gain from their efforts. (Freshﬁelds Bruckhaus Deringer, #3)53 As a professional practice our people are at the very core of our business. It is crucial to the overall wellbeing of our business that our people are treated fairly, with respect and are given the opportunity to nurture and develop their skills. (Clarke Willmott, #72)54 These, and similar, statements suggest that CSR is, at least in part, about keeping people at the ﬁrm happy. We also see this understanding in how the Law Society, the representative body for solicitors in England and Wales, frames CSR in terms of reinvesting ‘some of the resources located in business into the wider and less privileged community, while also internally investing in a healthier, happier, more diverse and productive workforce.’55 The relationship between CSR programmes and employee satisfaction has been found in other areas of corporate activity outside of legal services.56 Some of the statements as to why law ﬁrms engage in CSR are tantalisingly vague. Our data As part of our review of law ﬁrm websites in the summer of 2014, we were interested in whether and how ﬁrms reported on CSR, whether those ﬁrms had a CSR section to their website and, if so, what those sections contained; whether (and for how long) formal CSR reports had been produced by the ﬁrms; and what the ﬁrms categorised as constituting CSR activity (pro bono, diversity, environment etc). We report our results in the sections which follow. By way of overview, Table 1 sets out the percentages of ﬁrms which had sep- arate sections on their websites for each of CSR, pro bono, equality and diversity, and environmental matters/sustainability. What is most striking, but is perhaps also unsurpris- ing, is the sliding scale from top to bottom in terms of which ﬁrms do what in the context of CSR. In total, 88 of the top 100 law ﬁrms had a CSR section to their websites. However, only 21 of these 88 ﬁrms (24%) explained on those websites, or in associated CSR policies or annual statements, why they reported on CSR and/or why they felt CSR was important for them as a ﬁrm. Where ﬁrms did give reasons (and these tended to be the larger ﬁrms), these reasons varied greatly and reﬂected a mix of business case and moral case arguments.47 For example, on the latter, Olswang (#33)48 comments that: In sharing what we have and what we know and by collaborating with others we make a positive and sustainable impact on the people with whom we do business, on the commu- nity and on the environment. In viewing ourselves in this way, we ensure that CR [corporate responsibility] inﬂuences all that we do and gain clarity of our role as a responsible business.49 Similarly, Mischon de Reya (#39) channels Mahatma Gandhi when it says that, ‘In November 2013 we launched our new social impact programme, developed to inspire LEGAL ETHICS 145 Table 1. Law ﬁrm website sections. Our data For example, in 2014 Pinsent Masons (#14) stated that, ‘there are many good reasons why we, as a business, need to be aware of the impact we have on our neighborhood’,57 but does not go on to say what these reasons are. Though the data to support this hypothesis does not exist in the public domain, we would go so far as to question whether some law ﬁrms are engaging with and reporting on CSR simply because their competitor ﬁrms (and/or their clients) are doing so, without really asking themselves why it is a good idea. Despite the majority of ﬁrms saying something about CSR, we question whether this is a result of S. VAUGHAN ET AL. 146 dedication by those ﬁrms to the principles of CSR and/or a knee jerk reaction to the implementation of CSR strategies by their competitors. As such, a public commitment to CSR by some law ﬁrms may have greater symbolic than substantive weight. Those ﬁrms which do have CSR sections on their websites vary greatly as to where those sections are found and what they are called. In some ﬁrms, CSR is one of the home page sections; for others in ‘About Us’ pages. This, in and of itself, is not particularly interesting, but does speak to the ease with which third party stakeholders (clients, potential clients, competitors, regulators, employees, potential lateral hires etc) can compare and contrast data. Of the top 10 ﬁrms, Slaughter and May (#10) stands out for being the ﬁrm where information on CSR is hardest to ﬁnd: its data are tucked away under the ‘Facts and Figures’ pages of the ﬁrm’s ‘Who We Are’ section.58 Most ﬁrms use the catch-all of CSR or ‘Corporate Responsibility’ but others, such as Freshﬁelds (#3) and Clifford Chance (#2), have framed these matters in terms of ‘Responsible Business,’59 and Linkla- ters (#4) talks of ‘Collective Responsibility’.60 We had expected, before undertaking our review of the ﬁrm websites, that the higher the ranking of the ﬁrm, the better its disclosures on CSR would be. This, however, was not necessarily the case. 58 <http://www.slaughterandmay.com/who-we-are/key-facts-and-ﬁgures.aspx> 59<http://www.freshﬁelds.com/en/global/who_we_are/Reporting_and_policy_global/> 60<http://www.linklaters.com/Responsibility/Pages/Index.aspx> 61<http://www.simmons-simmons.com/en/About-Us/Corporate-Responsibility> 62<http://www.osborneclarke.com/about-us/> 63<http://www.burges-salmon.com/who_we_are/corporate_responsibility/default.aspx> 64<http://www.freshﬁelds.com/uploadedFiles/Locations/Global/Who_we_are/Reporting/2005-6CSR.pdf> – one of the authors (Vaughan) was a trainee who helped to work on this report, authored by his supervising partner Paul Watchman. 65<https://www.slaughterandmay.com/media/1430833/corporate-responsibility-brochure.pdf> 66Instructive starting points are to be found in the work of Deborah L Rhode and of Scott Cummings. 67Judith L Maute, ‘Changing Conceptions of Lawyers’ Pro Bono Responsibilities: From Chance Noblesse Oblige to Stated Expectations’ (2002) 77 Tulane Law Review Association 91. Our data Simmons & Simmons (#15), for example, has a more voluminous and accessible CSR section to its website than Herbert Smith Freehills (#9).61 Osborne Clark (#31) says fewer than 150 words about CSR;62 whereas Burges Salmon (#46) has a series of relatively detailed pages on CSR.63 As far as we could determine, all bar two of the top ten UK law ﬁrms produce annual CSR reports, a tradition that began in 2005/06 with Freshﬁelds (#3).64 Most of these reports are around 20 pages long. However, the latest Freshﬁelds (#3) report comprises more than 50 pages and the latest Slaughter and May (#10) report is 40 pages long.65 This trend, for the production of CSR reports, is not seen with the smaller ﬁrms in the top 100, as Table 2 shows. p p In the sections that follow we look at speciﬁc aspects of CSR in law ﬁrms: pro bono and community giving; diversity; and the environment and sustainability. 68Robert Granﬁeld and Lynn Mather, ‘Pro Bono, The Public Good and the Legal Profession’ in Robert Granﬁeld and Lynn Mather (eds), Private Lawyers and the Public Interest (OUP 2009) 2. 69Deborah L Rhode, ‘Rethinking the Public in Lawyers’ Public Service’ (2009) 77(3) Fordham Law Review 1435, 1435. 70Hilary Sommerlad, ‘Researching and Theorizing the Processes of Professional Identity Formation’ (2007) 34(2) Journal of Law & Society 190. 71Scott L Cummings and Deborah L Rhode, ‘Managing Pro Bono: Doing Well by Doing Better’ (2010) 78 Fordham Law Review 2359. 72<http://www.americanlawyer.com/id=1202730400870/Pro-Bono-Report-2015-Treading-Water> 73Amit Jain Chauradia and Deepak Somaya, ‘Is Pro Bono Also Pro Lucrum? Doing Well by Letting Your Human Capital Do Good’ (2014) Academy of Management Proceedings. Copy on ﬁle with the authors. The suggestion in the paper is that law ﬁrms with longer histories of pro bono schemes have happier employees who in turn contribute more and better to the law ﬁrm’s ﬁnancial performance. 74<http://l2b.thelawyer.com/solicitor/news/arnold-and-porter-tops-thomson-reuters-pro-bono-index-but-only-26-ﬁrms-in- england-provided-data/3035637.article> 75Adam Grant, ‘Giving Time, Time after Time: Work Design and Sustained Employee Participation in Corporate Volunteer- ing’ (2012) 37(4) Academy of Management Review 589. Pro bono publico and ‘community giving’ By comparison with writing on law ﬁrms and CSR, there is a signiﬁcant body of scholar- ship on lawyers, ﬁrms and pro bono, more so in relation to North America than the UK.66 Since the mid-twentieth century, the phrase ‘pro bono public’ has signalled the giving of legal advice by lawyers without compensation.67 For some, there is recognition that lawyers have a professional responsibility to provide legal representation for free although, as Granﬁeld and Mather argue, ‘that recognition has been far stronger in theory than in LEGAL ETHICS 147 LEGAL ETHICS 147 Table 2. Firms with a CSR report available on their website as at August 2014. Law ﬁrm ranking Percentage with a CSR report available on their website as at August 2014 1–10 80% 11–25 27% 26–50 24% 51–100 4% Total 20% rms with a CSR report available on their website as at August 2014. king Percentage with a CSR report available on their website as at August 2014 practice.’68 Equally, Rhode argues that, ‘in practice, pro bono has never been only about what is good for the public. It has also been about what is good for lawyers: what will enhance their reputation, experience, contacts and relationships?’69 As such, pro bono might be thought to be a part of the cultural capital of lawyers in large ﬁrms.70 Cummings and Rhode show how, for some law ﬁrms, the selection of pro bono cases has more to do with the pedagogic value of enhancing certain associate skill sets than with any inherent worthiness of the pro bono matter.71 In the United States, the trade publication American Lawyer publishes, on an annual basis, the commitment to pro bono by the top 200 US law ﬁrms (hours worked, nature of work undertaken etc).72 This is a useful and important dataset not found in England and Wales. Recent empirical work using this data has found that long term commitment by US law ﬁrms to pro bono is positively correlated with the ﬁnancial performance of the ﬁrm.73 Where the media has asked English ﬁrms for data on pro bono, the results have been disappointingly poor.74 Below, we draw out the data we were able to glean from the top 100 law ﬁrm websites in the summer of 2014. We split our data between what we might think of as ‘pure’ pro bono (i.e. Pro bono publico and ‘community giving’ the giving of free legal advice) and other chari- table acts that beneﬁt the public good but which do not involve the giving of legal advice. 68Robert Granﬁeld and Lynn Mather, ‘Pro Bono, The Public Good and the Legal Profession’ in Robert Granﬁeld and Lynn Mather (eds), Private Lawyers and the Public Interest (OUP 2009) 2. 76Boon and Whyte (n 34) 190. 77<http://www.dlapiperbreakintolaw.com/> 78<http://www.boodlehatﬁeld.com/the-ﬁrm/corporate-social-responsibility.aspx> 79For example, <https://sites-blm.vuturevx.com/54/448/july-2014/ﬁnal-version-16-july.asp> 80DLA Piper has pledged $6.5 million to UNICEF’s global child justice programme (<http://www.dlapiperprobono.com/ what-we-do/signature/pro-bono/unicef_sig.html>) 81<http://www.cliffordchance.com/about_us/corporate_responsibility/community_pro_bono.html> 82Pinsent Masons asserts that its foundation has made donations of over £260,000 to charities and projects that ‘inspire young lives’ since its formation in 2009. Link on website no longer live. Similarly, Mills & Reeve (#47) refers to over £80,000 being raised for a range of causes in recent years, then goes on to say that an annual donation of £5,000 is made to help young people from disadvantaged backgrounds gain access to higher education. It is not clear whether this is taken from the aforementioned £80,000 or is an additional sum. 83<http://www.tltsolicitors.com/about-us/corporate-social-responsibility/> 84<http://www.cliffordchance.com/about_us/corporate_responsibility.html> Community giving Law ﬁrms of all sizes engage in a diverse range of community giving activity, by which we mean donating money to charity and/or facilitating employee involvement in charitable activities, such as redecorating a charity’s premises, teaching school children to read, taking part in a sponsored event etc. This mirrors volunteering programs in the corporate sphere.75 Boon and Whyte suggest that community based activities by law ﬁrms open up the possibility of ‘giving back’ to transactional lawyers and non-fee earning staff who S. VAUGHAN ET AL. 148 might otherwise be unable to offer pro bono legal advice.76 Due to the lack of consistency in the style and content of CSR reporting by the law ﬁrms we reviewed, it has not been possible for us to gain an accurate account of how many ﬁrms undertook particular types of each community giving activity. However, some patterns can be gleaned. Certain causes appear to attract charitable attention from large law ﬁrms across the top 100. For example, and as Table 3 below shows, a focus on young people is a common denominator throughout the top 100 ﬁrms, seen at either end of the spectrum. DLA Piper (#1) runs a scheme entitled ‘Break into Law’,77 which provides mentoring, employ- ability skills workshops and scholarships (amongst other things) for young people consid- ering a career in the legal profession. At the other end, Boodle Hatﬁeld (#99) operates a reading scheme with a local primary school.78 Within the scope of community giving, a number of law ﬁrms also make ﬁnancial donations to charitable causes. The amount donated varies signiﬁcantly and can range from relatively small sums raised by employees through sponsored activities and ‘dress down days’,79 to signiﬁcant pecuniary contributions made by the ﬁrm or its charitable trust to global NGOs or charities.80 We found no consistency in terms of whether or how such donations are reported by ﬁrms. Some make reference to their charitable trusts on their websites and disclose the amount donated each year. 85For example, <http://www.dlapiperprobono.com/news/awards/>; <http://www.freshﬁelds.com/en/global/who_we_are/ Community_investment_global/>; <https://www.slaughterandmay.com/what-we-do/community-and-environment/ recent-news.aspx> 86Boon and Abbey (n 16). 87Boon and Abbey (n 16) 649. 88Boon and Abbey (n 16) 654. 89Boon and Whyte (n 34). Community giving For example, Clifford Chance (#2) states that the Clifford Chance Foundation has a budget of £1 million per year which is donated to support a host of local and global charities.81 Others, such as Pinsent Masons (#14), provide approximate ﬁgures and refer to total donations made over a period of years.82 A third group of ﬁrms, such as TLT (#54), are entirely silent on the subject of how much they donate to charity, simply mentioning that sponsored and fun- draising activities are undertaken by employees.83 Our review of law ﬁrm websites also suggests that the larger the ﬁrm, the more likely it is to publicly disclose its collaboration with clients on community giving activities. Table 4 below shows the number of ﬁrms that referred to collaborations with clients on the CSR pages of their website. Evidently, larger ﬁrms place a greater public emphasis on the value of client collaboration and it would appear that some recognise the commercial beneﬁts of such collaboration: We look for ways to strengthen our relationships with clients by collaborating around CR. By combining our efforts [with clients] we are able to mobilise greater resource than we could individually, and our shared experiences deepen our relationship in a way that is personally and professionally rewarding.84 (Clifford Chance, #2) LEGAL ETHICS 149 149 Table 3. ‘Community giving’ projects involving young people. Law ﬁrm ranking Percentage mentioning projects involving young people on their website 1–10 80% 11–25 60% 26–50 36% 51–100 39% Total 46% Table 4. Firms that mention client collaboration in CSR activities on their website. Law ﬁrm ranking Percentage mentioning client collaboration in CSR activities on their website 1–10 70% 11–25 20% 26–50 12% 51–100 4% The signiﬁcantly greater degree of client engagement in CSR activities reported amongst the largest ﬁrms, coupled with the higher proportion of annual reporting, suggests a coor- dinated and strategic approach to CSR amongst the top 10 ﬁrms which is not present even amongst their closest competitors, as is evidenced by the sharp decline in both immedi- ately outside of the top 10. Notwithstanding the above, many top 50 ﬁrms highlight their CSR achievements on their websites. Again, this is more prevalent in the largest ﬁrms, with 9 out of 10 of the top 10 mentioning awards received for community giving and/or pro bono initiatives.85 Pro bono Despite pro bono being a recognised and relatively well-established constituent part of the practices of large law ﬁrms, academic consideration of English law ﬁrms’ pro bono activity remains underdeveloped. A notable exception is found in the work of Andrew Boon. In 1997 Boon and Abbey published the results of a study that mapped the pro bono work undertaken at 61 large UK law ﬁrms.86 This work was carried out at a time when ‘a new culture of pro bono publico [was] in the process of formation’,87 with large law ﬁrms in particular beginning to adopt a coordinated approach to pro bono activity. Despite the authors citing this work as a ‘benchmark for future efforts’88 to consider the pro bono practices in UK law ﬁrms, we are not aware of any comparable work having been conducted subsequently. In later work, Boon and Whyte explore the ‘aetiol- ogy of the emerging culture of pro bono public.’89 Arguing for the importance of seeing law ﬁrm CSR in its social, political and organisational contexts, they demonstrate that client pressure and growing commitment by small groups of individuals inside large law ﬁrms had impacts on why and how large ﬁrms undertake pro bono. Their research shows that pro bono in the large ﬁrms, in the 1990s at least, was partly driven by 150 S. VAUGHAN ET AL. S. VAUGHAN ET AL. 150 Table 5. Firms that mention awards linked to community giving/pro bono activity on their website. Law ﬁrm ranking Percentage mentioning awards linked to community giving/pro bono activity 1–10 90% 11–25 67% 26–50 24% 51–100 2% enthusiasm at the bottom end (trainees, junior associates),90 and met by ‘clear leadership from the top.’91 Whether this would remain the case today is unclear,92 and is worthy of further qualitative exploration. Pro bono Certainly, for many large US law ﬁrms the ﬁnancial crisis has been shown to have had a signiﬁcant impact on their pro bono contributions.93 The Law Society of England and Wales conducts an annual survey which asks a random sample of circa 1500 solicitors about pro bono work that they have undertaken.94 This data has been collected in a consistent manner from 2010.95 It charts, amongst other matters: the number of solicitors to have undertaken pro bono work during the preceding 12 months; the proportion of solicitors providing pro bono work by size of ﬁrm; the perceived adequacy of opportunity for private practice solicitors to get involved in pro bono work; and the annual ﬁnancial value of private practice pro bono work. The survey does not identify any of the participants by ﬁrm, meaning that no conclusions can be drawn as to the institutional approach to pro bono activity.96 However, a comparison between the 2010 and the 2014 surveys reveals that the percentage of solicitors to have undertaken some pro bono during the 12 months prior to each survey has remained constant at 42%.97 It is worth noting that the deﬁnition of pro bono used by the surveyors in 2010 differs from that used in 2014. In 2010, pro bono was broadly deﬁned as the ‘delivery of free legal ser- vices to individuals, organisations and communities in need’.98 From 2012, the narrower deﬁnition, contained in The Pro Bono Protocol,99 is instead used: y 92In a 2005 piece, Boon refers to an anecdote about lawyers from a magic circle ﬁrm being discouraged from doing pro bono work. See Andrew Boon, ‘From Public Service to Service Industry: the Impact of Socialisation and Work on the Motivation and Values of Lawyers’ (2005) 12(2) International Journal of the Legal Profession 229, 230. y ( ) ( ) g , 93Scott L Cummings and Rebecca L Sandefur ‘Beyond the Numbers: What We Know – and Should Know – About American Pro Bono’ (2013) 7 Harvard Law & Policy Review 83, 84. 94For the 2014 survey see, <http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/pro- bono-work-pc-holder-survey-2014/> 95Surveys on pro bono activity were also conducted in 2002, 2007 and 2009 by The Law Society of England and Wales. In 2002 and 2009 data on pro bono was collected via surveys which included questions on a broader range of issues affect- ing legal practitioners. 90Echoing ﬁndings by Webley. See Lisa Webley, ‘Pro Bono and Young Solicitors: Views from the Front Line’ (2001) 3(2) Legal Ethics 152. 91Boon and Whyte (n 34) 173. 92 y 92In a 2005 piece, Boon refers to an anecdote about lawyers from a magic circle ﬁrm being discouraged from doing pro bono work. See Andrew Boon, ‘From Public Service to Service Industry: the Impact of Socialisation and Work on the Motivation and Values of Lawyers’ (2005) 12(2) International Journal of the Legal Profession 229, 230. 90Echoing ﬁndings by Webley. See Lisa Webley, ‘Pro Bono and Young Solicitors: Views from the Front Line’ (2001) 3(2) Legal Ethics 152. 91Boon and Whyte (n 34) 173. 92In a 2005 piece, Boon refers to an anecdote about lawyers from a magic circle ﬁrm being discouraged from doing pro bono work. See Andrew Boon, ‘From Public Service to Service Industry: the Impact of Socialisation and Work on the Motivation and Values of Lawyers’ (2005) 12(2) International Journal of the Legal Profession 229, 230. 93Scott L Cummings and Rebecca L Sandefur ‘Beyond the Numbers: What We Know – and Should Know – About American Pro Bono’ (2013) 7 Harvard Law & Policy Review 83, 84. 94For the 2014 survey see, <http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/pro- bono-work-pc-holder-survey-2014/> 95Surveys on pro bono activity were also conducted in 2002, 2007 and 2009 by The Law Society of England and Wales. In 2002 and 2009 data on pro bono was collected via surveys which included questions on a broader range of issues affect- ing legal practitioners. The 2007 report was commissioned to inform the publicity if National Pro Bono week and there- fore focused solely on pro bono. Copies of the reports of these surveys are on ﬁle with the authors. For the results of the 2010 survey see <http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/solicitors-pro- bono-work—omnibus-survey-2010/> 96The 2007 survey did ask some questions at ﬁrm level, thereby offering a ﬁrm level perspective. This practice has not been continued, therefore it is impossible to chart any changes in attitudes or practices at that level. 97The 2010 report refers to 42% of practising certiﬁcate holders undertaking pro bono in the previous 12 months. The 2014 report simply refers to ‘solicitors’. It is not clear whether the constituents of these groups differ. 98<http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/solicitors-pro-bono-work- omnibus-survey-2010/>. For a wider commentary on the challenges in deﬁning pro bono, see Andrew Boon and Avis Whyte, ‘Pro Bono Publico’ (2000) 150 New Law Journal 1564. 99Best practice guidelines endorsed by the Law Society, Bar Council and CILEx: <http://lawworks.org.uk/pro-bono- y 95Surveys on pro bono activity were also conducted in 2002, 2007 and 2009 by The Law Society of England and Wales. In 2002 and 2009 data on pro bono was collected via surveys which included questions on a broader range of issues affect- ing legal practitioners. The 2007 report was commissioned to inform the publicity if National Pro Bono week and there- fore focused solely on pro bono. Copies of the reports of these surveys are on ﬁle with the authors. For the results of the 2010 survey see <http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/solicitors-pro- bono-work—omnibus-survey-2010/> y ( ) ( ) g , 93Scott L Cummings and Rebecca L Sandefur ‘Beyond the Numbers: What We Know – and Should Know – About American Pro Bono’ (2013) 7 Harvard Law & Policy Review 83, 84. ( ) y , 94For the 2014 survey see, <http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/pro- bono-work-pc-holder-survey-2014/> 100<http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/pro-bono-work-pc-holder- survey-2013/> 101On which, see the various accounts in: Robert Granﬁeld and Lynn Mather, Private Lawyers and the Public Interest (OUP 2009). 102<http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/pro-bono-work-pc-holder- survey-2014/> 103LawWorks is the registered operating name of the Solicitors Pro Bono Group. 104This report also used the Pro Bono Protocol deﬁnition of Pro Bono: <http://www.lawworks.org.uk/tmp_downloads/ w87n101n14c35h108x83k18q138q119x71l98q96x125t91n79/lawworks-pro-bono-survey-report-2012-ﬁnal.pdf> 105<http://lawworks.org.uk/tmp_downloads/k78w57k26g23q65y112j98t105g59t60j49p78m38f32q106/lawworks-2013- pro-bono-report-ﬁnal.pdf> 106For example, see <http://www.dacbeachcroft.com/db812f0308d94210a2112bfc97c2f6b6/voluntary-work-charitable- giving> Pro bono The 2007 report was commissioned to inform the publicity if National Pro Bono week and there- fore focused solely on pro bono. Copies of the reports of these surveys are on ﬁle with the authors. For the results of the 2010 survey see <http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/solicitors-pro- bono-work—omnibus-survey-2010/> y 96The 2007 survey did ask some questions at ﬁrm level, thereby offering a ﬁrm level perspective. This practice has not been continued, therefore it is impossible to chart any changes in attitudes or practices at that level. 0 report refers to 42% of practising certiﬁcate holders undertaking pro bono in the previous 12 months. The 2014 The 2010 report refers to 42% of practising certiﬁcate holders undertaking pro bono in the previous 12 months. The 2014 report simply refers to ‘solicitors’. It is not clear whether the constituents of these groups differ. p p g g p p report simply refers to ‘solicitors’. It is not clear whether the constituents of these groups differ. 98 98<http://www.lawsociety.org.uk/policy-campaigns/research-trends/research-publications/solicitors-pro-bono-work- omnibus-survey-2010/>. For a wider commentary on the challenges in deﬁning pro bono, see Andrew Boon and Avis Whyte, ‘Pro Bono Publico’ (2000) 150 New Law Journal 1564. p y g p y p g p p omnibus-survey-2010/>. For a wider commentary on the challenges in deﬁning pro bono, see Andrew Boon and Avis Whyte, ‘Pro Bono Publico’ (2000) 150 New Law Journal 1564. omnibus-survey-2010/>. For a wider commentary on the challenges in deﬁning pro bono, see Andrew Boon and Avis Whyte, ‘Pro Bono Publico’ (2000) 150 New Law Journal 1564. y 99Best practice guidelines endorsed by the Law Society, Bar Council and CILEx: <http://lawworks.org.uk/pro-bono- protocol> 99Best practice guidelines endorsed by the Law Society, Bar Council and CILEx: <http://lawworks.org.uk/pro-bono- protocol> GAL ETHICS 151 151 151 LEGAL ETHICS 151 151 LEGAL ETHICS Legal advice or representation provided by lawyers in the public interest including to indi- viduals, charities and community groups who cannot afford to pay for that advice or rep- resentation and where public funding is not available. Pro bono Legal work is Pro Bono Legal Work only if it is free to the client, without payment to the lawyer or law ﬁrm (regardless of the outcome) and provided voluntarily either by the lawyer or his or her ﬁrm.100 We wonder whether this revised deﬁnition was put in place to avoid lawyers and ﬁrms claiming that work in which they were unable to recover fees was undertaken pro bono (as has happened in the US).101 The average (self disclosed) number of pro bono hours worked by solicitors surveyed by the Law Society increased from 45 hours per annum in 2010 to 52 hours in 2014 and the (self-) estimated ﬁnancial value of pro bono work by private practitioners increased from £476 million (2.3% of total gross fee income) in 2010 to £601 million (equivalent to approximately 2.8% of total turnover generated by solicitors’ ﬁrms) in 2014.102 In 2012, LawWorks, an organisation based in London that acts as a facilitator between law ﬁrms and clients seeking pro bono legal advice,103 conducted its ﬁrst annual pro bono survey.104 The 2013 survey details the responses from 119 law ﬁrms and outlines, amongst other things, which level of fee earners undertake pro bono; who is receiving pro bono assistance; and how law ﬁrms recognise time spent on pro bono matters. The report notes that its sample size is not large enough to be representative, but it does nonetheless give some insights. We learn, for example, that law ﬁrm pro bono work is predominantly carried out by solicitors and trainees (rather than partners or other employees); amongst larger ﬁrms 86% of respondents reported that their organisations actively encourage par- ticipation in pro bono work; and that two thirds of the largest organisations recognise pro bono in chargeable hours targets.105 Unfortunately, the report does not deﬁne what is meant by ‘the largest organisations’ and so we cannot meaningfully compare these ﬁnd- ings with our own data. From our summer 2014 review, Table 6 indicates how many of the top 100 include a distinct section on their website relating to their pro bono work. Once again, the difference between the top 10 and other large ﬁrms is striking. In some cases, ﬁrms without a distinct section for pro bono do make some other mention of it on their website. 107<http://www.birketts.co.uk/about-us/csr.aspx> 108For example, Anderson Strathern (#97) (<http://www.andersonstrathern.co.uk/about/corporate-social-responsibility/>). 109<http://www.cliffordchance.com/about_us/corporate_responsibility/community_pro_bono/access-to-justice.html> 110<http://www.freshﬁelds.com/uploadedFiles/Locations/Global/Who_we_are_new/CR_Reporting/CR_Report_2013.pdf> Pro bono Here, however, pro bono is often dealt with brieﬂy and under a generic heading, such as ‘community’, and is detailed alongside other CSR initiat- ives such as volunteering.106 In Table 7 we compare the number of ﬁrms that include a CSR section on their website with the number to include a distinct pro bono section. It is clear that whilst the majority of ﬁrms exhibit a commitment to CSR, notably fewer outside of the top 10 publicly categorised pro bono activity as a distinct area of their CSR activity. 152 S. VAUGHAN ET AL. 152 S. VAUGHAN ET AL. Table 6. Pro bono content on law ﬁrm websites. Law ﬁrm ranking Percentage with distinct pro bono section on website 1–10 100% 11–25 53% 26–50 56% 51–100 57% Percentage with distinct pro bono section on website Table 7. CSR and pro bono content on websites. Law ﬁrm ranking Percentage with distinct CSR section on website Percentage with distinct pro bono section on website 1–10 100% 100% 11–25 93% 53% 26–50 84% 56% 51–100 82% 57% 107<http://www.birketts.co.uk/about-us/csr.aspx> 108 Trends in pro bono activity On a per capita basis, that breaks down as just 8.89 hours of time spent on pro bono per employee LEGAL ETHICS 153 or only 17.29 hours per lawyer per annum.111 Similarly, in its ‘Corporate Responsibility Report 2013′, Clifford Chance (#2) disclosed that it had undertaken 55,348 pro bono hours globally in 2012–2013, or 18.3 hours per full time employee.112 These numbers, while now at least two years out of date, fall far below the aspirational targets to which that some ﬁrms now publicly ascribe. For example, at the time of writing, 27 ﬁrms (including Freshﬁelds (#3) and Clifford Chance (#2)) have signed up to ‘The Col- laborative Plan’, a group of law ﬁrms aimed at improving access to justice through pro bono in the UK, which requires ﬁrms to agree to an aspirational target of 25 hours of pro bono work, per fee earner per year.113 In the United States, Rule 6.1 of the American Bar Association’s Module Rules of Professional Conduct sets out that, ‘Every lawyer has a professional responsibility to provide legal services to those unable to pay. A lawyer should aspire to render at least (50) hours of pro bono publico legal services per year.’114 We have no comparable aspiration in the rules of professional conduct for solicitors in England and Wales, and would question whether this lack of aspiration (substantiated and promoted by a regulatory or representative body) remains appropriate. The amount of time (when disclosed) that ﬁrms spent on pro bono/community giving by ﬁrms outside of the top 10 is signiﬁcantly less than their top 10 counterparts. The lowest number of hours spent per annum on pro bono activities by a top 10 law ﬁrm in 2012– 2013 was 10,099 (CMS Cameron McKenna #8), followed by 25,000 (Linklaters #4). Ever- sheds (#11) disclosed just 3,092 pro bono hours in the year 2012–13, although they did report 10,540 of community giving activities.115 Nine ﬁrms in the top 10 disclosed hours on pro bono and/or community. These 9 said that they gave, in total, 653,420 hours of their time.116 Only a further 12 ﬁrms in the rankings 11–100 disclosed their pro bono and/or community hourages. Trends in pro bono activity We encountered a number of difﬁculties in seeking to analyse the data available on pro bono practices by large English law ﬁrms. An initial challenge was that not all ﬁrms differ- entiate between pro bono work and community giving more generally, or do not provide sufﬁcient information on their websites to allow us to understand the nature or amount of pro bono work undertaken. Birketts LLP (#81) is a typical example: We are members of the Suffolk, Norfolk and Cambridgeshire branches of the Pro-Help Group, one of many schemes created by the ‘Business in the Community’ organisation. The aim is to provide a broad range of professional services to locally based charities and the voluntary sector free of charge and the group’s members include local ﬁrms of solicitors, accountants, surveyors and a host of other professionals. We also participate in a local pro- bono initiative with the object of providing legal assistance and support to small local groups principally concerned with the interests of ethnic minorities and the disadvantaged.107 Clearly, the ﬁrm does undertake some pro bono activity, but it is not possible to ascertain how much or what type of work is carried out. Other ﬁrms provide even less detail, out- lining a general commitment to CSR, community giving and/or pro bono, without giving any examples of projects or work undertaken.108 International ﬁrms tend to tailor their pro bono offerings to the different countries in which they operate. For example, Clifford Chance (#2) offers advice sessions at a law centre in London, whilst in Hong Kong, Tokyo and Sydney it provides casework support for asylum seekers.109 Despite running local pro- jects, ﬁgures given by ﬁrms detailing the extent of pro bono participation are often global. For example, Freshﬁelds (#3) reported in its 2012–2013 ‘Responsible Business Report’ that, across the ﬁrm, it worked on 431 matters for 253 clients, recording 43,212 pro bono hours.110 However, we do not know how this breaks down between the ﬁrm’s many ofﬁces. Furthermore, the same report states that Freshﬁelds employed 4,859 employees globally during the same year, over 2,500 of whom were lawyers. 111Using the ﬁgure of 2,500 lawyers in our calculation. 112<http://www.cliffordchance.com/content/dam/cliffordchance/CR2013/CRR2013.pdf> 113<http://www.trust.org/spotlight/Collaborative-Plan-for-Pro-Bono-uk/?tab=methodology> 114See <http://www.americanbar.org/groups/probono_public_service/policy/aba_model_rule_6_1.html> 115<http://viewer.zmags.com/publication/4db70ae2#/4db70ae2/2> 116Some of these were global ﬁgures, for example, DLA Piper quoted 207,000 of pro bono work globally (<http://www. dlapiperwin.com/about/corporate-responsibility.html>). Other ﬁrms quoted a minimum ﬁgure, for example, Herbert Smith Freehills state that they gave more than 52,000 hour of pro bono advice in 2013 (<http://www. herbertsmithfreehills.com/about-us/pro-bono-and-citizenship>) and in some cases was difﬁcult to determine whether ﬁgures given for community giving were inclusive of pro bono hours, or were in addition to pro bono hours, see for example, p 27 of CMS Cameron McKenna’s Corporate Social Responsibility Report 2013: <http://www.cms-cmck.com/ Corporate-Social-Responsibility-Report-2013—Joining-the-Dots-07-11-2013> 117Of these, some ﬁgures were again minimums, for example, the 2013 Gateleys CSR Report stated ‘During 2013 we donated over 150 hours of pro bono activities’. Others were admitted to be estimates. For example, Stewarts Law state on their website that ‘We estimate that we provide circa 10,000 hours of free legal advice and assistance on an annual basis’: <http://www.stewartslaw.com/about-us/corporate-social-responsibility.aspx> 118See <http://www.americanlawyer.com/id=1202730400870/Pro-Bono-Report-2015-Treading-Water#ixzz3hBGdbAer> 119Boon and Whyte (n 34) 189. Trends in pro bono activity These totalled 62,023.5 hours.117 By comparison, in the US the top 200 law ﬁrms (self) reported performing a total of 4.75 million domestic pro bono hours in 2014, and 11 of those US ﬁrms reported an average of 100 or more domestic pro bono hours per lawyer.118 This UK/US disparity has been present for some time.119 However, and as set out in Table 8, the majority of English ﬁrms in the top 100 did not disclose any statistics at all regarding the extent of their pro bono work, making a true comparison between ﬁrms and/or between jurisdictions impossible. g p j p Many ﬁrms grouped time employees spent on community giving and pro bono activities together when disclosing total hours: for example, DWF (#20) disclosed in its 154 S. VAUGHAN ET AL. Table 8. Hard data on pro bono and community giving contributions. Law ﬁrm ranking Percentage of ﬁrms to disclose total hours spent on pro bono and/or community giving activities Percentage of ﬁrms to disclose hours spent on community giving/pro bono activities by non-fee earners/support staff Percentage of ﬁrms to disclose total ﬁnancial value of pro bono and/or community giving activity 1–10 90% 20% 50% 11–25 20% 0% 13% 26–50 24% 0% 8% 51–100 6% 0% 0% 2012–2013 ‘Annual CR Review’ that 8,397 hours had been given to pro bono and commu- nity activities by fee earners.120 Other ﬁrms failed to state whether the ﬁgures disclosed related to community giving, pro bono or both. As a result, we have grouped these disclos- ures together in Table 8. Only ﬁve ﬁrms out of the entire top 100 provided a distinct ﬁgure for number of hours spent on community giving activities, as a separate category from pro bono.121 Relatedly, outside of the top 10 ﬁrms, very few disclosed the ﬁnancial value of their pro bono or community giving activities. Our review of law ﬁrm websites suggests a signiﬁcant amount of philanthropic endea- vor on the part of the legal profession in England and Wales based in the largest ﬁrms. However, a number of ﬁrms elide their pro bono and community giving efforts. We would suggest that this is a mistake. We were surprised by the lack of data across the top 100 on the speciﬁc pro bono efforts of law ﬁrms. 120<http://www.dwf.co.uk/media/582033/DWF-CR-Review-2012-13.pdf> 121Linklaters (<http://www.linklaters.com/pdfs/cr/CR_REPORT_2014.pdf>); Eversheds (<http://viewer.zmags.com/ publication/4db70ae2#/4db70ae2/1>); Burges Salmon (<http://www.burges-salmon.com/who_we_are/corporate_re- sponsibility/volunteering_and_trusteeships/>); Shoosmiths (<http://www.shoosmiths.co.uk/ﬁles/Shoosmiths_CR_ report_2012-13.pdf>); and CMS Cameron McKenna (although it is not clear whether these are in fact two distinct ﬁgures) (<http://www.cms-cmck.com/Corporate-Social-Responsibility-Report-2013-Joining-the-Dots-07-11-2013>). 122Steven A Boutcher, ‘The Institutionalisation of Pro Bono in Large Law Firms’ in Robert Granﬁeld and Lynn Mather (eds), Private Lawyers and the Public Interest (OUP 2009) 135–155; Scott L Cummings, ‘The Politics of Pro Bono’ (2004) 52 UCLA Law Review 1. 123Cummings and Sandefur (n 94) 85. 124For a recent review and commentary on these literatures, see Hilary Sommerlad and Louise Ashley, ‘Diversity and Inclusion’, in Laura Empson and others (eds), Oxford Handbook of Professional Service Firms (OUP 2015) Chapter 20. 125Julie Ashdown, ‘Shaping Diversity and Inclusion Policy with Research’ (2005) 83(5) Fordham Law Review 2249, 2249. Trends in pro bono activity This is because other work has suggested that pro bono has become ‘institutionalised’ in large law ﬁrms in the US.122 Is the same not also true for English ﬁrms? We were also surprised by the lack of disclosure of hard data (i.e. number of hours committed) and by the lack of public commitment by ﬁrms to each of their lawyers or staff engaging in a minimum amount of pro bono work each year. Even where ﬁrms do disclose hard data, it might be important to question both the content and the impact of their pro bono efforts (including the quality and efﬁciency of what they are doing).123 g ( ) 124For a recent review and commentary on these literatures, see Hilary Sommerlad and Louise Ashley, ‘Diversity and Inclusion’, in Laura Empson and others (eds), Oxford Handbook of Professional Service Firms (OUP 2015) Chapter 20. 125Julie Ashdown, ‘Shaping Diversity and Inclusion Policy with Research’ (2005) 83(5) Fordham Law Review 2249, 2249. 123Cummings and Sandefur (n 94) 85. 124 126Law Society, ‘Annual Statistical Report 2000′ (Law Society, 2000) para 9.7. 127Black Solicitors Network, ‘Diversity League Table 2013′ (BSN, Law Society and Bar Council 2013) 128Law Society, ‘Annual Statistical Report 2014′ (Law Society 2014). 129Law Society, ‘Private Practice Solicitors’ Salaries 2014’ (Law Society 2014). 130Hilary Sommerlad et al, Diversity in the Legal Profession in England and Wales: a Qualitative Study of Barriers and Individual Choices (London, Legal Services Board, 2010), see <http://www.legalservicesboard.org.uk/what_we_do/Research/ Publications/pdf/lsb_diversity_in_the_legal_profession_ﬁnal_rev.pdf> 131LSB, Increasing Diversity and Social Mobility in the Legal Workforce: Transparency and Evidence – LSB Response Document (Legal Services Board 2011) Annex B. 132‘Direct’ as opposed to the general equality and diversity obligations imposed via the various codes of conduct. 133LSB, Increasing Diversity and Social Mobility in the Legal Workforce: Transparency and Evidence – Consultation Paper on Proposals (Legal Services Board 2010). 134Legal Services Act 2007, s 1(f). 135Legal Services Board, Increasing Diversity and Social Mobility in the Legal Workforce: Transparency and Evidence – LSB Response Document & Accompanying Statutory Guidance issued under Section 162 of the Legal Services Act 2007 (Legal Services Board 2011) Annex B. 136See <http://www.sra.org.uk/diversitydata/> 137ABSs are ﬁrms where either a non-lawyer is a manager of the ﬁrm, or has an ownership-type interest in the ﬁrm. 138<http://www.sra.org.uk/mysra/mysra.page> 139<http://www.sra.org.uk/diversitydata/?301#Collection_5> Equality, diversity and inclusion For the last three decades, academics in the UK and elsewhere have engaged in work criti- quing the lack of diversity in the legal profession.124 This has largely focussed on issues of gender and ethnicity and, more recently, on social background. The picture painted is of a profession that has been labelled ‘male, pale [white] and stale [elderly]’.125 Entry to the profession largely reﬂects the wider population: since 1989 female trainee solicitors LEGAL ETHICS LEGAL ETHICS LEGAL ETHICS 155 have outnumbered males,126 and the proportion of black, Asian and minority ethnic (BAME) trainees has risen signiﬁcantly in the last decade.127 However, such diversity is not reﬂected at partner/senior level,128 and women and minority lawyers are paid less and frequently work in smaller, less prestigious ﬁrms and areas of practice.129 This lack of diversity is thought to be due to a series of formal and informal barriers over the lifetime of a lawyer, the operation of law ﬁrm culture, and the importance of personal relationships to career progression.130 Our data on diversity As Table 9 below shows, 78 of the top 100 law ﬁrms we reviewed in the summer of 2014 had a diversity section somewhere (no matter how small) on their websites. These web- pages disclose a raft of EDI initiatives undertaken by ﬁrms141 but, as we will come to show, hard data on law ﬁrm workforce diversity was lacking for a number of ﬁrms despite the obligation imposed by the LSB and SRA to publish. Previous work by one of the authors of this paper has shown how there has been an increase in the number of law ﬁrms disclosing data on workforce diversity between 2010 (i.e. before the reporting rule was introduced) and 2014.142 Table 10 shows these differences. What is not clear is to what extent the increase in disclosures is linked to the LSB’s reporting rule. Equally, what this table does not show, and what we come to discuss below, is the quality of those disclosures. In general, the report card for law ﬁrm diversity disclosures would read: ‘can do better.’ While 78 of the top 100 law ﬁrms talk about diversity on their websites, and 58 of the top 100 ﬁrms disclosed some diversity data in 2014, the content of the websites and the spread of data disclosed is, on the whole, rather poor. As Table 10 highlights, we were unable to ﬁnd workforce diversity data for a number of large UK law ﬁrms on their websites, despite the regulatory requirement to publish. In total, we were unable to ﬁnd disclosed data for 42 ﬁrms out of the top 100.143 Our research over the summer of 2014 was particularly laborious, challenging and frustrating when it came to EDI disclosures. Different ﬁrms report diversity data in different ways and on different parts of their websites. As noted earlier, this discretion is permitted by the reg- ulators. We would suggest this is a mistake. When conducting our review, we found it interesting that some law ﬁrms appear to place public emphasis on certain diversity characteristics over others. For example, Clifford Chance (#2) has a page on ‘Diversity’ on its website which contains the following, The ﬁrm has a global Diversity Committee, which considers policy, reporting and initiatives. There is also a thriving network of local and global groups that raise awareness and foster understanding around speciﬁc diversity issues. 140Ibid. 141Including lawyer mentoring, afﬁnity groups, membership of networking and representative groups for women and min- ority lawyers. 142Steven Vaughan, ‘Going Public: Diversity Disclosures by Large UK Law Firms’ (2015) 83(5) Fordham Law Review 101. 143Accepting that the inability to ﬁnd disclosed workforce data might be due to researcher error (and not to non-compli- ance with the LSB’s reporting rule), we contacted, via email, each of the ﬁrms for which we could not see the relevant statistics. From 42 emails sent to ﬁrms, 12 responses were received. The reporting of EDI Of all of the aspects of CSR covered in this paper, equality, diversity and inclusion (EDI) is the only area in which law ﬁrms in England and Wales have a legal obligation to collect and report data. In July 2011, the Legal Services Board (LSB) introduced a rule requiring the collection of data on workforce diversity, and the publication of that data, by the legal profession.131 This was the ﬁrst (and, indeed, is the only) direct regulatory intervention taken as regards diversity in the legal profession.132 It was introduced in order to allow the LSB to better meet its own regulatory objective,133 contained within the Legal Services Act 2007, to ‘encourage[e] an independent, strong, diverse and effective legal profession.’134 As a consequence of the LSB’s rule (contained within ‘statutory guidance’),135 law ﬁrms regulated by the Solicitors Regulation Authority are now required to annually collect, report and publish data about the diversity make-up of their workforce.136 The require- ments apply to all ﬁrms regulated by the SRA, including sole practices and ‘alternative business structures’.137 Everyone working at the ﬁrm should be covered by the workforce diversity data collection exercise, including owners of the ﬁrm and all other qualiﬁed and non-qualiﬁed staff. Firms are required to input their aggregated diversity data into the organisation’s diversity data section on the SRA’s website (known as ‘mySRA’).138 Firms are also required to publish a summary of their workforce diversity data.139 Impor- tantly, for this paper, there is no prescription as to where or how the workforce data must be published by law ﬁrms. The regulator sets out that: 156 S. VAUGHAN ET AL. The SRA has not prescribed the manner or format in which a ﬁrm is required to publish a summary of their workforce data. It could be published on the ﬁrm’s website, at the ﬁrm’s ofﬁces or in one of the ﬁrm’s publications. It is only a summary of the data that needs to be published and ﬁrms can present the data in a variety of ways.140 The SRA has not prescribed the manner or format in which a ﬁrm is required to publish a summary of their workforce data. It could be published on the ﬁrm’s website, at the ﬁrm’s ofﬁces or in one of the ﬁrm’s publications. The reporting of EDI It is only a summary of the data that needs to be published and ﬁrms can present the data in a variety of ways.140 As we will come to see below, this lack of prescription is a stumbling block to allowing third parties to undertake meaningful comparisons of the diversity of different ﬁrms. 144<http://www.cliffordchance.com/about_us/corporate_responsibility/people.html> 145<http://www.freshﬁelds.com/en/united_states/who_we_are/Inclusivness_and_diversity/> 146<http://www.linklaters.com/pdfs/mkt/london/Linklaters_Diversity_Statistics%202014.pdf> 147<http://www.simmons-simmons.com/en/About-Us/Corporate-Responsibility/Inclusion-and-Diversity> 148(p8) <http://www.hoganlovells.com/ﬁles/Uploads/Documents/Citizenship%20Report%202013_E.pdf> 149<http://careers.hoganlovells.com/local/UK/aboutus/Documents/UK%20Diversity%20Proﬁle%20-%20August%202011. pdf> Our data on diversity Arcus is our global network for lesbian, gay, bisexual and transgendered (LGBT) colleagues and their friends that aims to encourage an inclusive, integrated culture within the ﬁrm. The group organises regular events as well as an annual art exhibition in the London and New York ofﬁces. They also actively support LEGAL ETHICS 157 LGBT charities – such as the Human Dignity Trust, Non-Gendered and the Albert Kennedy Trust – through fundraising and pro bono work.144 Table 9. Law ﬁrms with EDI webpages. Law ﬁrm ranking Percentage with EDI section on website 1–10 100% 11–25 93% 26–50 72% 51–100 72% Table 10. Workforce EDI data disclosures (2014 and 2010 comparison). Law ﬁrm ranking Workforce EDI data disclosure (2014) Workforce EDI data disclosure (2010) 1–10 90% 90% 11–25 87% 53% 26–50 52% 52% 51–100 46% 16% Table 9. Law ﬁrms with EDI webpages. Law ﬁrm ranking Percentage with EDI section on website 1–10 100% 11–25 93% 26–50 72% 51–100 72% Percentage with EDI section on website Table 10. Workforce EDI data disclosures (2014 and 2010 comparison). Law ﬁrm ranking Workforce EDI data disclosure (2014) Workforce EDI data disclosure (2010) 1–10 90% 90% 11–25 87% 53% 26–50 52% 52% 51–100 46% 16% LGBT charities – such as the Human Dignity Trust, Non-Gendered and the Albert Kennedy Trust – through fundraising and pro bono work.144 Why is there public disclosure of a global network for LGBT members of the ﬁrm, and not for any other group? Similarly Freshﬁelds (#3), in its ‘Diversity and Inclusiveness’ pages, speaks of one speciﬁc initiative for female lawyers (the ‘strategic excellence program’) and another for LGBT lawyers (the ‘halo program’), but not for anyone else.145 Do BAME lawyers at Freshﬁelds (#3) not merit their own initiative? Of the top 10 ﬁrms, Linklaters (#4) stands a clear head and shoulders above the other ﬁrms as regards disclosure on EDI: not only does the ﬁrm report publicly on the gender, ethnicity, sexual orientation, religious belief, social background and disability status of their employees, but they also provide this information in respect of applicants to their internship/vacation scheme and graduate training contract programmes.146 This is com- y g g y their employees, but they also provide this information in respect of applicants to their internship/vacation scheme and graduate training contract programmes.146 This is com- mendable and, we would suggest, is a model of good practice which other ﬁrms should follow. Our data on diversity So too, we think, is the colour coding approach (red, amber, green) taken by Simmons & Simmons (#15) to denote success with its diversity goals, and the willingness of the ﬁrm to ‘red light’ some of its less successful, or not yet achieved, policies and approaches.147 We would suggest the LSB look to the approach taken by Linklaters (#4) should it decide to revise the parameters of its reporting rule. The ﬁrm’s webpages on Diversity are also notable for their ease of access and clarity. We found that the size of ﬁrm is not necessarily an indication of a robust public approach to disclosure on EDI: for example, Hogan Lovells (#6) has only a tiny snapshot of diversity (one quarter of one page) in its 2013 CSR report,148 and a 2011 report on diversity data (which we were only able to ﬁnd via Google and the search term ‘Hogan Lovells diversity statistics’).149 Allen & Overy (#5) only disclosed diversity data in respect of gender, ethnicity and part time working in 2014, but have now produced a S. VAUGHAN ET AL. 158 more detailed report in 2015 across a large number of diversity characteristics.150 For this paper, we had been hoped to break down diversity reporting by ﬁrm, and to offer up tables (comparable with those in the section above on pro bono) of comparisons. This has, however, proven impossible. The top 100 law ﬁrms we reviewed report on diversity in such a variety of ways that we been unable to draw together the data in any meaningful way. This has been enormously frustrating. To give just one example, Norton Rose Fulb- right (#7) gives aggregate disability data for everyone in its London ofﬁce in the year 2014 (i.e. for the entire ofﬁce workforce, and not broken down into partners/associates/trainees/ staff etc);151 CMS Cameron McKenna gives disability data broken down into partners, associates, trainees, ‘other legal’, secretaries and business services but only for the year 2011;152 and a whole raft of other ﬁrms provide no data whatsoever on disability. We are, as such, strongly in favour of the regulators amending the reporting rule to require some uniformity in how law ﬁrms disclose their diversity data so that meaningful com- parisons and analysis can be undertaken. Our data on diversity Despite the fact that the majority of ﬁrms are saying something public about diversity, and despite there being a regulatory requirement for disclosure, what is said is often very limited indeed. This is disappointing. 150<http://www.allenovery.com/SiteCollectionDocuments/UK%20diversity%20metrics.pdf> 151<http://www.nortonrosefulbright.com/uk/corporate-responsibility/diversity-and-inclusion/our-statistics/> 152<http://www.cms-cmck.com/Hubbard.FileSystem/ﬁles/Publication/fa57a9a2-86ba-4395-bc35-9bc94b3ae57c/ Presentation/PublicationAttachment/812b462f-ab3f-4c28-82c4-3dc1783e7894/Social%20Mobility%20and%20Diversity %20Statistics%202011.pdf> 153Even Fallan, ’Explaining the Variation in Adoption Rates of the Information Content of Environmental Disclosure’ (2015) 11(2) Journal of Accounting & Organizational Change 247. 154Matthias S Fifka, ‘Corporate Responsibility Reporting and its Determinants in Comparative Perspective – a Review of the Empirical Literature and a Meta-analysis’ (2013) 22 Business Strategy and the Environment 1. 155Daniel Tschopp and Michael Nastanski, ‘The Harmonization and Convergence of Corporate Social Responsibility Report- ing Standards’ (2014) 125 Journal of Business Ethics 147. 156Marvin Mikesell, ‘Geographical Perspectives in Anthropology’ (1974) 57 Annals of the Association of American Geogra- phers 617. 157Ron Johnston, Geography & Geographers (Oxford University Press 1979). 158Jill Solomon and others, ‘Impression Management, Myth Creation and Fabrication in Private Social and Environmental Reporting: Insights from Erving Goffman’ (2013) 38(3) Accounting, Organizations and Society 195; Thomas P Lyon and John Maxwell, ‘Greenwash: Corporate Environmental Disclosure under Threat of Audit’ (2011) 20 Journal of Economics & Man- agement Strategy 3. 159For a review of these, see Tschopp and Nastanski (n 156). 150<http://www.allenovery.com/SiteCollectionDocuments/UK%20diversity%20metrics.pdf> 151<http://www.nortonrosefulbright.com/uk/corporate-responsibility/diversity-and-inclusion/our-statistics/> 152<http://www.cms-cmck.com/Hubbard.FileSystem/ﬁles/Publication/fa57a9a2-86ba-4395-bc35-9bc94b3ae57c/ Presentation/PublicationAttachment/812b462f-ab3f-4c28-82c4-3dc1783e7894/Social%20Mobility%20and%20Diversity %20Statistics%202011.pdf> 153Even Fallan, ’Explaining the Variation in Adoption Rates of the Information Content of Environmental Disclosure’ (2015) 11(2) Journal of Accounting & Organizational Change 247. 154Matthias S Fifka, ‘Corporate Responsibility Reporting and its Determinants in Comparative Perspective – a Review of the Empirical Literature and a Meta-analysis’ (2013) 22 Business Strategy and the Environment 1. 155Daniel Tschopp and Michael Nastanski, ‘The Harmonization and Convergence of Corporate Social Responsibility Report- ing Standards’ (2014) 125 Journal of Business Ethics 147. 156Marvin Mikesell, ‘Geographical Perspectives in Anthropology’ (1974) 57 Annals of the Association of American Geogra- phers 617. 157Ron Johnston Geography & Geographers (Oxford University Press 1979) p y y p 1<http://www.nortonrosefulbright.com/uk/corporate-responsibility/diversity-and-inclusion/our-statistics/> 2 159For a review of these, see Tschopp and Nastanski (n 156). g p y g p y 158Jill Solomon and others, ‘Impression Management, Myth Creation and Fabrication in Private Social and Environmental Reporting: Insights from Erving Goffman’ (2013) 38(3) Accounting, Organizations and Society 195; Thomas P Lyon and John Maxwell, ‘Greenwash: Corporate Environmental Disclosure under Threat of Audit’ (2011) 20 Journal of Economics & Man- agement Strategy 3. 159 f h h d k ( ) 150<http://www.allenovery.com/SiteCollectionDocuments/UK%20diversity%20metrics.pdf> 157Ron Johnston, Geography & Geographers (Oxford University Press 1979). agement Strategy 3. 159 f h 160Fifka (n 155) 3. 161Susan Sturm, ‘Law’s Role in Addressing Complex Discrimination’ in Robert Nelson and Laura Beth Nielson (eds) Handbook of Employment Research: Rights and Realities (Springer 2005). 162<http://www.harpermacleod.co.uk/community/> 163<http://wedlakebell.com/environment.html> 164<https://www.ashurst.com/about-ashurst.aspx?id_Content=2642> 165<http://www.legalsectoralliance.co.uk/> Environment and sustainability disclosures Disclosures by law ﬁrms on environmental and sustainability matters were the poorest of all of the elements of CSR reviewed, both in terms of numbers of ﬁrms reporting on these matters and on the quality of the disclosures made. Just over a third of the top 100 ﬁrms did not have separate environment/sustainability sections to their websites. Where these sections were present, they varied considerably (see Table 11). Most ﬁrms split their environmental impacts into two main areas: carbon footprint and sustainable procurement. In this respect, these law ﬁrms are similar to corporates (and, as such, interested in energy usage, waste and recycling, responsible travel etc). A number of ﬁrms (for example, DLA Piper (#1) and Freshﬁelds (#3)) also point to their work advising governments and clients on global environmental initiatives such as the Kyoto Protocol. In this way, law ﬁrms and their lawyers have the potential to act as ‘norm intermediaries’,161 developing law that has the ability to improve the public good. What is striking, however, is the lack of comment by law ﬁrms on how matters on which they act (i.e. things done by their clients) impact, potentially negatively, on the environment – so, for example, Firm X advising Client Y on fracking, or enabling Client Z to buy a series of fossil fuel power stations in a developing nation. Most ﬁrms split their environmental impacts into two main areas: carbon footprint and sustainable procurement. In this respect, these law ﬁrms are similar to corporates (and, as such, interested in energy usage, waste and recycling, responsible travel etc). A number of ﬁrms (for example, DLA Piper (#1) and Freshﬁelds (#3)) also point to their work advising governments and clients on global environmental initiatives such as the Kyoto Protocol. In this way, law ﬁrms and their lawyers have the potential to act as ‘norm intermediaries’,161 developing law that has the ability to improve the public good. What is striking, however, Hard data on environmental matters (e.g. carbon footprint, waste reduction) is dis- closed by ﬁrms on an ad hoc basis. In some ﬁrms, the public disclosures on environmental matters are more aspirational (vague) mission statements than anything else. For example, Harper McLeod (#98) states: Our environmental policy recognises our clear commitment to improving our business prac- tices to make a positive impact on the environment. The environment and sustainability Large corporate actors in the West have reported on environmental matters since the 1970s,153 although a sustained emphasis on environmental disclosures as part of CSR was not apparent until the 1990s.154 In their work,155 Tschopp and Nastanski argue that the push for environmental disclosures was driven partly because of public concerns about environmental problems in the late 1960s,156 and partly because technological advances and the consequent rapid depletion of resources led to a demand for more accountability.157 Despite early (and sustained) claims of ‘greenwash’ (where business is said to disclose certain environmental data for selﬁsh or reputational reasons only),158 environmental reporting has become a core aspect of CSR, and a number of international standards exist in this area.159 Over time, there has been a shift in reporting from disclos- ures on ‘environmental’ matters to broader ‘sustainability’ disclosures, and a correspond- ing geographic broadening to include reporting by businesses based in emerging and LEGAL ETHICS 159 developing nations.160 Below, we present our data on environment and sustainability dis- closures on the law ﬁrm websites we reviewed. developing nations.160 Below, we present our data on environment and sustainability dis- closures on the law ﬁrm websites we reviewed. Environment and sustainability disclosures Through an environmental scan and audit process, and working in partnership with our key stakeholder groups, we recognise, evaluate and adopt ‘best practice’ methods to minimise the impact of our business on the environment.162 However, the ﬁrm does not publish its environmental policy, nor is further data given on the ‘best practice methods’ that the ﬁrms uses to minimise its impacts. The lack of hard data on environmental matters is more commonly seen the further down the list of the top 100 law ﬁrms that one goes, although there are some notable exceptions: see, for example, the relatively expansive data given by Wedlake Bell (#85),163 and the very limited data given by Ashurst (#13).164 What seems to make (some) difference is whether or not a ﬁrm is a member of the Legal Sustainability Alliance (LSA, formerly known as the Legal Sector Alliance).165 So, for example, while Michelmores (#95) says 160 S. VAUGHAN ET AL. S. VAUGHAN ET AL. 160 Table 11. Law ﬁrm environment/sustainability website sections. Law ﬁrm ranking Percentage with environment/sustainability sections on website 1–10 90% 11–25 87% 26–50 64% 51–100 50% very little indeed about environmental issues, the ﬁrm does report that in the LSA survey 2012 it had emissions per employee of 2.31 tonnes CO2.166 very little indeed about environmental issues, the ﬁrm does report that in the LSA survey 2012 it had emissions per employee of 2.31 tonnes CO2.166 The LSA was formed in 2007 originally with 18 law ﬁrm members that were each ‘com- mitted to working collaboratively to take action on climate change by reducing their carbon footprint and adopting environmentally sustainable practices.’167 There are now than 300 LSA law ﬁrm members,168 and ‘Every year [the LSA] encourages members to report on their environmental performance during the previous 12 months using [its] bespoke online reporting platform.’169 However, in 2014, only 66 out of the more than 300 LSA members reported data in this way.170 This is despite the fact that one of the prin- ciples (P6) to which LSA member ﬁrms sign up states that they will ‘Report on [their] pro- gress and be accountable’.171 This disconnect is striking, and we wonder whether some ﬁrms sign up to the LSA for presentational or PR reasons, and not because of a ﬁrm com- mitment to reducing environmental impacts. 166<http://www.michelmores.com/about-us/corporate-social-responsibility> 167<http://www.legalsectoralliance.co.uk/about> 168LSA, Annual Report 2014 (Legal Sector Alliance 2014) 5. See <http://www.legalsectoralliance.com/ﬁles/lsa-annual-report- 2014.pdf> 169Ibid, 4. 170LSA (n 169) 6. 171<http://www.legalsectoralliance.co.uk/principles> 172LSA (n 169) 33–35. 173Martin Jones and Jill Solomon, ‘Social and Environmental Report Assurance: Some Interview Evidence’ (2010) 34(1) Accounting Forum 20. 174<http://www.weightmans.com/about-us/environmental-action/> 67<http://www.legalsectoralliance.co.uk/about> 68 175<http://www.herbertsmithfreehills.com/about-us/pro-bono-and-citizenship> 176<http://www.herbertsmithfreehills.com/practice-areas/environment-planning-and-communities> 177<http://www.herbertsmithfreehills.com/insights/legal-brieﬁngs/minding-the-specter-of-csr-reporting-liability> 178For these rankings, see <http://www.chambersandpartners.com/11814/25/editorial/1/1> 179Ibid. 180<http://www.allenovery.com/corporate-responsibility/Documents/CS%20brochure.pdf> 181<http://www.allenovery.com/corporate-responsibility/environment/Pages/default.aspx> 182<http://www.cliffordchance.com/content/dam/cliffordchance/CR2013/CRR2013.pdf> 183<http://www.pinsentmasons.com/en/about-us/starﬁsh/our-communities-environment/> Environment and sustainability disclosures With two outliers – DLA Piper (#1) and Eversheds (#11) – the 2014 LSA report clearly shows that the larger the law ﬁrm the higher the carbon footprint per employee, with Clifford Chance (#2) having the highest footprint of all of the 66 ﬁrms that reported data.172 The power and potential of external audit in relation to environmental reporting has been seen with businesses more generally.173 A number of ﬁrms, across the rankings state that they are carbon neutral (e.g. Olswang (#33), Freshﬁelds (#3), Mischon de Reya (#39) and Forsters (#75)). This seems to have been achieved, in part at least, by the purchase of carbon offsets, which other law ﬁrms have instead directly rejected as a policy approach: Our strategy is to reduce our consumption of resources and carbon emissions before consid- ering carbon offset, which could in turn create a carbon neutral working environment. As there is some scepticism over the integrity of carbon offsetting, we are aiming to reduce the amount of pollution and waste that we generate, while continuing to monitor our CO2- emissions.174 (Weightmans, #44) A number of ﬁrms reference environmental targets, but some do not say what these are and/or whether (or the extent to which) they have been achieved; and not one ﬁrm explains (as far as we can see) why, when the numerical targets are published, those LEGAL ETHICS LEGAL ETHICS 161 particular targets have been chosen. As such, it is difﬁcult for a third party to say whether those targets are realistic and/or progress has been made. Similarly, even where data is made available by a ﬁrm, its coverage tends to be patchy. Of the largest ﬁrms, Herbert Smith Freehills (#9) stands out for its lack of data on environmental matters. The following is all that is said, under a website heading of ‘Pro Bono and Citizenship’: We are committed to reducing our environmental impact. This includes targets for reducing energy and carbon, food waste and increasing recycling in our ofﬁces around the world as well as encouraging more environmentally friendly travel options.175 The Herbert Smith Freehills (#9) targets are not disclosed and, as noted earlier in this paper, the ﬁrm does not produce annual CSR reports. Environment and sustainability disclosures This is particularly striking given the ﬁrm lists an ‘Environment, Planning and Communities’ team as one of its practice areas,176 and the ﬁrm (in February 2015) gave the following advice to its clients on CSR: With CSR programs the new normal, consumers and investors increasingly factoring CSR performance into their decision-making, and companies eager to display ‘good corporate citizenship,’ strict CSR due diligence is required to manage legal risks. More than ever, words must match deeds.177 Interestingly, there does not appear to be any connection between the extent to which a law ﬁrm reports on environmental/sustainability issues and how highly that ﬁrm is ranked for its environmental law expertise.178 So, for example, whereas Allen & Overy (#5) is ranked in the top band for environmental law in Chambers & Partners,179 the ﬁrm says very little indeed about environmental matters in its CSR report,180 and equally little in its CSR webpages.181 The ﬁrm also does not disclose any hard data about impacts and/or emissions. By contrast, in its 2013 CSR report, Clifford Chance (#2) (ranked as Band 2 for environmental law) discloses an amount of hard data on its environmental performance;182 and Pinsent Masons (#14) (ranked Band 4 for environ- mental law work) publishes its full environmental policy.183 Interestingly, there does not appear to be any connection between the extent to which a law ﬁrm reports on environmental/sustainability issues and how highly that ﬁrm is ranked for its environmental law expertise.178 So, for example, whereas Allen & Overy (#5) is ranked in the top band for environmental law in Chambers & Partners,179 the ﬁrm says very little indeed about environmental matters in its CSR report,180 and equally little in its CSR webpages.181 The ﬁrm also does not disclose any hard data about impacts and/or emissions. By contrast, in its 2013 CSR report, Clifford Chance (#2) (ranked as Band 2 for environmental law) discloses an amount of hard data on its environmental performance;182 and Pinsent Masons (#14) (ranked Band 4 for environ- mental law work) publishes its full environmental policy.183 Conclusions The public commitment of the top 100 law ﬁrms in England and Wales to CSR is wide- spread. The majority say something to the wider world about CSR. However, what is said varies signiﬁcantly. This is, perhaps, unsurprising. What is more surprising is that so few ﬁrms explain why they are committed to CSR. As general trend, the lower the ranking of the law ﬁrm, the less they say about CSR and the less likely they are to explain their CSR motivations, although there are some notable exceptions to the rule at both ends of the rankings. Consistently, the top ten ﬁrms outperform lower ranked ﬁrms on all elements of CSR. We might speculate as to why this is so. Such might reﬂect greater resources S. VAUGHAN ET AL. 162 devoted to CSR by those ﬁrms (a corollary of their size, turnover and reach) and/or a greater desire to make additional work-related offerings to their employees in the form of CSR activities. It might reﬂect the client base of those ﬁrms (who may be more inter- ested than other clients in CSR). These largest ﬁrms might also be more susceptible to approaches taken by, and the impacts of cultural differences from, US law ﬁrms. Finally, it may be that competition for work is ﬁercest between these top ten ﬁrms com- pared with other ﬁrms.184 We question whether current approaches by law ﬁrms to CSR really reﬂect the nuances (and speciﬁc impacts and responsibilities) of law ﬁrms as organisations. As providers of legal advice law ﬁrms are in a privileged position to incorporate their legal services into their CSR offerings. We would suggest that recent practice may have been (to varying degrees) symbolic; more about competition and about demand side pressures (i.e. appealing to clients), than a substantive, altruistic commitment to CSR. There may not have been any real attempt to, ‘translate the ideals of professionalism into concrete institutional forms’ of CSR.185 We were struck, for example, by the widespread references to awards and external recognition for CSR on many law ﬁrm websites. This, we would suggest, is part of the ‘media-isation’ of legal practice, and one concrete aspect of the wider pan-promotionalism of contemporary culture. One might argue that this is linked to the use of CSR for commercial advantage. 184Cummings and Sandefur (n 87) offer up market competition as one driver for pro bono. 185Cummings (n 123) 7. 186Boon and Whyte (n 34) 184. 187Boon and Abbey (n 16) 637. 188In this paper, we have not engaged in the debate over exactly how large ﬁrms should help to alleviate unmet legal need. See n 11. We do not suggest that in reconﬁguring their CSR activities to increase pro bono efforts corporate law ﬁrms ever can, or should, seek ﬁll the justice gap left by recent cuts to legal aid, as has recently been suggested by the Lord Chan- cellor – see <http://www.telegraph.co.uk/news/uknews/law-and-order/11693145/Michael-Gove-Wealthy-lawyers- should-do-more-free-work-for-the-justice-system.html> 189Cummings and Sandefur (n 94) 95. 190David Luban, Lawyers and Justice: An Ethical Study (Princeton, NJ: Princeton University Press 1988); David Luban, ‘The Noblesse Oblige Tradition in the Practice of Law’ (1988) 41 Vanderbilt Law Review 717. 190David Luban, Lawyers and Justice: An Ethical Study (Princeton, NJ: Princeton University Press 1988); David Luban, ‘The Noblesse Oblige Tradition in the Practice of Law’ (1988) 41 Vanderbilt Law Review 717. 184Cummings and Sandefur (n 87) offer up market competition as one driver for pro bono. 185Cummings (n 123) 7. 186 Conclusions Such undermines altruism as an ethical ideal, and weakens the level of commit- ment to CSR,186 part of an history of practice in large commercial ﬁrms which sees, ‘industry, initiative,responsibilityand success [put]over benevolence, altruismand,possibly, justice.’187 Where ﬁrms do make disclosures on CSR, these tend to group around the following three areas: (i) pro bono and community giving; (ii) diversity and inclusion; and (iii) environmental matters. For a number of ﬁrms, little or no distinction is made between pro bono (i.e. the giving of free legal advice) and wider ‘community giving’. We suggest that this is a mistake, and potentially relegates the important role that lawyers in large corporate ﬁnance ﬁrms can play in alleviating unmet legal need.188 We accept here that the extent of unmet legal need is vast, and that the contributions of large ﬁrms via pro bono initiatives can only ever hope to go a tiny part of the way towards meeting that need. We also accept the argument put forward by Cummings and Sandefur that, ‘lawyers’ pro bono service is importantly related to conditions in legal services markets and in the markets for lawyers’,189 and, as such, there will always be signiﬁcant variation in where and how pro bono is enacted by lawyers. However, and like Luban, we see a correlation between the monopolistic privileges granted to lawyers and the corollary expectations that the state might have of those lawyers.190 As Boon and LEGAL ETHICS 163 Whyte have demonstrated, the traditions and ethics of lawyers are ‘rich in the ideology of public service.’191 Despite there being regulatory intervention by the LSB as regards the collection and reporting of diversity data by law ﬁrms (and other lawyers), the quality of disclosures (in terms of the amount, nature and breadth of data reported on) varies to such an extent that we have been unable to draw any meaningful comparisons or conclusions. This is both disappointing and frustrating, and we have suggested that regulatory reform is necessary. Such reform would see some consistency to reporting introduced by the LSB and/or SRA, and the current discretion given to ﬁrms on disclosure reigned in signiﬁcantly. Our expectation is that consistent reporting and the ability to compare performance between ﬁrms would go some way to improve diversity, especially if third parties (regulators, clients and others) undertook systematic comparisons of competitor ﬁrms. Conclusions Environmental reporting by the top 100 ﬁrms was particularly poor, despite these ﬁrms having potentially signiﬁcant environmental footprints, and despite advising their clients on potentially environmentally deleterious matters. p y y To borrow a phrase from Cummings, the story of law ﬁrms and CSR is still being written.192 As part of that unfolding narrative there is a question as to the extent to which CSR should be a matter wholly for law ﬁrms to decide for themselves or the extent to which CSR requires some form of regulatory intervention.193 Certainly, as dis- cussed above, diversity as one aspect of CSR is an area in which the regulators of legal ser- vices in England and Wales have taken some initiative and introduced rules. In the same jurisdiction a number of bodies corporate have CSR reporting obligations under statute.194 We leave this paper with the question, supported by the data we have offered up and which we plan to explore in further work, why law ﬁrms should not be treated the same, as regards CSR reporting, as the corporate clients they serve? Acknowledgements We are very grateful to Tina Martin and Lloyd Brown for their research assistance with this project, and to Robert Lee, Richard Moorhead and Hilary Sommerlad for comments on an earlier draft. The usual disclaimer applies. 191Boon and Whyte (n 34) 175. This might extend to the way in which the lawyer advises their clients. As Cummings points out, ‘One version of this public service ideal places the lawyer in the role of mediating between client interests and public goals, advising clients to pursue the course of action that is not simply privately beneﬁcial, but socially just’ (n 123) 9. 192Cummings (n 123) 148. 193In an era of ‘regulatory red tape’ challenges, where legal services regulation thought to be superﬂuous or anti-competi- tive is done away with, we accept that the appetite for additional rules may be limited. 194Company Act 2006, s414A-D. Funding Part of Steven’s time on this project was paid for via his ESRC ‘Future Research Leader’ grant (refer- ence ES/K00834X/1). ORCID Steven Vaughan http://orcid.org/0000-0002-6984-3757 191Boon and Whyte (n 34) 175. This might extend to the way in which the lawyer advises their clients. As Cummings points out, ‘One version of this public service ideal places the lawyer in the role of mediating between client interests and public goals, advising clients to pursue the course of action that is not simply privately beneﬁcial, but socially just’ (n 123) 9. 192Cummings (n 123) 148. 193In an era of ‘regulatory red tape’ challenges, where legal services regulation thought to be superﬂuous or anti-competi- tive is done away with, we accept that the appetite for additional rules may be limited. 194Company Act 2006, s414A-D.
https://openalex.org/W4392438776	https://periodicos.saude.sp.gov.br/RIAL/article/download/36854/35070	Portuguese	null	Identificação simultânea dos ácidos benzóico, cítrico e láctico em refrigerantes, por cromatografia em camada delgada	Revista do Instituto Adolfo Lutz	1,985	cc-by	2,031	Reo, Inet, Adolfo Lut«, 45 (1/2) :81-84, 1985. Reo, Inet, Adolfo Lut«, 45 (1/2) :81-84, 1985. INTRODUÇÃO três ácidos nas concentrações comumente em- pregadas em refrigerantes 2. Algumas fábricas de refrigerantes estão alternando o uso de acidulante entre os ácidos cítrico e láctico, e empregando o ácido ben- zóíco como conservador. Diversos métodos para identificação por cromatografia destes ácidos foram descritos 3. 4. 5, 7, 8, 9, 10, 11, 12, 13, 14 e por nós testados, encontrando-se, assim, certa dificuldade em uma única análise para uma identificação simultânea dos três ácidos, com Rf bem distinguíveis. O ácido benzóico pode ser determinado quantitativamente com adaptação deste méto- do ao descrito nos métodos oficiais de análise da AOAC 1. RIALA6/600 RIALA6/600 RICCIARDI, A.J.; FERREIRA, J.F. & CARLUCCI, N.T. - Identificação simultânea dos ácidos benzóico, cítrico e láctíco em refrigerantes, por cromatografia em camada delgada. Re», Inet; Adolfo Lutz, 45 (1/2) :81-84, 1985. RESUMO: Foi descrita a metodologia utilizada para detectar os ácidos benzóico, cítrico e láctico em refrigerantes comerciais. A cromatografia em camada delgada foi escolhida por ser um método simples e sensível, quando comparado com outros méto- dos analíticos. DESCRITORES: acidulantes, conservadores em refrigerantes, detecção; refri- gerantes, identificação dos ácidos benzóico, cítrico e láctico por cromatografia em camada delgada. Material No período de junho de 1982 a janeiro de 1985, foram examinadas noventa amostras de diversos refrigerantes (tabela p. seguinte), encaminhados pelos Centros de Saúde da Coordenadoria de Saúde da Comunidade da Secretaria de Estado da Saúde do Estado de São Paulo, recolhidas nos supermercados e nos distribuidores de bebidas da região de Ribeirão Preto, SP. Em 1979, LEVENBERGER et alii 6 apre- sentaram um trabalho quantificando a saca- rina e os ácidos benzóico e sórbico, por cro- matografia líquida de alta resolução, mas, por ser este um método que exige um certo investimento, não foi testado, sendo nossa principal preocupação o desenvolvimento de uma técnica rápida e de baixo custo, o que foi conseguido através da cromatografia em camada delgada. * Realizado no Laboratório I de Ribeirão Preto, SP, do Instituto Adolfo Lutz, São Paulo, SP. ** Do Instituto Adolfo Lutz. IDENTIFICAÇÃO SIMULTÂNEA DOS ÁCIDOS BENZóICO, CíTRICO E LÁCTICO EM REFRIGERANTES, POR CROMATOGRAFIA EM CAMADA DELGADA * Almir José RICCIARDI Jacob Fernando FERREIRA Nina Turko CARLUCCI ** Método O método descrito permitiu a extração des- tes ácidos com acetato de etila e posterior identificação através da cromatografia em camada delgada, utilizando uma mistura de clorofórmio e ácido acético como fase móvel e uma mistura de indicadores como revelador, o que tornou possível a identificação destes Extração: a 50 ml do refrigerante, após a eliminação do gás carbônico, foram adiciona- dos 5 ml de HCI (1 + 3) e 25 ml de acetato de etila. Após agitação, em funil de sepa- ração, a camada superior de acetato de etila foi recolhida e concentrada até quase a secura, com aquecimento em banho-maria. 81 ~ Cl ~~, G -.q> .... ~.~ 'l.õgc;; ~ •... •• (t) •• ~o;::> Ir.I~ ~ :to:. . " •..• 0"" «>",t;l ~g~ . ~ ~ t;l ...., ..... ~.~ '" . ..• •.. 1». i:l "" ~.. !!'I\'> "'0 ° :> '" \'d n t" <l d g O 1»0 ..•.- ° . llQ•.•Z 10 , t:!!t-i 10' '"S _ "" " '" I» i:l gtt 10 •••• ;:l. n" ;o I» '" "" 1»' '"° r;q'" I» ~. ""S 10 I'l .;:; ~I»' ~ ; • ;o •...• "" ;! ° " '" :"" ~" n ).Õ: "",o <> '" '::;:c:r <> <Di:l ~ S: ••• <> .•~ .,.0 TABELA 00 I>:) Identificação dOI!ácidos benzóico, cítrico, láctico e fosfórico em amostras de refrigerantes analisados pelo Laboratório I de Ribeirão Preto, SP, do Instituto Adolfo Lute, São Paulo, SP, no período de junho de 1982 a janeiro de 1985 do Instituto Adolfo Lute, São Paulo, SP, no período de junho de 1982 a janeiro de 1985 Aditivos detectados N.o de Ácido láctico Ácido cítrico Ácido benzóico Ácido fosfórico Produto amostras examinadas Número de Percentagem Número de Percentagem Número de Percentagem Número de Percentagem amostras de amostras amostras de amostras amostras de amostras amostras de amostras Refrigerantes 25 4 16 20 80 24 96 5 20 de guaraná Refrigerantes 13 4 31 9 69 12 92 6 46 de maçã I Refrigerantes de frutas cítricas (limão, 30 5 17 25 83 21 70 5 17 laranja e tangerina) Refrigerantes 12 O O 3 25 7 58 12 100 de cola Refrigeran tes de 10 3 30 8 80 9 90 O O "tutti-frutti" Total 90 16 18 65 72 73 81 28 I 31 -~--- -- , G -.q> .... ~.~ 'l.õgc;; ~ •... Método •• (t) •• ~o;::> Ir.I~ ~ :to:. . " •..• 0"" «>",t;l ~g~ . ~ ~ t;l ...., ..... ~.~ '" . ..• •.. 1». i:l "" ~.. !!'I\'> "'0 ° :> '" \'d n t" <l d g O 1»0 ..•.- ° . llQ•.•Z 10 , t:!!t-i 10' '"S _ "" " '" I» i:l gtt 10 •••• ;:l. n" ;o I» '" "" 1»' '"° r;q'" I» ~. ""S 10 I'l .;:; ~I»' ~ ; • ;o •...• "" ;! ° " '" :"" ~" n ).Õ: "",o <> '" '::;:c:r <> <Di:l ~ S: ••• <> .•~ .,.0 RICCIAl:\Üi, A.J.; FERREIRA, J.F. & CARLUCCI, N.T. -- Identificação simultânea dos ácidos benzóico cítrico e láctico em refrigerantes, por cromatografia em camada delgada. Re1,. Inst. Adolfo Lutz, 45 (1/2) :81-84, 1985. janeiro de 1985, pelo Laboratório I de Ribei- rão Preto, SP, do Instituto Adolfo Lutz, São Paulo, SP, 18% das amostras apresentaram ácido láctico, sendo que os refrigerantes de cola foram os únicos em que não foi encon- trado este ácido; porém, constatou-se a pre- sença de ácido fosfórico na totalidade destes refrigerantes. O ácido fosfórico, por ser um acidulante usado em refrigerantes, foi por nós identificado e encontrado em 31% de todas as amostras analisadas. Em mais de 15% dos refrigerantes rotulados como refri- gerantes de frutas cítricas (como limão, laranja e tangerina), não foram adicionados sucos naturais, observação constatada pela ausência de betaína. Cromatoçraiia: em placas de camada del- gada, com silica gel G (STAHL tipo 60), soluções-padrões de ácido benzóico (0,01 % foram aplicados o extrato concentrado e p/v), ácido lático (0,02% p/v) e ácido cítrico (0,03% p/v). Após esse processo, foi desen- volvido o cromatograma, utilizando-se como fase móvel clorofórmio e ácido acêtico glacial (27 + 3). Como revelador foi empregada a solução vermelho de metila 0,12 g, verde de bromo- cresol 0,04 g e álcool etílico até completar 100 ml. Os ácidos foram revelados como manchas vermelho-róseas sobre fundo roxo-avermelha- do, com os seguintes Rf aproximados: ácido cítrico 0,0, ácido láctico 0,5 e ácido benzóico 1,0. Analisando ainda a tabela, observamos que em 81% dos refrigerantes examinados foi constatada a presença de ácido benzóico, adi- cionado como conservador na forma de ben- zoato de sódio. Dentre as marcas de refrige- rantes analisadas constatamos que um único fabricante não usa conservadores em seus produtos, utilizando-se do método da pasteu- rização. CONCLUSÃO A cromatografia em camada delgada reve- lou ser um método eficiente para detectar os ácidos cítrico, benzóico e láctico em refrige- rantes, por ser simples e sensível quando com- parado com outros métodos analíticos. Método O ácido benzóico pôde ser quantificado adaptando-se este método ao descrito nos mé- todos oficiais de análise da AOAC 1 com as seguintes modificações: adição de 0,1 ml de extrato diluído com 5 ml de acetato de etila nas cromatoplacas, paralelamente com 0,1 ml de uma solução padrão de ácido benzóíco a 0,1%. Desenvolver o cromatograma até uma altura de 10 em, empregando-se como fase móvel clorofórmio e ácido acético glacial (27 + 3). Secar e raspar dois círculos de 2 em de diâmetro centrados na linha de frente e perpendiculares aos pontos de aplicação da amostra e do padrão e transferir para tubos de centrífuga. Transferir para outro tubo de centrífuga um círculo com 2 em de diâmetro, raspado da placa em um local onde não foi desenvolvido o cromatograma, para servir de branco. A cada tubo adicionar 10 ml de álcool etílico e agitar por trinta segundos. Centrifu- gar até o sobrenadante tornar-se límpido. Decantar o sobrenadante em uma cubeta e ler no espectrofotômetro a 272 nm, Calcular a con- centração de ácido benzóico na amostra. Para verificar a sensibilidade do método, foram preparadas as seguintes soluções pa- drões: ácido benzóico a 0,01 % p/v, ácido lâc- tico a 0,02% p/v e ácido cítrico a 0,03% p/v, soluções estas com concentrações dez vezes menores dos que os valores máximos permiti- dos por lei (Decreto Federal n.? 55.871, de 26-3-19652), identificando-se perfeitamente as respectivas manchas. RESULTADOS E DISCUSSÃO Agradecemos ao Dr. Paulo de Almeida Vi- dai pela colaboração prestada na realização deste trabalho. De acordo com a tabela, dos 90 refrigeran- tes analisados no período de junho de 1982 a RIALA6/~ RICCIARDI, A.J.; FERREIRA, J.F. & CARLUCCI, N.T. - Simultaneous identi- fication of benzoic acíd, citric acid, and lactie acid in commercíal 80ft drínks, by thín-layer chromatography. Rev. Inst. Adolfo Lute, 45 (1/2) :81-84, 1985. ABSTRACT: A description is made of a procedure for the detectíon of citric acíd, benzoic acid and lactic acid in commercial soft drinks. Thin-layer chrornato- graphy was found to be a simple and sensitive method compared with standard analytical methods. DESCRIPTORS: acidifers and preservatives in 80ft drinks, detection; soft drinks, identification of benzoic acíd, citric acid and lactie acid by thín-layer clrro- matography. 83 JdCCIARDI, A.J.; FERRE IRA, J.F. & CARLUCCI, N.T. - Identificação simultânea dos ácidos benzéíco, cítrico e lactíco em refrigerantes, por cromatografia em camada delgada. Rev, Inst. Adolfo Lut«, 45 (1/2) :81-84, 1985. JdCCIARDI, A.J.; FERRE IRA, J.F. & CARLUCCI, N.T. - Identificação simultânea dos ácidos benzéíco, cítrico e lactíco em refrigerantes, por cromatografia em camada delgada. Rev, Inst. Adolfo Lut«, 45 (1/2) :81-84, 1985. Recebido para publicação em 5 de agosto de 1985. JdCCIARDI, A.J.; FERRE IRA, J.F. & CARLUCCI, N.T. - Identificação simultânea dos ácidos benzéíco, cítrico e lactíco em refrigerantes, por cromatografia em camada delgada. Rev, Inst. Adolfo Lut«, 45 (1/2) :81-84, 1985. REFERtNCIAS BIBLIOGRÁFICAS 7. LONG, A.G. - The separation of acids by paper partition chromatography. J. Chamo ss«, 3:2197-201, 1951. 1. ASSOCIATION OF OFFICIAL ANALYTI- CAL CHEMISTS - Ofticial Methods oi analysis of the Association of Ofticial Analitical Chemists. 14th ed. Washington, DC, AOAC, 1984. p. 377-8. 8. LUGG, J.W. & OVERELL, B.T. apud LEDE- RER, E. & LEDERER, M. - Cromato- grafia. 2.a ed. Buenos Aires, Ateneo, 1960. p. 214. 2. BRASIL. Leis, decretos etc, - Decreto n.? 55.871, de 26 de março de 1965. Diário Oficial, Brasília, 9 abro 1965. Modifica o Decreto n.? 50.040, de 24 jan. 1961, refe- rente a normas reguladoras do emprego de aditivos para alimentos, alterada pelo Decreto n,? 691, de 13 de março de 1962. 9. MIKES, O. - Laboratory handbook of chromatography methods. London, Van Nostrand, 1966. p. 97. 10. OPIENSKA-BLAUTH, J.; SAKLAWSKA- -SZYMONOWA, M. - Partition paper chromatography of some organic acids. Nature, 168:511-2, 1951. 3. BRAUN, D. & GEENEN, H. - Dunnschícht- chromatographie von Carbonsauren. J. Chromatogr., 7: 56-9, 1962. 11. RANDERATH, K. - Cromatografia de capa fina. 2.a ed. Bilbao, Urmo, 1965. p. 263,5. 4. BUSH, M.L.; MONTGOMERY, R. & POR- TER, W.L. - Identification of organic acids on paper chromatograms. Anal. Chem., 24:489-91, 1952. 12. RILEY, R.T. & MIX, M.C. - Thin-Iayer separa tion of citric acid cycle in termedía- tes, lactic acid and the amino acids tau- rine. J. Chromatogr., 189 :286-8, 1980. 13. SÁO PAULO. Instituto Adolfo Lutz - Nor- mas Analíticas do Instituto Adolf o Lutz. v. 1: métodos químicos e físicos para aná- lise de alimentos. 2.a ed. São Paulo, 1976. p. 76. 5. CHEFTEL, R.I. et alii apud LEDERER, E. & LEDERER, M. - Chromatografia. 2.a ed. Buenos Aires, Ateneo, 1960. p. 214. 6. LEUENBERGER, U.; GAUCH, R. & BAUM- GARTNER, E. - Determination of food preservatives and saccha'rin by hígh-per- formance Iiquid chromatography. J. Chro- matogr., 173:343-8, 1979. 14. STARK, J.B.; GOODBAN, A.E. & OWENS, H.S. - Paper chromatography of organic acids. Anal. Chem., 23 :413-5, 19;:;1. Recebido para publicação em 5 de agosto de 1985. 84 84
https://openalex.org/W4393983967	https://ejournal.itn.ac.id/index.php/jati/article/download/8357/5028	Indonesian	null	KLASIFIKASI PENERIMA BANTUAN BPNT MENGGUNAKAN ALGORITMA K-NN	JATI	2,024	cc-by-sa	3,776	KLASIFIKASI PENERIMA BANTUAN BPNT MENGGUNAKAN ALGORITMA K-NN Hanna Syajida, Ade Irma Purnamasari, Tati Suprapti Teknik Informatika, STMIK IKMI Cirebon Jalan Perjuangan No.10B, Karyamulya, Kesambi, Kota Cirebon, Jawa Barat 45131, Indonesia hannasyajida6@gmail.com Jalan Perjuangan No.10B, Karyamulya, Kesambi, Kota Cirebon, Jawa Barat 45131, Indonesia hannasyajida6@gmail.com JATI (Jurnal Mahasiswa Teknik Informatika) JATI (Jurnal Mahasiswa Teknik Informatika) Vol. 8 No. 1, Februari 2024 ABSTRAK Program Keluarga Harapan (PKH) dan Bantuan Pangan Non-Tunai (BPNT) merupakan dua program bantuan sosial yang bertujuan untuk meningkatkan kesejahteraan masyarakat yang berada dalam kondisi ekonomi rendah. PKH fokus pada bantuan tunai bagi keluarga miskin, sementara BPNT memberikan bantuan pangan dalam bentuk non-tunai. Meskipun memiliki tujuan yang sama, yaitu mengentaskan kemiskinan dan memberikan bantuan kepada keluarga yang membutuhkan, seringkali terdapat overlap antara penerima bantuan dari kedua program ini. perlu dilakukan analisis lebih lanjut untuk mengidentifikasi pola-pola dalam data penerima bantuan dan memahami kesamaan karakteristik antara kedua program. Penelitian ini bertujuan untuk melakukan klasifikasi penerima bantuan Program Keluarga Harapan (PKH) dan Bantuan Pangan Non-Tunai (BPNT) menggunakan algoritma K-NN. Klasifikasi ini dilakukan dengan mengelompokkan penerima bantuan berdasarkan kriteria tertentu, seperti pendapatan keluarga, jumlah tanggungan, status rumah tinggal, pekerjaan, dinding rumah, lantai rumah, kendaraan. Penelitian ini menggunakan pendekatan kuantitatif dengan mengumpulkan data kriteria penerima bantuan PKH dan BPNT. Hasil dari penelitian memberikan kontribusi penting dalam meningkatkan akurasi dan efisiensi dalam penentuan penerima manfaat dari program ini di masa mendatang. ini dapat memberikan wawasan yang lebih baik terkait pola distribusi bantuan dan membantu penyelenggara program dalam menentukan kriteria penerima bantuan BPNT. Program ini dapat mencapai tujuan utama penerima bantuan, yaitu mengurangi tingkat kemiskinan dan meningkatkan kesejahteraan masyarakat yang berhak menerima manfaat dari program pemerintah. 1. PENDAHULUAN Selain itu, penelitian ini juga menunjukkan bahwa atribut yang paling berpengaruh dalam menentukan kelayakan penerima BST adalah memiliki bantuan lebih lanjut, pekerjaan, pendidikan, kartu kemiskinan, dan sejumlah besar tanggungan [1]. Dalam penelitian ini, akan menggunakan algoritma K-NN sebagai metode untuk klasifikasi pengelompokkan penerima bantuan berdasarkan kriteria tertentu seperti pendapatan keluarga, jumlah tanggungan, status rumah tinggal, pekerjaan, dinding rumah, lantai rumah, kendaraan. Pendekatan kuantitatif dipilih dengan mengumpulkan data kriteria penerima bantuan dari PKH dan BPNT. Implementasi algoritma ini akan dilakukan dengan bantuan perangkat Rapidminer, yang akan membantu dalam menemukan nilai yang akurat, penelitian ini untuk meningkatkan akurasi penyaluran Bahan Pangan Non Tunai (BPNT) kepada masyarakat berpenghasilan rendah di Kecamatan Beber. Dalam prosesnya akan melalui tahapan Pengumpulan Data, Pemrosesan Data, Pelatihan Model K-NN, Evaluasi Hasil, Penyusunan Laporan, Presentasi Hasil. Dari Penelitian yang dilakukan oleh [2] membahas mengenai klasifikasi penerima dan bukan penerima pangan non tunai di kelurahan utama menggunakan teknik data mining. Metode yang digunakan meliputi algoritma Naïve Bayes dan algoritma Decision Tree untuk perbandingan. Tujuan bantuan penelitian adalah untuk memberikan evaluasi terhadap program pangan pemerintah. Hasil penelitian mencakup penggunaan RapidMiner untuk membuat model data mining, dengan algoritma Naïve Bayes memiliki tingkat akurasi sebesar 58.29%, sedangkan algoritma Decision Tree memiliki tingkat akurasi sebesar 73.97%. Selain itu, penelitian juga mencakup hasil pengujian T-Test antara kedua algoritma yang menunjukkan adanya perbedaan yang signifikan. p , Penelitian ini memiliki potensi implikasi yang signifikan dalam bidang Informatika dan pengelolaan bantuan sosial. Para praktisi di lapangan akan dapat menggunakan pendekatan ini untuk meningkatkan efisiensi dalam penyaluran Bahan Pangan Non Tunai (BPNT). Masyarakat akan dapat lebih tepat dalam menentukan siapa yang sebenarnya memerlukan bantuan, sehingga bantuan dapat disalurkan dengan lebih efektif. Selain itu, hasil penelitian ini juga akan bermanfaat bagi peneliti dalam bidang Data Mining dan Algoritma K-NN. Pemahaman lebih mendalam terhadap kriteria masyarakat berpendapatan rendah, membuka peluang untuk perbaikan kebijakan lebih lanjut yang dapat mengatasi masalah kemiskinan, solusi dari penelitian dapat menanggulangi permasalahan overlap antara penerima bantuan PKH dan BPNT, sehingga distribusi bantuan dapat dilakukan dengan lebih tepat sasaran dan efektif. Penelitian ini juga dapat memberikan kontribusi yang signifikan dalam meningkatkan efektivitas program bantuan sosial dan mengarah pada upaya lebih baik dalam mengatasi kemiskinan di tingkat masyarakat berpendapatan rendah. Program ini dapat mencapai tujuan utama penerima bantuan, yaitu mengurangi tingkat kemiskinan dan meningkatkan kesejahteraan masyarakat yang berhak menerima manfaat dari program pemerintah. 1. PENDAHULUAN penerima program bantuan BPNT untuk mengatasi kemiskinan di Indonesia.untuk meningkatkan akurasi klasifikasi penerima program bantuan BPNT dengan menggunakan Algoritma K-NN[2]. Kemiskinan merupakan permasalahan yang terdapat di seluruh negara di dunia. Keadaan kemiskinan adalah situasi di mana seseorang kekurangan sumber daya finansial yang diperlukan untuk memenuhi kebutuhan dasar dalam kehidupan. Kemiskinan sering terjadi di negara-negara yang sedang mengalami perkembangan dan menjadi salah satu permasalahan yang kompleks untuk diselesaikan[1]. Adapun studi yang berhubungan dengan peneliatian dengan metode klasifikasi Algoritma K- Nearest Neighbor dapat digunakan untuk klasifikasi status sosial ekonomi di masyarakat. Dapat dilihat hasil dari perhitungan Algoritma K-Nearest Neighbor memang akurat dengan melihat data terdekatnya. Kekurangan dalam metode ini adalah Algoritma K- Nearest Neighbor perlu menentukan nilai dari parameter K, dan pembelajaran oleh jarak tidak jelas mengenai jenis jarak dan atribut apa yang harus digunakan untuk memperoleh hasil terbaik serta biaya komputasi yang cukup tinggi karena diperlukan perhitungan jarak dari setiap data training dengan data testingnya. Penelitian yang dilakukan[3]. Meskipun PKH dan BPNT memiliki sasaran yang serupa, yakni keluarga dengan kondisi ekonomi rendah, seringkali terjadi overlap antara penerima bantuan dari kedua program tersebut. Hal ini dapat menyebabkan redundansi dan kurangnya efisiensi dalam distribusi bantuan. Oleh karena itu, perlu dilakukan analisis lebih lanjut untuk mengidentifikasi pola-pola dalam data penerima bantuan dan memahami kesamaan kriteria antara kedua program tersebut. Tujuan Utama penelitian ini bertujuan untuk menerapkan algoritma K-Nearest Neighbor dalam mengklasifikasikan jenis penerimaan bantuan di Kecamatan Beber [1]. Penggunaan algoritma K- Nearest Neighbor dapat membantu pemerintah Kecamatan Beber dalam menentukan penerima bantuan yang tepat berdasarkan bantuan PKH dan BPNT. Klasifikasi ini dilakukan dengan mengelompokkan penerima bantuan berdasarkan kriteria tertentu, seperti pendapatan keluarga, jumlah tanggungan, status rumah tinggal, pekerjaan, dinding rumah, lantai rumah, kendaraan, dalam klasifikasi Dengan mengidentifikasi pola-pola dalam data penerima bantuan guna memahami kesamaan karakteristik antara kedua program tersebut. untuk memperoleh pemahaman yang lebih mendalam mengenai overlap atau kesamaan penerima bantuan antara PKH dan BPNT, penelitian ini bermaksud memberikan kontribusi dalam meningkatkan efisiensi penyelenggaraan program bantuan sosial dengan membantu menentukan kriteria penerima bantuan BPNT secara lebih tepat. penelitian ini dapat memberikan wawasan yang lebih baik terkait pola distribusi bantuan sosial, sehingga dapat memberikan 608 Vol. 8 No. 1, Februari 2024 JATI (Jurnal Mahasiswa Teknik Informatika) rekomendasi yang lebih akurat dalam perencanaan dan implementasi program bantuan. berhasil dikembangkannya model klasifikasi menggunakan algoritma C4.5 untuk menilai kelayakan penerima BST di Desa Keramas, Gianyar Bali. Model ini berhasil mencapai akurasi sebesar 97,83%. 3. METODE PENELITIAN Pengguna BPNT diklasifikasikan menggunakan algoritma K-NN dalam prosedur sistematis yang mencakup pengumpulan data, pemrosesan, pelatihan model, dan evaluasi hasil. Gambar 1 memberikan penjelasan mengenai langkah-langkah prosedur yang akan diterapkan dalam penelitian ini sebagai berikut: Penggunaan Metode Algoritma K-Nearest Neighbor untuk pendataan status sosial ekonomi masyarakat, serta pengujian keakuratan data menggunakan aplikasi RapidMiner Studio . Tujuannya adalah untuk memaksimalkan klasifikasi dan pendataan status sosial ekonomi masyarakat. Gambar 1. Metode Penelitian 2. TINJAUAN PUSTAKA Penelitian yang dilakukan oleh [8] Tujuan Utama penelitian ini bertujuan untuk menerapkan algoritma K-Nearest Neighbor dalam mengklasifikasikan jenis penerimaan bantuan di Kecamatan Beber. Penggunaan algoritma K-Nearest Neighbor dapat membantu pemerintah Kecamatan Beber dalam menentukan penerima bantuan BPNT yang tepat berdasarkan bantuan PKH. 2.3. Klasifikasi Klasifikasi merupakan suatu proses yang bertujuan untuk menemukan model atau fungsi yang dapat menjelaskan serta menggambarkan konsep atau kelas data, klasifikasi dapat dilihat sebagai suatu proses untuk mendeklarasikan suatu objek data sebagai salah satu jenisnya[7]. 2.4. RapidMiner RapidMiner adalah platform perangkat lunak ilmu data yang dikembangkan oleh perusahaan ternama sama dengan yang menyediakan lingkungan terintegrasi untuk persiapan data, pembelajaran mesin, pembelajaran dalam penambangan teks, dan analisis prediktif[5]. Hasil yang didapatkan dalam Penelitian yang dilakukan oleh [7] penggunaan metode Weighted Naïve Bayes dengan Laplace Smoothing untuk mengklasifikasikan kelayakan penerima bantuan Program Keluarga Harapan (PKH). Tujuan dari penelitian ini adalah untuk memberikan prediksi klasifikasi kelayakan penerima bantuan PKH sehingga dapat membantu perangkat desa dalam menentukan warga desa yang layak menerima bantuan tersebut. 2.2. Algoritma K-Nearest Neighbors (KNN) g g K-nearest neighbors atau knn adalah algoritma yang berfungsi untuk melakukan klasifikasi suatu data berdasarkan data pembelajaran (train data sets), yang diambil dari nilai K data terdekat (nearest neighbors). Dengan nilai k merupakan banyaknya tetangga terdekat. Tujuan utama K-NN untuk mengklasifikasikan suatu data ke dalam salah satu kelas berdasarkan mayoritas kelas dari k-tetangga terdekatnya. Dengan kata lain, KNN mencoba mencari jarak terdekat antara data yang dievaluasi dengan K- NN terdekatnya dalam data pelatihan [6]. p y g Hasil yang didapatkan dalam studi yang dilakukan oleh [6] merupakan aplikasi Algoritma C4.5 Kantor Desa Martoba Pematangsiantar untuk menentukan pola tata letak file. Setelah pengumpulan data selesai, dilakukan analisis data. Data yang digunakan adalah data primer yang dikumpulkan melalui observasi dan wawancara. Alat analisis data yang digunakan adalah Microsoft Excel dan RapidMiner. Metode yang digunakan adalah Algoritma C4.5, yang merupakan perpanjangan dari Algoritma ID3 dan menggunakan prinsip pohon keputusan. Hasil akhir dari penerapan Algoritma C4.5 dibagi menjadi dua tahap, yaitu proses perhitungan semi manual menggunakan Microsoft Excel dan penyesuaian hasil perhitungan melalui pengujian data menggunakan software RapidMiner. Berdasarkan pengolahan data menggunakan RapidMiner, sistem memiliki tingkat akurasi sebesar 100%. Dari hasil deskripsi pohon keputusan yang terbentuk, dapat disimpulkan bahwa penggunaan Data Mining Algoritma C4.5 dapat digunakan untuk menggali data dan menarik kesimpulan dengan menggunakan pohon keputusan. 1. PENDAHULUAN j y p y g g Penelitian yang dilakukan oleh [3] Penelitian ini membahas tentang penggunaan metode klasifikasi Naïve Bayes untuk mengidentifikasi rumah tangga miskin di desa Lepak, Lombok. Tujuannya adalah untuk memberikan informasi yang dapat membantu pihak desa dalam merancang strategi untuk meningkatkan kesejahteraan masyarakat. Hasil penelitian ini menunjukkan bahwa metode klasifikasi Naïve Bayes mampu mengklasifikasikan rumah tangga miskin dengan tingkat akurasi sebesar 96.63%. Penelitian ini fokus pada klasifikasi penerima bantuan sosial. Tujuannya untuk mengetahui siapa penerima bantuan yang berhak menerima bantuan dan mana yang tidak. Metodologi penelitian meliputi pengumpulan data, preprocessing data, penerapan metode K-Means, dan analisis hasil. Analisisnya berdasarkan data penerima bansos, dan dengan menggunakan metode K-Means, penerima dikelompokkan menjadi dua klaster, yaitu klaster 1 mewakili penerima yang layak menerima bantuan, dan klaster 2 mewakili penerima yang tidak layak menerima bantuan. Kesimpulan yang diperoleh dari analisis adalah bantuan tersebut sampai pada penerima yang dituju, karena mayoritas penerimanya adalah mereka yang benar-benar membutuhkan bantuan, seperti buruh yang berpenghasilan rendah dan tidak memiliki aset [4]. Penelitian yang dilakukan oleh [4] peneliti ini membahas mengenai memanfaatkan Algoritma Naïve Bayes untuk mengkategorikan penerima manfaat Program Keluarga Harapan (PKH) di Desa Sarang Helang Kecamatan Sei Kepayang Timur. Tujuannya adalah untuk membantu Dinas Sosial dalam menentukan penerima bantuan PKH yang tepat sasaran. penelitian ini berhasil mengimplementasikan Algoritma Naïve Bayes untuk mengklasifikasikan penerima Program Keluarga Harapan (PKH) dengan tingkat akurasi yang tinggi. Penelitian ini menggunakan metode data mining K-Medoids untuk mengelompokkan lapangan pekerjaan utama pada penduduk usia 15 tahun ke atas. Dalam penelitian ini, terdapat 17 lapangan pekerjaan yang dikelompokkan menjadi dua cluster, yaitu cluster tinggi dan cluster rendah. Cluster tinggi terdiri dari 3 lapangan pekerjaan utama, sedangkan cluster rendah terdiri dari 14 lapangan pekerjaan utama. Dengan menggunakan metode K-Medoids, penelitian ini dapat kelayakan penerima Bantuan Sosial Tunai (BST) untuk membuat saran dalam memilih apakah akan menerima bantuan sosial lebih lanjut. Penelitian ini 609 Vol. 8 No. 1, Februari 2024 JATI (Jurnal Mahasiswa Teknik Informatika) memberikan informasi yang berguna dalam mengatur penduduk berdasarkan lapangan pekerjaan utama mereka, menurut penelitian yang diteliti oleh [5]. 2.1. Data Mining Data mining adalah proses menganalisa data dari perspektif atau penilaian yang berbeda dan menyimpulkannya menjadi informasi penting yang dapat digunakan, data mining dapat disebut sebagai proses untuk menemukan korelasi atau pola dari ratusan atau ribuan field dari sebuah basis data relasional yang besar [5]. Gambar 1. Metode Penelitian 610 JATI (Jurnal Mahasiswa Teknik Informatika) Vol. 8 No. 1, Februari 2024 No. Uraian Keterangan Kendaraan Polynomial Alamat Polynomial Hasil Binomial No. Uraian Keterangan Kendaraan Polynomial Alamat Polynomial Hasil Binomial 4.4. Transformation Gambar 3. Operator Set Role pada Rapidminer Saat mengubah data untuk tujuan penelitian untuk mengubah tipe atribut penerima BPNT menjadi label, digunakan operator set role. Parameter pada Set Role yang digunakan bisa dilihat pada tabel dibawah ini. Gambar 5. Set Role Tabel 2. Parameter dan atribut yang dipilih pada operator select Atribut No. Uraian Keterangan 1 Record 1224 2 Special Atribut 2 3 Reguler Atribut 10 4 Attribut : Nik Real Nama Polynomial Bansos Polynomial Penghasilan Polynomial Status Rumah Polynomial Tanggungan Integer Pekerjaan Polynomial Tabel 2. Parameter dan atribut yang dipilih pada operator select Atribut Gambar 5. Set Role 4.1. Data Informasi tersebut terkandung dalam file Excel bernama Dataset PKH diperoleh dari Kantor Kecamatan Beber yang berjumlah sebanyak 1530, dataset pkh.xlsx yang sudah di impor terdiri dari 12 atribut seperti pada tabel dibawah ini. 4.2. Selection Untuk menentukan label pada dataset, digunakan operator Set Role seperti tampak pada gambar dibawah ini. Dari hasil Result dataset sudah tidak memiliki lagi titik data yang hilang. Kumpulan data diperiksa secara langsung melalui catatan untuk melihat apakah konsisten atau tidak, dan ini menunjukkan bahwa kumpulan data berisi data yang konsisten dengan nilainya. ni. Gambar 3. Operator Set Role pada Rapidminer 4.3. Pre-Processing Proses cleaning atau pembersihan data yang missing atau memiliki nilai yang tidak konsisten pada langkah preprocessing. Sebelum melakukan proses ini, dilakukan analisis terlebih dahulu apakah atribut pada dataset yang dipilih memiliki nilai missing atau tidak serta memiliki data yang konsisten atau tidak. Dari hasil result dari statistik dataset seperti tampak pada gambar 4.4 dibawah ini, diketahui bahwa tidak ada atribut yang memiliki nilai missing. Untuk memeriksa konsisten atau tidak konsistennya dataset yang digunakan diperiksa per-record secara langsung dan menunjukan bahwa dataset memiliki data yang konsisten terhadap nilainya. Tabel 1. Dataset Penerima PKH NO Atribut Tipe Data 1. NIK Real 2. Nama Polynomial 3. Bansos Polynomial 4. Penghasilan Polynomial 5. Status Rumah Polynomial 6. Pekerjaan Polynomial 7. Jumlah Tanggungan Integer 8. Lantai Polynomial 9. Dinding Polynomial 10. Kendaraan Polynomial 11. Alamat Polynomial 12. Hasil Binomial Gambar 4. Data Statistic Agar dataset ada di repository Rapidminer, maka perlu dilakukan import data Dataset PKH.xlsx menggunakan operator ReadExcel. Gambar 2. Operator ReadExcel pada Rapidminer Gambar 2. Operator ReadExcel pada Rapidminer Gambar 4. Data Statistic 4. HASIL DAN PEMBAHASAN 4.1. Data 4.6. Evaluation Dengan menggunakan akurasi sebesar 93.14% sebagai ukuran performa keseluruhan model, kita dapat melihat evaluasi lebih rinci dengan mempertimbangkan metrik-metrik lain seperti presisi, recall, untuk setiap penerima BPNT. Berdasarkan hasil evaluasi model yang telah diberikan sebelumnya: Gambar 8 Hasil Visualisasi Penerima BPNT 1) Presisi (Precision) a. Kelas “Ya” : 95.42% b. Kelas “Tidak” : 63.64% 2) Recall a. Kelas “Ya” : 97.13% b. Kelas “Tidak” : 51.85% 1) Presisi (Precision) a. Kelas “Ya” : 95.42% b. Kelas “Tidak” : 63.64% b. Kelas “Tidak” : 51.85% Gambar 8 Hasil Visualisasi Penerima BPNT Temuan evaluasi ini menawarkan pemahaman yang lebih mendalam mengenai fungsi model di setiap kelas penerima BPNT. Meskipun akurasi model secara keseluruhan sangat tinggi, berkonsentrasi pada presisi dan perolehan kembali mungkin menunjukkan area yang perlu ditingkatkan, terutama jika kelas tertentu memiliki kebutuhan unik. Pada Gambar 8 merupakan representasi kategorisasi penerima BPNT di Kecamatan Beber terbagi menjadi dua kategori yaitu Ya dan Tidak. Gambar 9 Hasil Accuracy Gambar 9 Hasil Accuracy 4.5. Data Mining Langkah-langkah dalam proses pencarian atau eksplorasi Mencari informasi yang signifikan dari ukuran pasti kumpulan data yang telah dipilih dengan menggunakan teknik atau metode tertentu merupakan tahap pertama dalam penelitian ini. Metode K-NN digunakan untuk mengklasifikasikan data kelompok 611 JATI (Jurnal Mahasiswa Teknik Informatika) Vol. 8 No. 1, Februari 2024 berdasarkan peruntukan penerima bantuan BPNT. Perangkat lunak yang disebut RapidMiner dan algoritma K-NN digunakan untuk memproses data. Model data mining yang dibuat untuk penelitian ini adalah sebagai berikut: Terlihat dari data kinerja bahwa model atau sistem klasifikasi telah dievaluasi menggunakan dua kategori penerima BPNT: Ya dan Tidak 1) Accuracy (Akurasi) Nilai Akurasi: 93.14% Akurasi mengukur sejauh mana model mampu memprediksi dengan benar, diukur dalam persentase. Dalam hal ini, model memiliki tingkat akurasi sebesar 93.14%, yang merupakan persentase dari jumlah prediksi yang benar dari keseluruhan data. 1) Accuracy (Akurasi) Nilai Akurasi: 93.14% Akurasi mengukur sejauh mana model mampu memprediksi dengan benar, diukur dalam persentase. Dalam hal ini, model memiliki tingkat akurasi sebesar 93.14%, yang merupakan persentase dari jumlah prediksi yang benar dari keseluruhan data. Gambar 6 Model Proses Data Mining 2) Weighted Mean Precision (Presisi Rata- Rata Terbobot) Nilai Precision: 79.53% dengan Bobot: 1, 1. Precision mengukur sejauh mana prediksi positif model benar-benar relevan. Weighted Mean Precision dihitung dengan memberikan bobot pada masing-masing kelas. Sejumlah 79.53% dari prediksi positif model relevan. Gambar 6 Model Proses Data Mining Dari hasil Gambar 6 penerapan dari Algoritma K- NN dalam penelitian ini untuk mengklasifikasikan data penerima BPNT. Gambar 7 Hasil Exampleset 3) Weighted Mean Recall (Sensitivitas Rata- Rata Terbobot) Nilai Recall: 74.49% dengan Bobot: 1, 1. Recall mengukur sejauh mana model dapat mendeteksi semua instance positif. Dalam hal ini, Weighted Mean Recall dihitung dengan memberikan bobot pada masing-masing kelas. Sejumlah 74.49% dari instance positif YA dan TIDAK berhasil dideteksi oleh model. Gambar 7 Hasil Exampleset Berdasarkan Gambar 7, hasil prediksi kualitas udara dibagi menjadi dua kelompok berdasarkan hasil sample set yang merupakan keluaran dari exampleset kinerja operator. Klasifikasi tersebut adalah sebagai berikut: Ya dan Tidak 4.7. Bagaimana menentukan kriteria penerima BPNT berdasarkan penerima PKH? Dalam penelitian ini, dilakukan analisis terhadap kriteria penerima bantuan dari Program Keluarga Harapan (PKH) dan Bantuan Pangan Non-Tunai Gambar 9 Hasil Accuracy Gambar 9 Hasil Accuracy 612 Vol. 8 No. 1, Februari 2024 JATI (Jurnal Mahasiswa Teknik Informatika) [3] L. Qadrini and S. A, "Decision Tree dan Adaboostpada Klasifikasi Penerima Program Bantuan (BPNT). Kriteria yang dievaluasi meliputi pendapatan keluarga, jumlah tanggungan, status rumah tinggal, pekerjaan, dinding rumah, lantai rumah, dan kendaraan. Penelitian ini menggunakan Algoritma K- NN yang diterapkan untuk mengklasifikasikan penerima bantuan BPNT berdasarkan karakteristik yang ditemukan pada penerima bantuan PKH. K-NN merupakan metode klasifikasi yang memanfaatkan informasi dari tetangga terdekat untuk memprediksi kategori suatu data. [4] N. W. Oktha dan N. Widya Utami, “Klasifikasi Penentuan Penerima Bantuan Sosial Tunai (BST) Menggunakan Algoritma C4.5 Di Desa Keramas, Gianyar, Bali,” Jurnal Informatika, Teknologi dan Sains, pp. 101-107, 2022. [5] C. Agus dan F. Rizky Maulana, “Algoritma Naïve Bayes Untuk Klasifikasi Penerima Bantuan Pangan Non Tunai ( Studi Kasus Kelurahan Utama ),” Techno.Com, pp. 321-331, 2019. 4.8. Bagaimana cara kita mengetahui penerima itu layak atau tidaknya menerima bantuan? [6] W. Pura dan D. Ayu Lara, “Penerapan Naive Bayes dalam Mengklasifikasikan Masyarakat Miskin di Desa Lepak,” Geodika: Jurnal Kajian Ilmu dan Pendidikan Geografi, pp. 123-132, 2021. Mengevaluasi dari kriteria penerima bantuan, langkah-langkah klasifikasi menggunakan K-NN melibatkan evaluasi kriteria seperti pendapatan keluarga, jumlah tanggungan, status rumah tinggal, pekerjaan, dinding rumah, lantai rumah, dan kendaraan. Ini memungkinkan penentuan apakah seseorang layak atau tidak menerima bantuan BPNT berdasarkan kriteria yang telah ditetapkan. [7] L. Ganda Rady dan . A. Anggrawan, “Pengelompokan Penerima Bantuan Sosial Masyarakat dengan Metode K-Means,” MATRIK : Jurnal Manajemen, Teknik Informatika dan Rekayasa Komputer, pp. 205-214, 2021. 5. KESIMPULAN DAN SARAN [8] N. Pulungan dan Suhada, “PENERAPAN ALGORITMA K-MEDOIDS UNTUK MENGELOMPOKKAN PENDUDUK,” KOMIK(Konferensi Nasionaal Teknologi Informasi dan Komputer), pp. 329-334, 2019 Berdasarkan hasil dan pembahasan penelitian ini, mencakup nilai akurasi sebesar 93.14%, weighted mean precision sebesar 79.53%, dan weighted mean recall sebesar 74.49%. Akurasi mengukur sejauh mana model dapat memprediksi dengan benar, sedangkan precision dan recall memberikan informasi tentang sejauh mana model mampu memberikan prediksi yang relevan dan mendeteksi instance positif. Penelitian ini membuktikan bahwa penggunaan algoritma K-NN dapat efektif digunakan untuk mengklasifikasikan penerima bantuan BPNT berdasarkan karakteristik yang ada pada penerima bantuan PKH. Hasil evaluasi kinerja model menunjukkan tingkat akurasi yang tinggi, menunjukkan potensi model untuk digunakan dalam penentuan penerima bantuan di masa mendatang. Dengan demikian, penelitian ini memiliki implikasi penting dalam peningkatan efisiensi dan efektivitas program bantuan sosial di Indonesia. Untuk studi berikutnya, disarankan untuk Eksplorasi pengembangan model yang lebih kompleks atau kombinasi algoritma untuk meningkatkan kinerja dan akurasi prediksi. Mungkin ada model machine learning lain yang dapat memberikan hasil yang lebih baik [9] W. Lidysari dan H. Staria Tambunan, “Penerapan Data Mining Dalam Menentukan Kelayakan Penerima Bantuan Sosial Pemko Dengan Algoritma C4 . 5 ( Kasus Kantor Kelurahan Martoba ),” KESATRIA Jurnal Penerapan Sistem Informasi (Komputer dan Manajemen), pp. 53- 61, 2022 [10] D Utami dan P. Aisyiyah , “Klasifikasi Kelayakan Penerima Bantuan Program Keluarga Harapan (Pkh) Menggunakan Metode Weighted Naïve Bayes Dengan Laplace Smoothing,” JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika), pp. 1373-1384, 2022. [11] D. Irawan dan D. Riswanto, “Penerapan Algoritma K-Nearest Neighbor (KNN) Untuk Mengklasifikan Jenis Penerimaan Bantuan,” JTKSI (Jurnal Teknologi Komputer dan Sistem Informasi), 2023. [12] L. Qadrini dan S. A, “Decision Tree dan Adaboostpada Klasifikasi Penerima Program Bantuan Sosial,” Decision Tree Dan Adaboost Pada Klasifikasi Penerima Program Bantuan Sosial, pp. 1959-1966, 2021. DAFTAR PUSTAKA [1] W. and N. , "Perbandingan Performa Cluster Model Algoritma K-Means Dalam Mengelompokkan Penerima Bantuan Program Keluarga Harapan," Jurnal Sistem Informasi Bisnis, vol. 1, pp. 20-28, 2003 [13] R. Fitra dan I. Rusdi, “Penerapan Metode Algoritma K-Nearest Neighbor Menggunakan Rapidminer Studio Pada Klasifikasi Status Sosial Ekonomi Studi Kasus : Kelurahan Kapuk Muara Rt 010 Rw 04,” Smart Comp: Jurnalnya Orang Pintar Komputer, pp. 653-660, 2022. [2] D. Irawan and D. Riswanto, "Penerapan Algoritma K-Nearest Neighbor (KNN) Untuk Mengklasifikan Jenis Penerimaan Bantuan," JTKSI (Jurnal Teknologi Komputer dan Sistem Informasi), 2023. [14] A. Asri Awaliyah dan W. Handoko , “Implementasi Metode Naive Bayes Untuk Klasifikasi Penerima Program Keluarga 613 Vol. 8 No. 1, Februari 2024 JATI (Jurnal Mahasiswa Teknik Informatika) Harapan,” J-Com (Journal of Computer), pp. 21- 26, 2022 [19] R. Fitra dan I. Rusdi, “Penerapan Metode Algoritma K-Nearest Neighbor Menggunakan Rapidminer Studio Pada Klasifikasi Status Sosial Ekonomi Studi Kasus : Kelurahan Kapuk Muara Rt 010 Rw 04,” Smart Comp: Jurnalnya Orang Pintar Komputer, vol. 11, no. 4, pp. 653-660, 2022. [15] F. Kurnia dan J. Kurniawan , “Klasifikasi Penduduk Miskin Penerima PKH Menggunakan Metode Naïve Bayes dan KNN,” Jurnal Kajian dan Terapan Matematika, pp. 75-84, 2022. [16] A. N. Shahifatun dan H. Mulyo, “Klasifikasi penerima bantuan program rehabilitasi rumah tidak layak huni menggunakan algoritme K- Nearest Neighbor Classification of beneficiaries for the rehabilitation of uninhabitable houses using the K-Nearest Neighbor algorithm,” Jurnal Teknologi dan Sistem Komputer, pp. 32-37, 2022. [20] F. Kurnia dan J. Kurniawan, “Klasifikasi penerima bantuan program rehabilitasi rumah tidak layak huni menggunakan algoritme K- Nearest Neighbor Classification of beneficiaries for the rehabilitation of uninhabitable houses using the K-Nearest Neighbor algorithm,” Jurnal Teknologi dan Sistem Komputer, vol. 10, no. 1, pp. 32-37, 2022. [17] E. Rahma dan M. Fauzan, “Klasifikasi Penduduk Miskin Penerima PKH Menggunakan Metode Naïve Bayes dan KNN,” Jurnal Kajian dan Terapan Matematika, pp. 75-84, 2022. [21] W. dan N. , “Perbandingan Performa Cluster Model Algoritma K-Means Dalam Mengelompokkan Penerima Bantuan Program Keluarga Harapan,” Jurnal Sistem Informasi Bisnis, vol. 1, pp. 20-28, 2003. [18] A. Ifon Purnama dan A. Aziz, “Penerapan Data Mining Untuk Mengklasifikasi Penerima Bantuan Pkh Desa Wae Jare Menggunakan Metode Naïve Bayes,” Kurawal - Jurnal Teknologi, Informasi dan Industri, pp. 173-180, 2020 [22] Fajriana, “Classification of Determination the Recipients of the Program,” JITE (Journal of Informatics and Telecommunication Engineering), vol. 6, no. January, pp. 403-413, 2023 [23] 614
https://openalex.org/W3101961216	http://jurnal.unsil.ac.id/index.php/jes/article/download/1949/1361	Indonesian	null	APAKAH PEMBIAYAAN BANK SYARIAH DAN TINGKAT PENDIDIKAN BERPENGARUH TERHADAP PERTUMBUHAN EKONOMI PADA ENAM PROVINSI DI PULAU JAWA ?	Jurnal ekonomi syariah/Jurnal Ekonomi Syariah	2,020	cc-by	6,682	Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 ABSTRAK Penelitian kami dengan tujuan untuk menganalisa hubungan kausalitas antara pembiayaan bank syariah serta tingkat pendidikan terhadap tingkat pertumbuhan ekonomi. Dimana ipertumbuhan ekonomi Indonesia pada 10 tahun terakhir ini dinilai cukup tinggi. Sedangkan pertumbuhan pembiayaan perbankan syariah sebagai lembaga keuangan yang dapat imembantu ipertumbuhan iekonomi idinilai ilamban. iJuga iterdapat ifakta ibeberapa iprovinsi idi iIndonesia, idimana itingkat ipendidikannya itergolong imasih irendah, inamun ipertumbuhan iekonominya imengalami ipeningkatan iyang iterbilang itinggi, iseperti ihalnya iMaluku idan iPapua iyang imengalami ipertumbuhan iekonomi itertinggi iditahun i2018, idisusul iSulawesi idiurutan ikedua idan ipulau iJawa idiurutan ike itiga. i iMetode iyang ikami igunakan pada ipenelitian iini iadalah ianalisis idata ipenel imenggunakan ipendekatan irandom ieffect imethod. Kata iKunci: ipembiayaan, itingkat ipendidikan, Ipdb ABSTRACT Our iresearch iaimed ito ianalyze icausality irelations ibetween iSharia ibank ifinancing iand ieducational ilevel ito ieconomic igrowth. iAt ithe isame itime, iIndonesia igrowth iof ieconomic iin ithe ilast idecade iis iconsidered ihigh imeanwhile, ithe iSharia ibank ifinancing iis iconsidered ilow. iAnd ithere iis ia ifact ithat ia ifew iprovinces iwhich itheir ieducational ilevel iis iconsidered ilow, ihowever itheir ieconomic igrowth iincrease ihigh, isuch ias iMaluku iand iPapua, iwhich ihad ithe ihighest ieconomic igrowth iin i2018, ifollowed iby iCelebes iand iJava iin ithe isecond iand ithe ithird iposition irespectively. iWe iused ipanel idata ianalysis iwith ithe iapproach iof ithe irandom ieffect imethod iin ithis ipaper. Keywords: iFinance, iEducational iLevel, iGDP Keywords: iFinance, iEducational iLevel, iGDP iAPAKAH iPEMBIAYAAN iBANK iSYARIAH iDAN iTINGKAT iPENDIDIKAN iBERPENGARUH iTERHADAP iPERTUMBUHAN iEKONOMI iPADA iENAM iPROVINSI iDI iPULAU iJAWA? Muhammad iJagat iDermawan Jurusan iIlmu iEkonomi idan iKeuangan iIslam, iUPI iBandung, Email: mjagatd@upi.edu PENDAHULUAN Produk Domestik Bruto (PDB) yang mewakili pertumbuhaniekonomiiIndonesia pada 10 tahun terakhir ini cukupistabilibahkan dapat dikategorikan itinggi ijika iperbandingannya \| 31 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X D i Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 iitu idengan inegara iberkembang iyang ilainnya. iDengan imenggunakan idata iyang idipublikasikan ioleh iBadan iPusat iStatistik i(BPS) itingkat ipertumbuhan iekonomi iIndonesia iberada ipada ikisaran i5% idalam i10 itahun iterakhir. iitu idengan inegara iberkembang iyang ilainnya. iDengan imenggunakan idata iyang idipublikasikan ioleh iBadan iPusat iStatistik i(BPS) itingkat ipertumbuhan iekonomi iIndonesia iberada ipada ikisaran i5% idalam i10 itahun iterakhir. Disisi ilain ipertumbuhan ibank isyariah i(sektor ikeuangan) iyang imenjadi isatu ifaktor idiantara iyang ilainnya, idianggap imampu imendukung itumbuh ikembang iekonomi imemperlihatkan imasih ibelum imenunjukkan iperan iyang idiharapkan. iMenurut ilaporan, ipada ikuartal iIII-2018 idan isebelumnya, ikinerja ibank iumum isyariah i(BUS) idari isisi ipembiayaannya imengalami ipelambatan. iApabila idisempitkan iberdasarkan ikategori iusaha, iBank iUmum iKelompok iUsaha i(BUKU) iI iyang imengalami ipenurunan ipaling iburuk. iBUS iBUKU iI ipada iJuli i2018 imencata ipada isektor iindustri imengalami ipenyusutan ipembiayaan iyang icukup idrastis imenjadi inegatif i10,54%. iSementara iitu, ipada isetahun isebelumnya, ipembiayaan ipada iBUKU iI isyariah imengalami ipertumbuhan isebesar i6,33%. iKemudian ibaik iBUKU iII imaupun iBUKU iIII isyariah imengalami ipenyusutan ipertumbuhan ipembiayaan imenjadi i3,35% idan i8,51%. iJika idibandingkan idengan itahun isebelumnya iyang imampu inaik ihingga i8,17% idan i8,51% ipada iJuli i2017 i(“Agustus, ipertumbuhan ikredit iperbankan imelambat,” in.d.). Di ibawah iini idisajikan itabel ipertumbuhan iekonomi iIndonesia idan ipembiayaan ibank isyariah idari itahun i2009-2018. TABEL i1 i i i i i iDATA iPERTUMBUHAN iEKONOMI iINDONESIA iDAN iPEMBIAYAAN iBANK iSYARIAH i2009-2018 Tahun PDB PEMBIAYAAN iBANK iSYARIAH 2009 1303666.38 1354873.656 2010 1395689.34 1718442.258 2011 1472742.2 2268258.074 2012 1561416.15 3016344.149 2013 1606367.61 3975781.385 2014 1682633.63 4423009.167 2015 1770952.22 5130200.519 2016 1857257.63 6026102.005 2017 1951182.17 7076138.437 2018 2046638.88 8370792.826 i iSumber i: iBPS idan iSPS idata idiolah ikembali \| 32 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X p Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 Bahkan ibila imelihat idata iStatistik iPerbankan iSyariah i(SPS) imilik iOtoritas iJasa iKeuangan i(OJK), ikeadaan ipembiayaan ibermasalah i(non-performing ifinancing/NPF) ikonvensional ilebih irendah idaripada isyariah. iPada iakhir iMaret, itingkat iNPF iada ipada i3,44%, idan idisisi ilain, ikredit ibermasalah iperbankan ikonvensional i(NPL) iada ipada i2,5%. iJika imelihat iperiode isebelumnya, itingkat ipembiayaan ibermasalah iperbankan isyariah ilebih ibesar ilagi. iContohnya ipada iakhir i2017 iyang imencapai i4,76% iataupun i2016 iyang imencapai i4,42% i(Hastuti, in.d.). iDari idata idiatas isangat imenarik iuntuk iditeliti ikarena ipembiayaan iperbankan iIslam idi iIndoensia imengalami ipertumbuhan iyang itidak imenggembirakan. iSementara ipertumbuhan iekonomi idalam isepuluh itahun iterakhir iini imengalami ipeningkatan. iAdapun ipembiayaan iperbankan isyariah iyang itelah idisalurkan idi ienam iprovinsi iyang iada idi ipulau iJawa iselama iperiode i2009-2018, iseperti iterlihat ipada itabel idi ibawah iini: TABEL i2 i i i iTOTAL iPEMBIAYAAN iBANK iSYARIAH TABEL i2 i i i i i iTOTAL iPEMBIAYAAN iBANK iSYARIAH PADA iENAM iPROVINSI iDI iPULAU iJAWA i2009-2018 Tahun Total/tahun Pertumbuhan 2009 1171172 2010 1475844 4.33% 2011 1970371 5.67% 2012 2622773 5.50% 2013 3414213 5% 2014 3882912 2.33% 2015 4470992 2.50% 2016 5144778 2.50% 2017 5997963 2.83% 2018 7086335 3% Sumber i: iOJK iSPS idata idiolah iKembali Dari idata itersebut imenunjukkan itotal ipembiayaan iperbankan isyariah iyang idikeluarkan idi ienam iprovinsi iyang iada idi ipulau iJawa ipertumbuhannya iberpluktuasi. iPertumbuhan ipada itahun i2011 iadalah iyang itertinggi idengan ikisaran i5.67% idan ipertumbuhan ipada itahun i2015 idan i2016 iadalah iyang iterendah idengan ikisaran ihanya i2.50%. TABEL i4 ANGKA iPARTISIPASI iMURNI i(APM) iSM, iSMK, iMA, iDAN iPAKET iC PADA iENAM iPROVINSI iDI iPULAU iJAWA TAHUN DKI JABAR BANTEN JATENG DIY JATIM 2009 50.43 38.59 38.77 44.53 58.69 48.26 2010 50.57 38.84 39.61 45 59.35 48.6 2011 49.91 42.45 46.24 47.17 59.25 49.29 2012 54.25 51.24 53 51.11 63.54 52.36 2013 55.4 52.25 53.28 51.81 64.86 53.3 2014 58.79 56.48 56.87 58.11 68.46 60 2015 59.04 56.73 57.04 58.27 68.6 60.31 2016 59.3 56.92 57.21 58.49 68.96 60.76 2017 59.54 57.22 57.88 59.2 69.66 61.49 2018 60.01 57.33 58.72 59.31 70.22 61.51 Sumber iBPS idata idiolah ikembali Data idi iatas iadalah idata idengan ijenjang ipendidikan iSekolah iMenengah i(SM), iSekolah iMenengah iKejuruan i(SMK), iMadrasah iAliyah i(MA) idan ipaket iC i(jenjang ipendidikan isetingkat iSekolah iMenengah. iDimana iangka iparisipasi imurni i(APM) ipada ienam iprovinsi idi ipulau iJawa imenunjukkan idiatas i50. iMeskipun ipada itahun i2009 idan i2010 iprovinsi iJawa iBarat, iBanten idan iJawa iTengah iberada idibawah i50. i p Doi : 10.37058/jes.v5i2.1949 Berikut itabel ipertumbuhan iekonomi idi ipulau iJawa idalam iperiode ipeneletian, iyang idiambil idari itahun i2009 ihingga itahun i2018: \| 33 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 TABEL i3 PERTUMBUHAN iEKONOMI iATAS iDASAR iHARGA iKONSTAN PADA iENAM iPROVINSI iDI iPULAU iJAWA Miliar iRupiah TAHUN DKI JABAR BANTEN JATENG DIY JATIM 2009 1002808,73 847749,46 252384,72 590181,11 61308,07 i926895,91 2010 1075183,48 906685,76 271465,28 623224,62 64678,97 990648,84 2011 1147558,23 965622,06 290545,84 656268,13 68049,87 1054401,77 2012 1222527,92 1028409,74 310385,59 691343,12 71702,45 1124464,64 2013 1296694,57 1093543,55 331099,11 726655,12 75627,45 1192789,80 2014 1373389,13 1149216,06 349351,23 764959,15 79536,08 1262684,50 2015 1454563,85 1207232,34 368377,20 806765,09 83474,45 1331376,10 2016 1540078,20 1275527,64 387824,35 849313,20 87688,20 1405561,04 2017 1635855,75 1342953,38 409959,69 894050,47 92300,66 1482147,59 2018 1736788,05 1418695,95 423775,36 911613,96 94023,30 1521720,93 Sumber iBPS idata idiolah ikembali Miliar iRupiah TAHUN DKI JABAR BANTEN JATENG DIY JATIM 2009 1002808,73 847749,46 252384,72 590181,11 61308,07 i926895,91 2010 1075183,48 906685,76 271465,28 623224,62 64678,97 990648,84 2011 1147558,23 965622,06 290545,84 656268,13 68049,87 1054401,77 2012 1222527,92 1028409,74 310385,59 691343,12 71702,45 1124464,64 2013 1296694,57 1093543,55 331099,11 726655,12 75627,45 1192789,80 2014 1373389,13 1149216,06 349351,23 764959,15 79536,08 1262684,50 2015 1454563,85 1207232,34 368377,20 806765,09 83474,45 1331376,10 2016 1540078,20 1275527,64 387824,35 849313,20 87688,20 1405561,04 2017 1635855,75 1342953,38 409959,69 894050,47 92300,66 1482147,59 2018 1736788,05 1418695,95 423775,36 911613,96 94023,30 1521720,93 Sumber iBPS idata idiolah ikembali Dari idata idi iatas imenunjukkan ibahwa ipertumbuhan iekonomi ipada ienam iprovinsi idi ipulau iJawa isetiap itahunnya iselalu imeningkat. iNamun ihal iyang imenarik iadalah iprovinsi iDKI iJakarta iyang imenunjukkan ipertumbuhan iekonomi iyang iselalu itertinggi idiantara ienam iprovinsi itersebut, ipadahal iprovinsi iDKI iJakarta iberdasarkan idata iSPS ipembiayaan iyang idikucurkan iperbankan isyariah irelative isedikit idibanding idengan iprovinsi ilainnya iseperti iJawa iBarat, iBanten, iJawa iTengah, iDIY idan iJawa iTimur. Faktor ilain iyang idapat imempengaruhi ipertumbuhan iekonomi iadalah ikualitas isumber idaya imanusia. iPendidikan imerupakan isalah isatu icara iyang idapat imeningkatkan ikualitas iSDM itersebut. iUntuk imengetahui iseberapa ibanyak ipenduduk iyang imemanfaatkan ifasilitas ipendidikan isalah isatunya idapat idilihat idari ipersentase ipenduduk imenurut ipartisipasi isekolah idengan imelihat iAngka iPartisipasi iMurni i(APM) iwilayah itersebut. iAngka ipartisipasi imurni i(APM) iadalah iperbandingan iantara isiswa iusia isekolah itertentu ipada ijenjang ipendidikan idengan ipenduduk iusia iyang isesuai idan idinyatakan idalam ipersentase. iKriterianya iadalah isemakin itinggi iAPM iberarti isemakin ibanyak ianak iusia isekolah iyang ibersekolah isesuai iusia iresmi idi ijenjang ipendidikan itertentu. i(“APK-APM,” in.d.). Adapun iangka ipartisipasi imurni i(APM) ipada ienam iprovinsi idi ipulau iJawa idalam ikurun iwaktu i2009 isampai idengan i2018 iseperti iterlihat ipada itabel idibawah iini: \| 34 Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 ikeuangan idapat imempengaruhi ipertumbuhan iekonomi idengan imemperbanyak ikapasitas ibank iuntuk idapat imemperbanyak ikredit ikepada isektor iriil i(Olufemi iAdeyeye, i2015). ikeuangan idapat imempengaruhi ipertumbuhan iekonomi idengan imemperbanyak ikapasitas ibank iuntuk idapat imemperbanyak ikredit ikepada isektor iriil i(Olufemi iAdeyeye, i2015). Menurut iMegginson, iteori iintermediasi i(Intermediation iTheory) imembahas itentang ieksistensi ilembaga iperantara ikeuangan, iterutama ibank iumum iatau ibank ikomersial ikarena ibank imemiliki ikeunggulan ikompetitif iyang ijelas iatas ipasar imodal iuntuk isemua ipendanaan-pendanaan iperusahaan. iPerantara ikeuangan idan ipasar ikeuangan iadalah idua ilembaga ipenting, iyang iberkontribusi iterhadap ialokasi isumber idaya iyang ioptimal idalam isuatu iperekonomian i(Jaelani, i2015). iTeori iintermediasi ikeuangan isangat iterkait imakro iekonomi, imenurut iAndries ibahwa iteori iintermediasi ikeuangan imodern imembahas ibagaimana iintermediasi ikeuangan imempengaruhi iekonomi isecara ikeseluruhan i(Jaelani, i2015). ikeuangan idapat imempengaruhi ipertumbuhan iekonomi idengan imemperbanyak ikapasitas ibank iuntuk idapat imemperbanyak ikredit ikepada isektor iriil i(Olufemi iAdeyeye, i2015). Menurut iMegginson, iteori iintermediasi i(Intermediation iTheory) imembahas itentang ieksistensi ilembaga iperantara ikeuangan, iterutama ibank iumum iatau ibank ikomersial ikarena ibank imemiliki ikeunggulan ikompetitif iyang ijelas iatas ipasar imodal iuntuk isemua ipendanaan-pendanaan iperusahaan. iPerantara ikeuangan idan ipasar ikeuangan iadalah idua ilembaga ipenting, iyang iberkontribusi iterhadap ialokasi isumber idaya iyang ioptimal idalam isuatu iperekonomian i(Jaelani, i2015). iTeori iintermediasi ikeuangan isangat iterkait imakro iekonomi, imenurut iAndries ibahwa iteori iintermediasi ikeuangan imodern imembahas ibagaimana iintermediasi ikeuangan imempengaruhi iekonomi isecara ikeseluruhan i(Jaelani, i2015). Sementara iberdasarkan isalah isatu istudi iyang iterkenal iyang idilakukan ioleh iDenison iyang imenganalisis ifaktor iyang imengakibatkan i i iperkembangan idi iNegara imaju idi iantara itahun i1950-1962. iKesimpulan ikajiannya itersebut iadalah, i ipertambahan ibarang-barang imodal ihanya imewujudkan i25 ipersen idari ipertumbuhan iekonomi idi iAmerika iSerikat, i18 ipersen idari ipertumbuhan iekonomi idi iEropa iBarat idan i21 ipersen idari ipertumbuhan iekonomi iyang iterjadi idi iInggris, imenunjukkan ibahwa i iteknologi idan iperkembangan iketerampilan iyang imenjadi ifaktor iutama iyang imewujudkan ipertumbuhan iekonomi i(Sukirno, i2006). iSedangkan imenurut iArsyad ibahwa ifaktor- faktor iyang idapat imempengaruhi ipertumbuhan iekonomi i idiantaranya iadalah i1. iAkumulasi imodal, itermasuk isemua iinvestasi ibaru iyang iberwujud itanah, iperalatan ifisik i(mesin-mesin), idan isumber idaya imanusia i(human iresources) i2. iPertumbuhan ipenduduk, i3. iKemajuan iteknologi, i4. iSumber idaya iinstitusi i(sistem ikelembagaan) i(Arsyad, i2016). Sedangkan iteori imodal imanusia imenjelaskan iproses idimana ipendidikan imemiliki ipengaruh ipositif ipada ipertumbuhan iekonomi. iTeori iini itelah imendominasi iliteratur ipembangunan iekonomi idan ipendidikan idari ipasca iperang idunia ikedua isampai ipada itahun i70-an. LITERATUR iREVIEW iDAN iHIPOTESIS Sejak berdirinya dan berkembangnya iindustri iperbankan isebagai ilembaga ipemberi imodal itidak isedikit ipeneliti iyang imeneliti itentang ipengaruh isektor ikeuangan idapat imenjadi ipendorong ipertumbuhan iekonomi. Menurut iGhali, iCopelman, iGraff, iFritzer, iAllen, iLee, idan iValey, ibahwa idari ihasil istudinya imereka imendukung ihipotesis ibahwa isektor ikeuangan imenjadi ikatalistaor idalam ipertumbuhan iekonomi i(supply ileading ihypothesis) i(Inggrid, i2006). iSementara istudi iyang idilakukan iPatrick idi iNegeria, idimana idari ihasil ipenelitiannya imenunjukkan ibahwa isupply ileading ihypothesis imemiliki ipengaruh iyang ilemah ipada ipertumbuhan iekonomi idi iNigeria. iHal iini iwajar ikarena iperkembangan isektor ikeuangan idi iNigeria ibelum ibegitu iberkembang iseperti idi inegara-negara imaju. iSehingga ipertumbuhan iekonomi idi iNigeria ilebih ibanyak ididorong ioleh isektor iriil. iNamun idalam ikesimpulannya imenyatakan ibahwa iperkembangan isektor \| 35 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 iPara ipelopornya iantara ilain iadalah ipemenang ihadian iNobel iilmu iekonomi iGary iBecker idari iUniversitas iChicago, iAmerika iSerikat, iEdward iDenison idan iTheodore iSchultz, iyang ijuga ipemenang ihadiah inobel iekonomi iatas ipenelitiannya itentang imasalah iini. iArgumensi iyang idisampaikan ioleh ipendukung iteori iini iadalah \| 36 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 imanusia iyang imemiliki itingkat ipendidikan ilebih itinggi, iyang idiukur ijuga idengan ilamanya iwaktu isekolah, iakan imemiliki ipekerjaan idan iupah iyang ilebih ibaik idibandingkan idengan iyang ipendidikannya ilebih irendah. iApabila iupah imencerminkan iproduktivitas, imaka isemakin ibanyak iorang iyang imemiliki ipendidikan itinggi, imaka iakan isemakin itinggi iproduktivitas, isehingga ihasilnya iekonomi inasional iakan ibertumbuh ilebih itinggi i(Nugroho, i2014). iDari iuraian idi iatas ijelaslah ibahwa ijumlah ipenduduk, iserta itenaga ikerja iyang iterampil idan iterdidik isangat iberpengaruh iterhadap ipertumbuhan iekonomi isuatu inegara. iNamun iternyata iberbeda ikenyatannya idengan ibeberapa iprovinsi idi iIndonesia, idimana iterdapat ibeberapa iprovinsi iyang itingkat ipendidikan idan iketerampilannya itergolong irelative irendah ibila idibandingkan idengan iprovinsi ilainnya iyang ada di iIndonesia, inamun ipertumbuhan iekonominya itertinggi idi iIndonesia iseperti ipertumbuhan iekonomi idi iwilayah iMaluku idan iPapua imencapai i9,7 ipersen ihingga iakhir itahun i2018. iAngka iitu iterpaut ijauh idari ipertumbuhan iekonomi iwilayah iJawa iyang iberkisar i5,7 ipersen. iPertumbuhan iekonomi itertinggi ikedua idisumbang idari iSulawesi idi i6,8 ipersen. iDisusul ioleh iJawa, iSumatra, iKalimantan, iserta iBali idan iNusa iTenggara i(“Pertumbuhan iEkonomi iMaluku idan iPapua iDiproyeksi iLampaui iJawa,” in.d.). iDisinilah ihal iyang isangat imenarik iuntuk iditeliti imengapa ipertumbuhan iekonomi idipulau ijawa idalam ibeberapa itahun iterakhir iini iteringgal ioleh iwilayah iMaluku, iPapua idan iSulawesi ipadahal ibila idilihat idari itingkat ipendidikannya iwilayah ipulau iJawa ilebih itinggi idibandingkan idengan iMaluku, iPapua idan iSulawesi. Pertumbuhan iEkonomi Suatu iperekonomian idikatakan imengalami ipertumuhan iekonomi ijika ijumlah iproduksi ibarang idan ijasanya imeningkat. iPertumbuhan iatau ipemerataan? iTerlepas idari imana iyang ilebih ipenting, iyang ipasti ipertumbuhan iekonomi isangat ipenting idan idibutuhkan. iSebab, itanpa ipertumbuhan itidak iakan iterjadi ipeningkatan ikesejahteraan, ikesempatan ikerja, iproduktifitas idan idistribusi ipendapatan. iPertumbuhan iekonomi ijuga ipenting iuntuk imempersiapkan iperekonomian imenjalani itahapan ikemajuan iselanjutnya i(Rahardja i& iManurung, i2014). iPertumbuhan iekonomi ibiasa idiartikan isebagai iupaya imencapai itingkat ipertumbuhan ipendapatan iperkapita i(income iper icapita) iyang \| 37 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 iberkelanjutan iagar inegara idapat imemperbanyak ioutput iyang ilebih icepat idibandingkan ilaju ipertumbuhan ipenduduk i(Todaro i& iSmith, i2006). iberkelanjutan iagar inegara idapat imemperbanyak ioutput iyang ilebih icepat idibandingkan ilaju ipertumbuhan ipenduduk i(Todaro i& iSmith, i2006). ubungan iantara iPembiayaan iPerbankan iSyariah idan iPertumbuhan iEkonomi Menurut iKiryanto iPembangunan iekonomi idi isuatu inegara isangat itergantung ipada iperkembangan idinamis idan ikontribusi inyata idari isektor iperbankan. iKetika isektor iperbankan iterpuruk iperekonomian inasional ijuga iikut iterpuruk. iDemikian ipula isebaliknya, iketika iperekonomian imengalami istagnasi isektor iperbankan ijuga iterkena iimbasnya idimana ifungsi iintermediasi itidak iberjalan inormal i(Baroroh, i2012). Menurut iAli iRama, ibahwa iperkembangan isektor ikeuangan iberpengaruh iterhadap ipertumbuhan iekonomi, ibegitu ipula isebaliknya. iJika isektor ikeuangan imengalami ipertumbuhan iyang ibaik imaka iakan isemakin ibanyak isumber ipembiayaan iyang idapat idialokasikan ike isektor-sektor iekonomi iproduktif i(Rama, i2013). Arah iketerkaitan iantara isektor ikeuangan idengan ipertumbuhan iekonomi idapat idijelaskan idalam idua ihipotesis, iyaitu isupply-leading idan idemand-following, iseperti iyang itelah idikemukakan ioleh iPatrick, iHipotesis isupply-leading imengedepankan iarah ihubungan idari iperkembangan isektor ikeuangan iterhadap ipertumbuhan iekonomi, iyang iberarti ibahwa ipembangunan iinstitusi idan ipasar ikeuangan iakan imeningkatkan ipenawaran ijasa ikeuangan iyang iakan imengarah ipada ipertumbuhan iekonomi iriil i i(Maski, i2010). Pembiayaan iPerbankan iSyariah Perbankan iIslam isebagai ilembaga ikeuangan, ikeberadaannya isangat idiperlukan idalam irangka imenopang ipermodalan ibagi ipara ipelaku iusaha. iDengan iberagam ijenis ipembiayannya, iperbankan iIslam idiharapkan ibisa imenjadi isolusi ibagi imasyarakat iyang imemerlukan ipembiayaan idengan itidak imemakai ikonsep ibunga, iyang idipandang isebagai iriba iyang imesti idijauihi. iMeskipun imenurut iAyub iIndustri iperbankan idan ikeuangan iIslam isekarang iini imenghadapi ibeberapa itantangan. iTantangan iterbesar iadalah imengembangkan iproduk iinvestasi idan ipembiayaan iyang itidak ihanya isesuai idengan iprinsip iekonomi isyariah, itapi ijuga imemenuhi ikebutuhan iperdagangan, ibisnis, idan iindustri iyang iterus iberkembang isecara iberubah, ibaik idalam isektor iswasta imaupun isektor ipublik. iHanya idengan icara iinilah imereka idapat imemenuhi itantangan iuntuk imembangun ikeyakinan idan imeningkatkan iintegritas iinstitusi ifinansial iIslam i(Ayub, i2013). Pembiayaan imerupakan isalah isatu itugas ipokok ibank, iyaitu ipemberian ifasilitas ipenyediaan idana iuntuk imemenuhi ikebutuhan ipihak-pihak iyang imerupakan idefisit iunit. iMenurut isifat ipenggunaanya, ipembiayaan idapat idibagi imenjadi idua ihal iberikut. 1. Pembiayaan iproduktif, iyaitu ipembiayaan iyang iditujukan iuntuk imemenuhi ikebutuhan iproduksi idalam iarti iluas, iyaitu iuntuk ipeningkatan iusaha, ibaik iusaha iproduksi, iperdagangan, imaupun iinvestasi. 2. Pembiayaan ikonsumtif, iyaitu ipembiayaan iyang idigunakan iuntuk imemenuhi ikebutuhan ikonsumsi, iyang iakan ihabis idigunakan iuntuk imemenuhi ikebutuhan. 2. Pembiayaan ikonsumtif, iyaitu ipembiayaan iyang idigunakan iuntuk imemenuhi ikebutuhan ikonsumsi, iyang iakan ihabis idigunakan iuntuk imemenuhi ikebutuhan. Menurut ikeperluannya, ipembiayaan iproduktif idapat idibagi imenjadi idua ihal iberikut. 1. Pembiayaan imodal ikerja, iyaitu ipembiayaan iuntuk imemenuhi ikebutuhan: i(a) ipeningkatan iproduksi, ibaik isecara ikuantitatif, iyaitu ijumlah ihasil iproduksi, imaupun isecara ikualitatif, iyaitu ipeningkatan ikualitas iatau imutu ihasil iproduksi; idan i(b) iuntuk ikeperluan iperdagangan iatau ipeningkatan iutility iof iplace idari isuatu ibarang. \| 38 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 2. Pembiayaan iinvestasi, iyaitu iuntuk imemenuhi ikebutuhan ibarang-barang imodal i(capital igoods) iserta ifasilitas-fasilitas iyang ierat ikaitannya idengan iitu i(Antonio, i2001). Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 ipendidikan idan ipengalaman iyang imenentukan ibagi ipembangunan iekonomi idan ipolitik isuatu inegara. iPembentukan imodal imanusia ikarenanya idikaitkan idengan iinvestasi ipada imanusia idan ipengembangannya isebagai isuatu isumber iyang ikreatif idan iproduktif. iMenurut iSchultz, iada ilima icara ipengembangan isumber idaya imanusia:”(1) ifasilitas idan ilayanan ikesehatan, ipada iumumnya idiartikan imencakup isemua ipengeluaran iyang imempengaruhi iharapan ihidup, ikekuatan idan istamina, itenaga iserta ivitalitas irakyat; i(2) ilatihan ijabatan, itermasuk imagang imodel ilama iyang idiorganisasikan ioleh iperusahaan; i(3) ipendidikan iyang idiorganisasikan isecara iformal ipada itingkat idasar, imenengah idan itinggi; i(4) iprogram istudi ibagi iorang idewasa iyang itidak idiorganisasikan ioleh iperusahaan, itermasuk iprogram iekstension ikhususnya ipada ipertanian; i(5) imigrasi iperorangan idan ikeluarga iuntuk imenyesuaikan idiri idengan ikesempatan ikerja iyang iselalu iberubah i(Jhingan, i2016). ipendidikan idan ipengalaman iyang imenentukan ibagi ipembangunan iekonomi idan ipolitik isuatu inegara. iPembentukan imodal imanusia ikarenanya idikaitkan idengan iinvestasi ipada imanusia idan ipengembangannya isebagai isuatu isumber iyang ikreatif idan iproduktif. iMenurut iSchultz, iada ilima icara ipengembangan isumber idaya imanusia:”(1) ifasilitas idan ilayanan ikesehatan, ipada iumumnya idiartikan imencakup isemua ipengeluaran iyang imempengaruhi iharapan ihidup, ikekuatan idan istamina, itenaga iserta ivitalitas irakyat; i(2) ilatihan ijabatan, itermasuk imagang imodel ilama iyang idiorganisasikan ioleh iperusahaan; i(3) ipendidikan iyang idiorganisasikan isecara iformal ipada itingkat idasar, imenengah idan itinggi; i(4) iprogram istudi ibagi iorang idewasa iyang itidak idiorganisasikan ioleh iperusahaan, itermasuk iprogram iekstension ikhususnya ipada ipertanian; i(5) imigrasi iperorangan idan ikeluarga iuntuk imenyesuaikan idiri idengan ikesempatan ikerja iyang iselalu iberubah i(Jhingan, i2016). Faktor iSumber iDaya iManusia i(SDM). iSama ihalnya idengan iproses ipembangunan, ipertumbuhan iekonomi ijuga idipengaruhi ioleh iSDM. iSumber idaya imanusia imerupakan ifaktor ipenting idalam iproses ipembangunan, icepat ilambatnya iproses ipembangunan itergantung ikepada isejauh imana isumber idaya imanusianya iselaku isubjek ipembangunan imemiliki ikompetensi iyang imemadai iuntuk imelaksanakan iproses ipembangunan idengan imembangun iinfrastruktur idi idaerah-daerah i(Putra, i2018). Tingkat iPendidikan Gagasan iinvestasi ipada imodal imanusia iadalah ibetul-betul ibaru. iDalam iproses ipertumbuhan iekonomi, ilazimnya iorang ilebih imenekankan iarti ipenting iakumulasi imodal ifisik. iSekarang imakin idisadari ibahwa ipertumbuhan ipersediaan imodal inyata isampai ibatas-batas itertentu itergantung ipada ipembentukan imodal imanusia iyaitu i“proses ipeningkatan ipengetahuan, iketerampilan idan ikemampuan iseluruh irakyat isuatu inegara” i(Jhingan, i2016). Menurut iF.H.Harbison ibahwa ipengertian ipembentukan imodal imanusia iadalah i“proses imemperoleh idan imeningkatkan ijumlah iorang iyang imempunyai ikeahlian, \| 39 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Hubungan iAntara iPendidikan idan iPertumbuhan iEkonomi Tenaga ikerja iadalah isalah isatu ifaktor iyang imempengaruhi ipendapatan inasional. iDan iselama iini ikita ihanya imemperhatikan isegi ikuntitasnya isaja, ikita iberanggapan, ibahwa ikalau ijumlah itenaga ikerja imeningkat, imaka ijumlah iproduktifitasnya ijuga imeningkat. iPernyataan itersebut itidak iseluruhnya ibenar, ikarena iwalupun ijumlah itenaga ikerja iitu itidak iberubah, itetapi ibila ikualitas idari itenaga ikerja itersebut ilebih ibaik iatau imeningkat, imaka itingkat iproduksi ijuga iakan imengalami ipeningkatan. iSedangkan ikemampuan iuntuk ibekerja iseseorang idipengaruhi ioleh ikeadaan ikesehatan idan ikecakapannya, iketerampilan idan ikeahliannya. iSelanjutnya itingkat ikesehatan idipengaruhi ioleh ikeadaan igizi idan ilingkungannya, isedangkan ikecakapan, iketerampilan, idan ikeahlian idipengaruhi ioleh itingkat ipendidikan ibaik iformal imaupun iinformal i(Putra, i2018). iBahkan idari ipengalaman inegara-negara imaju i iterbukti ibahwa ifaktor \| 40 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 iyang ipaling iberpengaruh iterhadap ikemajuan iekonomi iadalah ikualitas isumber idaya imanusia i(Rahardja i& iManurung, i2014). i i p Doi : 10.37058/jes.v5i2.1949 iberpengaruh isecara isignifikan iterhadap iPDB iIndonesia ipada itingkat isignifikansi i5 ipersen. iSedangkan ikoefsien iregresi iyang isudah idistandarisasi imenunjukkan iangka i0,962 iyang iberarti isetiap ikenaikan i1 ipoin iAMH iakan imenaikkan iPDB iIndonesia isebesar i0,962 imilyar irupiah i(Nugroho, i2014). iberpengaruh isecara isignifikan iterhadap iPDB iIndonesia ipada itingkat isignifikansi i5 ipersen. iSedangkan ikoefsien iregresi iyang isudah idistandarisasi imenunjukkan iangka i0,962 iyang iberarti isetiap ikenaikan i1 ipoin iAMH iakan imenaikkan iPDB iIndonesia isebesar i0,962 imilyar irupiah i(Nugroho, i2014). 3. Ali iRama. iDengan ijudul ipenelitian iPerbankan iSyariah idan iPertumbuhan iEkonomi iIndonesia. iDimana ipenelitiannya imenguji ihubungan iantara iperbankan isyariah, ipasar imodal, iperdagangan, idan iinflasi ipengaruhnya iterhadap ipertumbuhan iekonomi. iHasil ipenelitiannya imenunjukkan ibahwa iterdapat ihubungan ikesimbangan ijangka ipanjang iantara ibank isyariah, ipasar imodal, iperdagangan idan iinflasi iterhadap ipertumbuhan iekonomi. iBerdasarkan iuji ikausalitas iGranger imenunjukkan ibahwa iterdapat ihubungan idua iarah iantara ipertumbuhan iekonomi idengan ibank isyariah idi iIndonesia i(Rama, i2013). Dari iberbagai iteori idan ihasil ipenelitian idi iatas imaka idapat idibuat imodel ihubungan ivariabel ipenelitian isebagai iberikut: Gambar i1 Paradigma iPenelitian Hipotesis iPenelitian: Gambar i1 Paradigma iPenelitian Paradigma iPenelitian Penelitian iTerdahulu Adapun ipenelitian iterdahulu idiambil idari ijurnal iilmiah, ikarena ijurnal isebagai isumber iyang isangat ipenting idalam imemberikan iinformasi iterkini. iArtikel idi idalam ijurnal iakademik isecara iumum isudah idireview ioleh ipara iahlinya iyang imenandakan iartikel itersebut isudah iditeliti ikembali ioleh iahlinya isebelum iditerima iuntuk idipublikasikan i(Sekaran i& iBougie, i2016). iAdapun isumber ijurnal idalam ipenelitian iini iadalah: i i i 1. Penelitian iyang idilakukan ioleh: iSalahuddin iEl iAyubi, iLukytawati iAnggraeni, iAlmira iDyah iMahiswary. iDengan ijudul, iPengaruh iBank iSyariah iterhadap iPertumbuhan iEkonomi iIndonesia. iAda ihasil iestimasi iVECM iterdapat ihubungan ipada ijangka ipanjang. iSalah isatu iVariabel iyang isignifikan iberpengaruh iterhadap iGDP iadalah itotal ipembiayaan iperbankan isyariah. i iVariabel itotal ipembiayaan iperbankan isyariah ipada ilag ipertama imemiliki ihubungan ipositif iterhadap iGDP iriil idalam ijangka ipanjang, idan isignifikan isecara istatistik ipada itaraf inyata i1 ipersen iyaitu isebesar i3.362112. iArtinya, ipada isaat iterjadi ipeningkatan ipenyaluran ipembiayaan isyariah i1persen ipada ilag ipertama, iakan iterjadi ipeningkatan iterhadap iGDP iriil isebesar i3.362112 ipersen. iAlasannya iadalah ipada isaat iterjadi ipeningkatan itotal ipembiayaan iyang idisalurkan, iakan iberdampak ipada ipeningkatan imodal ipada iusaha-usaha ikemudian ihal itersebut imenyebabkan ipeningkatan ipada iperekonomian isektor iriil. iPeningkatan ipada iperekonomian isektor iriil iberarti iterjadi ipeningkatan iaktivitas iekonomi, iyang imana iakan imeningkatkan ipertumbuhan iekonomi i(El iAyubi, iAnggraeni, i& iMahiswary, i2017). 2. Nugroho iDengan ijudul ipeneletian, iPengaruh iPendidikan iTerhadap iPertumbuhan iEkonomi. iData ipendidikan iyang idi iproxy idengan iAngka iMelek iHuruf i(AMH), idan iertumbuhan iekonomi idiproxy idengan iProduk iDomestik iBruto i(PDB). iHasil ianalisis iuji iasumsi iklasik idan ikoefisien ideterminasi i(R2) idengan imeng-gunakan iprogram iSPSS imenunjukkan ibahwa ikoefisien ideterminasi iadalah i0,926 iyang iberarti i92,6 ipersen inaik iturunnya iPDB iIndonesia ibisa idijelaskan ioleh iAngka iMelek iHuruf. iSedangkan i6,4 ipersen ioleh ifaktor-faktor ilain iseperti ivariabel ilain idan ibahwa iAMH \| 41 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X p Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 Hipotesis iPenelitian: Hipotesis iPenelitian: H1 i i i: Terdapat ipengaruh iyang isignifikan iantara ipembiayaan iperbankan isyariah iterhadap ipertumbuhan iekonomi ipada ienam iprovinsi idi iPulau iJawa. H1 i i i: Terdapat ipengaruh iyang isignifikan iantara ipembiayaan iperbankan isyariah iterhadap ipertumbuhan iekonomi ipada ienam iprovinsi idi iPulau iJawa. H2 i i i: Terdapat ipengaruh iyang isignifikan iantara itingkat ipendidikan iterhadap ipertumbuhan iekonomi ipada ienam iprovinsi idi iPulau iJawa H3 i i i: Secara isimultan terdapat pengaruh antara variableipembiayaanibank syariah dan itingkatipendidikan terhadap ipertumbuhan iekonomi, ipada ienam provinsi idi iPulau iJawa \| 42 \| 42 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 p Doi : 10.37058/jes.v5i2.1949 Uji if idilakukan idalam irangka imenguji ikeberartian ikoefisien iarah iregresi idari ipengaruh ivariable-variabel ibebas i(independen) iterhadap ivariable iterikat i(dependen) isecara isimultan i(bersamaan). iDengan imenentukan itingkat isignifikan i(α) iyaitu isebesar i5% iberdasarkan inilai iprobabilitas. iDengan ikriteria ikeputusan isebagai iberikut: Uji if idilakukan idalam irangka imenguji ikeberartian ikoefisien iarah iregresi idari ipengaruh ivariable-variabel ibebas i(independen) iterhadap ivariable iterikat i(dependen) isecara isimultan i(bersamaan). iDengan imenentukan itingkat isignifikan i(α) iyaitu isebesar i5% iberdasarkan inilai iprobabilitas. iDengan ikriteria ikeputusan isebagai iberikut: a. Jika inilai iprobabilitas i> i0,05 imaka iH1 iditerima, idan iHo iditolak, iyang i artinya ivariabel ibebas i(independent) isecara iserentak iatau ibersama-sama i mempengaruhi ivariabel iterikat i(dependent) isecara isignifikan b. Jika inilai iprobabilitas i< i0,05 imaka iH1 iditolak, idan iHo iditerima, iyang i artinya ivariabel ibebas i(independent) isecara iserentak iatau ibersama-sama i tidak imempengaruhi ivariabel iterikat i(dependent) isecara isignifikan. 1. Pengujian iStatistik Dari ihasil iuji iChow idan iuji iHousman iyang itelah idilakukan iuntuk ipemilihan imodel, imodel iyang iterbaik idalam ipenelitian iini iadalah iRandom iEffect iModel (REM). iMaka imodel iREM iakan idipergunakan idalam iuji-uji iselanjutnya. Tabel i5. Hasil iUji iHausman Test iSummary Chi-Sq. iStatistic Chi-Sq. id.f. Prob. i Cross-section irandom 1.049581 2 0.5917 Cross-section irandom ieffects itest icomparisons: Variable Fixed i i Random i Var(Diff.) i Prob. i X1PMBY 0.167449 0.172374 0.000046 0.4662 X2PDDKN 11153.262690 10803.954359 182785.212403 0.4139 2. iPengujian iUji iAsumsi iKlasik Uji iNormalitas idengan iMetode iRandom iEffect. Tabel i5. Hasil iUji iHausman Test iSummary Chi-Sq. iStatistic Chi-Sq. id.f. Prob. i Cross-section irandom 1.049581 2 0.5917 Cross-section irandom ieffects itest icomparisons: Variable Fixed i i Random i Var(Diff.) i Prob. i X1PMBY 0.167449 0.172374 0.000046 0.4662 X2PDDKN 11153.262690 10803.954359 182785.212403 0.4139 2. iPengujian iUji iAsumsi iKlasik Uji iNormalitas idengan iMetode iRandom iEffect. Tabel i5. Hasil iUji iHausman METODE Penelitian iini imenggunakan idata ipanel. iYaitu, igabungan iantara idata itime iseries i(runtut iwaktu) idan idata icross isection. iDalam imencari ihubungan iketiga ivariabel ipenelitian idigunakan ianalisis iregresi iberganda idengan iterlebih idahulu imelakukan iUji iChow, iUji iHousman idan iUji iLagrange iMultiplier, isehingga imenghasilkan imodel iregresi idata ipanel iyang itepat iuntuk idipergunakan idalam imelakukan iuji-uji iselanjutnya. Uji istatistik iyang idilakukan idalam imencari ihubungan iantara ivaribel iindependen idan idependen iadalah idengan imelakukan iuji isignifikansi. iDimana iuji-t idilakukan idalam irangka imengetahui ihubungan isecara iparsial, idan iuji-f idilakukan iuntuk imengetahui ipengaruh isecara isimultan ikedua ivariabel iindependen iterhadap ivariabel idependen. iDan ilangkah iselanjutnya iadalah iuji iKoefisien ideterminasi i(R2). Uji iR2 iatau iuji ideterminasi idilakukan idalam irangka imenjelaskan iseberapa ibesar ivariasi idari ivariabel iterikat i(dependen) idapat iditerangkan ioleh ivariabel ibebas i(independen). iJika ikoefisien ideterminasi isama idengan inol i i(R2 i= i0) imaka ivariasi idari ivariabel iterikat itidak idapat iditerangkan isama isekali ioleh ivariabel ibebas, idan ijika iR2 i= i1 imaka ivariabel iterikat isecara ikeseluruhan idapat iditerangkan ioleh ivariabel ibebas, idengan ikata ilain ijika iR2 i= i1 imaka isemua ititik ipengamatan iberada itepat ipada igaris iregresi. iSetelah itahapan idi iatas ilangkah iselanjutnya iadalah imelakukan: 1. Uji it iatau ipengujian isecara iparsial Uji it idigunakan iuntuk imenguji ikeberartian ikoefisien iarah iregresi idari ipengaruh ivariable ibebas i(independen) iterhadap ivariable iterikat i(dependen). iDengan imenentukan itingkat isignifikan i(α) iyaitu isebesar i5% iberdasarkan inilai iprobabilitas. iDengan ikriteria ikeputusan isebagai iberikut: a. Jika inilai iprobabilitas i> i0,05 imaka iH1 iditerima, idan iHo iditolak, iyang i artinya ivariabel ibebas i(independent) imempengaruhi ivariabel i terikat i(dependent) isecara isignifikan b. Jika inilai iprobabilitas i< i0,05 imaka iH1 iditolak, idan iHo iditerima, iyang i artinya ivariabel ibebas i(independent) itidak imempengaruhi i variabel i terikat i(dependent) isecara isignifikan. 2. Uji iF iatau ipengujian isecara isimultan \| 43 \| 43 Cross-section irandom ieffects itest icomparisons: Menurut iGujarati, ibahwa iuji iasumsi inormalitas isecara ieksplisit ibiasa idilakukan iuntuk imenganalisis isampel iyang iterbatas iatau isedikit i(Gujarati, i2013).iDibawah iini \| 44 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 idisajikan ihasil iuji inormalitas idari idata ipertumbuhan iekonomi, ikesempatan ikerja, ikesejahteraan imasyarakat iserta ipertumbuhan iperbankan isyariah: idisajikan ihasil iuji inormalitas idari idata ipertumbuhan iekonomi, ikesempatan ikerja, ikesejahteraan imasyarakat iserta ipertumbuhan iperbankan isyariah: Tabel i6. Hasil iUji iNormalitas i Tabel i6. Hasil iUji iNormalitas i Hasil idiatas imenunjukkan ibahwa ivariabel-variabel ipenelitian imenunjukan iberdistribusi inormal. iKarena ip-value iJarque iBera ilebih ibesar idari itaraf isignifikansi i(=5%) imaka idistribusi iresidual ipersamaan iregresi iberdistribusi inormal, iatau ijika inilai iprobabilitas idari iJB i< i0,05 iberarti iresidual imodel itidak iterdistribusi inormal i(Athoillah, i2015). i Uji iMultikolinearitas Menurut iGujarati, ibahwa igejala imultikolinearitas idan isignifikansi idari ivariabel iyang idipergunakan idapat idilihat idari i iR2 inya. iJika iR2 i itinggi itetapi iterdapat ibanyak ivariable iyang itidak isignifikan imaka ididuga iterjadi imultikolinearitas ipada ivariable- variabel itersebut i(Gujarati, i2013). iPada ipenelitian iini, ihasil idari iR2 isebesar i0.967311 iatau i97% idan isemua ivariable isignifikan. iMaka ivariable idalam imodel ipenelitian iini idapat idiambil ikesimpulan iterbebas idari imultikolinearitas. 3. iPengujian iKriteria iStatistik Ringkasan ihasil ipengujian idapat iterlihat ipada itabel idi ibawah: Variabel Koefisien t-statistik Probability Kesimpulan C 90600.81 0.554534 0.5816 Signifikan Pembiayaan iBank iSyariah i 0.167449 3.379253 0.0014 Signifikan Tingkat iPendidikan 11153.26 3.329593 0.0016 Signifikan R-squared F-statistik Prob i(F-statistik) 0.967311 219.8206 0.000000 Sumber: ihasil iolah idata ieviews i7 i Uji iAutokorelasi Untuk imendeteksi iautokorelasi, i(Athoillah i: i2015) idapat idilakukan iuji istatistik imelalui iuji iDurbin-Watson i(DWtest). iDasar ipengambilan ikeputusan iada itidaknya iautokorelasi iadalah isebagai iberikut: \| 45 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X p Doi : 10.37058/jes.v5i2.1949 p Doi : 10.37058/jes.v5i2.1949 1. Bila inilai iDW iterletak idi iantara ibatas iatas iatau iupper ibound i(du) idan i(4-du) imaka ikoefisien iautokorelasi i= i0, iberarti itidak iautokorelasi. 1. Bila inilai iDW iterletak idi iantara ibatas iatas iatau iupper ibound i(du) idan i(4-du) imaka ikoefisien iautokorelasi i= i0, iberarti itidak iautokorelasi. 2. Bila inilai iDW ilebih irendah idari ipada ibatas ibawah iatau ilower ibound i(dl) imaka ikoefisien iautokorelasi i> i0, iberarti iada iautokorelasi ipositif. 2. Bila inilai iDW ilebih irendah idari ipada ibatas ibawah iatau ilower ibound i(dl) imaka ikoefisien iautokorelasi i> i0, iberarti iada iautokorelasi ipositif. 3. Bila iDW ilebih ibesar idari i(4-dl) imaka ikoefisien iautokorelasi i< i0, iberarti iada iautokorelasi inegative. 3. Bila iDW ilebih ibesar idari i(4-dl) imaka ikoefisien iautokorelasi i< i0, iberarti iada iautokorelasi inegative. Tabel i7 Hasil iUji iAutoKorelasi Dari ihasil iuji iautokorelasi idi iatas itidak imenunjukkan iadanya iautokorelasi ikarena id idi iatas iada idiantara idl idan idu ipada itabel iDurbin-Watson. 4. iKoefisien iDeterminasi Berdasarkan ihasil ianalisis iregresi iberganda ipengaruh ipembiayaan ibank isyariah, idan itingkat ipendidikan iterhadap ipertumbuhan iekonomi ipada ienam iprovinsi idi iPulau iJawa idiperoleh inilai ikoefisien ideterminasi iatau iR2 isebesar i0.967311. iNilai iini \| 46 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X imenunjukkan ibahwa ivariable-variabel iindependen itelah imemberikan ikontribusi isebesar i96,73 ipersen idalam imempengaruhi ivariabel iY i(pertumbuhan iekonomi) ipada ienam iprovinsi idi iPulau iJawa. iSedangkan isisanya isebesar i3,27 ipersen idipengaruhi ioleh ivariabel-variabel ilain iyang itidak iditeliti. 6. i iUji iF iatau ipengujian isecara isimultan Dari ihasil ianalisis idata imenunjukkan ibahwa iF-hitung iyang idiperoleh iadalah isebesar i219.8206 idengan iprobabilitas iyang idiperoleh ilebih ikecil idari iα=5% iyang iditetapkan i(0,000<0,05). iDengan idemikian iHo iditolak, imaka ivariabel ipembiayaan ibank isyariah idan itingkat ipendidikan isecara ibersama-sama i(simultan) iberpengaruh isecara isignifikan iterhadap ipertumbuhan iekonomi i i i ipada ienam iprovinsi idi iPulau iJawa. 5. iUji it iatau ipengujian isecara iparsial Dari ihasil ianalisis imenunjukkan ibahwa ipembiayaan ibank isyariah idengan inilai it-hitung isebesar i3.379253 idengan iprobabilitas isebesar i0.0014(0.0014<0,05) imaka iHo iditolak isehingga ipembiayaan ibank isyaraiah iberpengaruh isignifikan iterhadap ipertumbuhan iekonomi. iNilai ikoefisien isebesar i0,168 imenunjukkan ibahwa ivariabel ipembiayaan ibank isyaraiah i imemiliki iarah ihubungan ipositif idan isignifikan iterhadap ipertumbuhan iekonomi. i Variabel itingkat ipendidikan idengan inilai it-hitung isebesar i3.329593 idan iprobabilitas isebesar i0.0016 ijika idibandingkan idengan iα=5% iyang iditetapkan i(0.0016<0,05) imaka iHo iditolak isehingga ivariabel itingkat ipendidikan iberpengaruh isignifikan iterhadap ipertumbuhan iekonomi. iNilai ikoefisien isebesar i11153.26 imenunjukkan ibahwa ivariabel itingkat ipendidikan imemiliki iarah ihubungan ipositif idan isignifikan iterhadap ipertumbuhan iekonomi. I Pengaruh iPertumbuhan iEkonomi, iKesempatan iKerja, idan iKesejahteraan iMasyarakat iTerhadap iPertumbuhan iPerbankan iSyariah i Persamaan imodel iyang ididapat idari iestimasi ipengaruh ipembiayaan ibank isyariah idan itingkat ipendidikan iberpengaruh iterhadap ipertumbuhan iekonomi, iadalah isebagai iberikut: iY=0.167449(X1) i+ i11153.26 i(X2). Persamaan itersebut idapat iinter-prestasikan \| 47 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X ibahwa isetiap ikenaikan i0.167449(X1) idan ikenaikan i11153.26(X2), imaka iY iakan inaik isatu isatuan. iAtau idengan ikata ilain ijika ipembiayaan ibank isyariah inaik isebesar i10 ipersen imaka iakan imenyebabkan ipertumbuhan iekonomi inaik isebesar i1,67 isatuan. iDan ijika itingkat ipendidikan inaik isebesar i10 ipersen imaka ipertumbuhan iekonomi iakan inaik isebesar i111.530 isatuan. Dengan demikianihasil penelitian ini mendukung hasilipenelitianiyangidilakukanioleh Salahuddin El Ayubi, Lukytawati Anggraeni, Almira Dyah Mahiswary, yang berjudul, Pengaruh Bank Syariah terhadap Pertumbuhan Ekonomi Indonesia. Dimana hasil penelitiannya bahwa pembiayaan yang dilakukan perbankan syariah dapat mempengaruhi pertumbuhan ekonomi,ijugaisejalanidenganipenelitianiyang dilakukanioleh AliiRama, dengan judul, Perbankan Syariah dan Pertumbuhan Ekonomi Indonesia dimana sector keuangan dapat mempengaruhi pertumbuhan ekonomi melalui kemampuannya memberikan pinjaman sehingg isekto iriil dapat tumbuh. Sementara hasil penelitian pengaruh variable tingkat pendidikan berpengaruh terhadap pertumbuhan ekonomi mendukung hasil penelitian yang dilakukan oleh Nugroho SBM. Dengan judul penelitian, Pengaruh Pendidikan Terhadap Pertumbuhan Ekonomi. Data ipendidikan dengan Angka Melek Huruf (AMH) sebagai ukurannya, dan pertumbuhan ekonomi dengan Produk Domestik Bruto (PDB) sebagai ukurannya. Menunjukkan hasil analisis uji asumsi klasik bahwa AMH berpengaruh secaraisignifikan terhadapiPDB Indonesia.iHasil ujiikoefisien determinasii(R2)imenunjukkan bahwa naik turunnya PDBiIndonesiaibisaidijelaskaniolehiAngkaiMelekiHuruf, danidari setiap kenaikani1 poin AMH akanimenaikkaniPDB Indonesia sebesari0,962imilyarirupiah. Saran 1. Untuk imeningkatkan ipertumbuhan iekonomi ipemerintah ihendaknya imemberikan iruang igerak idan iperhatian iyang ilebih ibesar ilagi ikepada iperbankan isyariah, isehingga iperbankan isyariah ilebih icepat idalam ipertumbuhannya iyang ipada iakhirnya iperbankan isyariah imampu imelayani ikeperluan imasyarakat idengan imemberikan ipembiayaan iatau ipermodalan ikepada imasyarakat isehingga isektor iriil imasyarakat ilebih itumbuh idan iberkembang. i 1. Untuk imeningkatkan ipertumbuhan iekonomi ipemerintah ihendaknya imemberikan iruang igerak idan iperhatian iyang ilebih ibesar ilagi ikepada iperbankan isyariah, isehingga iperbankan isyariah ilebih icepat idalam ipertumbuhannya iyang ipada iakhirnya iperbankan isyariah imampu imelayani ikeperluan imasyarakat idengan imemberikan ipembiayaan iatau ipermodalan ikepada imasyarakat isehingga isektor iriil imasyarakat ilebih itumbuh idan iberkembang. i 2. Pemerintah ihendaknya ilebih ibanyak imengalokasikan ikeuangannya ipada isektor ipendidikan isehingga itingkat ipendidikan imasyarakat imeningkat. iKarena iterbukti itingkat ipendidikan imasyarakat idapat imempengaruhi ikualitas isumber idaya imanusia i(SDM), iyang iakhirnya idapat imempengaruhi ipertumbuhan iekonomi. I 3. Pembiayaan iperbankan isyariah idan itingkat ipendidikan imasyarakat iterbukti iberpengaruh isecara isignifikan iterhadap ipertumbuhan iekonomi. iNamun idalam ipenelitian iini itidak idirinci imengenai ipembiayaan iyang idisalurkan iapakah ipembiayaan iuntuk ikonsumtif iataukah iuntuk ikredit iusaha ikecil idan imenengah i(UKM) iyang idapat imeningkatkan ipertumbuhan iekonomi. iUntuk iitu isebaiknya iada ipenelitian ilanjutan iterkait ipengaruh ipembiayaan iperbankan isyariah i iterhadap ipertumbuhan iUKM idan idampaknya iterhadap ipertumbuhan iekonomi. I KESIMPULAN 1. Pembiayaan iperbankan isyariah iberpengaruh isignifikan iterhadap ipertumbuhan iekonomi ipada ienam iprovinsi idi iPulau iJawa, iselama ikurun iwaktu ipenelitian itahun i2009 isampai idengan itahun i2018 2. Tingkat ipendidikan iberpengaruh isignifikan iterhadap ipertumbuhan iekonomi ipada ienam iprovinsi idi iPulau iJawa, iselama ikurun iwaktu ipenelitian itahun i2009 isampai idengan itahun i2018 3. Pembiayaan iperbankan isyariah idan itingkat ipendidikan imasyarakat ipada ienam iprovinsi idi ipulau iJawa isecara iserentak i(simultan) iberpengaruh isignifikan iterhadap 3. Pembiayaan iperbankan isyariah idan itingkat ipendidikan imasyarakat ipada ienam iprovinsi idi ipulau iJawa isecara iserentak i(simultan) iberpengaruh isignifikan iterhadap \| 48 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 ipertumbuhan iekonomi iselama ikurun iwaktu ipenelitian itahun i2009 isampai idengan itahun i2018 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Arsyad, iL. i(2016). iEkonomi ipembangunan. iYogyakarta: iBag. iPenerb. iSekolah iTinggi iIlmu iEkonomi-YKPN. to o, . S. ( 00 ). ank sya iah: a i teo i ke p aktik. Ja a ta: Ge a sa . APK-APM. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttp://apkapm.data.kemdikbud.go.id/ Agustus, ipertumbuhan ikredit iperbankan imelambat. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttps://keuangan.kontan.co.id/news/agustus-pertumbuhan-kredit-perbankan- melambat-1 nio, iM. iS. i(2001). iBank isyariah: iDari iteori ike ipraktik. iJakarta: iGema iInsani. Ayub, iM. i(2013). iUnderstanding iIslamic iFinance. iJakarta: iGramedia iPustaka iUt Baroroh, iU. i(2012). iAnalisis iSektor iKeuangan iTerhadap iPertumbuhan iEkonomi iRegional idi iWila- iyah iJawa, iPendekatan iModel iLevine. iJurnal iEtikomi, i11(2). El iAyubi, iS., iAnggraeni, iLukytawati., i& iMahiswary, iA. iD. i(2017). iPengaruh i Bank iSyariah iterhadap iPertumbuhan iEkonomi iIndonesia. iJurnal iAl- Muzara’ah, i5(2), i88–106. REFERENSI Agustus, ipertumbuhan ikredit iperbankan imelambat. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttps://keuangan.kontan.co.id/news/agustus-pertumbuhan-kredit-perbankan- melambat-1 Antonio, iM. iS. i(2001). iBank isyariah: iDari iteori ike ipraktik. iJakarta: iGema iInsani. APK-APM. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttp://apkapm.data.kemdikbud.go.id/ \| 49 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Ayub, iM. i(2013). iUnderstanding iIslamic iFinance. iJakarta: iGramedia iPustaka iUtama. Gujarati, iD. iN. i(2013). iDasar-Dasar iEkonometrika. iErlangga. Gujarati, iD. iN. i(2013). iDasar-Dasar iEkonometrika. iErlangga. Hastuti, iR. iK. i(n.d.). iKetua iLPS: iKondisi iPerbankan iSyariah iCukup iBerat. iRetrieved iOctober i30, i2019, ifrom iSyariah iwebsite: ihttps://www.cnbcindonesia.com/syariah/20190606191828-29-76983/ketua-lps- kondisi-perbankan-syariah-cukup-berat Inggrid. i(2006). iSektor iKeuangan idan iPertumbuhan iEkonomi idi iIndonesia: iPendekatan iKausalitas idalam iMultivariate iVector iError iCorrection iModel i(VECM). i50 JURNAL iMANAJEMEN iDAN iKEWIRAUSAHAAN, i8(1). iRetrieved ifrom ihttp://www.petra.ac.id/~puslit/journals/dir.php?DepartmentID=MAN Jaelani. i(2015). iStudi iEfisiensi iBank iUmum idi iIndonesia iTahun i2002-2013, i(Komparasi iFaktor-faktor iPenjelas iEfisiensi iAntara iBank iUmum iKonvensional idengan iBank iUmum iSyariah). iProgram iStudi iIlmu iManajemen, iSekolah iPasca iSarjana iUniversitas iPendidikan iIndonesia, iBandung. Jhingan, iM. i(2016). iEkonomi iPembangunan idan iPerencanaan. iJakarta: iRajaGrafindo iPersada. Latumaerissa, iJ. iR. i(2015). iPerekonomian iIndonesia idan idinamika iekonomi iglobal. iJakarta. Maski, iG. i(2010). iAnalisis iKausalitas iantara iSektor iKeuangan idan iPertumbuhan iEkonomi idi iIndonesia. iJurnal iEkonomi iDan iPembangunan iIndonesia, i10(2), i143–158. ihttps://doi.org/10.21002/jepi.v10i2.117 Nugroho. i(2014). iPengaruh iPendidikan iTerhadap iPertumbuhan iEkonomi. iMedia iEkonomi idan iManajemen, i29(2). Olufemi iAdeyeye, iP. i(2015). iDoes iSupply- Leading iHypothesis ihold iin ia iDeveloping iEconomy? iA iNigerian iFocus. iProcedia iEconomicsang iFinance. iRetrieved ifrom ihttps://core.ac.uk/download/pdf/82469979.pdf \| 50 Jurnal Ekonomi Syariah Vol. 5. No. 2. November 2020 p-ISSN 2548-5032 e-ISSN 2714-769X Doi : 10.37058/jes.v5i2.1949 Todaro, iM. iP., i& iSmith, iS. iC. i(2006). iEconomic iDevelopment. iJakarta: iErlangga. Doi : 10.37058/jes.v5i2.1949 Doi : 10.37058/jes.v5i2.1949 Pengangguran i& iTenaga iKerja idi iIndonesia i\| iIndonesia iInvestments. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttps://www.indonesia- investments.com/id/keuangan/angka-ekonomi-makro/pengangguran/item255? Pertumbuhan iEkonomi iMaluku idan iPapua iDiproyeksi iLampaui iJawa. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttps://www.cnnindonesia.com/ekonomi/20181127161912-532- 349628/pertumbuhan-ekonomi-maluku-dan-papua-diproyeksi-lampaui-jawa Putra, iWindhu. i(2018). iPerekonomian iIndonesia, iPenerapan iBeberapa iTeori iEkonomi i Pembangunan idi iIndonesia. iDepok: iRajaGrafindo iPersada. Rahardja, iP., i& iManurung, iM. i(2014). iTeori iEkonomi iMakro. iJakarta: iFakultas i Ekonomi iUniversitas iIndonesia. Rama, iA. i(2013). iPerbankan iSyariah idan iPertumbuhan iEkonomi iIndonesia. iJurnal iSignifikan, i2(1). Sekaran, iU., i& iBougie, iR. i(2016). iResearch iMethods iFor iBusiness: iA iSkill iBuilding iApproach. iJohn iWiley i& iSons. Sukirno, iS. i(2006). iEkonomi ipembangunan: iProses, imasalah, idan idasar ikebijakan. iKencana i(Prenada iMedia). Tenaga iKerja iAsing iMasuk iRI iMelesat i38%, iTerbanyak iAsal iChina. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttps://finance.detik.com/berita-ekonomi-bisnis/d- 4699585/tenaga-kerja-asing-masuk-ri-melesat-38-terbanyak-asal-china Tertinggi iSejak i2014, iBPS: iEkonomi iIndonesia i2018 iTumbuh i5,17 iPersen. i(n.d.). iRetrieved iOctober i30, i2019, ifrom ihttps://www.menpan.go.id/site/berita- terkini/berita-daerah/tertinggi-sejak-2014-bps-ekonomi-indonesia-2018-tumbuh-5-17- persen odaro, iM. iP., i& iSmith, iS. iC. i(2006). iEconomic iDevelopment. iJakarta: iErlangga. Todaro, iM. iP., i& iSmith, iS. iC. i(2006). iEconomic iDevelopment. iJakarta: iErlangga. \| 51
https://openalex.org/W4392837910	https://ejournal.unp.ac.id/index.php/sendratasik/article/download/121562/108865	Indonesian	null	Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah Kota Padang: Tinjauan Koreografi	Jurnal Sendratasik (edisi elektronik)	2,024	cc-by-sa	4,362	Kata Kunci: Tari Piriang Lenggang Manatiang; Sanggar Anjuang Siriah; Proses Koreografi; Aspek Bentuk Leri Yano Wardana1; Nerosti 2; Leri Yano Wardana1; Nerosti 2; 1, 2 Prodi Pendidikan Tari Universitas Negeri Padang, Indonesia. (*) (e-mail) leriyanowardana23@gmail.com1, nerosti@fbs.unp.ac.id2, Vol. 12 Nomor 4 Th. 2023, Hal 544-554 ISSN: Online 2302-3201 (online) http://ejournal.unp.ac.id/index.php/sendratasik/user Diterima 23 01, 2023; Revisi 10 01, 2024; Terbit Online 10 01, 20 Vol. 12 Nomor 4 Th. 2023, Hal 544-554 ISSN: Online 2302-3201 (online) http://ejournal.unp.ac.id/index.php/sendratasik/user Diterima 23 01, 2023; Revisi 10 01, 2024; Terbit Online 10 01, 20 Vol. 12 Nomor 4 Th. 2023, Hal 544-554 ISSN: Online 2302-3201 (online) http://ejournal.unp.ac.id/index.php/sendratasik/user Diterima 23 01, 2023; Revisi 10 01, 2024; Terbit Online 10 01, 202 Vol. 12 Nomor 4 Th. 2023, Hal 544-554 ISSN: Online 2302-3201 (online) http://ejournal.unp.ac.id/index.php/sendratasik/user Diterima 23 01, 2023; Revisi 10 01, 2024; Terbit Online 10 01, 2024 Abstract This research aims to describe and analyze the choreography of the Piriang Lenggang Manatiang Dance at the Anjuang Siriah Studio, Padang City. This type of research is qualitative research using descriptive analysis methods. The instrument for this research is the researcher himself and is assisted by supporting instruments in the form of writing instruments, books, recording equipment and audio-visual equipment. Data collection techniques were carried out using literature study, observation, interviews and documentation. The steps for analyzing data are data collection, data reduction, data presentation and conclusion. The results of the research show that the choreography process for the Piriang Lenggang Manatiang Dance starts from determining the idea/theme, then exploration, improvisation and composition. The choreography process produces the following form aspects: movement, space design, time design, energy design, top design, dramatic design, floor design, group composition, equipment and music. The movements of the Piriang Lenggang Manatiang Dance are the saayun steps, buai ayun movements, lenggang mantiang movements, manyabiak movements, rampak movements, saayun saangkat movements, Baginjai movements, ayun ma ayun movements, lenggang sway movements. The spatial aspect in the Piriang Lenggang Manatiang Dance uses straight, oblique and curved lines. The group compositions performed by dancers include unison (union), alternating, broken, alternating (canon), and balanced (balance). The Piriang Lenggang Manatiang Dance Costume is a garment that was developed and created by the Baju Kurong. The musical instruments used are Gandang, Bansi, Sarunai, Canang and Talempong Pacik. Meanwhile, the property used is piriang tujuah. Keywords: Piriang Lenggang Manatiang Dance; Anjuang Siriah Studio; Choreography Process; Aspects of Form Keywords: Piriang Lenggang Manatiang Dance; Anjuang Siriah Studio; Choreography Process; Aspects of Form Keywords: Piriang Lenggang Manatiang Dance; Anjuang Siriah Studio; Choreography Process; Aspects of Form Abstrak Penelitian ini bertujuan untuk mendeskripsikan dan menganalisis koreografi Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah Kota Padang. Jenis penelitian ini adalah penelitian kualitatif dengan menggunakan metode deskriptif analisis. Instrumen penelitian ini adalah peneliti sendiri dan dibantu dengan instrumen pendukung berupa alat tulis, buku, alat perekam, dan alat audio visual. Teknik pengumpulan data dilakukan dengan cara studi pustaka, observasi, wawancara dan dokumentasi. Langkah-langkah menganalisis data adalah pengumpulan data, reduksi data, penyajian data dan penarikan kesimpulan. Hasil penelitian menunjukkan bahwa proses koreografi Tari Piriang Lenggang Manatiang berawal dari penentuan ide/tema, kemudian eksplorasi, improvisasi, dan komposisi. Dari proses koreografi yang dilakukan menghasilkan aspek bentuk sebagai berikut: gerak, desain ruang, desain waktu, desain tenaga, desain atas, desain dramatik, desain lantai, komposisi kelompok, perlengkapan, dan musik. Gerak Tari Piriang Lenggang Manatiang adalah gerak langkah saayun, gerak buai ayun, gerak lenggang mantiang, gerak manyabiak, gerak rampak, gerak saayun salangkah, gerak baginjai, gerak ayun ma ayun, gerak lenggang lenggok. Aspek ruang dalam Tari Piriang Lenggang Manatiang memakai garis lurus, serong dan melengkung. Komposisi kelompok yang dilakukan penari diantaranya serempak (union), selang-seling (alternate), terpecah (broken), bergantian (canon), berimbang (balence). Kostum Tari Piriang Lenggang Manatiang merupakan baju yang dikembangkan dan dikresikan dari baju kurung. Alat musik yang digunakan adalah Gandang, Bansi, Sarunai, Canang dan Talempong Pacik. Sedangkan properti yang digunakan adalah piriang tujuah. Jurnal Sendratasik licensed under Creative Commons Attribution-ShareAlike 4.0 International License. Pendahuluan Seni tari adalah ungkapan perasaan atau ekspresi seorang manusia yang dituangkan dalam bentuk gerak yang berirama sesuai dengan iringan musik. Seni tari memiliki fungsi yang hampir sama dengan kesenian lainnya yaitu sebagai media untuk menyalurkan ekspresi dan sarana komunikasi pada penikmat seni melalui gerakan. Seni Tari juga merupakan warisan budaya harus dilestarikan dan dikembangkan agar tidak padam, sehingga generasi selanjutnya masih bisa mengenal beragam tarian yang ada di Indonesia. Tari sebagai bagian dari budaya merupakan kajian literal yang selalu dieksplorasi manusia untuk menemukan bentuk yang terbaik. Hal ini memungkinkan pengungkapan adab yang teratur dan bernilai tinggi yang menjadi tolok ukur budaya leluhur kelangsungan hidup manusia (Nerosti, 2020) Suryadiningrat dalam (Mulyani, 2016) mengungkapkan bahwa seni tari adalah gerak tubuh manusia yang disusun sedemikian rupa selaras dengan irama musik dan mempunyai tujuan tertentu. Tari adalah bagian dari seni, bagian dari budaya, dan merupakan “suatu kegiatan manusia yang diekspresikan melalui gerak dan ekspresi yang direncanakan, terstruktur dan terpola secara jelas, cerita, tetapi mereka memiliki nilai, estetika, logis dan etis (Astuti, 2016; Indrayuda, 2012).. Seni adalah cerminan dari kepercayaan atau pendapat manusia yang menciptakannya, termasuk alasan dibalik penciptaan karya seni dan makna keindahan yang terkandung 545 DOI: 10.24036/js.v12i4.121562 ISSN: 2302-3201 ari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah – Leri Yano Wardana & Nerosti dalam karya seni (Sugiartha, 2018; Nerosti, 2020). Setiap kesenian tradisional dalam suatu masyarakat memiliki beberapa arti dan fungsi dalam masyarakat tersebut (Anggraini, 2020). Tari tradisional dibagi menjadi dua jenis menurut nilai seninya, yaitu tari rakyat dan tari klasik. Tarian rakyat berkembang di kalangan masyarakat dengan gerakan yang sederhana, iringan musik yang sederhana, kostum dan dandanan yang sangat sederhana. Tari klasik berkembang di kalngan masyarakat bangsawan. Tarian ini sekarang terdapat di lingkungan Keraton yang ada di Jawa yaitu Yogyakarta dan Surakarta, dengan gerak yang mempunyai artistik yang tinggi dengan pakem tertentu dan geraknya terstandar (Nerosti, 2019). Selain tari tradisional terdapat tari kreasi yang sekarang lebih berkembang di tengah masyarakat. Tari kreasi merupakan tari yang diciptakan berdasarkan dasar-dasar tari yang sudah ada sehingga menjadi karya tari baru. Tari kreasi banyak diciptakan oleh sanggar- sanggar kesenian yang ada di Minangkabau terutama di Kota Padang. Tari kreasi merupakan tarian ada dasarnya memiliki pola tradisi yang aturannya sudah baku dengan mengkreasikan tari tradisional kedalam bentuk baru (Rahmida, 2008). Setiap daerah di Indonesia memiliki tarian yang khas salah satunya adalah Sumatera Barat. Pendahuluan Sumatera Barat terkenal melahirkan banyak koreografer hebat dan ternama yang menciptakan bermacam-macam jenis tarian salah satunya tari kreasi. Di Kota Padang memiliki beberapa sanggar yang mengembangkan seni pertunjukan, salah satunya adalah Sanggar Anjuang Siriah. Sanggar ini berdiri pada tahun 2012 yang dipimpin oleh Rosnelly Ramli alumni dari SMKN 7 Padang Jurusan Teater tahun 1992. Sanggar Anjuang Siriah beralamatkan di Jalan Lumba-lumba No. 28 Pasar Pagi Rimbau Kaluang Kecamatan Padang Barat, Sumatera Barat. Sanggar Anjuang Siriah aktif sebagai wadah untuk membina generasi muda dalam melestarikan kesenian-kesenian tradisional seperti: seni musik dan seni tari. Di sanggar ini terdapat beberapa tarian yaitu Tari Galombang, Tari Indang Badarak, Tari Kipas Galatiak Rang Mudo. Salah satunya adalah Tari Piriang Lenggang Manatiang. Tarian-tarian ini diciptakan pada tahun 2013 pada bulan yang berbeda dan penata musik dalam sanggar ini yaitu Delri Endri, S.Pd (Wawancara, Rosnelly Ramli 21 Juli 2022) selaku pendiri sanggar. Peneliti tertarik untuk meneliti tarian tersebut dari segi koreografinya, baik dari segi penataan gerak, pengolahan kostum, musik, dan pola lantai. Selain tarian ini termasuk ke dalam garapan tari entertainmen yaitu tari yang dipertunjukkan untuk satu event, baik pada pesta perkawinan maupun pada acara resmi lainnya. Alasan lain dari pemilihan tari ini adalah peneliti pernah terlibat langsung dan aktif menarikan Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah. Dari kegiatan yang peneliti tekuni tersebut, peneliti memiliki pengalaman yang sangat bermanfaat dilingkungan sanggar. Belajar bagaimana berproses dengan sesama penari dan sesosialisasi dengan orang baru, dan peneliti dapat belajar tari- tarian yang ada di Sanggar Anjuang Siriah. Tari Piriang Lenggang Manatiang menceritakan suatu rapat nagari niniak mamak dengan masyarakat setempat, dan keramah-tamahan muda-mudi di Ranah Minang yang bersemangat dalam menjamu tamunya di suatu perhelatan dalam nagari. Tarian ini juga mengambarkan tentang ketegasan dan kewibawaan niniak-mamak dalam menyelesaikan persoalan dalam nagari, dengan prinsip yang salah dikatakan salah, yang benar dikatakan benar. Ekspresi dan kemampuan pertahanan diri juga terggambar dengan menari di atas pecahan kaca. Penari dengan lincahnya meloncat-loncat di atas pecahan kaca sesuai tempo 546 Jurnal Sendratasik Vol. 12 No. 4. Pendahuluan Thn.2023 musik yang dinamis ditambah dengan irama dendang yang dinyanyikan dengan riang oleh pemusik, dengan isi pantun sebagai berikut: musik yang dinamis ditambah dengan irama dendang yang dinyanyikan dengan riang oleh pemusik, dengan isi pantun sebagai berikut: Banyak lah bana piriang ditatiang Langkok jo samba nan di dalamnyo Kami bamain lai Tari Piriang Jo anjuang siriah namo sanggarnyo Santianglah bana kanduang manatiang Suok jo kida lai capek juo Elok-eloklah manari piriang Mamijak kaco jan sampai luko Tarian ini menggunakan properti Piring, tanpa menggunakan alat bantu untuk mengikat piring di jari. Jumlah penari dalam Tari Piriang Lenggang Manatiang yaitu ada 5 orang di antaranya 3 penari perempuan dan 2 penari laki-laki sesuai kebutuhan dan besarnya area tempat menari. Dalam tarian ini terdapat 10 motif gerak dengan formasi pola lantai yang berbeda di setiap geraknya. Gerakan dalam tarian ini dilakukan secara bervariasi dengan arah hadap yang berbeda pada setiap motif gerak. Level, volume, dan intensitas gerak juga bervariasi, ditarikan secara berkualitas. Kostum yang digunakan pun tidak membuat penari terganggu saat menarikan gerak yang cukup unik dalam menggerakkan anggota tubuh dengan tetap memegang piring di kedua tangan penari. Musik dalam Tari Piriang Lenggang Manatiang yang ditata oleh penata musiknya, selain bertujuan mempertegas gerak dengan tempo musik juga memperkuat suasana tari, sehingga penampilan tari lebih meriah. Tarian ini ditampilkan pada acara pesta perkawinan atau pada acara penyambutan tamu terhormat dan sudah sering mengisi acara pada beberapa event baik acara pertunjukan daerah dan Nasional seperti acara Festival Siti Nurbaya pada tahun 2017 dan acara Apeksi Tanjuang Pinang pada tahun 2019, sejak itu hingga sekarang tari ini selalu ditampilkan pada pesta perkawinan di Kota Padang. a. Penemuan Ide Tari Ide tari merupakan pokok pikiran atau pokok permasalahan yang disampaikan melalui bahasa gerak. Judul sebagai identitas karya tari yang dihasilkan yang ditetapkan atas pertimbangan tema tarinya dan konsep bentuk tari yang ditawarkan. Menurut Nerosti (2021:23) proses berkarya merupakan refleksi keindahan, yang mucul dari pengendapan dan interpretasi seseorang terhadap apa yang dilihat dan diserap dari alam, sehingga dituangkannya ke dalam karya seni. Dengan menggunakan intelektual dan insting kesenimannya, seorang koreografer mesti mampu menangkap getar keindahan alam di sekitarnya, dan selanjutnya menjadikan sebuah pengalaman pribadi tentang keindahan yang dituangkan ke dalam karya tari. Menurut Rosnelly (wawancara, Oktober 2022), Tari Piriang Lenggang Manatiang pertama kali ditarikan pada tahun 2013 pada pesta perkawinan warga masyarakat di Kota Padang. Dari situ koreografer kemudian mengimplementasikan bentuk dan konsep Anjuang di Siria Studio. Saat ini Sanggar Anjuang Siriah tumbuh dan berkembang serta mengakar kuat di hati masyarakat Kota Padang. Tari Piriang Lenggang Manatiang merupakan salah satu tari kreasi yang bersumber dari gerak tokoh adat, dan termasuk dalam tari hiburan atau tari yang dipentaskan hanya sebagai tari hiburan. Tari Piriang Lenggang Manatiang karya Sanggar Anjuang Siriah merupakan salah satu tarian yang diciptakan oleh koreografer. Tarian ini merupakan bentuk tarian yang berkembang yang tidak lepas dari gerak dasar tari Minangkabau. Pada awalnya, tarian ini hanyalah semacam keceriaan bagi koreografer untuk menciptakan sebuah tarian dalam bentuk gerakan yang indah dan ritmis. Rosnelly kemudian mencoba melakukan gerak tari piriang yang berbeda dengan gerak dasar minangkabau. Selain itu, koreografer menciptakan tarian tersebut hanya sebagai tarian rekreasi yang tidak menyimpang dari aturan dan norma yang ada, karena penari wanita tidak menari seperti gerakan pria. Ada etika dalam bergerak dan ada etika dalam berpakaian. Koreografer juga bertujuan untuk melestarikan ciri khas tari Minangkabau yaitu tari piriang agar tari piriang tidak hilang begitu saja (wawancara dengan Rosnelly Ramli, Oktober 2022). Sesuai dengan sinopsis Tari Piriang Lenggang Manatiang, yaitu: “Tarian ini menceritakan keramah tamahan muda mudi di ranah Minang dalam menjamu tamunya dalam suatu perhelatan dalam nagari dan juga menceritakan ketegasan dan kewibawaan niniak mamak dalam menyelesaikan permasalahan dalam nagari, yang salah dikatakan salah dan yang benar dikatakan salah. Kemudian menari di atas pecahan kaca melambangkan musuh pantang dicari dan bertemu pantang dielakan”. Sehingga koreografer tari memutuskan ide untuk membuat sebuah tarian yang menggambarkan tokoh Niniak Mamak di Minangkabau. Metode Jenis penelitian yang dipakai adalah penelitian kualitatif dengan memakai metode analisis deskriptif. Subyek yang dipelajari adalah Tari Piriang Lenggang Manatiang Sanggar Anjuang Siriah Kota Padang. Alat dalam penelitian ini adalah peneliti sendiri, dilengkapi dengan alat tulis, buku, alat perekam, alat audio visual dan alat bantu lainnya. Teknik pengumpulan data dilakukan melalui; 1)studi pustaka terkait objek dengan mencari buku serta penelitian relevan; 2)observasi, dilakukan dengan melakukan secara lansung terhadap Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah; 3)wawancara dilakukan kepada Delri Endri dan Rosnelly Ramli selaku pendiri sanggar; 4)serta dokumentasi. Teknik pengumpulan data adalah metode atau strategi untuk memperoleh data yang diperlukan untuk menjawab pertanyaan penelitian. Setelan pengumpulan data kemudian dilakukan analisis data dengan teknik; reduksi data, penyajian data serta menarik kesimpulan. (Moleong, 2012) 547 ISSN: 2302-3201 DOI: 10.24036/js.v12i4.121562 Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah – Leri Yano Wardana & Nerosti Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah – Leri Yano Wardana & Nerosti a. Penemuan Ide Tari Kemudian gerakan tidak selalu lahir hanya dari pemikiran Rosnelly Ramli, tetapi ide-ide tersebut lahir dari para penarinya dan juga dari musiknya, yang memungkinkan gerakan-gerakan eksploratif muncul dari ketiadaan (wawancara, Rosnelly Ramli, Oktober 2022). 548 Jurnal Sendratasik Vol. 12 No. 4. Thn.2023 d. Komposisi Komposisi tari, juga dikenal sebagai koreografi, adalah bagian penting dari tarian, terutama tarian kelompok, dan melibatkan seni membuat atau merancang struktur dan alur sehingga menjadi pola gerak. Koreografer tari mulai menggabungkan semua gerakan yang berbeda yang digunakan sebagai hasil dari proses penelitian, improvisasi dan evaluasi. Ragam geraknya tentu disesuaikan dengan tema tariannya, yakni keramahan para pemuda Ranah Minang yang gemar menghibur para tamu pada upacara desa (Wawancara dengan Rosnelly Ramli, 2 Oktober 2022). Selain itu, lintasan diarahkan pada formula sesuai tema, bentuk, struktur, ritme dan tempo karya, disesuaikan dengan tema dan mempertimbangkan transisi dari satu genre ke genre lainnya. c. Improvisasi Manatiang dilakukan secara tiba-tiba oleh penari karena usaha kreatif mereka untuk merespon situasi adegan atau suasana panggung. Improvisasi dilakukan oleh manajemen tari yaitu secara spontan mencoba atau mencari bentuk gerak yang dapat digunakan. Selain itu, improvisasi juga merupakan tahapan dimana gerak tari dipilih dan dikembangkan sesuai dengan tema yang ditetapkan pada karya tari tersebut. Improvisasi guru tari dalam Tari Piriang Lenggang Manatiang ini merupakan eksperimen dengan bentuk-bentuk gerak yang berasal dari gerak-gerak baru tetapi dikembangkan sesuai kreativitas penata. Improvisasi Tari Piriang Lenggang Manatiang bukan hanya sekedar improvisasi bebas, melainkan improvisasi yang diilhami oleh gerak- gerak tari Piriang. (wawancara Rosnelly Ramli, 2 Oktober 2022). b. Eksplorasi Eksplorasi adalah proses penjelajahan dan pencarian motif-motif gerak melalui berbagai cara yang dilakukan pada saat melakukan proses garap gerak tari piriang langgang manatiang. Rosnelly Ramli melakukan wawancara dengan petua-petua adat di suatu nagari upaya untuk mendapatkan berbagai informasi tentang bagaimana berjalannya suatu perhelatan yang dilaksanakan. Bisa dilihat dari keramah tamahan muda mudi yang bersemangat dalam menyambut tamu yang akan datang, kemudian warga sekitar, niniak mamak, dan muda-mudi menghadiri pesta panen padi dan sekaligus mengadakan rapat dalam nagari untuk membicarakan tentang kemajuan kampung dan membahas persoalan yang terjadi pada saat pesta panen padi. Tari Piriang Lenggang Manatiang memiliki arti, “Mengangkat piriang atau menghidangkan makanan”. Kegiatan yang dilakukan pada acara adat waktu menjamu makan, pekerjaan ini dilakukan oleh kaum perempuan bila tamu yang datang adalah perempuan maka yang akan menghidangkan makanan adalah kaum perempuan begitupun dengan sebaliknya. c. Improvisasi 2. Bentuk Koreografi Tari Piriang Lenggang Manatiang Koreografi Tari Piriang Lenggang Manatiang terdiri dari aspek bentuk, antara lain gerak, desain ruang, desain waktu, desain tenaga, komposisi kelompok, penari, kostum, musik, dan properti. Soedarsono (1977) dalam (Nerosti, 2021) yang berisikan tentang koreografi berasal dari bahasa Yunani yaitu chorea dan graphia, artinya ‘tari bersama’ dan koor/graphia artinya ‘penulisan’. (Nerosti, 2021) juga menjelaskan tentang proses 549 DOI: 10.24036/js.v12i4.121562 ISSN: 2302-3201 ari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah – Leri Yano Wardana & Nerosti koreografi yang mempunyai tahapan pemilihan dan pembentukan tarian menjadi sebuah gerak. Gerak Tari Piriang Lenggang Manatiang terdiri dari; 1)gerak langkah saayun, 2)gerak buai ayun, 3)gerak lenggang mantiang, 4)gerak manyabiak, 5)gerak rampak, 6)gerak saayun salangkah, 6)gerak baginjai, 7)gerak ayun ma ayun, 8)gerak lenggang lenggok Gerak Tari Piriang Lenggang Manatiang juga dapat dilihiat dari aspek ruang, waktu dan tenaga. Pada aspek ruang terdapat garis tubuh, volume, arah hadap, level dan fokus pandang. Kemudian pada aspek waktu terdapat tempo, dan ritme. Sedangkan pada aspek tenaga terdapat intensitas, tekanan dan kualitas. Aspek ruang dalam Tari Piriang Lenggang Manatiang memakai garis lurus, serong dan melengkung, seperti volume besar dan sedang, selanjutnya ada arah hadap yaitu lurus, depan, belakang dan serong. Kemudian level ada tinggi dan sedang, dan terakhir ada fokus pandangan yaitu depan belakang dan serong. Aspek waktu Tari Piriang Lenggang Manatiang pada gerak Langkah Saayun terdapat tempo dan ritme yang lambat, selanjutnya pada gerak Buai Baayun, terdapat tempo dan ritme yang cepat. Kemudian pada gerak Lenggang Manatiang sampai dengan gerak Ayun Maayun terdapat tempo dan ritem yang sedang. Terakhir pada gerak Lenggang Lenggok dan gerak Panutuik terdapat tempo dan ritme yang cepat. Aspek tenaga Tari Piriang Lenggang Manatiang pada gerak Langkah Saayun, Buai Ayun, Saayun Salangkah dan Gerak Baginjai memiliki intesitas sedikit dan tekanan yang banyak. Selanjutnya pada gerak Lenggang Manatiang sampai dengan gerak Panutuik memiliki intesitas dan tekanan yang banyak. Suasana dramatik pada Tari Piriang Lenggang Manatiang diawali dengan suasana tenang, dimana suasana awal saat prosesi gerak sambah awal, Yang manggambarkan permohonan diberi perlindungan saat menginjak kaca pada atraksi akhir. Lalu berlanjut pada bagian dua tarian, yang menunjukan suasana menghibur saat manarikan Tari Piring Lenggang Manatiang, suasana menggambarkan bentuk ketegasan dan kewibawaan niniak mamak diranah minang. Kemudian berangsur naik pada bagian tiga dan empat tarian dengan suasana klimaks penu energi, yang menggambarkan suasana kegembiraan muda mudi yang bersemangat dalam menjamu tamu di suatu perhelatan di Nagari. 2. Bentuk Koreografi Tari Piriang Lenggang Manatiang Dibagian ini juga dibuat salah satu atraksi yaitu menginjak pecahan kaca piring , yang menjadi bagian yang disenangi oleh banyak penonton. Dalam Tari Piriang Lenggang Manatiang ini terdapat 10 macam bentuk pola lantai yang dikembangkan oleh garis lurus dan lengkung. Pola lantai garis lurus yang membentuk horizontal gerak Langkah Saayun, gerak Buai Ayun, gerak Saayun Salangkah, gerak Baginjai, gerak Ayun Maayun, Gerak Lenggang Lenggok, sedangkan pada pola lantai garis lengkung terdiri dari gerak Rampak dan gerak Manyabiak. Komposisi kelompok pada Tari Piriang Lenggang Manatiang adalah: 1)gerak lenggang manatiang, 2)gerak langkah saayun, 3)gerak lengang lenggok, 4)gerak manyabiak, 5)gerakan saayun salangkah. a. Pada gerakan lenggang manatiang desain yang terlihat adalah desain berimbang (balance), karena para penari melakukan gerakan yang sama terlihat pada bentuk tangan yang sama-sama ke atas. Namun saja posisi penari laki-laki dan perempuan berada di level yang berbeda. a. Pada gerakan lenggang manatiang desain yang terlihat adalah desain berimbang (balance), karena para penari melakukan gerakan yang sama terlihat pada bentuk tangan yang sama-sama ke atas. Namun saja posisi penari laki-laki dan perempuan berada di level yang berbeda. 550 Jurnal Sendratasik Vol. 12 No. 4. Thn.2023 Gambar 1. Gerak Berimbang (Balance) (Dok. Leri Yano Wardana, Oktober 2022) b. Pada gerak Langkah Saayun dilakukan dengan cara serempak (union), dikarenakan penari laki-laki dan perempuan melakukan gerakan yang sama Gambar 2. Gerak Serempak (Union) (Dok. Leri Yano Wardana, Oktober 2022) c. Pada gerakan lengang lenggok desain yang digunakan adalah bergantian (canon), dikarenakan pada gerakan ini dilakukan secara bergantian. Gambar 1. Gerak Berimbang (Balance) (Dok. Leri Yano Wardana, Oktober 2022) b. Pada gerak Langkah Saayun dilakukan dengan cara serempak (union), dikarenakan penari laki-laki dan perempuan melakukan gerakan yang sama Gambar 1. Gerak Berimbang (Balance) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 1. Gerak Berimbang (Balance) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 1. Gerak Berimbang (Balance) (Dok. Leri Yano Wardana, Oktober 2022) b. Pada gerak Langkah Saayun dilakukan dengan cara serempak (union), dikarenakan penari laki-laki dan perempuan melakukan gerakan yang sama Gambar 2. Gerak Serempak (Union) (Dok. Leri Yano Wardana, Oktober 2022) c. Pada gerakan lengang lenggok desain yang digunakan adalah bergantian (canon), dikarenakan pada gerakan ini dilakukan secara bergantian. Gambar 3. Gerak Bergantian (Canon) (Dok. Leri Yano Wardana, Oktober 2022) b. Pada gerak Langkah Saayun dilakukan dengan cara serempak (union), dikarenakan penari laki-laki dan perempuan melakukan gerakan yang sama Gambar 2. Gerak Serempak (Union) (Dok. Leri Yano Wardana, Oktober 2022) b. DOI: 10.24036/js.v12i4.121562 2. Bentuk Koreografi Tari Piriang Lenggang Manatiang Pada gerak Langkah Saayun dilakukan dengan cara serempak (union), dikarenakan penari laki-laki dan perempuan melakukan gerakan yang sama b. Pada gerak Langkah Saayun dilakukan dengan cara serempak (union), dikarenakan penari laki-laki dan perempuan melakukan gerakan yang sama Gambar 2. Gerak Serempak (Union) (Dok. Leri Yano Wardana, Oktober 2022) c. Pada gerakan lengang lenggok desain yang digunakan adalah bergantian (canon), dikarenakan pada gerakan ini dilakukan secara bergantian. Gambar 2. Gerak Serempak (Union) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 2. Gerak Serempak (Union) (Dok. Leri Yano Wardana, Oktober 2022) c. Pada gerakan lengang lenggok desain yang digunakan adalah bergantian (canon), dikarenakan pada gerakan ini dilakukan secara bergantian. Gambar 3. Gerak Bergantian (Canon) (Dok. Leri Yano Wardana, Oktober 2022) c. Pada gerakan lengang lenggok desain yang digunakan adalah bergantian (canon), dikarenakan pada gerakan ini dilakukan secara bergantian. Gambar 3. Gerak Bergantian (Canon) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 3. Gerak Bergantian (Canon) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 3. Gerak Bergantian (Canon) (Dok. Leri Yano Wardana, Oktober 2022) 551 ISSN: 2302-3201 ISSN: 2302-3201 DOI: 10.24036/js.v12i4.121562 ari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah – Leri Yano Wardana & Nerosti d. Pada gerakan manyabiak terlihat penari menggunakan desain selang-seling (alternate), dilihat dari bentuk pola lantai yang digunakan yaitu merupakan deain selang-seling, dan gerakan-gerakan yang dilakukan secara bergantian. d. Pada gerakan manyabiak terlihat penari menggunakan desain selang-seling (alternate), dilihat dari bentuk pola lantai yang digunakan yaitu merupakan deain selang-seling, dan gerakan-gerakan yang dilakukan secara bergantian. Gambar 4. Gerak Selang Seling (Alternate) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 4. Gerak Selang Seling (Alternate) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 4. Gerak Selang Seling (Alternate) (Dok. Leri Yano Wardana, Oktober 2022) e. Pada gerakan saayun salangkah penari menggunakan desain terpecah (broken), dilihat dari bentuk pola lantai yang digunakan yaitu penari laki-laki dan perempuan saling membuat kelompoknya masing-masing. Gambar 5. Gerak Terpecah (Broken ) (Dok. Leri Yano Wardana, Oktober 2022) e. Pada gerakan saayun salangkah penari menggunakan desain terpecah (broken), dilihat dari bentuk pola lantai yang digunakan yaitu penari laki-laki dan perempuan saling membuat kelompoknya masing-masing. Gambar 5. Gerak Terpecah (Broken ) (Dok. Leri Yano Wardana, Oktober 2022) Gambar 5. Gerak Terpecah (Broken ) (Dok. Leri Yano Wardana, Oktober 2022) Kostum Tari Piriang Lenggang Manatiang merupakan busana yang didesain dan dibuat dari bahan dasar Baju Kuning beludru. Celana yang digunakan dalam tarian ini adalah galembong galembong yang dibalut dengan atasan sasampiang. Kesimpulan Tari Piriang Lenggang Manatiang merupakan tari kreasi di Minangkabau. Tarian ini menceritakan suatu rapat nagari niniak mamak dengan masyarakat setempat, dan keramah tamahan muda mudi di ranah minang yang bersemangat dalam menjamu tamunya di suatu perelatan dalam nagari. Proses koreografi Tari Piriang Lenggang Manatiang terdiri dari ide/tema, eksplorasi, improvisasi, dan komposisi. Maka dari proses yang dilakukan koreografer menghasilkan aspek bentuk sebagai berikut: gerak, desain ruang, desain waktu, desain tenaga, desain atas, desain dramatik, desain lantai, komposisi kelompok, perlengkapan, musik. Tari Piriang Lenggang Manatiang adalah tari kelompok, gerak Tari Piriang Lenggang Manatiang bersumber dari gerak dasar pencak silat dan gerak yang menggambarkan aktivitas orang ke sawah. Nama-nama gerak dalam Tari Piriang Lenggang Manatiang yaitu: Gerak langkah saayun, gerak buai ayun, gerak lenggang mantiang, gerak manyabiak, gerak rampak, gerak saayun salangkah, gerak baginjai, gerak ayun ma ayun, gerak lenggang lenggok Iringan pada Tari Piriang Lenggang Manatiang adalah dendang, gandang, sarunai, bansi, canang, dan talempong pacik. Kostum yang digunakan merupakan kostum yang sudah dimodifikasikan yaitu menggunakan baju yang berbahan bludru dan celana galembong, kemudian untuk penari perempuan menggunakan aksesoris di bagian kepalanya. Tari Piriang Lenggang Manatiang menggunakan properti Piring atau piriang dalam Bahasa Minangkabau. Selain itu, terdapat konsepsi isi yang menggambarkan karakter dan unsur filosofis Minang yang tercermin pada bentuk tarian yang menyajikan pola berkelompok dan senantiasa bergerak bersama dan saling melengkapi. Dengan ciri-ciri yang demikian, tarian ini termasuk dalam rumpun tari kreasi. 2. Bentuk Koreografi Tari Piriang Lenggang Manatiang Kemudian mereka mengenakan ikat pinggang dan tali bahu, asesoris lengkap seperti bros, kalung dan anting- anting, batang bambu dan bunga. Tari Piriang Lenggang Manatiang menggunakan ciri berupa piring besar berjumlah 7 buah, pada tarian ini piring melambangkan rasa syukur atas hasil panen yang melimpah, ritual panen diletakkan di atas piring kemudian dibawakan dengan gerakan yang dinamis. Selain itu piring lambang kemakmuran dan kesejahteraan masyarakat Minangkabau (Fathonah, 2019). Instrumen yang digunakan adalah gandang, bansi, sarunai, canang dan talempong pacik. 552 Jurnal Sendratasik Vol. 12 No. 4. Thn.2023 Referensi Anggraini, D. A., & Astuti, F. (2020). Nilai–Nilai Pendidikan Karakter Dalam Gerak Tari Piring Di Sanggar Pelangi Ranah Minang Di Kota Padang. Jurnal Sendratasik, 10(1), 156-167. Astuti, F. (2016). Menumbuhkan Nilai-Nilai Kearifan Lokal Melalui Karya Tari Koreografer Perempuan Di Sumatera Barat: Suatu Tinjauan Gender. Humanus, 12(2), 53. https://doi.org/10.24036/jh.v15i1.6412 Fathonah, S., Paramita, S., & Utami, L. S. S. (2019). Makna Pesan dalam Tari Tradisional (Analisis Deskriptif Kualitatif Makna Pesan dalam Kesenian Tari Piring). Koneksi, 3(1), 99-104. Indrayuda. (2012). Eksistensi Tari Minangkabau Padang: UNP Press. Jurnal Sendratasik Vol. 12 No. 4. Thn.2023 Indrayuda. (2012). Eksistensi Tari Minangkabau Padang: UNP Press. Izaty, R., & Nerosti, N. (2019). Bentuk Garapan Tari Piring Lampu Togok Kreasi Sanggar Lubuak Nan Tigo Kota Solok. Jurnal Sendratasik, 8(1), 53-61. 553 DOI: 10.24036/js.v12i4.121562 ISSN: 2302-3201 553 ISSN: 2302-3201 DOI: 10.24036/js.v12i4.121562 Tari Piriang Lenggang Manatiang di Sanggar Anjuang Siriah – Leri Yano Wardana & Nerosti Moleong, L. (2012). Metodologi Penelitian Kualitatif . Bandung: PT Remaja Rosdakarya. Mulyani, Novi. (2016). Pendidikan Seni Tari Anak Usia Dini. Yogyakarta: Gava Media. erosti. (2021). Mencipta Dan Menulis Skripsi Tari. Depok: Rajawali Press. Putri, M. S., & Nerosti, N. (2020). Analisis Gerak Dan Karakter Tari Kain Di Pauah V Kecamatan Pauah Kota Padang. Jurnal Sendratasik, 9(4), 203-211. Rahmida, Setiawati. (2008). Seni Tari. Jakarta: Direktorat Pembinaan Sekolah Menengah Kejuruan, Direktorat Jendral Manajemen Pendidikan Dasar Dan Menengah, Departemen Pendidikan Nasional. Sugiartha, I. G. A. (2018). Bentuk dan Konsep Estetik Musik Tradisional Bali. Panggung, 23(1), 46–60. https://jurnal.isbi.ac.id/index.php/panggung/article/view/14/18 Windy, U. I., & Nerosti, N. (2020). Pewarisan Silek Pauh Di Perguruan Silaturahmi Di Kelurahan Kalumbuk Kecamatan Kuranji Kota Padang. Jurnal Sendratasik, 9(1), 57-63. 554 Jurnal Sendratasik Vol. 12 No. 4. Thn.2023
https://openalex.org/W3194776740	https://europepmc.org/articles/pmc8442117?pdf=render	English	null	An in vitro model for the cultivation of polymicrobial biofilms under continuous-flow conditions	F1000Research	2,021	cc-by	17,996	approved] Thomas James O'Brien 1, Marwa Mohsen Hassan 2, Freya Harrison2, Martin Welch1 1Department of Biochemistry, University of Cambridge, Cambridge, CB2 1QR, UK 2School of Life Sciences, University of Warwick, Coventry, CV4 7AL, UK Open Peer Review Reviewer Status Invited Reviewers 1 2 version 1 13 Aug 2021 report report Daniel R. Neill , University of Liverpool, Liverpool, UK 1. Lindsay R. Kalan, University of Wisconsin- Madison, Madison, USA 2. Any reports and responses or comments on the article can be found at the end of the article. Open Peer Review Reviewer Status Invited Reviewers 1 2 version 1 13 Aug 2021 report report Daniel R. Neill , University of Liverpool, Liverpool, UK 1. Lindsay R. Kalan, University of Wisconsin- Madison, Madison, USA 2. Any reports and responses or comments on the article can be found at the end of the article. Open Peer Review Reviewer Status Invited Reviewers 1 2 version 1 13 Aug 2021 report report Daniel R. Neill , University of Liverpool, Liverpool, UK 1. Lindsay R. Kalan, University of Wisconsin- Madison, Madison, USA 2. Any reports and responses or comments on the article can be found at the end of the article. First published: 13 Aug 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 Latest published: 13 Aug 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 v1 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Abstract The airways of people with cystic fibrosis (CF) are often chronically colonised with a diverse array of bacterial and fungal species. However, little is known about the relative partitioning of species between the planktonic and biofilm modes of growth in the airways. Existing in vivo and in vitro models of CF airway infection are ill-suited for the long-term recapitulation of mixed microbial communities. Here we describe a simple, in vitro continuous-flow model for the cultivation of polymicrobial biofilms and planktonic cultures on different substrata. Our data provide evidence for inter-species antagonism and synergism in biofilm ecology. We further show that the type of substratum on which the biofilms grow has a profound influence on their species composition. This happens without any major alteration in the composition of the surrounding steady-state planktonic community. Our experimentally-tractable model enables the systematic study of planktonic and biofilm communities under conditions that are nutritionally reminiscent of the CF airway microenvironment, something not possible using any existing in vivo models of CF airway infection. Lindsay R. Kalan, University of Wisconsin- Madison, Madison, USA 2. Lindsay R. Kalan, University of Wisconsin- Madison, Madison, USA 2. Any reports and responses or comments on the article can be found at the end of the article. Any reports and responses or comments on the article can be found at the end of the article. Keywords This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. How to cite this article: O'Brien TJ, Hassan MM, Harrison F and Welch M. An in vitro model for the cultivation of polymicrobial biofilms under continuous-flow conditions [version 1; peer review: 2 approved] F1000Research 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 First published: 13 Aug 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 Corresponding author: Martin Welch (mw240@cam.ac.uk) Author roles: O'Brien TJ: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Hassan MM: Investigation, Methodology, Resources, Writing – Review & Editing; Harrison F: Methodology, Resources, Writing – Review & Editing; Welch M: Conceptualization, Funding Acquisition, Project Administration, Supervision, Writing – Review & Editing Competing interests: No competing interests were disclosed. Grant information: This work was supported by a studentship (NC/P001564/1) from the NC3Rs to support TO’B, and consumables support from a UK Cystic Fibrosis Trust Venture and Innovation award and from the British Lung Foundation. FH and MH were supported by a grant from the MRC (ref. MR/R001898/1). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2021 O'Brien TJ et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. How to cite this article: O'Brien TJ, Hassan MM, Harrison F and Welch M. An in vitro model for the cultivation of polymicrobial biofilms under continuous-flow conditions [version 1; peer review: 2 approved] F1000Research 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 First published: 13 Aug 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 Keywords y Cystic fibrosis (CF), polymicrobial, biofilms, in vitro models, 3Rs This article is included in the NC3Rs gateway. Page 1 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Corresponding author: Martin Welch (mw240@cam.ac.uk) Author roles: O'Brien TJ: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Hassan MM: Investigation, Methodology, Resources, Writing – Review & Editing; Harrison F: Methodology, Resources, Writing – Review & Editing; Welch M: Conceptualization, Funding Acquisition, Project Administration, Supervision, Writing – Review & Editing Competing interests: No competing interests were disclosed. Grant information: This work was supported by a studentship (NC/P001564/1) from the NC3Rs to support TO’B, and consumables support from a UK Cystic Fibrosis Trust Venture and Innovation award and from the British Lung Foundation. FH and MH were supported by a grant from the MRC (ref. MR/R001898/1). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2021 O'Brien TJ et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. How to cite this article: O'Brien TJ, Hassan MM, Harrison F and Welch M. An in vitro model for the cultivation of polymicrobial biofilms under continuous-flow conditions [version 1; peer review: 2 approved] F1000Research 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 First published: 13 Aug 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 Corresponding author: Martin Welch (mw240@cam.ac.uk) Author roles: O'Brien TJ: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Hassan MM: Investigation, Methodology, Resources, Writing – Review & Editing; Harrison F: Methodology, Resources, Writing – Review & Editing; Welch M: Conceptualization, Funding Acquisition, Project Administration, Supervision, Writing – Review & Editing Competing interests: No competing interests were disclosed. Grant information: This work was supported by a studentship (NC/P001564/1) from the NC3Rs to support TO’B, and consumables support from a UK Cystic Fibrosis Trust Venture and Innovation award and from the British Lung Foundation. FH and MH were supported by a grant from the MRC (ref. MR/R001898/1). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2021 O'Brien TJ et al. Introduction Cystic fibrosis (CF) is a life-limiting genetic disorder estimated to affect 70,000 people worldwide (Cystic Fibrosis Foundation). Although a systemic multi-organ disease, the most striking manifestation of CF is chronic obstruction of the airways via an overproduction of viscous, nutrient-rich airway secretions. These secretions block the airways and predispose people with CF to life-long microbial infection. These infections are frequently polymicrobial and comprise both bacterial and fungal species (Ahmed et al., 2019; Boutin et al., 2015; Carmody et al., 2015; Hogan et al., 2016; Jorth et al., 2019; Mahboubi et al., 2016; Rogers et al., 2010; Sibley and Surette, 2011; Zhao et al., 2012). CF airway infections contribute towards a decline in pulmonary function and it is estimated that ~90% of persons with CF succumb to respiratory failure as a direct result of microbial infection (Chmiel and Davis, 2003; Elborn, 2016; Lubamba et al., 2012; Lyczak et al., 2002; Rajan and Saiman, 2002; Sibley et al., 2006). Recent years have seen increasing recognition that inter-species interactions between the airway microbiota may play a role in modulating the behaviour, virulence, and even the response to therapeutic intervention (Antonic et al., 2013; Armbruster et al., 2016; Baldan et al., 2014; Barnabie and Whiteley, 2015; Beaume et al., 2015; Briaud et al., 2019; Dalton et al., 2011; Diggle et al., 2007; Elias and Banin, 2012; Hotterbeekx et al., 2017; Hibbing et al., 2010; Korgaonkar et al., 2013; Mastropaolo et al., 2005; O'Brien and Fothergill, 2017; Peters et al., 2012; Weimer et al., 2010). Moreover, co-culturing bacterial species in vitro and in vivo has been shown to lead to significant alterations in the expression of core essential genes (Ibberson et al., 2017; Ibberson and Whiteley, 2020). One particularly important unaddressed question relates to the impact of co-habiting species on the biology of Pseudomonas aeruginosa (PA). PA is a common inhabitant of the CF airways, and a model organism for the study of biofilms. However, the paucity of experimentally tractable in vivo or in vitro models of CF infection has severely hampered the in-depth and longitudinal study of such polymicrobial communities (O'Brien and Welch, 2019b). It has been estimated that at least 60% of bacterial infections in the western world involve the formation of biofilms (Fux et al., 2005) and CF is no exception. One crucial feature of microbial biofilms is their increased ability to bypass effective immune clearance and resist antimicrobial action. Practical benefits • Studying changes in ecology of polymicrobial biofilm formation during co-culture on different solid substrata. Potential applications • Studying changes in gene expression and behaviour of polymicrobial biofilms comprising of different species combinations. • Studying response of polymicrobial biofilms and planktonic communities against treatment with antimicrob for the development and validation of polymicrobial biofilm dispersal/treatment regimens. • Studying response of polymicrobial biofilms and planktonic communities against treatment with antimicrobials for the development and validation of polymicrobial biofilm dispersal/treatment regimens. Corresponding author: Martin Welch (mw240@cam.ac.uk) Corresponding author: Martin Welch (mw240@cam.ac.uk) Author roles: O'Brien TJ: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Hassan MM: Investigation, Methodology, Resources, Writing – Review & Editing; Harrison F: Methodology, Resources, Writing – Review & Editing; Welch M: Conceptualization, Funding Acquisition, Project Administration, Supervision, Writing – Review & Editing First published: 13 Aug 2021, 10:801 https://doi.org/10.12688/f1000research.55140.1 Page 2 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Research highlights Scientific benefits • Allows for direct comparison of biofilm and planktonic microbial lifestyles. • Allows for longitudinal and real-time analysis of interspecies interactions among polymicrobial biofilms. 3Rs benefits • Reduces the need for vertebrate infection models when studying microbe-microbe interactions in the context of cystic fibrosis. Practical benefits • Inexpensive and simple to operate. • Chemically defined culture conditions, allows for the reproducible study of subtle interspecies interactions. • Experimentally tuneable, allows for different species combinations or disease states to be studied. Current applications • Studying changes in ecology of polymicrobial biofilm formation during co-culture on different solid substrata. Potential applications • Studying changes in gene expression and behaviour of polymicrobial biofilms comprising of different species combinations. • Studying response of polymicrobial biofilms and planktonic communities against treatment with antimicrobials for the development and validation of polymicrobial biofilm dispersal/treatment regimens. • Allows for direct comparison of biofilm and planktonic microbial lifestyles. • Allows for longitudinal and real-time analysis of interspecies interactions among polymicrobial biofilms. Introduction CF mice do not develop spontaneous airway infections (Bayes et al., 2016; Cash et al., 1979; van Heeckeren et al., 2006), and artificially- induced infections of the airways are rapidly cleared if the inoculated microbial species are not immobilised using agar/ agarose/alginate beads (Chattoraj et al., 2010; Hoffmann et al., 2005; Moser et al., 2009; Munder et al., 2011). As with CF ferrets, this prevents the long-term study of polymicrobial communities and interspecies interactions. Due to inherent limitations in the way in which the data is reported, the exact number of animals used for research into CF airway infections is difficult to ascertain. Yet a systematic review of the literature up to 2015 has found 12,304 publications discussing the use of CF animal models. Of those that are primary research articles (799 publications), 636 report the use of CF mice, rats, pigs, ferrets or zebrafish (with the remaining publications not reporting the use of any genetic CF models) (Leenaars et al., 2020). Given that this literature review only accounts for publications up to 2015, and given the recent surge of interest into understanding the polymicrobial communities associated with CF airway infections, the number of CF animals used in research will have certainly increased over the last six years. Despite the importance of biofilms in CF pathology, and their well-known link with increased patient mortality rates, there is currently no suitable in vivo model for the study of polymicrobial biofilm communities in the context of CF. Furthermore, none of the aforementioned 636 studies report the use of CF models to study polymicrobial biofilm communities. Practical limitations when sampling the airway microbiota in infected animal models mean that the animals must be sacrificed and their lungs excised before histopathological examination. Not only are there significant ethical implica- tions associated with these approaches, but the longitudinal/long-term study of microbial populations becomes imprac- tical. More recently, ex vivo models of CF have redressed some of these issues, enabling the study of microbial biofilm lifestyles reminiscent of those observed in situ (Harrington et al., 2020; Sweeney et al., 2021). However, owing to the intense competition observed among microbial species, existing ex vivo models do not yet permit the study of more than one microbial species at a time. This is a significant experimental limitation when studying long-term polymicrobial infection scenarios such as CF. Introduction Indeed, some bacterial biofilms are up to 1000 more resistant to antimicrobial intervention compared with their planktonic cell counterparts (Parsek, 2003). The formation of biofilm-like aggregates in the CF airways is often cited as a reason why therapeutic strategies aimed at eradicating keystone CF pathogens such as PA fail (Bjarnsholt et al., 2009; Döring et al., 2011; Elias and Banin, 2012; Folkesson et al., 2012; Leekha et al., 2011; Lopes et al., 2012; Lopes et al., 2014; Mowat et al., 2011). However, our understanding of how Page 3 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 inter-species interactions alter biofilm physiology remains limited. This is important because agonistic and antagonistic interactions between species almost certainly confer a significant selection pressure, thereby driving adaptive divergence in members of the polymicrobial community (Markussen et al., 2014; Schick and Kassen, 2018; Winstanley et al., 2016). inter-species interactions alter biofilm physiology remains limited. This is important because agonistic and antagonistic interactions between species almost certainly confer a significant selection pressure, thereby driving adaptive divergence in members of the polymicrobial community (Markussen et al., 2014; Schick and Kassen, 2018; Winstanley et al., 2016). There are three key vertebrate models of CF available to researchers: the CF mouse, the CF ferret and the CF pig (reviewed recently by O'Brien and Welch, 2019b). Although the porcine airways share a remarkable degree of genetic and structural homology with human airways (Judge et al., 2014; Rogers et al., 2008), the cost and technical/ethical complexity of using porcine models ensures they are rarely used in research. CF ferrets are also rarely used, since they develop severe airway infections soon after birth and subsequently succumb to respiratory failure (Hoffman and Hajjar, 2018; Sun et al., 2010), although this situation may change as in utero administration of cystic fibrosis transmembrane conductance regulator (CFTR) modulators can rescue this severe phenotype (Sun et al., 2019). By contrast, the CF mouse is widely used, with >14 different distinct models available (Guilbault et al., 2007). However, and despite the widespread availability of these murine CF models, their usefulness for studying chronic airway infection is limited. Introduction Hence, the in vitro model we describe here provides a novel tool for the study of mixed- species biofilms that is not currently possible using existing in vivo CF models. The primary limitations of in vitro infection models, compared with in vivo models, is a lack of host cells and a functional immune system (which may contribute towards clearing microbial infections in situ). However, it should be noted that impaired immune clearance of microbes in an inherent feature of CF, which somewhat mitigates the lack of a functional immune system in in vitro models (Cohen and Prince, 2012; Bonfield and Chmiel, 2017). Although in vitro models do not capture all aspects of human disease pathophysiology, they do provide an ideal tool for studying interspecies interactions between microbes (O'Brien and Welch, 2019b). Emerging evidence suggests that chemical, not spatiotem- poral, factors have the greatest impact on driving changes in microbial lifestyles (Lopes et al., 2017; Quinn et al., 2015). Hence, the defined and experimentally-perturbable nature of in vitro models are an attractive option for studying microbial behaviour in a reductive manner. Furthermore, the development of artificial sputum medium (ASM), closely mimicking the nutritional composition of CF airway secretions, provides an unparalleled opportunity for the study of polymicrobial populations under conditions that chemically recapitulate the CF microenvironment (Haley et al., 2012; Sousa et al., 2018; Sriramulu et al., 2005; Turner et al., 2015; Frapwell et al., 2018). However, simply mixing CF airway- associated species together in ASM and hoping for the best is not a recipe for success. This is because the co-cultures display compositional instability, and the initially diverse community rapidly become dominated by just one or a few species. To remedy this, we developed an in vitro model of CF (O'Brien and Welch, 2019a). Importantly, our model permits planktonic microbial communities of three distinctly different species associated with CF airway infections (PA, Staphylococcus aureus (SA) and Candida albicans (CA)) to be maintained, indefinitely, as a stable steady-state community. In this report we build upon our existing model and describe a simple, yet versatile, method of culturing polymicrobial biofilms on different solid substrata. Methods Methods Methods for the model development Microbial strains Methods for the model development i bi l i All microbial strains used in this work are shown in Table 1. Strains were routinely cultured in lysogeny broth (LB) (Formedium) on a 40 cm diameter rotating drum with mild aeration (0.5 rotations per second) at 37°C overnight. Continuous-flow culture vessel and biofilm container Artificial sputum medium (ASM) was used as the main growth medium for all experiments and was prepared as previously described (O'Brien and Welch, 2019a). The continuous-flow culture system has also been previously described (O'Brien and Welch, 2019a). Both are described in more detail in the protocol below. Briefly, the culture vessel consists of a 100 mL Duran flask, fitted with an assembled 4-port HPLC GL80 screw cap (Duran). A 24-channel IPC ISM934C standard-speed digital peristaltic pump (Ismatec) was used to deliver sterile ASM from a media reservoir at a defined flowrate (Q) through 1.5 mm bore sterilin silicon tubing (Fisher Scientific) to the culture vessel. A different channel of the same pump was used to remove waste culture into a discard jar at the same flowrate (Figure 1). Biofilms Figure 1. Schematic diagram of the continuous-flow culture vessel containing biofilm container. The main culture vessel (centre) is a 100 mL Duran bottle fitted with a 4-port HPLC GL80 screwcap lid, containing four sealable inlet/outlet ports from which the biofilm container is suspended. A multichannel peristaltic pump delivers fresh media (ASM) into the culture vessel from a reservoir (left), and also removes waste culture into a discard jar (right) at the same rate of flow (Q). Arrows show the direction of media flow. The culture vessel and media reservoirs are incubated at 37°C and the contents are kept homogenous through gentle stirring (100 rpm). The value of Q depends on the microbial species being cultured within the vessel. Table 1. Microbial strains used in this study. Strain Description Reference PAO1 Pseudomonas aeruginosa, spontaneous chloramphenicol-resistant derivative. Used worldwide as a laboratory reference strain (isolated Melbourne, 1954) (Holloway, 1955) ATCC 25923 Staphylococcus aureus Rosenbach (ATCC®25923D-5™), methicillin sensitive clinical isolate. Laboratory reference strain lacking recombinases and mecA (isolated Seattle, 1945) (Treangen et al., 2014) SC5314 Candida albicans, clinical isolate commonly used as a wild-type laboratory reference strain (isolated New York, 1980’s) (Gillum et al., 1984) Page Table 1. Microbial strains used in this study. Table 1. Microbial strains used in this study. Introduction Page 4 of 24 Page 4 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 The model described in this work permits the simultaneous cultivation of steady-state planktonic and biofilm commu- nities, allowing for direct comparisons to be made between these two modes of microbial growth. In principle, any combination of microbial species associated with CF airway infections could be cultured using the setup. As such, not only does our system reduce the need for in vivo CF infection models for studying microbe-microbe interactions; it enables the real-time, longitudinal study of polymicrobial communities in an experimentally reproducible, controlled setup. It has not escaped our notice that this model system is also well-suited for the road-testing of interventions aimed at preventing the formation of biofilms in chronic CF airway infections. Preparation of biofilm substratum Agar plates (2.5% w/v agar in milliQ water) were poured to a depth of 5 mm. Using a sterile number 21 scalpel and a set of sterile 12.7 cm curved forceps, cubes of approximately 5 mm x 5 mm were cut from the plate and transferred into the sterile biofilm container. Biofilm work using EVPL sections was performed in collaboration with Dr Freya Harrison at the Agar plates (2.5% w/v agar in milliQ water) were poured to a depth of 5 mm. Using a sterile number 21 scalpel and a set of sterile 12.7 cm curved forceps, cubes of approximately 5 mm x 5 mm were cut from the plate and transferred into the sterile biofilm container. Biofilm work using EVPL sections was performed in collaboration with Dr Freya Harrison at the Figure 2. Container for culturing biofilms under continuous-flow conditions. The biofilm container constructed from stainless steel gauze for this study. Biofilm containers consisted of a cylinder, 35 mm in length and 25 mm in diameter (Ø), connected to two stainless steel arms (45 mm in length), which suspend the container from the HPLC screw-port lid. (A) Side view of the container with dimensions. (B) Bottom view of the container. Single strands of stainless-steel wire were threaded across the bottom of the container to form a mesh and prevent the biofilm substratum from falling through. A small gap (~3 mm Ø) was also included for the outlet tube to be threaded through to allow the removal of culture media during incubation. (C) View of the 4-port HPLC screw cap lid. The left port is the media inlet, the right port is the media outlet and the two remaining ports are threaded with a single piece of rubber tubing (1.5 mm bore) to hold the arms of the mesh container. (D) The biofilm container in situ in the assembled setup. Note that the container is completely submerged in the growth medium but is separated from the magnetic stir bar to prevent interference with the continual stirring of the culture vessel. Pag Figure 2. Container for culturing biofilms under continuous-flow conditions. The biofilm container constructed from stainless steel gauze for this study. Biofilm containers consisted of a cylinder, 35 mm in length and 25 mm in diameter (Ø), connected to two stainless steel arms (45 mm in length), which suspend the container from the HPLC screw-port lid. Methods Strain Description Reference PAO1 Pseudomonas aeruginosa, spontaneous chloramphenicol-resistant derivative. Used worldwide as a laboratory reference strain (isolated Melbourne, 1954) (Holloway, 1955) ATCC 25923 Staphylococcus aureus Rosenbach (ATCC®25923D-5™), methicillin sensitive clinical isolate. Laboratory reference strain lacking recombinases and mecA (isolated Seattle, 1945) (Treangen et al., 2014) SC5314 Candida albicans, clinical isolate commonly used as a wild-type laboratory reference strain (isolated New York, 1980’s) (Gillum et al., 1984) Figure 1. Schematic diagram of the continuous-flow culture vessel containing biofilm container. The main culture vessel (centre) is a 100 mL Duran bottle fitted with a 4-port HPLC GL80 screwcap lid, containing four sealable inlet/outlet ports from which the biofilm container is suspended. A multichannel peristaltic pump delivers fresh media (ASM) into the culture vessel from a reservoir (left), and also removes waste culture into a discard jar (right) at the same rate of flow (Q). Arrows show the direction of media flow. The culture vessel and media reservoirs are incubated at 37°C and the contents are kept homogenous through gentle stirring (100 rpm). The value of Q depends on the microbial species being cultured within the vessel. Page 5 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 were allowed to develop in the continuous-flow culture vessel on two types of solid substratum; agar chunks and ex vivo porcine lung tissue (EVPL) sections, which have been previously reported to closely mimic the airway surface environment (Harrison et al., 2014; Harrison and Diggle, 2016; Sweeney et al., 2021; Harrington et al., 2020). These substrata were suspended in bespoke cylindrical biofilm containers (25 mm diameter 35 mm length) constructed from stainless steel wire gauze (Fisher Scientific) (Figure 2A,B). Containers were then suspended in the culture vessel hanging from a piece of silicon tubing threaded through two of the unused HPLC ports (Figure 2C). The entire culture system was maintained at 37°C and the liquid contents of the vessel were kept homogenous by stirring (100 rpm) using a round magnetic stir bar (length 20 mm, diameter 8 mm). were allowed to develop in the continuous-flow culture vessel on two types of solid substratum; agar chunks and ex vivo porcine lung tissue (EVPL) sections, which have been previously reported to closely mimic the airway surface environment (Harrison et al., 2014; Harrison and Diggle, 2016; Sweeney et al., 2021; Harrington et al., 2020). Methods These substrata were suspended in bespoke cylindrical biofilm containers (25 mm diameter 35 mm length) constructed from stainless steel wire gauze (Fisher Scientific) (Figure 2A,B). Containers were then suspended in the culture vessel hanging from a piece of silicon tubing threaded through two of the unused HPLC ports (Figure 2C). The entire culture system was maintained at 37°C and the liquid contents of the vessel were kept homogenous by stirring (100 rpm) using a round magnetic stir bar (length 20 mm, diameter 8 mm). Preparation of biofilm substratum (A) Side view of the container with dimensions. (B) Bottom view of the container. Single strands of stainless-steel wire were threaded across the bottom of the container to form a mesh and prevent the biofilm substratum from falling through. A small gap (~3 mm Ø) was also included for the outlet tube to be threaded through to allow the removal of culture media during incubation. (C) View of the 4-port HPLC screw cap lid. The left port is the media inlet, the right port is the media outlet and the two remaining ports are threaded with a single piece of rubber tubing (1.5 mm bore) to hold the arms of the mesh container. (D) The biofilm container in situ in the assembled setup. Note that the container is completely submerged in the growth medium but is separated from the magnetic stir bar to prevent interference with the continual stirring of the culture vessel. Page 6 of 24 Page 6 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Figure 3. Ex vivo porcine bronchiole. A single bronchiole excised from porcine lung tissue after cleaning and a single wash in DMEM/RPMI solution. All alveolar and vascular tissue was removed using a razor blade and dis- section scissors. After cleaning, bronchiole tissue was cut into 5 mm 5 mm squares for use as solid substratum to promote biofilm growth. Figure 3. Ex vivo porcine bronchiole. A single bronchiole excised from porcine lung tissue after cleaning and a single wash in DMEM/RPMI solution. All alveolar and vascular tissue was removed using a razor blade and dis- section scissors. After cleaning, bronchiole tissue was cut into 5 mm 5 mm squares for use as solid substratum to promote biofilm growth. University of Warwick, using previously described methods (Harrison and Diggle, 2016). Briefly, fresh pig lungs were collected from the butcher (John Taylor, Earlsdon, UK) and processed within the hour. To remove surface contaminants, a 25 cm palette knife was heated until red-hot using a Bunsen burner and briefly tapped (< 1 s) on the area to be dissected. Working with aseptic technique, a sterile razor blade was used to excise the bronchiole and remove all alveolar tissue; a cleaned bronchiole can be seen in Figure 3. Sample collection For characterisation of biofilm populations, pieces of substratum were aseptically removed at the indicated times and transferred to 500 μL PBS in 24-well microtiter plates. Loosely attached planktonic cells were removed from the substratum by briefly swirling the plates. The substratum was then transferred to a second well in the same plate and washed once more. Finally, after a third wash, the samples were transferred into 2 mL bead beating tubes containing 1 mL PBS and eighteen metal beads (2.38 mm diameter, Qiagen). The tubes were them agitated in a FastPrep-24 5G benchtop homogeniser (MP Biomedicals) for 40 s at 4 m s1 to liberate the cells. CFU mL1 counts were performed as described below. Three separate substratum pieces were sampled at each time point (yielding three biological replicates per timepoint, per independent biological experiment). Samples of the planktonic culture (1 mL) were removed directly from the culture vessel using a sterile serological pipette. Independent biological experiments were then performed over separate independent weeks using fresh microbial cultures and pieces of biofilm substratum without any deviation from the methods detailed in this article. Preparation of biofilm substratum Bronchioles were washed in Dulbecco's Modified Eagle Medium (DMEM)/ Roswell Park Memorial Institute 1640 Medium (RPMI) solution (50:50 ratio, 40 mL) and cut into 5 mm wide strips with sterile dissection scissors. The strips were washed in DMEM/RPMI and cut into 5 mm 5 mm squares. EVPL sections were washed again in DMEM/RPMI, then transferred to a petri dish containing 40 mL ASM and irradiated in UV lightbox for 5 min before being aseptically transferred to the biofilm container [Full dissection methods for the production of EVPL tissue sections are also demonstrated in an open access video protocol (Harrington et al., 2021)]. Culture vessel inoculation and incubation Overnight cultures of the microbial strains (grown in LB as described above) were washed three times in phosphate- buffered saline (PBS) prior to inoculating the culture vessel. Pre-warmed ASM (37°C, 100 mL) was added to the culture vessel and inoculated with the required combination of microbial species. The optical density (at 600 nm; OD600 nm) of the washed microbial cultures was then measured using a spectrophotometer (Eppendorf BioSpectrometer kinetic) and each species was introduced into the culture vessel to achieve a starting OD600 nm of 0.05. The culture vessel was then incubated for three hours with stirring prior to starting the flow of medium (Q = 145 μL min1). CFU mL1 enumeration Three independent samples of biofilm substratum from the same culture vessel were removed per timepoint (corresponding to three biological replicates). For each biological replicate, three independent serial dilutions were made and plated out for CFU enumeration (constituting three technical replicates per each piece of substratum sampled), non-blinded CFU mL1 counts were then recorded as the average of the technical replicates. Two independent samples of culture supernatant were removed from each culture vessel per timepoint (corresponding to two technical replicates) to enumerate planktonic CFU mL1 counts of each species. selective agar were used to enumerate microbial cell counts of the different species present in a single sample: PA was isolated using Pseudomonas isolation agar (PIA, Oxoid), SA was isolated using mannitol salt agar (MSA, Oxoid) and CA was isolated onto BiGGY agar (Oxoid). When enumerating cell counts from the polymicrobial cultures, the agar plates used to isolate PA and SA were further supplemented with 5 μg mL1 itraconazole to inhibit the growth of CA. The different selective media only permit the growth of a single microbial species of interest and inhibit the growth of the other species present. This enables CFU mL1 counts for each species in the polymicrobial culture to be determined with confidence [note that when studying other co-cultures containing other combinations of microbial species, different types of selective media may be required for the selective enumeration of other species]. All plates were incubated at 37°C. PIA and MSA plates were incubated overnight (16 h) and BiGGY agar plates were incubated for 24 h. Three independent samples of biofilm substratum from the same culture vessel were removed per timepoint (corresponding to three biological replicates). For each biological replicate, three independent serial dilutions were made and plated out for CFU enumeration (constituting three technical replicates per each piece of substratum sampled), non-blinded CFU mL1 counts were then recorded as the average of the technical replicates. Two independent samples of culture supernatant were removed from each culture vessel per timepoint (corresponding to two technical replicates) to enumerate planktonic CFU mL1 counts of each species. We tested whether there was any significant difference in cell counts determined on non-selective vs selective agar, and there was not (Figure 4 and underlying data (O'Brien et al., 2021)). CFU mL1 enumeration Colony forming units (CFU) were determined using the single plate-serial dilution spotting method, as described previously (Thomas et al., 2015). Briefly, 10-fold serial dilutions of the microbial cultures were made in sterile PBS and 20 μL of each dilution was spotted (approximately 2 μL per drop) onto the appropriate selective agar. Three types of Page 7 of 24 Page 7 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Figure 4. Comparison of cell titres on selective and non-selective agar. The figure shows the viable cell counts (expressed as CFU mL1) of overnight single-species cultures of P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) plated on non-selective agar media (black bars) and selective agar media (grey bars). Data represent the mean standard deviation from two technical replicates collected from three independent biological experiments. P > 0.05 is considered as not significantly different (ns). Figure 4. Comparison of cell titres on selective and non-selective agar. The figure shows the viable cell counts (expressed as CFU mL1) of overnight single-species cultures of P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) plated on non-selective agar media (black bars) and selective agar media (grey bars). Data represent the mean standard deviation from two technical replicates collected from three independent biological experiments. P > 0.05 is considered as not significantly different (ns). selective agar were used to enumerate microbial cell counts of the different species present in a single sample: PA was isolated using Pseudomonas isolation agar (PIA, Oxoid), SA was isolated using mannitol salt agar (MSA, Oxoid) and CA was isolated onto BiGGY agar (Oxoid). When enumerating cell counts from the polymicrobial cultures, the agar plates used to isolate PA and SA were further supplemented with 5 μg mL1 itraconazole to inhibit the growth of CA. The different selective media only permit the growth of a single microbial species of interest and inhibit the growth of the other species present. This enables CFU mL1 counts for each species in the polymicrobial culture to be determined with confidence [note that when studying other co-cultures containing other combinations of microbial species, different types of selective media may be required for the selective enumeration of other species]. All plates were incubated at 37°C. PIA and MSA plates were incubated overnight (16 h) and BiGGY agar plates were incubated for 24 h. CFU mL1 enumeration To test this, we performed serial dilutions of overnight microbial cultures (routinely grown in LB as described above) and plated the same dilution series onto LB-agar (non- selective) and the appropriate selective agar for the cultured species [see notes above]. Plates were then incubated in the same static incubator at 37°C. Plates containing PA or SA incubated for 16 h and plates containing CA incubated for 24 h (as described above). Two independent serial dilutions of the same overnight culture were prepared and plated in parallel from the same overnight culture (constituting two technical replicates) and three independent biological replicates for each microbial species were performed across separate days using fresh microbial cultures and freshly prepared agar plates. Protocol for the use of this model Here we describe step-by-step the procedure used to prepare ASM and to set up/inoculate the continuous-flow culture vessel for the study of polymicrobial biofilm and planktonic communities. Reagents and equipment used in this study are listed in Table 2 and Table 3, respectively. 1. Step 1 Assembly of the biofilm containers. This can be done any time prior to making ASM. A wire “bucket” was crafted by hand from stainless steel wire gauze (Fischer Scientific). Figure 2 shows an image of a completed container and dimensions. Wire cutters are used to trim a mesh of stainless-steel gauze to a rectangle approximately 150 mm x 40 mm. The gauze was then carefully rolled into a cylinder with an approximate diameter of 25 mm. The protruding cut ends of the wire mesh on the side of the cylinder were threaded through the adjacent mesh to secure the side of the cylinder in place. The height of the bucket was then adjusted to 35 mm via trimming with the wire cutters. Spare lengths of wire thread were then woven across the base of the container to generate a bucket-like structure (Figure 2B). Note that we ensured that there was a hole in the bottom of the bucket large enough to fit the out-flow tubing (~ 3mm diameter) (Step 6). To keep the stainless- steel container from interfering with the magnetic stirrer, and to allow easy removal of substratum material during cultivation, we suspended the bucket via two wire “arms” from a loop of silicone tubing threaded through two unused ports of the HPLC screw cap lid (Figure 2C). The wire arms were made by entwining three strands of stainless-steel wire and threading these through the mesh at the top of the bucket. The protruding portion of the arms were bent to allow the bucket assembly to be draped over the silicone tubing support loop and trimmed in length to 45 mm. The correct length of these container arms is important to ensure that the bucket is completely submerged in the culture medium throughout incubation (Figure 2D). 1. Step 1 Assembly of the biofilm containers. This can be done any time prior to making ASM. A wire “bucket” was crafted by hand from stainless steel wire gauze (Fischer Scientific). Figure 2 shows an image of a completed container and dimensions. Protocol for the use of this model Add ammonium chloride (0.124 g), potassium chloride (1.116 g), sodium chloride (3.032 g) and 3-(N-morpholino) propanesulfonic acid (MOPS, 2.092 g) to a 1 L beaker and dissolve in 250 mL milliQ water with gentle stirring (50 rpm). Then add the remaining salts and amino acids from the pre- made stock solutions as shown in Table 4. Next, combine the dissolved DNA and mucin solutions and add to the beaker with gentle mixing (50 rpm) for 5 min, or until the solution is homogenous. Adjust to pH 6.8 with 0.1 M acetic acid or 0.1 M potassium hydroxide. Following this, prepare the remaining media solutions and add them to the beaker according to Table 5 alongside 5 mL egg yolk emulsion. Once fully homogenous, adjust the total volume to 1 L with milliQ water. The ASM is then filter sterilised using a 1 L disposable Stericup filter unit (0.22 μM pore size) attached to a large diaphragm vacuum pump. Note that the filtering process is slow and may take 2–3 days to complete. Half-way through filtering we recommend carefully decanting (into a clean beaker) the unfiltered ASM from the Stericup unit and rinsing the top of the filter membrane with several mL of sterile PBS (that is then discarded) and then continuing with the filtration to speed up this process. 4. Step 4 Assembly of continuous-flow culture vessel (day 3). While the ASM is being filter-sterilised, assemble the continuous-flow culture vessel as outlined below with inclusion of the biofilm bucket constructed in Step 1. The continuous-flow culture vessel consists of a 100 mL Duran flask fitted with an assembled GL-80 4-port HPLC screw cap (Duran). Two lengths of 1.5 mm bore sterlin silicone tubing (Fisher Scientific) were fed into two of the HPLC ports to act as an inlet/outlet (respectively) for the medium. To prevent contamination of the media reservoir with motile bacteria, ensure that only the outlet tube is in contact with the culture medium when the vessel is filled with ASM. A short piece of silicone tubing was threaded through the two-remaining unused HPLC ports (diagonal from one another) and secured tightly before sealing any gaps around the HPLC ports with parafilm. The biofilm bucket (assembled in Step 1) was hooked onto the suspended tubing and, using forceps, the outlet tube was gently pulled through the meshed hole in the base of the bucket (see step 1). Statistical analysis All biofilm data are represented as the mean standard deviation (SD) of three separate biological replicates (collected simultaneously from the same culture vessel at each point of sampling) per timepoint across three independent biological experiments (conducted across different weeks) that were performed using fresh: ASM, microbial cultures and pieces of biofilm substrata [note that experiments using agar chunks or EVPL sections were performed independently of one another]. All planktonic data are represented as the mean standard deviation (SD) of two technical replicates (collected simultaneously from the same culture vessel) per timepoint across three independent biological experiments. Planktonic and biofilm CFU mL1 counts were performed in parallel from the same culture vessel. All statistical analysis was performed using GraphPad Prism version 8.2.0, with P < 0.05 being considered statistically significant [statistical analysis can also be performed using R/Python packages stats/SciPy, respectively]. Paired group two-tailed t-tests were used to analyse: changes in CFU mL1 counts of the different species present in single- or mixed-species biofilm samples collected at T = 24 h or 96 h; differences in CFU mL1 counts of the microbial species present on either agar chunks or EVPL sections; and for comparisons of cell viability on selective and non-selective media. Changes in planktonic Page 8 of 24 Page 8 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 CFU mL1 counts across the timepoints for each experiment were analysed using one-way repeated measures ANOVA. Differences in planktonic CFU mL1 counts in cultures containing the different biofilm substrata across the timepoints for each experiment were analysed using two-way repeated measures ANOVA. Protocol for the use of this model Wire cutters are used to trim a mesh of stainless-steel gauze to a rectangle approximately 150 mm x 40 mm. The gauze was then carefully rolled into a cylinder with an approximate diameter of 25 mm. The protruding cut ends of the wire mesh on the side of the cylinder were threaded through the adjacent mesh to secure the side of the cylinder in place. The height of the bucket was then adjusted to 35 mm via trimming with the wire cutters. Spare lengths of wire thread were then woven across the base of the container to generate a bucket-like structure (Figure 2B). Note that we ensured that there was a hole in the bottom of the bucket large enough to fit the out-flow tubing (~ 3mm diameter) (Step 6). To keep the stainless- steel container from interfering with the magnetic stirrer, and to allow easy removal of substratum material during cultivation, we suspended the bucket via two wire “arms” from a loop of silicone tubing threaded through two unused ports of the HPLC screw cap lid (Figure 2C). The wire arms were made by entwining three strands of stainless-steel wire and threading these through the mesh at the top of the bucket. The protruding portion of the arms were bent to allow the bucket assembly to be draped over the silicone tubing support loop and trimmed in length to 45 mm. The correct length of these container arms is important to ensure that the bucket is completely submerged in the culture medium throughout incubation (Figure 2D). 2. Step 2 Preparation of ASM (day 1). Add 5 g of Type II mucin from porcine stomach to 250 mL phosphate buffered saline (pH 7.4) and leave to dissolve overnight with stirring (~400 rpm) at 4°C. On the same day, add 4 g of deoxyribonucleic acid from salmon sperm to 250 mL autoclaved milliQ water and leave to dissolve overnight in a shaking water bath (200 rpm, 30°C). While the mucin/DNA is dissolving, prepare the amino acid and salt/buffer stock solutions (50 mL of each) according to Table 4. Most amino acid stock solutions can be kept for one month in the dark at 4°C. However, stock solutions of tyrosine, threonine, cysteine, phenylalanine and histidine must be made freshly for each batch of ASM. 3. Step 3 Preparation of ASM (day 2). Protocol for the use of this model Equipment Supplier [Catalogue Number] 100 mL Duran Flask Merck [Z232076] 4-port HPLC GL80 Screw Cap Fisher Scientific [10583913] 24-channel IPC ISM934C Standard-speed Digital Peristaltic Pump Cole-Parmer [WZ-78001-42] 1.5 mm Bore Sterilin Silicon Tubing ThermoFisher [TSR0150150P] Stainless Steel Wire Gauze Fisher Scientific [12958950] Velp 6-position Multiposition Digital Stirrer Cole-Parmer [F203A0179] Round White PTFE Magnetic Stir Bar, Length 20 mm x Diameter 8 mm Merck [HS120548] Swann-Morton Number 21 Carbon Steel Scalpel Blade Fisher Scientific [11778363] Epredia Shandon 12.7 cm Cartilage Fine Point Curved Thumb Forceps Fisher Scientific [15307805] Epredia Shandon 16.5 cm Straight Dissecting Scissors (Sharp Ended) Fisher Scientific [15202290] 25 cm Eisco Palette Knife Spatula Fisher Scientific [S80827] Azpack Carbon Steel Razor Blades Fisher Scientific [11904325] Microspec 5 μL Nichrome Microbial inoculation Loop Fisher Scientific [15712175] 30 mL Polystyrene Universal Tube (Sterile) Starlab [E1412-3011] Eppendorf BioSpectrometer Kinetic Eppendorf [6136000819] 2.38 mm Diameter Metal Beads Qiagen [13118-400] 2 mL PowerBead Beating Tubes Qiagen [13116-50] FastPrep-24 5G benchtop homogeniser MP Biomedicals [116005500] Stericup Quick Release-GP Sterile Vacuum Filtration System Merck [S2GPU10RE] Nunc Cell-Culture Treated 24-well Plate ThermoFisher [142475] Sterile 90 mm 15 mm Petri Dishes Merk [Z717223] Bemis Parafilm M Laboratory Wrapping Fisher Scientific [HS234526B] was wrapped in aluminium foil and the entire assembly was autoclaved. The setup was allowed to completely dry overnight (in a drying cabinet) before use. was wrapped in aluminium foil and the entire assembly was autoclaved. The setup was allowed to completely dry overnight (in a drying cabinet) before use. 5. Step 5 Preparation of biofilm substratum (day 3). (i) Using agar chunks as a solid substratum for biofilm formation, prepare the substratum by suspending 2.5 g of agar in 100 mL milliQ water. Autoclave. Once cooled, but still warm (~65°C), pour the agar to a depth of 5 mm in a sterile 90 mm petri dish. Once set, the agar plates can be wrapped in parafilm and kept at 4°C until the day of use. (ii) Ex vivo porcine lung (EVPL) tissue sections can also be used as an alternative substratum for biofilm formation. Please refer to Step 7 (day 4) for the protocol detailing the inclusion of EVPL tissue sections in the model. 6. Step 6 Preparation of overnight microbial cultures (day 3). Place a nichrome 5 μL microbial inoculation loop into the blue flame of a Bunsen burner for 5 s. Protocol for the use of this model A clean magnetic stir bar was placed in the culture vessel and the fully-assembled lid was fitted. Finally, the inlet tubing Page 9 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Table 2. Reagents used in this study. Reagent Supplier [Catalogue Number] Mucin from porcine stomach (Type II) Merck [M2378] Deoxyribonucleic acid from salmon sperm Merck [31149] Phosphate buffered saline Oxoid [BR0014G] L-serine Merck [S4500] L-glutamic acid hydrochloride Merck [G2128] L-proline Merck [P8865] L-glycine Merck [G7126] L-alanine Merck [A7627] L-valine Merck [V0500] L-methionine Merck [M9625] L-isoleucine Merck [I2752] L-leucine Merck [L8000] L-ornithine hydrochloride Merck [W419001] L-lysine hydrochloride Merck [W384712] L-arginine monohydrochloride Merck [A5131] L-tryptophan Merck [T0254] L-asparagine monohydrate Merck [A8381] L-tyrosine Merck [T3754] L-threonine Merck [T8625] L-cystine dihydrochloride Merck [C6727] L-phenylalanine Merck [P2126] L-histidine monohydrochloride monohydrate Merck [H8125] Sodium phosphate monobasic anhydrous Fisher Scientific [7558-80-7] Sodium phosphate dibasic anhydrous Fisher Scientific [7558-79-4] Potassium nitrate Fisher Scientific [7757-79-1] Potassium sulfate Fisher Scientific [7778-80-5] Ammonium chloride Merck [12125-02-9] Sodium chloride Fisher Scientific [7647-14-5] Potassium chloride Fisher Scientific [7447-40-7] MOPS [3-(N-morpholino)-propanesulfonic acid] Melford [1132-61-2] D-(+)-glucose (dextrose) Merck [G7021] L-(+)-lactic acid Merck [L1750] Calcium chloride dihydrate Merck [10035-04-8] Magnesium chloride hexahydrate Merck [7791-18-6] Iron (II) sulfate heptahydrate Merck [7782-63-0] N-acetyl-D-glucosamine Merck [A8625] Egg yolk emulsion Millipore [17148] Pseudomonas isolation agar Oxoid [CM0559] Mannitol salt agar Oxoid [CM0085] BiGGY-agar Oxoid [CM0589] Agar Formedium [009002-18-0] Lysogeny broth Formedium [LBX0102] Dulbecco's Modified Eagle Medium Merck [D5030] Roswell Park Memorial Institute 1640 Medium Merck [R8758] Page 1 Page 10 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Table 3. Equipment used in this study. was wrapped in aluminium foil and the entire assembly was autoclaved. The setup was allowed to completely dry overnight (in a drying cabinet) before use. Table 3. Equipment used in this study. Protocol for the use of this model Allow the loop to cool then pick a single microbial colony from an agar plate and use to inoculate 10 mL of LB in a sterile 30 mL universal tube. Repeat for all microbial strains to be inoculated into the culture vessel, then place the tubes on a rotating drum (200 rpm) and incubate at 37°C overnight. 7. Step 7 Preparation of EVPL biofilm substratum (day 4). This work wasconducted in collaboration with Dr Freya Harrison at the University of Warwick following their previously published protocol for the preparation of EVPL sections (Harrison et al., 2014) Prior to culture vessel assembly and inoculation (see steps 8-9 below), fresh pig lungs were collected from the butchers and processed within the hour. To remove surface contam- inants, a 25 cm palette knife was heated using a Bunsen burner and briefly tapped (< 1 s) on the area to be dissected. A sterile razor blade was used to excise the bronchiole and remove all alveolar tissue. Bronchioles were then washed in DMEM/RPMI solution (40 mL, 50:50 ratio) and cut into 5 mm wide strips with sterile dissection scissors. The strips were washed in DMEM/RPMI and cut into 5 mm 5 mm squares. EVPL sections 7. Step 7 Preparation of EVPL biofilm substratum (day 4). This work wasconducted in collaboration with Dr Freya Harrison at the University of Warwick following their previously published protocol for the preparation of EVPL sections (Harrison et al., 2014) Prior to culture vessel assembly and inoculation (see steps 8-9 below), fresh pig lungs were collected from the butchers and processed within the hour. To remove surface contam- inants, a 25 cm palette knife was heated using a Bunsen burner and briefly tapped (< 1 s) on the area to be dissected. A sterile razor blade was used to excise the bronchiole and remove all alveolar tissue. Bronchioles were then washed in DMEM/RPMI solution (40 mL, 50:50 ratio) and cut into 5 mm wide strips with sterile dissection scissors. The strips were washed in DMEM/RPMI and cut into 5 mm 5 mm squares. EVPL sections Page 11 of 24 Page 11 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Table 4. Amino acid and salt stock solutions and volumes needed to make 1 L ASM. Protocol for the use of this model The sample was homogenised in a Fast-Prep 5G benchtop homogeniser (MP biomedicals) for 40 s at 4 m s1. Using a pipette, and taking care to avoid any solid pieces of biofilm substratum, the cell supernatant was removed and CFU mL1 counts were determined as described in the Methods section above. Protocol for the use of this model Chemical Mass to add (g) Stock volume (mL) Stock conc (M) Stock to add to beaker (mL) Final Conc (mM) Notes L-serine 0.525 50 0.1 14.46 1.446 - L-glutamic acid hydrochloride 0.918 50 0.1 15.492 1.549 - L-proline 0.576 50 0.1 16.612 1.661 - L-glycine 0.375 50 0.1 12.032 1.203 - L-alanine 0.445 50 0.1 17.8 1.78 - L-valine 0.586 50 0.1 11.172 1.117 - L-methionine 0.746 50 0.1 6.332 0.633 - L-isoleucine 0.656 50 0.1 11.212 1.121 - L-leucine 0.656 50 0.1 16.092 1.609 - L-ornithine hydrochloride 0.843 50 0.1 6.76 0.676 - L-lysine hydrochloride 0.913 50 0.1 21.28 2.128 - L-arginine monohydrochloride 1.054 50 0.1 3.06 0.306 - L-tryptophan 1.021 50 0.1 0.132 0.013 Prep in 0.2 M NaOH L-asparagine monohydrate 0.666 50 0.1 8.272 0.827 Prep in 0.5 M NaOH L-tyrosine 0.906 50 0.1 8.02 0.802 Prep in 1.0 M NaOH Make fresh L-threonine 0.596 50 0.1 10.72 1.072 Make fresh L-cystine dihydrochloride 0.788 50 0.1 1.6 0.16 Make fresh L-phenylalanine 0.826 50 0.1 5.3 0.53 Make fresh L-histidine monohydrochloride monohydrate 1.048 50 0.1 5.192 0.519 Make fresh Sodium phosphate monobasic anhydrous 1.380 50 0.2 8.125 1.3 - Sodium phosphate dibasic anhydrous 1.420 50 0.2 6.252 1.25 - Potassium nitrate 5.056 50 1 0.348 0.348 - Potassium sulfate 2.178 50 0.25 1.084 0.271 - Table 4. Amino acid and salt stock solutions and volumes needed to make 1 L ASM. Table 5. Reagents to make 1 L ASM after addition of the mucin-DNA solution and pH adjustment to pH 6.8. Chemical Mass to add (g) Stock volume (mL) Stock conc (M) Stock to add to beaker (mL) Final Conc (mM) Notes D-(+)-glucose (dextrose) 4.504 25 1 1.2 3 - L-(+)-lactic acid 2.252 25 1 9.3 9.3 pH stock to 7 with NaOH Calcium chloride dihydrate 3.67535 25 1 1.754 1.754 - Magnesium chloride hexahydrate 5.08275 25 1 0.606 0.606 - Iron (II) sulfate heptahydrate 0.05 50 0.0036 1 0.0036 Make fresh N-acetyl-D-glucosamine 1.383 25 0.25 1.2 0.3 - Page 12 of 24 Table 5. Reagents to make 1 L ASM after addition of the mucin-DNA solution and pH adjustment to pH 6.8. Table 5. Reagents to make 1 L ASM after addition of the mucin-DNA solution and pH adjustment to pH 6.8. Protocol for the use of this model Page 12 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 were washed again in DMEM/RPMI, then transferred to a petri dish containing 40 mL ASM and irradiated in UV lightbox for 5 min before being aseptically transferred to the biofilm container (as in Step 8). 8. Step 8 Assembly of the culture vessel (day 4). Working in a microbiological safety cabinet, cut the agar into 5 mm 5 mm cubes using sterile (autoclaved) scalpel/forceps and aseptically transfer these into the wire mesh container so that they sit loose (not stacked) at the bottom of the bucket. The number of substratum pieces used for an experiment may differ, depending on the type of experiment, length of incubation and number of time points to be sampled. In the current work, we simply report the microbial composition of polymicrobial biofilms at two time points (T = 24 h and 96 h). Hence, only six pieces of substratum were incubated in the culture vessel (three for each time point). However, we note that each biofilm container can easily hold 30+ pieces of substratum. Next, aseptically transfer 100 mL of fresh ASM to the culture vessel and secure the lid, ensuring that the biofilm container does not interfere with the magnetic stir bar and is fully submerged in the ASM. The inlet tubing was carefully unwrapped from the foil and placed into the remaining ASM (forming the media reservoir) with the top sealed well with sterile parafilm to prevent contamination. Finally, place the end of the outlet tubing into a discard jar and connect the inlet/outlet tubing to a peristaltic pump. Figure 1 shows a schematic diagram of the assembled continuous-flow culture vessel. The entire system can then be primed with media. The setup was pre-warmed at 37°C prior to microbial inoculation. 8. Step 8 Assembly of the culture vessel (day 4). Working in a microbiological safety cabinet, cut the agar into 5 mm 5 mm cubes using sterile (autoclaved) scalpel/forceps and aseptically transfer these into the wire mesh container so that they sit loose (not stacked) at the bottom of the bucket. The number of substratum pieces used for an experiment may differ, depending on the type of experiment, length of incubation and number of time points to be sampled. Protocol for the use of this model In the current work, we simply report the microbial composition of polymicrobial biofilms at two time points (T = 24 h and 96 h). Hence, only six pieces of substratum were incubated in the culture vessel (three for each time point). However, we note that each biofilm container can easily hold 30+ pieces of substratum. Next, aseptically transfer 100 mL of fresh ASM to the culture vessel and secure the lid, ensuring that the biofilm container does not interfere with the magnetic stir bar and is fully submerged in the ASM. The inlet tubing was carefully unwrapped from the foil and placed into the remaining ASM (forming the media reservoir) with the top sealed well with sterile parafilm to prevent contamination. Finally, place the end of the outlet tubing into a discard jar and connect the inlet/outlet tubing to a peristaltic pump. Figure 1 shows a schematic diagram of the assembled continuous-flow culture vessel. The entire system can then be primed with media. The setup was pre-warmed at 37°C prior to microbial inoculation. 9. Step 9 Culture vessel inoculation and incubation (day 4). In this report we describe the formation of polymicrobial biofilms of Pseudomonas aeruginosa (PA), Staphylococcus aureus (SA) and Candida albicans (CA), although in principle, any combination of microbial species could be used for co-culture. Overnight cultures (prepared on day 3) of each species were washed three times in sterile PBS prior to inoculation. Once washed, the optical density (OD600) of the cultures was measured and each species was introduced into the culture vessel to achieve a starting OD600 of 0.05. The culture vessel was incubated at 37°C, with 100 rpm stirring using a round PTFE magnetic stir bar (8 mm diameter 20 mm length), for three hours prior to starting the flow of medium. For the current study, a continuous flow rate of 145 μL min1 was applied for the remaining period of incubation (96 h). From our previous study (O'Brien and Welch, 2019a), we note that the flow rate may need to be experimentally optimised if the goal is to maintain steady-state cultures of other species or strains. 10. Step 10 Sample removal and viable cell counts (Day 4–9). Protocol for the use of this model At the desired time point(s) (in this study, T = 24 h and 96 h for the biofilms and T = 24, 48, 72 and 96 h for the planktonic samples) remove 1 mL aliquots of the liquid culture using a sterile serological pipette, or pieces of the biofilm substratum from the biofilm bucket, as necessary. For enumeration of colony forming units (CFU mL1) from the biofilm samples, remove pieces of solid substratum using sterile forceps, taking care to flame and cool the forceps between the collection of individual samples. Pieces of the substratum were aseptically transferred to 500 μL sterile PBS in a 24-well plate. To remove any loosely attached planktonic cells, gently swirl the plates for approximately 5 s then transfer the substratum to another 500 μL of fresh PBS. Repeat the process two more times. A washed piece of substratum was then transferred to a 2 mL bead beating tube containing 1 mL PBS and eighteen 2.38 mm diameter metal beads (Qiagen). The sample was homogenised in a Fast-Prep 5G benchtop homogeniser (MP biomedicals) for 40 s at 4 m s1. Using a pipette, and taking care to avoid any solid pieces of biofilm substratum, the cell supernatant was removed and CFU mL1 counts were determined as described in the Methods section above. 10. Step 10 Sample removal and viable cell counts (Day 4–9). At the desired time point(s) (in this study, T = 24 h and 96 h for the biofilms and T = 24, 48, 72 and 96 h for the planktonic samples) remove 1 mL aliquots of the liquid culture using a sterile serological pipette, or pieces of the biofilm substratum from the biofilm bucket, as necessary. For enumeration of colony forming units (CFU mL1) from the biofilm samples, remove pieces of solid substratum using sterile forceps, taking care to flame and cool the forceps between the collection of individual samples. Pieces of the substratum were aseptically transferred to 500 μL sterile PBS in a 24-well plate. To remove any loosely attached planktonic cells, gently swirl the plates for approximately 5 s then transfer the substratum to another 500 μL of fresh PBS. Repeat the process two more times. A washed piece of substratum was then transferred to a 2 mL bead beating tube containing 1 mL PBS and eighteen 2.38 mm diameter metal beads (Qiagen). Results Characterisation of the biofilms formed in continuous flow conditions (i) Agar as a solid substratum We first sought to discern changes in the microbial composition of single-species and polymicrobial biofilms grown on agar chunks as a substratum (Figure 5 and underlying data (O'Brien et al., 2021)). Viable cell counts were determined on selective agar plates after rinsing each 5 mm agar chunk (with attached biofilm) and resuspending the attached cells in 1 mL PBS. When examining the single-species biofilm populations, titres of attached SA and CA were similar (ca. 106–107 CFU mL1) after 24 h incubation, whereas titres of attached PA were >10-fold greater (at just under 108 CFU mL1). However, over the next 72 h, PA titres increased only slightly (albeit significantly) to >108 CFU mL1, and SA titres rose to around the same level (i.e. to a titre 2 logs greater than the titre at the 24 h time point). By contrast, CA titres remained essentially unchanged at the 96 h time point compared with the 24 h time point. These data suggest CA biofilms establish rapidly with little overall change in cell titres over time. By contrast, PA and SA biofilms also establish quickly, but continue to grow significantly between sampling points. Page 13 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Figure 5. Cell titres derived from biofilms adhered to agar chunks. P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) cell counts (expressed as CFU mL-1) adhered to 2.5% (w/v) agar chunks after incubation for 24 h (black bars) or 96 h (white bars) in the continuous-flow culture model. Asterisks indicate significant (P < 0.01, P < 0.001) differences in CFU mL-1 counts between the 24 h and 96 h time points. Data represent the mean standard deviation of three independent biological experiments in which three pieces of biofilm substratum were sampled per culture vessel per timepoint. Figure 5. Cell titres derived from biofilms adhered to agar chunks. P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) cell counts (expressed as CFU mL-1) adhered to 2.5% (w/v) agar chunks after incubation for 24 h (black bars) or 96 h (white bars) in the continuous-flow culture model. Asterisks indicate significant (P < 0.01, P < 0.001) differences in CFU mL-1 counts between the 24 h and 96 h time points. Results Characterisation of the biofilms formed in continuous flow conditions (i) Agar as a solid substratum Data represent the mean standard deviation of three independent biological experiments in which three pieces of biofilm substratum were sampled per culture vessel per timepoint. The situation was different in the polymicrobial cultures. Here, the titres of PA, SA, and CA present on the biofilm substrata removed sequentially from the same culture vessel remained essentially unchanged at the 24 h and 96 h sampling points (at around 107 CFU mL1 for PA and SA, and around 105 CFU mL1 for CA). These data suggest that growth in a polymicrobial culture constrains the population size of individual species in a biofilm. (ii) EVPL tissue sections as a solid substratum We next examined biofilm formation on a different biofilm substratum, EVPL tissue (in place of agar chunks) (Figure 6 and underlying data (O'Brien et al., 2021)). No PA, SA or CA cells could be isolated from uninfected EVPL sections Figure 6. Cell titres derived from biofilms adhered to porcine bronchiole tissue. P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) cell counts (expressed as CFU mL1) adhered to sections of ex vivo porcine lung tissue after incubation for 24 h (black bars) or 96 h (white bars) in the continuous-flow culture model. Asterisks indicate significant (P < 0.001, *P < 0.0001) differences in CFU mL1 counts between the 24 h and 96 h time points. Data represent the mean standard deviation of three independent biological experiments in which three pieces of biofilm substratum were sampled per culture vessel per timepoint. Figure 6. Cell titres derived from biofilms adhered to porcine bronchiole tissue. P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) cell counts (expressed as CFU mL1) adhered to sections of ex vivo porcine lung tissue after incubation for 24 h (black bars) or 96 h (white bars) in the continuous-flow culture model. Asterisks indicate significant (P < 0.001, ***P < 0.0001) differences in CFU mL1 counts between the 24 h and 96 h time points. Data represent the mean standard deviation of three independent biological experiments in which three pieces of biofilm substratum were sampled per culture vessel per timepoint. Page 14 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Figure 7. Cell titres of planktonic cultures. P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. Results Characterisation of the biofilms formed in continuous flow conditions (i) Agar as a solid substratum albicans SC5314 (CA) cell counts (expressed as CFU mL1) in the planktonic fraction of single-species and polymicrobial cultures incubated in ASM in the continuous-flow model containing (A) agar chunks or (B) ex vivo porcine lung tissue sections as the solid substrata. Data represent the mean standard deviation of two technical replicates collected per timepoint from three independent biological experiments. P > 0.05 is considered as not significantly different (ns). Figure 7. Cell titres of planktonic cultures. P. aeruginosa PAO1 (PA), S. aureus 25923 (SA) and C. albicans SC5314 (CA) cell counts (expressed as CFU mL1) in the planktonic fraction of single-species and polymicrobial cultures incubated in ASM in the continuous-flow model containing (A) agar chunks or (B) ex vivo porcine lung tissue sections as the solid substrata. Data represent the mean standard deviation of two technical replicates collected per timepoint from three independent biological experiments. P > 0.05 is considered as not significantly different (ns). after any period of incubation. In the single-species biofilm populations, there was no appreciable difference in PA or SA titres adhered to the substratum at the initial point of sampling (107–108 CFU mL1 at 24 h). Adhered CA titres were consistently 10-fold lower (ca. 106 CFU mL1) at the same sampling point. After a further 72 h growth, PA and SA titres increased to >108 CFU mL1, whereas CA titres remained essentially unchanged compared with the titres measured at 24 h. These data indicate that the dynamics of SA biofilm formation varies on different substrata; SA clearly shows more robust initial colonisation of EVPL substrata compared with agar chunks. In the triple-species polymicrobial culture, there was no statistically significant difference in adhered PA or SA titres at the 24 h sampling point (both species achieving 106–107 CFU mL1 at 24 h). We note that this is 10-fold lower than the PA and SA titres at the same time point in the corresponding biofilms from axenic cultures. Similarly, CA titres in the polymicrobial biofilms were also 10-fold lower than the CA titres in axenically-cultured biofilms. However, and whereas adhered CA titres in the polymicrobial biofilms remained unchanged over the following 72 h, PA titres displayed a marked increase (attaining ca. 108 CFU mL1 by the 96 h time point) and SA titres displayed a marked decrease (falling to ca. 105 CFU mL1) in adhered cell counts. Results Characterisation of the biofilms formed in continuous flow conditions (i) Agar as a solid substratum We conclude that the population dynamics of each species can vary by orders of magnitude, depending on the nature of the biofilm substrata. Population structure of the planktonic fractions In parallel with the analysis of species titres in adhered biofilms, we also examined the corresponding titres of each species in the planktonic (i.e. non-attached) fraction (Figure 7 and underlying data (O'Brien et al., 2021)). Consistent with our previous findings (O'Brien and Welch, 2019a), steady-state microbial communities were established by 24 h incubation. There was no appreciable change in CFU mL1 counts of any species cultured as part of an axenic or polymicrobial population across any point of sampling. Furthermore, there was no discernible difference in CFU mL1 counts of the planktonic communities grown in the culture vessel flasks containing agar chunks (Figure 7A) or EVPL tissue sections (Figure 7B). These data indicate that the population dynamics of the biofilms is distinct from the population dynamics of the surrounding planktonic culture. Discussion In particular, we note that the species profiles in polymicrobial biofilms are consistent with both antagonistic and synergistic interactions. For example, irrespective of the substrate, adhered CA titres were lower in the polymicrobial biofilms compared with CA titres in biofilms from axenic CA cultures. The primary limitation of in vitro models of infection is a lack of spatial organisation and exclusion of host cells present when utilising in vivo infection models. We therefore attempted to redress this issue through the introduction of EVPL tissue sections as biofilm substrata. Porcine airways share a remarkable degree of structural homology with human airways (Judge et al., 2014; Rogers et al., 2008), and pig lungs are readily available from most butchers as a by-product of the food industry. Hence, the inclusion of ex vivo bronchiole tissue sections provide an ethically sustainable approach to introduce an element of host-microbe interaction and spatial organisation to the model (Harrington et al., 2020; Sweeney et al., 2021). Indeed, we note that growth on different solid substrata had a profound effect on the compositional dynamics of polymicrobial biofilms, especially with regards to S. aureus. SA is an effective early coloniser of the CF airways (Bogaert et al., 2004). When grown on EVPL sections, SA rapidly established a mono-species biofilm with significantly higher titres compared with the biofilm formed on agar chunks. This increased microbial attachment is consistent with the notion that SA adheres tightly to epithelial cells in the CF microenvironment (Josse et al., 2017), but is less effective at adhering to abiotic surfaces (agar chunks). However, and whereas SA titres on the agar chunk substrata were stable between the 24 h and 96 h sampling points, the SA titres on EVPL sections displayed a substantial decline over this period. By contrast, PA titres in the same biofilms increased, suggesting that the PA progressively displaces SA on the tissue surface. This reciprocal response (in terms of cell titres) by PA and SA was not observed when agar chunks were used as a substratum, suggesting that recognition of both airway epithelial cells and SA are required to drive the increased competitiveness by PA. Despite the introduction of some level of spatial organisation (in the form of EVPL tissue), the major physiological limitation of our model compared with in vivo systems is a lack of live host cells and the absence of any accompanying host-microbe interactions that may modulate microbial behaviour. Discussion The primary benefit of the model is that very stable polymicrobial communities can be rapidly established and maintained for extended periods of time, and that the planktonic and biofilm modes of growth can be studied concurrently in a single experiment. This enables the facile longitudinal study of polymicrobial communities in vitro, enabling experimental analyses that were previously not possible due to population instability. As such, the model allows researchers to explore previously inaccessible problems pertaining to microbial ecology, gene expression and metabolism in polymicrobial communities. For example, we are currently exploiting the system to examine how mixed-species biofilms such as those found in the CF airways form over time. We are also using the model to examine how the inclusion of different species or treatment with antimicrobial compounds affect the stability and dynamics of polymicrobial biofilms longitudinally. Alongside the ethical benefits and accessibility of an in vitro model for microbial culture, the experimentally tuneable nature of our model system provides several inherent benefits over existing in vivo CF models (O'Brien and Welch, 2019b). ASM is a chemically-defined synthetic growth medium that has been formulated to closely mimic the nutritional composition of CF airway secretions (Kirchner et al., 2012; Palmer et al., 2005; Palmer et al., 2007; Turner et al., 2015; Cornforth et al., 2018). This allows subtle interspecies interactions to be quantified, in real-time, with great reproduc- ibility. By contrast, the existing animal models display much greater variability, even within a single host species. Furthermore, culture conditions can be readily modified to enable the detailed study of how a particular variable (such as iron, for example) impinges upon the formation of polymicrobial biofilms (Mashburn et al., 2005; Palmer et al., 2005). This level of environmental control is near impossible to achieve with in vivo infection models. We also note that the experimentally facile nature of our system should permit the study of biofilm formation in other clinically-relevant polymicrobial infection scenarios such as non-CF bronchiectasis, asthma, or chronic obstructive pulmonary disorders (COPDs). This can be accomplished through the simple expedient of appropriately modifying the nutritional composition of the culture medium. The key finding in this report is that mixed-species biofilms comprised of three very different CF-associated pathogens (a Gram-positive species (SA), a Gram-negative species (PA), and a fungal species (CA)) display distinctly different compositional dynamics compared with their planktonic counterparts. Discussion Here we report a simple method for the robust and reproducible growth of single- and multi-species biofilms on different substrata. Our model system has been designed to mimic the nutritional environment associated with CF airways. We previously showed (O'Brien and Welch, 2019a) that the model can maintain very stable steady-state polymicrobial populations of planktonic cells, even among species that would ordinarily outcompete one another during ex situ co-cultivation. In the current work, we extend these findings to show that polymicrobial biofilms can be similarly maintained. Remarkably, we also show that the biofilm composition is far more dynamic than that of the surrounding planktonic culture, and can vary substantially, even when the planktonic species profile is stable. A major benefit of our model over existing in vivo approaches is that is inexpensive to set up and requires no specialist equipment or training to operate. Indeed, the only perceived barrier preventing wider uptake of the model is in preparation Page 15 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 of the ASM culture medium. Mitigating, this generally requires only a few minutes of effort each day (once the stock solutions have been made up). The primary benefit of the model is that very stable polymicrobial communities can be rapidly established and maintained for extended periods of time, and that the planktonic and biofilm modes of growth can be studied concurrently in a single experiment. This enables the facile longitudinal study of polymicrobial communities in vitro, enabling experimental analyses that were previously not possible due to population instability. As such, the model allows researchers to explore previously inaccessible problems pertaining to microbial ecology, gene expression and metabolism in polymicrobial communities. For example, we are currently exploiting the system to examine how mixed-species biofilms such as those found in the CF airways form over time. We are also using the model to examine how the inclusion of different species or treatment with antimicrobial compounds affect the stability and dynamics of polymicrobial biofilms longitudinally. of the ASM culture medium. Mitigating, this generally requires only a few minutes of effort each day (once the stock solutions have been made up). Discussion Nevertheless, and through our comparison of EVPL tissue and agar chunks as biofilm substrata, we note two things. First, the diminution in SA titres in polymicrobial biofilms after 96 h growth on EVPL tissue – a diminution not seen on agar chunk substrata – suggests that microbial interaction with the host tissue may up-regulate the localised production of virulence factors and enhance interspecies competition (Döring et al., 2011; Malhotra et al., 2019). The implication of this is that the growth of EVPL-associated biofilms in our continuous-flow model may enable the capture specific temporally-sensitive interactions between the host tissue and Page 16 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 infecting microbes. Second, agar is a more suitable substratum for examining the long-term, steady-state growth of polymicrobial biofilms. infecting microbes. Second, agar is a more suitable substratum for examining the long-term, steady-state growth of polymicrobial biofilms. The ability to maintain very stable (in terms of composition) long-term polymicrobial biofilms in our setup using agar chunks as a solid substratum can be exploited for several other applications. First and most obviously, the setup can be used to investigate the species specificity (for example) of novel antimicrobial or anti-biofilm agents. Using our system, it is possible to test both a sustained treatment regimen (achieved by adding the test compound(s) at the desired concentration to the media reservoir) or a short-term treatment regimen, via adding a compound directly to the culture vessel through the injection ports. The latter may loosely mimic the metabolism and excretion of antimicrobials in situ. This degree of temporal control is not possible using existing in vivo or in vitro models. Second, new species or strain variants can be introduced to pre-established polymicrobial biofilms (e.g. addition of PA to a PA-negative community), allowing facile examination of the impact made by new species or variants. This is directly analogous to the situation seen in CF, where the acquisition of new microbial species or variants can lead to major prognostic changes in the patient. It is important to note that whenever a new species is introduced, the medium flowrate (Q) may need to be re-optimized. If Q is too high, slower-growing organisms may be “washed out” and lost from the culture vessel. Discussion Conversely, if Q is too low, species will grow faster than the rate of media displacement, causing elements of the population to exhaust key nutrients and enter the stationary phase of growth. Underlying data Figshare: Underlying data for ‘An in vitro model for the cultivation of polymicrobial biofilms under continuous-flow conditions’. https://doi.org/10.6084/m9.figshare.14974479.v1 (O'Brien et al., 2021). This project contains the following underlying data: • EVPL Biofilms CFUs • EVPL Supernatant CFUs • Agar chunks Biofilm CFUs • Agar chunks Supernatant CFUs • Selective vs non-selective agar Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0 Armbruster CR, Wolter DJ, Mishra M, et al.: Staphylococcus aureus Protein A Mediates Interspecies Interactions at the Cell Surface of Pseudomonas aeruginosa. mBio. 2016; 7. PubMed Abstract\|Publisher Full Text\|Free Full Text Baldan R, Cigana C, Testa F, et al.: Adaptation of Pseudomonas aeruginosa in Cystic Fibrosis airways influences virulence of Staphylococcus aureus in vitro and murine models of co-infection. PLoS One. 2014; 9: e89614. PubMed Abstract\|Publisher Full Text\|Free Full Text Conclusions In summary, we present here a simple system for the study of CF-associated polymicrobial biofilms. Importantly, we have shown that polymicrobial biofilms display distinctly different population dynamics compared with the surrounding planktonic cells. Moreover, the compositional dynamics of the biofilms depends very much on the substratum employed. To further extend the utility of our model, we are currently adapting it for use with a wider range of CF-associated pathogens and testing the possibility of inoculating the system directly with CF patient-derived sputum. Ahmed B, Cox MJ, Cuthbertson L, et al.: Longitudinal development of the airway microbiota in infants with cystic fibrosis. Sci Rep. 2019; 9: 5143. PubMed Abstract\|Publisher Full Text\|Free Full Text Antonic V, Stojadinovic A, Zhang B, et al.: Pseudomonas aeruginosa induces pigment production and enhances virulence in a white phenotypic variant of Staphylococcus aureus. Infect Drug Resist. 2013; 6: 175–186. PubMed Abstract\|Publisher Full Text\|Free Full Text Data availability y Underlying data PubMed Abstract\|Publisher Full Text Carmody LA, Zhao J, Kalikin LM, et al.: The daily dynamics of cystic fibrosis airway microbiota during clinical stability and at exacerbation. Microbiome. 2015; 3: 12. PubMed Abstract\|Publisher Full Text\|Free Full Text Hoffmann N, Rasmussen TB, Jensen P, et al.: Novel mouse model of chronic Pseudomonas aeruginosa lung infection mimicking cystic fibrosis. Infect Immun. 2005; 73: 2504–2514. PubMed Abstract\|Publisher Full Text\|Free Full Text Cash HA, Woods DE, Mccullough B, et al.: A rat model of chronic respiratory infection with Pseudomonas aeruginosa. Am Rev Respir Dis. 1979; 119: 453–459. Hogan DA, Willger SD, Dolben EL, et al.: Analysis of Lung Microbiota in Bronchoalveolar Lavage, Protected Brush and Sputum Samples from Subjects with Mild-To-Moderate Cystic Fibrosis Lung Disease. PLoS One. 2016; 11: e0149998. 124 146. PubMed Abstract\|Publisher Full Text 6. PubMed Abstract\|Publisher Full Text Elborn JS: Cystic fibrosis. Lancet. 2016; 388: 2519–2531. PubMed Abstract\|Publisher Full Text Elias S, Banin E: Multi-species biofilms: living with friendly neighbors. FEMS Microbiol Rev. 2012; 36: 990–1004. PubMed Abstract\|Publisher Full Text Kirchner S, Fothergill JL, Wright EA, et al.: Use of artificial sputum medium to test antibiotic efficacy against Pseudomonas aeruginosa in conditions more relevant to the cystic fibrosis lung. J Vis Exp. 2012: e3857. PubMed Abstract\|Publisher Full Text Folkesson A, Jelsbak L, Yang L, et al.: Adaptation of Pseudomonas aeruginosa to the cystic fibrosis airway: an evolutionary perspective. Nat Rev Microbiol. 2012; 10: 841–851. PubMed Abstract\|Publisher Full Text Frapwell CJ, Howlin RP, Soren O, et al.: Increased rates of genomic mutation in a biofilm co-culture model of Pseudomonas aeruginosa and Staphylococcus aureus. BioRx. 2018. Publisher Full Text Fux CA, Costerton JW, Stewart PS, et al.: Survival strategies of infectious biofilms. Trends Microbiol. 2005; 13: 34–40. PubMed Abstract\|Publisher Full Text Folkesson A, Jelsbak L, Yang L, et al.: Adaptation of Pseudomonas aeruginosa to the cystic fibrosis airway: an evolutionary perspective. Nat Rev Microbiol. 2012; 10: 841–851. PubMed Abstract\|Publisher Full Text PubMed Abstract\|Publisher Full Text\|Free Full Text Korgaonkar A, Trivedi U, Rumbaugh KP, et al.: Community surveillance enhances Pseudomonas aeruginosa virulence during polymicrobial infection. Proc Natl Acad Sci U S A. 2013; 110: 1059–1064. PubMed Abstract\|Publisher Full Text\|Free Full Text Korgaonkar A, Trivedi U, Rumbaugh KP, et al.: Community surveillance enhances Pseudomonas aeruginosa virulence during polymicrobial infection. Proc Natl Acad Sci U S A. 2013; 110: 1059–1064. PubMed Abstract\|Publisher Full Text\|Free Full Text Leekha S, Terrell CL, Edson RS: General principles of antimicrobial therapy. Mayo Clin Proc. 2011; 86: 156–167. PubMed Abstract\|Publisher Full Text\|Free Full Text Leenaars CH, De Vries RB, Heming A, et al.: Animal models for cystic fibrosis: A systematic search and mapping review of the literature - References Page 17 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Gillum AM, Tsay EY, Kirsch DR: Isolation of the Candida albicans gene for orotidine-5'-phosphate decarboxylase by complementation of S. cerevisiae ura3 and E. coli pyrF mutations. Mol Gen Genet. 1984; 198: 179–182. Haley CL, Colmer-Hamood JA, Hamood AN: Characterization of biofilm- like structures formed by Pseudomonas aeruginosa in a synthetic mucus medium. BMC Microbiol. 2012; 12: 181. PubMed Abstract\|Publisher Full Text\|Free Full Text Harrington NE, Sweeney E, Alav I, et al.: Antibiotic Efficacy Testing in an ex vivo Model of Pseudomonas aeruginosa and Staphylococcus aureus Biofilms in the Cystic Fibrosis Lung. J Vis Exp. 2021. PubMed Abstract\|Publisher Full Text Bjarnsholt T, Jensen P, Fiandaca MJ, et al.: Pseudomonas aeruginosa biofilms in the respiratory tract of cystic fibrosis patients. Pediatr Pulmonol. 2009; 44: 547–558. ; PubMed Abstract\|Publisher Full Text PubMed Abstract\|Publisher Full Text Bogaert D, Van Belkum A, Sluijter M, et al.: Colonisation by Streptococcus pneumoniae and Staphylococcus aureus in healthy children. Lancet. 2004; 363: 1871–1872. ; PubMed Abstract\|Publisher Full Text Harrison F, Diggle SP: An ex vivo lung model to study bronchioles infected with Pseudomonas aeruginosa biofilms. Microbiology. 2016; 162: 1755–1760. Bonfield T, Chmiel JF: Impaired innate immune cells in cystic fibrosis: Is it really a surprise? J Cyst Fibros. 2017; 16: 433–435. PubMed Abstract\|Publisher Full Text y p J y PubMed Abstract\|Publisher Full Text Harrison F, Muruli A, Higgins S, et al.: Development of an ex vivo porcine lung model for studying growth, virulence, and signaling of Pseudomonas aeruginosa. Infect Immun. 2014; 82: 3312–3323. PubMed Abstract\|Publisher Full Text\|Free Full Text Boutin S, Graeber SY, Weitnauer M, et al.: Comparison of microbiomes from different niches of upper and lower airways in children and adolescents with cystic fibrosis. PLoS One. 2015; 10: e0116029. PubMed Abstract\|Publisher Full Text\|Free Full Text Hibbing ME, Fuqua C, Parsek MR, et al.: Bacterial competition: surviving and thriving in the microbial jungle. Nat Rev Microbiol. 2010; 8: 15–25. PubMed Abstract\|Publisher Full Text\|Free Full Text Briaud P, Camus L, Bastien S, et al.: Coexistence with Pseudomonas aeruginosa alters Staphylococcus aureus transcriptome, antibiotic resistance and internalization into epithelial cells. Sci Rep. 2019; 9: 16564. Hoffman LR, Hajjar AM: Ferreting Out the Role of Infection in Cystic Fibrosis Lung Disease. Am J Respir Crit Care Med. 2018; 197: 1243–1244. 6564. PubMed Abstract\|Publisher Full Text\|Free Full Text PubMed Abstract\|Publisher Full Text\|Free Full Text 1979; 119: 453 459. PubMed Abstract\|Publisher Full Text Chattoraj SS, Murthy R, Ganesan S, et al.: Pseudomonas aeruginosa alginate promotes Burkholderia cenocepacia persistence in cystic fibrosis transmembrane conductance regulator knockout mice. Infect Immun. 2010; 78: 984–993. PubMed Abstract\|Publisher Full Text\|Free Full Text Holloway BW: Genetic recombination in Pseudomonas aeruginosa. J Gen Microbiol. 1955; 13: 572–581. ; PubMed Abstract\|Publisher Full Text\|Free Full Text Chmiel JF, Davis PB: State of the art: why do the lungs of patients with cystic fibrosis become infected and why can't they clear the infection? Respir Res. 2003; 4: 8. Hotterbeekx A, Kumar-Singh S, Goossens H, et al.: In vivo and In vitro Interactions between Pseudomonas aeruginosa and Staphylococcus spp. Front Cell Infect Microbiol. 2017; 7: 106. PubMed Abstract\|Publisher Full Text\|Free Full Text p PubMed Abstract\|Publisher Full Text\|Free Full Text Cohen TS, Prince A: Cystic fibrosis: a mucosal immunodeficiency syndrome. Nat Med. 2012; 18: 509–519. PubMed Abstract\|Publisher Full Text\|Free Full Text Cohen TS, Prince A: Cystic fibrosis: a mucosal immunodeficiency syndrome. Nat Med. 2012; 18: 509–519. Ibberson CB, Stacy A, Fleming D, et al.: Co-infecting microorganisms dramatically alter pathogen gene essentiality during polymicrobial infection. Nat Microbiol. 2017; 2: 17079. PubMed Abstract\|Publisher Full Text\|Free Full Text Cornforth DM, Dees JL, Ibberson CB, et al.: transcriptome during human infection. Proc Natl Acad Sci U S A. 2018; 115: E5125–E5134. Ibberson CB, Whiteley M: The social life of microbes in chronic infection. Curr Opin Microbiol. 2020; 53: 44–50. PubMed Abstract\|Publisher Full Text\|Free Full Text Dalton T, Dowd SE, Wolcott RD, et al.: An in vivo polymicrobial biofilm wound infection model to study interspecies interactions. PLoS One. 2011; 6: e27317. Jorth P, Ehsan Z, Rezayat A, et al.: Direct Lung Sampling Indicates That Established Pathogens Dominate Early Infections in Children with Cystic Fibrosis. Cell Rep. 2019; 27: 1190–1204.e3. P bM d Ab t t\|P bli h F ll T t\|F F ll T t ; PubMed Abstract\|Publisher Full Text\|Free Full Text ; PubMed Abstract\|Publisher Full Text\|Free Full Text Diggle SP, Griffin AS, Campbell GS, et al.: Cooperation and conflict in quorum-sensing bacterial populations. Nature. 2007; 450: 411–414. PubMed Abstract\|Publisher Full Text Josse J, Laurent F, Diot A: Staphylococcal Adhesion and Host Cell Invasion: Fibronectin-Binding and Other Mechanisms. Front Microbiol. 2017; 8: 2433. DÖRing G, Parameswaran IG, Murphy TF: Differential adaptation of microbial pathogens to airways of patients with cystic fibrosis and chronic obstructive pulmonary disease. FEMS Microbiol Rev. 2011; 35: 124–146. 2017; 8: 2433. PubMed Abstract\|Publisher Full Text\|Free Full Text Judge EP, Hughes JM, Egan JJ, et al.: Anatomy and bronchoscopy of the porcine lung. A model for translational respiratory medicine. Am J Respir Cell Mol Biol. 2014; 51: 334–343. PubMed Abstract\|Publisher Full Text PubMed Abstract\|Publisher Full Text\|Free Full Text PubMed Abstract\|Publisher Full Text\|Free Full Text Mastropaolo MD, Evans NP, Byrnes MK, et al.: Synergy in polymicrobial infections in a mouse model of type 2 diabetes. Infect Immun. 2005; 73: 6055–6063. Sun X, Sui H, Fisher JT, et al.: Disease phenotype of a ferret CFTR-knockout model of cystic fibrosis. J Clin Invest. 2010; 120: 3149–3160. 6055 6063. PubMed Abstract\|Publisher Full Text\|Free Full Text Moser C, Van Gennip M, Bjarnsholt T, et al.: Novel experimental Pseudomonas aeruginosa lung infection model mimicking long- term host-pathogen interactions in cystic fibrosis. APMIS. 2009; 117: 95–107. Sun X, Yi Y, Yan Z, et al.: In utero and postnatal VX-770 administration rescues multiorgan disease in a ferret model of cystic fibrosis. Sci Transl Med. 2019; 11. Mowat E, Paterson S, Fothergill JL, et al.: Pseudomonas aeruginosa population diversity and turnover in cystic fibrosis chronic infections. Am J Respir Crit Care Med. 2011; 183: 1674–1679. PubMed Abstract\|Publisher Full Text Sweeney E, Harrington NE, Harley Henriques AG, et al.: An ex vivo cystic fibrosis model recapitulates key clinical aspects of chronic Staphylococcus aureus infection. Microbiology (Reading). 2021; 167. PubMed Abstract\|Publisher Full Text Munder A, Wölbeling F, Kerber-Momot T, et al.: Acute intratracheal Pseudomonas aeruginosa infection in cystic fibrosis mice is age-independent. Respir Res. 2011; 12: 148. PubMed Abstract\|Publisher Full Text\|Free Full Text O’Brien S, Fothergill JL: The role of multispecies social interactions in shaping Pseudomonas aeruginosa pathogenicity in the cystic fibrosis lung. FEMS Microbiol Lett. 2017; 364. O'Brien TJ, Hassan MM, Harrison F, et al.: An in vitro model for the cultivation of polymicrobial biofilms under continuous-flow conditions. Figshare. 2021. Turner KH, Wessel AK, Palmer GC, et al.: Essential genome of Pseudomonas aeruginosa in cystic fibrosis sputum. Proc Natl Acad Sci U S A. 2015; 112: 4110–4115. ; PubMed Abstract\|Publisher Full Text\|Free Full Text \| \| Palmer KL, Mashburn LM, Singh PK, et al.: Cystic fibrosis sputum supports growth and cues key aspects of Pseudomonas aeruginosa physiology. J Bacteriol. 2005; 187: 5267–5277. PubMed Abstract\|Publisher Full Text\|Free Full Text Parsek MRASPK: Bacterial Biofilms: An Emerging Link to Disease Pathogenesis. Annu Rev Microbiol. 2003; 57: 677–701. PubMed Abstract\|Publisher Full Text Peters BM, Jabra-Rizk MA, O'May GA, et al.: Polymicrobial interactions: impact on pathogenesis and human disease. Clin Microbiol Rev. 2012; Palmer KL, Mashburn LM, Singh PK, et al.: Cystic fibrosis sputum supports growth and cues key aspects of Pseudomonas aeruginosa physiology. J Bacteriol. 2005; 187: 5267–5277. PubMed Abstract\|Publisher Full Text\|Free Full Text PubMed Abstract\|Publisher Full Text Frapwell CJ, Howlin RP, Soren O, et al.: Increased rates of genomic mutation in a biofilm co-culture model of Pseudomonas aeruginosa and Staphylococcus aureus. BioRx. 2018. Publisher Full Text Leekha S, Terrell CL, Edson RS: General principles of antimicrobial therapy. Mayo Clin Proc. 2011; 86: 156–167. PubMed Abstract\|Publisher Full Text\|Free Full Text Fux CA, Costerton JW, Stewart PS, et al.: Survival strategies of infectious biofilms. Trends Microbiol. 2005; 13: 34–40. PubMed Abstract\|Publisher Full Text Leenaars CH, De Vries RB, Heming A, et al.: Animal models for cystic fibrosis: A systematic search and mapping review of the literature - Page 18 of 24 Page 18 of 24 Page 18 of 24 Page 18 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Lopes SP, Azevedo NF, Pereira MO: Developing a model for cystic fibrosis sociomicrobiology based on antibiotic and environmental stress. Int J Med Microbiol. 2017; 307: 460–470. PubMed Abstract\|Publisher Full Text ; PubMed Abstract\|Publisher Full Text Palmer KL, Aye LM, Whiteley M: Nutritional cues control Pseudomonas aeruginosa multicellular behavior in cystic fibrosis sputum. J Bacteriol. 2007; 189: 8079–8087. PubMed Abstract\|Publisher Full Text Sibley CD, Surette MG: The polymicrobial nature of airway infections in cystic fibrosis: Cangene Gold Medal Lecture. Can J Microbiol. 2011; 57: 69–77. Markussen T, Marvig RL, Gómez-Lozano M, et al.: Environmental heterogeneity drives within-host diversification and evolution of Pseudomonas aeruginosa. mBio. 2014; 5: e01592–e01514. PubMed Abstract\|Publisher Full Text\|Free Full Text Mashburn LM, Jett AM, Akins DR, et al.: Staphylococcus aureus serves as an iron source for Pseudomonas aeruginosa during in vivo coculture. J Bacteriol. 2005; 187: 554–566. Sriramulu DD, Lünsdorf H, Lam JS, et al.: Microcolony formation: a novel biofilm model of Pseudomonas aeruginosa for the cystic fibrosis lung. J Med Microbiol. 2005; 54: 667–676. g Publisher Full Text O'Brien TJ, Welch M: A Continuous-Flow Model for in vitro Cultivation of Van Heeckeren AM, Schluchter MD, Xue W, et al.: Response to acute lung infection with mucoid Pseudomonas aeruginosa in cystic fibrosis mice. Am J Respir Crit Care Med. 2006; 173: 288–296. PubMed Abstract\|Publisher Full Text\|Free Full Text O'Brien TJ, Welch M: Recapitulation of polymicrobial communities associated with cystic fibrosis airway infections: a perspective. Future Microbiol. 2019b; 14: 1437–1450. P bM d Ab \|P bli h F ll T Lindsay R. Kalan Department of Medical Microbiology and Immunology, School of Medicine and Public Health, University of Wisconsin-Madison, Madison, WI, USA The manuscript by O'brien et al. describes the development of an in vitro model for the study of polymicrobial biofilm, specifically focused on the CF lung environment. This work is timely as the importance of polymicrobial colonization, biofilm formation, and infection is increasingly recognized as important towards disease outcome. Here, the authors use artificial sputum in a continuous flow culture vessel to cultivate biofilms. They examined biofilm formation using two different substrates, agar or ex vivo porcine lung tissues. This allows the user to evaluate both biofilm growth on the substrate and measure planktonic cell populations in the surrounding media. The manuscript is well written, and the photographs of the apparatus are helpful to the reader who wishes to set this up in their own laboratory. The protocol provided is detailed and thorough although the production of ASM media seems cumbersome with some steps requiring first hand know-how that may introduce variability. For example, the filtration step is stated to take up to 2-3 days with the membrane being rinsed periodically, it is unclear how this might introduce contamination or impact the final media composition. I have no major comments, minor comments follow: I have no major comments, minor comments follow: The protocol states that each species in inoculated at an OD600 of 0.05. This could vary in the final cfu/mL depending on the species, especially fungi. It may be more informative to state the desired final inoculum in cfu/mL or total CFUs. ○ j The protocol states that each species in inoculated at an OD600 of 0.05. This could vary in the final cfu/mL depending on the species, especially fungi. It may be more informative to state the desired final inoculum in cfu/mL or total CFUs. ○ The substrates are washed three times after incubation and before final quantification of biofilm. I wonder if these washes are ever quantified. This may be important to determine if the wash step disrupts significant amounts of biofilm for organisms that are loosely adherent or reaching maturity and may dislodge due to sheer pressure. This could possibly be the case with S. aureus. ○ The substrates are washed three times after incubation and before final quantification of biofilm. I wonder if these washes are ever quantified. Version 1 © 2021 Kalan L. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Open Peer Review Current Peer Review Status: PubMed Abstract\|Publisher Full Text\|Free Full Text Parsek MRASPK: Bacterial Biofilms: An Emerging Link to Disease Pathogenesis. Annu Rev Microbiol. 2003; 57: 677–701. PubMed Abstract\|Publisher Full Text Peters BM, Jabra-Rizk MA, O'May GA, et al.: Polymicrobial interactions: impact on pathogenesis and human disease. Clin Microbiol Rev. 2012; Page 19 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Open Peer Review Lindsay R. Kalan This may be important to determine if the wash step disrupts significant amounts of biofilm for organisms that are loosely adherent or reaching maturity and may dislodge due to sheer pressure. This could possibly be the case with S. aureus. ○ Page 20 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 point is that the selective media does not impact growth and quantification, the legend should state the selective media used for each species including any antibiotics/antifungals in the media. point is that the selective media does not impact growth and quantification, the legend should state the selective media used for each species including any antibiotics/antifungals in the media. point is that the selective media does not impact growth and quantification, the legend should state the selective media used for each species including any antibiotics/antifung in the media. The authors state in the discussion that the model is amenable to longitudinal studies of polymicrobial biofilm for extended periods of time. However, this study only went out to hrs. It is not clear how long or how stable these communities may be over longer periods the model allows for study over several days and even weeks this is very powerful. This information must be determined for each community, however the authors could addres their experience with the three-member model community reported in this manuscript i the discussion. ○ Are a suitable application and appropriate end-users identified? Yes If applicable, is the statistical analysis and its interpretation appropriate? Yes Are the 3Rs implications of the work described accurately? Yes Is the rationale for developing the new method (or application) clearly explained? Yes Is the description of the method technically sound? Yes Are sufficient details provided to allow replication of the method development and its use by others? Yes If any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes Competing Interests: No competing interests were disclosed. Reviewer Expertise: Polymicrobial interactions, biofilm, chronic infection I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. The authors state in the discussion that the model is amenable to longitudinal studies of polymicrobial biofilm for extended periods of time. Lindsay R. Kalan However, this study only went out to 96 hrs. It is not clear how long or how stable these communities may be over longer periods. If the model allows for study over several days and even weeks this is very powerful. This information must be determined for each community, however the authors could address their experience with the three-member model community reported in this manuscript in the discussion. ○ Are a suitable application and appropriate end-users identified? Are a suitable application and appropriate end-users identified? Are the 3Rs implications of the work described accurately? Page 21 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 Reviewer Report 20 August 2021 Reviewer Report 20 August 2021 Daniel R. Neill Institute of Infection, Veterinary and Ecological Science, University of Liverpool, Liverpool, UK O’Brien and colleagues describe a novel model system for the study of polymicrobial biofilms, under conditions designed to reflect those of the cystic fibrosis airways. The authors have built on their previous study that described a continuous-flow model for cultivation of mixed microbial populations. The neat addition, described here, is a metal gauze cage that can be loaded with substrata on which microbial biofilm communities can form. In the examples presented, cubes of agar or ex vivo porcine bronchioles are used. Culture in artificial sputum media provides environmental conditions reflective of the CF airways. This is a very well presented manuscript, with a carefully considered introduction and sufficient methodological detail to enable other users to establish the model in their own labs. The discussion describes further applications for the model that all appear realistic. I agree with the authors’ assertion that this model has potential to reduce or replace animal use for study of polymicrobial biofilms in CF and related conditions, albeit dependent upon the experimental question under consideration. This is primarily a methods paper, so the experimental findings are understandably limited. Nonetheless, I thought the observation of steady planktonic microbial density but fluctuating biofilm densities/composition was an interesting one. There is clear potential for further informative science to be undertaken using this model system. I have no major criticisms of the work. I thought the study had been well planned, carefully executed and nicely presented. I have no doubt that others with interests in polymicrobial infections of the CF lung will find the model useful, and I can see it being adapted for study of infection in other chronic lung conditions, as the authors suggest. Below, I include a few minor comments that the authors may wish to consider: The introduction mentions the limits of in vitro systems as being the lack of host cells and an immune system. I would add that most fail to capture the spatial heterogeneity of the respiratory tract. Chemical, nutritional, and microbial conditions differ between anatomical sites within the upper and lower airways and between different foci of infection within lungs. These differences may influence the dynamics of infection, the balance of competing pathogens in the airway space and may also contribute to treatment failure. https://doi.org/10.5256/f1000research.58690.r91897 © 2021 Neill D. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Daniel R. Neill ○ Although clearly beyond the scope of this manuscript, I was interested as to how the authors think the model may need to be adapted to better reflect the environmental conditions of the CF lung now that patients are moving over to CFTR modulator therapies? ○ Are a suitable application and appropriate end-users identified? Although clearly beyond the scope of this manuscript, I was interested as to how the authors think the model may need to be adapted to better reflect the environmental conditions of the CF lung now that patients are moving over to CFTR modulator therapies? ○ Are a suitable application and appropriate end-users identified? Yes Daniel R. Neill ○ The introduction states that “it should be noted that impaired immune clearance of microbes is an inherent feature of CF, which somewhat mitigates the lack of a functional immune system in in vitro models.” I take the authors point on this, but it is an oversimplification. Inefficient ○ The introduction states that “it should be noted that impaired immune clearance of microbes is an inherent feature of CF, which somewhat mitigates the lack of a functional immune system in in vitro models.” I take the authors point on this, but it is an oversimplification. Inefficient ○ Page 22 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 immune clearance is not the same as having no immune response at all. There is still potential for interactions between soluble and cellular immune defences and bacterial pathogens to influence microbial phenotypes. The authors do partially address this point in their discussion, but I think this section of the intro is slightly misleading. For the presentation of results, I wondered whether CFU/cm2 of substratum might be more informative than per ml? If this is too challenging to quantify, then per gram of agar/EVPL would at least account for differences in preparation between individual cubes. In the section on the microbial densities within biofilm, I was interested in whether there was much difference between biological replicates (i.e. between cubes). Does PA outcompete SA every time or does it depend on which gets a foothold first? Do you ever find cubes that are completely dominated by a single pathogen? This information is available within the source data file, but would be easier to digest in a graph presenting the biological replicates for each species side by side. ○ The authors postulated explanations for the relative success of PA vs SA on EVPL are intriguing, but this section should acknowledge that only one strain of each species has been tested. PAO1 is fast growing compared to many PA isolates from CF and this may contribute to the observed outcomes. ○ The authors postulated explanations for the relative success of PA vs SA on EVPL are intriguing, but this section should acknowledge that only one strain of each species has been tested. PAO1 is fast growing compared to many PA isolates from CF and this may contribute to the observed outcomes. Is the rationale for developing the new method (or application) clearly explained? Yes Is the rationale for developing the new method (or application) clearly explained? Yes Are the 3Rs implications of the work described accurately? Yes Yes Is the rationale for developing the new method (or application) clearly explained? Yes Is the description of the method technically sound? Yes Are sufficient details provided to allow replication of the method development and its use by others? Yes If any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes Is the description of the method technically sound? Yes Are sufficient details provided to allow replication of the method development and its use by others? Yes If any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes Page 23 of 24 F1000Research 2021, 10:801 Last updated: 14 SEP 2021 I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com The benefits of publishing with F1000Research: Are the conclusions about the method and its performance adequately supported by the findings presented in the article? Competing Interests: I am a co-investigator on a multi-centre research grant on which Dr Harrison is also a named co-investigator and we are contributing work to separate work packages. Reviewer Expertise: Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact. My research is focussed on bacterial infections of the respiratory tract, including those of the CF lung. I also have interests in development of in vitro and in vivo infection models that reflect respiratory tract environments. I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. The benefits of publishing with F1000Research: The benefits of publishing with F1000Research: The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com Page 24 of 24
https://openalex.org/W3082881168	https://ghrp.biomedcentral.com/counter/pdf/10.1186/s41256-020-00166-z	English	null	The impact of the COVID-19 pandemic on firms: a survey in Guangdong Province, China	Global health research and policy	2,020	cc-by	7,076	RESEARCH Open Access © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The impact of the COVID-19 pandemic on firms: a survey in Guangdong Province, China Peng Zou, Di Huo and Meng Li* 1The International Monetary Fund (IMF) broadly defined the global economic growth rate to fall below 2.5% per year as a global recession. Zou et al. Global Health Research and Policy (2020) 5:41 https://doi.org/10.1186/s41256-020-00166-z Zou et al. Global Health Research and Policy (2020) 5:41 https://doi.org/10.1186/s41256-020-00166-z (2020) 5:41 Global Health Research and Policy Zou et al. Global Health Research and Policy https://doi.org/10.1186/s41256-020-00166-z Abstract Background: The COVID-19 pandemic has affected all sectors of the world economy and society. To understand the impact of the pandemic on firms in China and suggest public policies to deal with the negative effects, we investigated firms in Guangdong Province. Methods: The survey sample included 524 firms in 15 cities in Guangdong Province. We chose these firms from the list published by the government, considering the industrial characteristics of Guangdong province and firm size. The questionnaire was developed based on previous studies carried out by UNDP. It comprised four categories with a total of 17 questions. The executives of firms were contacted by telephone or WeChat, and were invited to answer self-administered questionnaires through an online survey platform. The data was analyzed by SPSS. Results: The following key findings were revealed through the survey: (1) 48.7% of firms maintained stability, and 35.1% experienced a halt in operation or faced closure; (2) Nearly 70–90% already exploit online commerce or are willing to do so, and also remote office work, and digital operations. (3) 46% believe that they will certainly incur losses for 2020, and 83.5% expect the city’s GDP to decrease. Conclusions: Firms in Guangdong Province have faced great challenges in the epidemic. Their production and operation activities have been limited, and they are facing significant risks. It is necessary to implement policies that would profoundly lower production costs for firms, help them survive this difficult period, and gradually return to normal business. Keywords: COVID-19 pandemic, Economic impact, Public policy, Health risks Keywords: COVID-19 pandemic, Economic impact, Public policy, Health risks * Correspondence: rilanhio@hit.edu.cn School of Management, Harbin Institute of Technology, No. 13 Fayuan District, Harbin 150001, China Background emergency and is likely to continue to cause serious public health damage including major infectious diseases, mass unexplained diseases, major food poisoning and other ser- ious public health issues [5]. Global economic growth is ex- pected to decrease continually [6, 7] and according to the an OECD forecast, the global GDP growth rate will drop to 2.4% in 2020 [3]. The risk of uncertainty and global reces- sion has increased greatly [8, 9],1 due to disruptions in the supply chain, shrinking demand for consumption and in- vestment, significant weakening of economic activities, and damaged market confidence. These factors are severely The current COVID-19 is a rapidly evolving global chal- lenge and like any pandemic, it weakens health systems, costs lives, and also poses great risks to the global economy and security [1–4]. According to data from World Health Organization (WHO) and Johns Hopkins University, as end- of June 2020, there have been ten million cases around the world, causing nearly 500,000 deaths in around 215 countries (https://www.arcgis.com/apps/opsdashboard/inde x.html). The COVID-19 pandemic is a public health * Correspondence: rilanhio@hit.edu.cn School of Management, Harbin Institute of Technology, No. 13 Fayuan District, Harbin 150001, China * Correspondence: rilanhio@hit.edu.cn School of Management, Harbin Institute of Technology, No. 13 Fayuan District, Harbin 150001, China 1The International Monetary Fund (IMF) broadly defined the global economic growth rate to fall below 2.5% per year as a global recession. Data processing and analysis Study setting and design Questionnaire development Questionnaire development The questionnaire was developed based on previous studies carried out by United Nations Development Program (UNDP) in China [11]. It was piloted among 10 CEOs and firm owners. The questionnaire was further revised based on their feedback along with from two senior experts [20]. The final questionnaire included 17 questions distributed in four categories: demographic characteristics (3 questions); the impact of the pandemic on firms (6 questions); firms action (5 questions); and firm perceptions (3 questions) (Appendix 1). As the basic unit of the national economy, the operation of firms is key to the development of the na- tional macro-economy. Therefore, it is necessary to understand the status of firms, their coping strategies, and the need for government policies to respond to the impact of the COVID-19 pandemic. Researchers have begun to analyze firms’ marketing innovation strategies and strategic responses to the crisis during the pandemic [16–18]. In addition, some industries have also been an- alyzed in order to propose more targeted countermea- sures for the pandemic [19]. However, the conclusions of these studies have not been empirically tested. Data collection h The survey was carried out from 10 April 2020 to 25 April 2020. Three researchers contacted the executives of firms via telephone or WeChat, and invited them to answer the self-administered questionnaires on “WJX”, an online sur- vey platform (https://www.wjx.cn/). Participation in the survey was fully voluntary and written consent was ob- tained from each participant. The objectives of the study, confidentiality of individual information, and other ethical considerations mentioned in the survey guidelines were explained to the participants prior to data collection. They were asked to answer as many of the questions as they could. However, if they were not sure about the answer, they could simply leave it blank. Altogether, 553 anonym- ous questionnaires were collected. In this context, in order to explore the impact of the COVID-19 pandemic on firms in China and to suggest how public policies might be formulated to deal with the crisis, we investigated firms in Guangdong Province (the province with the highest GDP in China). The purpose of this research was to investigate 1) the impact of the pan- demic on firms, 2) how they responded, and 3) and their expectations for the future. Some supporting policies are recommended. Study setting and design The research protocol was approved by the Institutional Review Committee of the School of Management, Harbin Institute of Technology, with approval number 2020–01. A survey was designed to be carried out in Guangdong Province. The data was entered into Excel for data documenta- tion. SPSS was applied to further analyze the data. A total of 29 participants were excluded from the sample because they answered less than 70% questions, and missing some key answer s. Descriptive statistics of demographic characteristics of the sample and each item of questions were employed to summarize the data. The minimum sample size was calculated based on the formula below: Sample Size ¼ z2σ2=E2 ð1Þ ð1Þ Results where E = 0.05 (margin of error); z = 1.96 (confidence level); and σ = 0.5. Demographic characteristics of the sample Zou et al. Global Health Research and Policy (2020) 5:41 Page 2 of 10 Zou et al. Global Health Research and Policy total GDP of Guangdong Province). These firms were selected from a list of firms published by the govern- ment, considering the industrial characteristics of Guangdong province and the size of firms. testing the resilience of many economies, the level of gov- ernance, and the effectiveness of international cooperation. The impact of the epidemic on China’s economy is significant [10–14]. In the first two months of 2020, China’s added value in terms of size, investment, and consumption fell by 13.5, 24.5, and 20.5% year-on-year, and the unemployment rate reached 6.2%, for the first quarter of 2020, China’s GDP growth rate fell by 6.8%, according to the data released by the National Bureau of Statistics of China on April 17 [15]. Demographic characteristics of the sample The effective response rate of the sample was 94.8% (524/553). The industries in the effective response inves- tigation were Information technology (IT) (18%), manu- facturing (34.6%), finance (12.4%), real estate (6.7%), and service and commercial industry (23.7%). Other indus- tries accounted for 4.8%. This distribution is in line with the Guangdong’s industrial characteristics. The size of the surveyed firms was relatively balanced. A total of 141 firms accounted for 26.9% of all the firms and have 50 Based on the above parameters, the estimated mini- mum sample size was 384. However, in order to improve the reliability of the data, the sample was increased to over 500. The survey sample included 524 firms distrib- uting in 15 cities including Guangzhou, Shenzhen, Dongguan, and Foshan (There are 21 cities in Guangdong Province and the GDP of the 15 cities included in our sample represent more than 90% of the Zou et al. Global Health Research and Policy (2020) 5:41 Page 3 of 10 employees or fewer, 162 firms (30.1%, 50–500 em- ployees, 129 firms (24.4%, 501–5000 employees), and 92 firms 17.6%, 5000 employees or more). Table 1 shows the demographic characteristics of the firms by number and percentage, including industry, size, and location. The impact of the pandemic on firms Although half of the firms maintained their operations and overall stability, many experienced a halt in opera- tions or faced closure for various reasons such as short- ages of materials and stock (Q1 and Q2). The expectations of firms A high percentage of firms believe that they will experi- ence financial losses this year, while a lower percentage think they will be able to make a profit (Q13). However, firms with more than 1000 employees appear to be in a relatively good position to survive, and some believe that they will certainly make a profit and have sufficient cash flows. Most firms barely maintained production, facing a shortage of materials or lack of supply (Q4). A total of 22.9% of firms had orders from domestic cus- tomers cancelled, and 63.9% had overseas customers who cancelled orders or failed to send supplies on time. More firms expected their local town to experience a decrease in GDP than those who thought that GDP will remain unchanged or even increase (Q14). Descriptive statistics can be found in Fig. 1. Most firms felt that the government needs reduce, ex- empt, or postpone social insurance, value added tax, in- come tax and other taxes (Q12). More than half of the firms expect that the government to stimulate consump- tion. They also need subsidies, for example, for rent, utilities, and post stabilization some firms expect to make a staged flexible salary. Smaller firms need more cash subsidies, while larger firms tend to get policies of extending loan repayment terms and debt forgiveness. How firms responded to the crisis Firms recognized the problems and devoted more ef- forts to R&D and innovation. However, the epidemic has delayed the launch of new products as well as the progress of projects in cooperation with other institu- tions, especially for the IT industry (Q7). Many firms have faced higher labor costs, which have forced them to consider reducing the number of employees, cutting wages, and postponing recruitment until the effects of the pandemic are over (Q5, Q6 and Q8). Manufacturing has been facing a higher proportion of employment difficulties, and the real estate industry has been shedding the jobs. Both IT and Manufactur- ing industries have also sought out loans as well as funding from shareholders (Q8). Descriptive statistics can be found Fig. 4. Demographic characteristics of the sample Table 1 Overview of the sample of firms Field of business Number of Firms Percentage Industry Sub-industry IT / Software and hardware services / E-commerce / Internet operations 94 17.94% Manufacture Main Manufacture 88 16.79% Electronic technologies / Semiconductors / Integrated circuits 14 2.67% Clothing / Textiles / Leather 7 1.34% Aerospace / Aviation / Energy / Chemical 10 1.91% Machinery / Equipment / Heavy industry 10 1.91% Electric appliances 5 0.95% Furniture / Crafts / Toys 6 1.15% FMCG (food / beverage / cosmetics) 12 2.29% Automobiles and spare parts 9 1.72% Medical / Nursing / Health / Sanitation 8 1.53% Instrumentation / Industrial Automation 5 0.95% Pharmaceutical / Bioengineering / Medical Equipment 15 2.86% 181 34.6% Service and Commercial Publishing / Printing / Packaging 5 0.95% Advertising / PR / Media / Art 4 0.76% Law 1 0.19% Accounting / Auditing 1 0.19% Traffic / Transportation / Logistics 11 2.10% Education / Training / Scientific research / Colleges 15 2.86% Trading / Import & Export 13 2.48% Wholesale / Retail 13 2.48% Communication / Telecommunications / Network equipment 21 4.01% Property Management / Commercial Centers 8 1.53% Agency / Consulting / Headhunting / Certification 12 2.29% 124 23.7% Finance Bank / Insurance / Securities / Investment Bank / Risk Fund 65 12.40% Real estate and Architecture Real estate development / Architectural engineering / Design 35 6.68% Other industry 25 4.77% Employees 50 and below 141 26.91% 51–100 54 10.31% 101–300 71 13.55% 301–500 37 7.06% 501–1000 42 8.02% 1001–4999 87 16.60% 5000 and above 92 17.56% Total 524 100% Service and Commercial Page 4 of 10 Page 4 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 Zou et al. Global Health Research and Policy (2020) 5:41 Zou et al. Global Health Research and Policy (2020) 5:41 Zou et al. Global Health Research and Policy (2020) 5:41 The expectations of firms Firms are suffering from at least one of the following pressures: employees’ salary and social insurance, rent, loss of orders, payment of accounts payable, and loan re- payment (Q3). For firms with fewer than 50 employees, rent payments seem to be the key pact. The impact of the pandemic on firms pressure of employees’ salary, insurances, rent payments, settlement of accounts payable, and loan repayment. Given the pandemic, returning to work was not the same as reaching production capacity. In fact, almost half of the firms were facing material shortages, some of them even said that they had run out of supplies (see an- swers to Q1 and Q2). In response to Q3 regarding oper- ational pressures, our survey revealed that firms were facing employment and costs. In addition, 22.9% of firms had had orders cancelled from domestic customers. Many firms responded that the market expansion was also affected, due to the impossibility of site visiting and face-to-face meetings with customers. As highlighted by the answers to Q4 (supply of raw materials, spare parts and other production and oper- ation materials), the business owners in the interviews reported that supply chain links - the producing and processing of raw material parts, and logistics etc. were affected by the epidemic situation in various regions, with different degrees of delay in delivery and work stop- pages while waiting for materials. The spread of epi- demic to outside China affected many export businesses. Firms reported cancellation of orders or delayed receipt of goods from overseas customers. Moreover, more than half of the firms reported a sharp increase in operating costs such as the cancellation of orders, delayed delivery, and production stagnation. Some of them had to deal with insolvency and compen- sation for breach of contract. Firms were also under the Discussion Most of the research on public health emergencies has fo- cused on changes in overall consumption trends, such as consumption behavior, changes in consumer decision- making and consumer psychology [21–24]. These studies show that pandemics reduce economic activity and em- ployment, and inhibit consumption, which is in line with our findings on the impact of the COVID-19 pandemic. However, such studies only analyzed the trend in overall consumption, while we investigated the firms themselves, thereby more comprehensively reflecting the impact of the COVID-19 pandemic on economic activities. A high percentage of firms have already moved their operations to online or plan to do so (Q9). Likewise, many firms have already started working re- motely and digital operations or plan to do so (Q10 and Q11). Descriptive statistics can be found in Fig. 2 and Fig. 3. Fig. 1 Descriptive statistics of Q1--Q4. Note: Appendix 1 contains details of the questions and options Fig. 1 Descriptive statistics of Q1--Q4. Note: Appendix 1 contains details of the questions and options Zou et al. Global Health Research and Policy (2020) 5:41 Page 5 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 Page 5 of 10 Fig. 2 Descriptive statistics of Q5-Q8. Note: Appendix 1 contains details of the questions and options Fig. 2 Descriptive statistics of Q5-Q8. Note: Appendix 1 contains details of the questions and options The expectations of firms The result of this investigation indicate that rebound consumption will not come. Consumption is mainly determined by demand and the ability to pay [29]. The ability to pay depends on the consumer’s current income and expectations of future income [30]. Since the epidemic has continued, almost half firms believe that there they will definitely incur losses this year. Q9 and Q10 indicate that, more than half of the firms believe that this pandemic has promoted the establish- ment of remote offices, remote recruitment, and remote business negotiation models, and the informatization and digitalization of firms have improved their ability to respond to major crises. A few firms also mentioned that the epidemic can eliminate competitors to a certain extent, by triggering the launch of new businesses, and accelerating changes in marketing approaches (such as community channel expansion). For example, the core business of clothing brand PeaceBird was offline retailing in large shopping malls. During the crisis, they trans- ferred employees from offline stores to online sales teams. They successfully completed this channel trans- formation and survived the crisis by using livestreaming and social media platforms [25]. Q13 asked “To what extent do you expect this pandemic will affect your firm’s development in the first quarter of 2020?” and nearly half of firms predicted that losses will directly affect their employees’ income and even work opportunities. More than 20% of firms have reduced the number of staff to cope with the epidemic. Moreover, respondents are also not optimistic about the GDP growth of the city where the firm is located (Q14). Among them, 83.5% expected a decreasing trend in their city’s GDP growth, and 38.6% predicted a signifi- cant decline. On the contrary, only a few respondents believed that GDP would remain unchanged or antici- pated that GDP would increase. Therefore, there is in- sufficient support for any consumption rebound given expectations for reduced income and spending power. The epidemic has strengthened the determination of some firms to turn crises into opportunities and invest in new industries. At the same time, it has strengthened consumers’ health awareness and changed their consumption behaviors [26–28]. Some firms reported that this trend has led to new business opportunities and has driven reforms in their marketing strategies. The interviewees indicated that there is a need to accelerate, for example, the development of online education, medical care, and 5G. How firms responded to the epidemic According to the answers for Q5, Q6 and Q8, the epidemic also negatively impacted the technological Fig. 3 Descriptive statistics of Q9-Q 11. Note: Appendix 1 contains details of the questions and options Fig. 3 Descriptive statistics of Q9-Q 11. Note: Appendix 1 contains details of the questions and options Zou et al. Global Health Research and Policy (2020) 5:41 Page 6 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 Page 6 of 10 Fig. 4 Descriptive statistics of Q12--Q14 Some firms have changed the original store-based marketing mode and integrated channels for interactive marketing. For example, the store is not open, but the sales people conduct live broadcasts, group buying and social marketing, to name a few. innovation and human resources in some firms. For example, the epidemic has delayed the launch of new products, new recruits, and the progress of projects in cooperation with other firms. According to Q7, the combination of this pandemic and the impact of the Sino US trade dispute prompted some firms to recognize the problems and to devote more efforts to R & D and innovation. Conclusions While many firms in Guangdong Province have main- tained overall stability, others have experienced a halt in their operations or faced closure. Almost all the firms in our survey are willing to transform into on- line marketing, remote office work and digital opera- tions. Half of the firms believe that there will be a certain loss this year, and a very high percentage of firms expected a decreasing trend in the city’s GDP growth. Options A. Very serious impact, leading to serious difficulties in business operations and bankruptcy B. Great impact: operations barely maintained C. Small impact, some difficulties in business operations, but overall stability D. No significant impact E. Positive impact, providing new opportunities for development E. Positive impact, providing new opportunities for development The expectations of firms Unfortunately, the Spring Festival, which took place at the height of this pandemic, would normally stimulate a high level of consumption [31] over a short period which is not protracted throughout the year [32]. In the long run, the epidemic will reduce people’s willingness to Page 7 of 10 Page 7 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 Zou et al. Global Health Research and Policy cultural, social and governing systems. A cross-sector analysis is also needed in order to get more specific sug- gestions for different industries in different regions. consume apart from necessities and immediate needs. In the short term, consumer psychology has not fully returned to normal [33], and establishing a work mode that is compatible with epidemic prevention and control takes time. Therefore, the endogenous power of com- pensatory and rebound consumption is insufficient. Policy implications On the basis of our survey and considering the firms’ ap- peal for government support, we believe that the follow- ing financial policies would stimulate consumption and help firms to survive. (1) Financial policies supporting production and operation. Firms expect public policies and measures such as the “tax exemption, reduction, postponement, return and compensation”, in order to substantially lower production costs for firms, and help firms survive the difficult period of production and operation, while gradually returning to normal business operations. There is a need to fully understand the impact of the epidemic on consumption and the difficulty of promot- ing the recovery of consumption, also in terms of how industry has been affected by this pandemic. Policies need to be introduced to profoundly lower production costs for firms, and help firms survive this difficult period, and gradually return to normal business operations. (2) Investment driving consumption. Investment should play a driving role in promoting consumption. Investment in fields such as public health facilities and health care should be increased to cultivate and expand consumer demand. Questionnaire Questionnaire Survey on firms affected by the COVID-19 Pandemic (3) Measures supporting and facilitating consumption. Holiday tourism products could be launched alongside a paid vacation system. Policies should be targeted to promote tourism and the ‘night economy’. 1. To what extent does has production and operations of your firm been affected by this pandemic? (single choice) 1. To what extent does has production and operations of your firm been affected by this pandemic? (single choice) (4) Role of industry organizations in promoting consumption. Industry organizations could release suppressed and frozen consumption by carrying out various activities such as shopping festivals and food festivals to stimulate consumption; and facilitate an environment for promoting online-to- offline integrated consumption, for example through preferential concessions, model innovation, policy stimulation, cooperation between banks and firms, strengthening services, and product innovation. (4) Role of industry organizations in promoting consumption. Industry organizations could release suppressed and frozen consumption by carrying out various activities such as shopping festivals and food festivals to stimulate consumption; and facilitate an environment for promoting online-to- offline integrated consumption, for example through preferential concessions, model innovation, policy stimulation, cooperation between banks and firms, strengthening services, and product innovation. Options A. Very serious impact, leading to serious difficulties in business operations and bankruptcy B. Great impact: operations barely maintained C. Small impact, some difficulties in business operations, but overall stability D. No significant impact E. Positive impact, providing new opportunities for development 2. What are the reasons for the suspension of production and operations of your firm? (multiple choice, up to 2 items) Options A. Shortage of production materials. B. Difficulty in developing market C. Impact of measures taken to respond to the pandemic Appendix Questionnaire Survey on firms affected by the COVID-19 Pandemic Limitation of this study Determined to invest more in technological innovation after being Questionnaire (Continued) better aware of the firm’s self-development problems during the pandemic 8. How are you currently or planning to cope with the cash flow shortage? (multiple choice). WHAT IF THEY DON’T HAVE a CASH FLOW SHORTAGE? Options A. Funding from existing shareholders B. Adding new shareholders C. Loans D. Delaying payment E. Cutting pay and jobs 9. Are you willing to transform to online commerce? Options A. Very unwilling B. Unwilling C. Reasonably willing D. Willing E. Very willing 10. What self-help measures has your firm taken so far? (multiple choice) Options A. Applied for financing B. Increased online operations C. Cut pay and jobs. D. Implemented a remote office (digital office) 11. What are the potentially positive impacts of the pandemic in your view? (multiple choice, up to 3 items) Options A. Promote the establishment of remote office work B. Enhance information and digital construction of firms C. Help to better realize firm’s shortcomings and solve existing problems better aware of the firm’s self-development problems during the pandemic 8. How are you currently or planning to cope with the cash flow shortage? (multiple choice). WHAT IF THEY DON’T HAVE a CASH FLOW SHORTAGE? Options A. Funding from existing shareholders B. Adding new shareholders C. Loans D. Delaying payment E. Cutting pay and jobs 4. What is the current situation regarding the supply of raw materials, spare parts and other production and operation materials in your firm? (single choice) A. Funding from existing shareholders A. Funding from existing shareholders B. Adding new shareholders C. Loans D. Delaying payment Options A. Total disruption of supply B. Supply shortage C. Supply barely maintains production D. Satisfactory supply E. Normal supply E. Cutting pay and jobs E. Cutting pay and jobs 9. Are you willing to transform to online commerce? Options A. Very unwilling B. Unwilling C. Reasonably willing D. Willing E. Very willing 5. Does your company plan to reduce or increase the number of employees? (single choice) 5. Does your company plan to reduce or increase the number of employees? (single choice) 10. What self-help measures has your firm taken so far? (multiple choice) 6. How has the pandemic affected recruitment? (multiple choice, up to 3 items) Options A. Increase in labor costs B. Unable to find a suitable recruitment channel C. Limitation of this study Our study has several limitations that provide opportun- ities for future research. First, this study can only be viewed as a preliminary study and more follow-up tra- cing investigations at different stages need to be con- ducted in order to monitor the continuing impact of the pandemic and the effectiveness of public policies and firms’ responses. Second, this study focuses on China which was among the first countries to recover from the pandemic, more surveys should be implemented in other countries to explore how COVID-19 has impacted other 2. What are the reasons for the suspension of production and operations of your firm? (multiple choice, up to 2 items) B. Difficulty in developing market Page 8 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 3. What are the main operating pressures that your firm is currently facing? (multiple choice, up to 3 items) Options A. Employee salaries, insurances B. Rent (Buildings, Equipment) C. Repayment of loans D. Payment of accounts payable E. Cancellation of orders 3. What are the main operating pressures that your firm is currently facing? (multiple choice, up to 3 items) Questionnaire (Continued) firm is currently facing? (multiple choice, up to 3 items) Options A. Employee salaries, insurances B. Rent (Buildings, Equipment) C. Repayment of loans D. Payment of accounts payable E. Cancellation of orders 4. What is the current situation regarding the supply of raw materials, spare parts and other production and operation materials in your firm? (single choice) Options A. Total disruption of supply B. Supply shortage C. Supply barely maintains production D. Satisfactory supply E. Normal supply 5. Does your company plan to reduce or increase the number of employees? (single choice) Options A. Reduce greatly(30–50%) B. Reduce slightly(10–30%) C. Remain basically the same D. Increase slightly(10–30%) E. Increase greatly(30–50%) 6. How has the pandemic affected recruitment? (multiple choice, up to 3 items) Options A. Increase in labor costs B. Unable to find a suitable recruitment channel C. Postponement or cancelation the existing recruitment plan D Transition to online recruitment 7. What is the clearest impact of the pandemic on your firm’s technological innovation? (multiple choice, up to 3 items) Options A R&D process may affect the launching process of new product B Unable to recruit suitable R&D personnel C. Unable to cooperate with other departments to carry out part of R & D work D. Limitation of this study Postponement or cancelation the existing recruitment plan D Transition to online recruitment D. Implemented a remote office (digital office) 11. What are the potentially positive impacts of the pandemic in your view? (multiple choice, up to 3 items) 16. What industry does your firm belong to? Options Chaozhou Dongguan Foshan Guangzhou Huizhou Jiangmen Maoming Shantou Shenzhen Yangjiang Zhanjiang Zhongshan Zhuhai IT / Software and hardware services / E-commerce / Internet operations Catering / Entertainment / Tourism / Hotel Publishing / Printing / Packaging Electronic technology / Semiconductor / Integrated circuit Law Real estate development / Architectural engineering / Design Clothing / Textile / Leather Advertising / PR / Media / Art Aerospace / Aviation / Energy / Chemical Accounting / Auditing Machinery / equipment / Heavy industry Home appliance industry Furniture / Crafts / Toys Traffic / Transportation / Logistics Education / Training / Scientific research / College FMCG (food / beverage / cosmetics) Trading / Import & Export Agriculture / Fishery / Forestry Wholesale / Retail Automobile and spare parts Communication / Telecommunications operations / Network equipment Property Management / Commercial Center Medical / Nursing / Health / Sanitation Instrumentation / Industrial Automation Bank / Insurance / Securities / Investment Bank / Risk Fund Pharmaceutical / Bioengineering / Medical Equipment Manufacture Consulting / Headhunting / Certification Others 17. How many employees are there in your company? Options 50 and below 51–100 101–300 301–500 501–1000 1001–4999 5000 and above Abbreviations COVID-19: Coronavirus disease 2019; OECD: Organization for Economic Co- operation and Development; GDP: Gross Domestic Product; WHO: World Health Organization; IMF: World Health Organization; CEO: Chief Executive A. Reduce, exempt or postpone value-added tax, income tax, in- surance premiums and other taxes B. Stimulate consumption B. Stimulate consumption C. Allow firms to implement a staged flexible salary method D. Provide subsidies for rent, utilities, post stabilization etc. 13. To what extent do you expect this pandemic will affect your firm’s development in the first quarter of 2020?(single choice) 13. To what extent do you expect this pandemic will affect your firm’s development in the first quarter of 2020?(single choice) Options A. Profits B. Balance of income and expenditure D. Losses E. Serious losses A. Bankruptcy 14. What has been the effect on your city’s economic (GDP) growth in the first quarter of this year? 17. How many employees are there in your company? Options 50 and below 51–100 101–300 301–500 501–1000 1001–4999 5000 and above Options Options A. Promote the establishment of remote office work 7. What is the clearest impact of the pandemic on your firm’s technological innovation? (multiple choice, up to 3 items) B. Enhance information and digital construction of firms B. Enhance information and digital construction of firms 7. What is the clearest impact of the pandemic on your firm’s technological innovation? (multiple choice, up to 3 items) Options A R&D process may affect the launching process of new product B Unable to recruit suitable R&D personnel C. Unable to cooperate with other departments to carry out part of R & D work D. Determined to invest more in technological innovation after being C. Help to better realize firm’s shortcomings and solve existing oblems C. Help to better realize firm’s shortcomings and solve existing problems Page 9 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 12. What policies do you expect the government will put in to place to help your firm overcome the difficulties? (multiple choice, up to 4 items) 16. What industry does your firm belong to? Options IT / Software and hardware services / E-commerce / Internet operations Catering / Entertainment / Tourism / Hotel Publishing / Printing / Packaging Electronic technology / Semiconductor / Integrated circuit Law Real estate development / Architectural engineering / Design Clothing / Textile / Leather Advertising / PR / Media / Art Aerospace / Aviation / Energy / Chemical Accounting / Auditing Machinery / equipment / Heavy industry Home appliance industry Furniture / Crafts / Toys Traffic / Transportation / Logistics Education / Training / Scientific research / College FMCG (food / beverage / cosmetics) Trading / Import & Export Agriculture / Fishery / Forestry Wholesale / Retail Automobile and spare parts Communication / Telecommunications operations / Network equipment Property Management / Commercial Center Medical / Nursing / Health / Sanitation Instrumentation / Industrial Automation Bank / Insurance / Securities / Investment Bank / Risk Fund Pharmaceutical / Bioengineering / Medical Equipment Manufacture Consulting / Headhunting / Certification Others Options A. Reduce, exempt or postpone value-added tax, income tax, in- surance premiums and other taxes B. Stimulate consumption C. Allow firms to implement a staged flexible salary method D. Provide subsidies for rent, utilities, post stabilization etc. 13. To what extent do you expect this pandemic will affect your firm’s development in the first quarter of 2020?(single choice) Options A. Profits B. Balance of income and expenditure D. Losses E. Serious losses A. Bankruptcy 14. What has been the effect on your city’s economic (GDP) growth in the first quarter of this year? Options A. Reduced significantly B. Reduced slightly C. Unchanged D. Increased slightly E. Increased significantly 15. Where is your firm located? Consent for publication Not applicable. 25. ChinaSSPP.com. The average daily retail sales of domestic clothing brand Peacebird exceeded 8 million during COVID-19 crisis. Available at http:// www.chinasspp.com/News/Detail/2020-2-11/446723.htm. Accessed 28 Apr 2020. Acknowledgements W ld lik h We would like to thank Dr. Hao Li and the anonymous reviewers for their valuable comments in improving the quality of this article. 16. Wang Y, Hong A, Li X, Gao J. Marketing innovations during a global crisis: a study of China firms’ response to COVID-19. J Bus Res. 2020;116:214–20. 17. Wenzel M, Stanske S, Lieberman MB. Strategic responses to crisis. Strateg Manag J. 2020;41:7–18. Abbreviations Abbreviations COVID-19: Coronavirus disease 2019; OECD: Organization for Economic Co- operation and Development; GDP: Gross Domestic Product; WHO: World Health Organization; IMF: World Health Organization; CEO: Chief Executive Page 10 of 10 Page 10 of 10 Page 10 of 10 Zou et al. Global Health Research and Policy (2020) 5:41 Officer; SPSS: Statistical Product and Service Solutions; R&D: Research and Development; UNDP: United Nations Development Program 14. Wang X, Sun T. China’s engagement in global health governance: a critical analysis of China’s assistance to the health sector of Africa. J Glob Health. 2014;4(1):793–804. 15. National bureau of statistics of china. Available at: http://www.stats.gov.cn/. Accessed 17 April 2020. nal bureau of statistics of china. Available at: http://www.stats.gov.cn/ Competing interests 26. Li JB, Yang A, Dou K. Chinese public’s knowledge, perceived severity, and perceived controllability of the COVID-19 and their associations with emotional and behavioral reactions, social participation, and precautionary behavior: A national survey. PsyArXiv Preprints. 2020. The authors declare that they have no competing interests. The authors declare that they have no competing interests. Received: 21 May 2020 Accepted: 7 July 2020 behavior: A national survey. PsyArXiv Preprints. 2020. 27. Xiang Y-T, Yang Y, Li W. Timely mental health care for the 2019 novel coronavirus outbreak is urgently needed. Lancet Psychiatry. 2020;7(3):228–9. Please contact the corresponding author for data requests. 23. Mizrachi I, Fuchs G. Should we cancel? An examination of risk handling in travel social media before visiting bola-free destinations. J Hosp Tour Manag. 2016;28:59–65. Funding Thi 20. Shah P, Shrestha R, Mao Z, et al. Knowledge, attitude, and practice associated with antibiotic use among university students: a survey in Nepal. Int J Environ Res Public Health. 2019;16(20):1–11. g This research is partially funded by a research grant from the National Natural Science Foundation of China (CN) under project No. 71972060 and 71672047. 21. Ordaz-Németh I, Arandjelovic M, Boesch L, et al. The socio-economic drivers of bushmeat consumption during the west African Ebola crisis. PLoS Negl Trop Dis. 2017;11(3):1–22. Ethics approval and consent to participate The research protocol was approved by the Institutional Review Committee School of Management, Harbin Institute of Technology, with the approval number 2020–01. 24. Peng EY-C, Lee M-B, Tsai S-T, et al. Population-based post-crisis psychological distress: an example from the SARS outbreak in Taiwan. J Formos Med Assoc. 2010;109(7):524–32. Authors’ contributions PZ i d f h d PZ conceived of the study and drafted the manuscript. DH performed the statistical analysis and helped to draft the manuscript. ML conceived of the study and participated in its design and coordination. The authors read and approved the final manuscript. 18. He H, Harris L. The impact of Covid-19 pandemic on corporate social responsibility and marketing philosophy. J Bus Res. 2020;116:176–82. 19. Pantanoa E, Pizzib G, Scarpib D, Dennisc C. Competing during a pandemic? Retailers’ ups and downs during the COVID-19 outbreak. J Bus Res. 2020; 116:209–13. 19. Pantanoa E, Pizzib G, Scarpib D, Dennisc C. Competing during a pandemic? Retailers’ ups and downs during the COVID-19 outbreak. J Bus Res. 2020; 116:209–13. References 28. Zhang J, Wu W, Zhao X. Recommended psychological crisis intervention response to the 2019 novel coronavirus pneumonia outbreak in China: a model of West China hospital. Precision Clin Med. 2020;1:1–12. 1. Chattu KV, Adisesh A, Yaya S. Canada’s role in strengthening global health security during the COVID-19 pandemic. Global Health Res Policy. 2020;5: 16–8. 29. Yan L, Liu MT, Chen X. An arousal-based explanation of affect dynamics. Eur J Mark. 2016;50(7–8):1159–84. 2. WHO. Health Security. Available at: https://www.who.int/health-security/en/. Accessed 24 Mar 2020. 2. WHO. Health Security. Available at: https://www.who.int/health-security/en/. Accessed 24 Mar 2020. 30. Di Muro F, Murray KB. An arousal regulation explanation of mood effects on consumer choice. J Consum Res. 2012;39(3):574–84. 3. Orlik T, Rush J, Cousin M, Hong J. Coronavirus could cost the global economy $2.7 trillion. Here’s how. 2020. Available at: https://www. bloomberg.com/graphics/2020-coronavirus-pandemic-global-economic-risk/. Accessed 13 Mar 2020. 3. Orlik T, Rush J, Cousin M, Hong J. Coronavirus could cost the global economy $2.7 trillion. Here’s how. 2020. Available at: https://www. bloomberg.com/graphics/2020-coronavirus-pandemic-global-economic-risk/. Accessed 13 Mar 2020. 31. Chen X, Li F, Nydegger L. Brief sensation seeking scale for Chinese – cultural adaptation and psychometric assessment. Personal Individ Differ. 2013;54(5): 604–9. 31. Chen X, Li F, Nydegger L. Brief sensation seeking scale for Chinese – cultural adaptation and psychometric assessment. Personal Individ Differ. 2013;54(5): 604–9. 4. Paul E, Brown G, Ridde V. COVID-19: time for paradigm shift in the nexus between local, national and global health. BMJ Glob Health. 2020;5(4):1–5. 32. Jia J, Lu X, Yuan Y, et al. Population flow drives spatio-temporal distribution of COVID-19 in China. Nature. 2020;582(7812):1–11. 32. Jia J, Lu X, Yuan Y, et al. Population flow drives spatio-temporal distribution of COVID-19 in China. Nature. 2020;582(7812):1–11. 5. The State Council of PRC. Regulations on emergency response to public health emergencies: Beijing, China Fangzheng Press; 2003. 33. Zhang L, Zhang C, Shang L. Sensation-seeking and domain-specific risk- taking behavior among adolescents: risk perceptions and expected benefits as mediators. Personal Individ Differ. 2016;101:299–305. 33. Zhang L, Zhang C, Shang L. Sensation-seeking and domain-specific risk- taking behavior among adolescents: risk perceptions and expected benefits as mediators. Personal Individ Differ. 2016;101:299–305. 6. van Wees SH, Holmer H. Global health beyond geographical boundaries: reflections from global health education. BMJ Glob Health. 2020;5(5):3. 7. Nagesh S, Chakraborty S. Saving the frontline health workforce amidst the COVID-19 crisis: challenges and recommendations. J Glob Health. 2020;10(1): 345. 7. Availability of data and materials p 22. Bowles J, Hjort J, Melvin T, et al. Ebola, jobs and economic activity in Liberia. J Epidemiol Community Health. 2016;70(3):271–7. Please contact the corresponding author for data requests. References Nagesh S, Chakraborty S. Saving the frontline health workforce amidst the COVID-19 crisis: challenges and recommendations. J Glob Health. 2020;10(1): 345. 8. Liu P, Guo Y, Qian X, Tang S, Li Z, Chen L. China’s distinctive engagement in global health. Lancet. 2014;384(9945):793–804. 9. Zhang M, Kim R. Occupational health and safety in China: from emergency response to Jiangsu chemical explosion to long-term governance improvement. J Glob Health. 2020;10:7–15. 10. Kim H. The sociopolitical context of the COVID-19 response in South Korea. BMJ Glob Health. 2020;5(5):8. 10. Kim H. The sociopolitical context of the COVID-19 response in South Korea. BMJ Glob Health. 2020;5(5):8. 11. United Nations Development Program in China Assessment report on impact of covid19 pandemic on Chinese enterprises. Available at: https:// www.cn.undp.org/content/china/zh/home/library/crisis_prevention_and_ recovery/assessment-report-on-impact-of-covid-19-pandemic-on- chineseente.html. Accessed 7 Apr 2020. 12. Tuangratananon T, Tang K, et al. China: leapfrogging to become a leader in global health. J Glob Health. 2019;9:010312. 12. Tuangratananon T, Tang K, et al. China: leapfrogging to beco 12. Tuangratananon T, Tang K, et al. China: leapfrogging to become a leader in global health. J Glob Health. 2019;9:010312. 13. Knight JB. China as a developmental state. World Econ. 2014;37(10):1335–47. 13. Knight JB. China as a developmental state. World Econ. 2014;37(10):1335–47.
https://openalex.org/W4234083195	https://www.researchsquare.com/article/rs-357329/latest.pdf	English	null	Experimental Studies on Natural Convection Open and Closed Solar Drying Using External Reflector	Research Square (Research Square)	2,021	cc-by	10,187	Experimental Studies on Natural Convection Open and Closed Solar Drying Using External Re§ector Abd Elnaby Kabeel Tanta University Experimental Studies on Natural Convection Open and Closed Solar Drying Using External Re§ector Abd Elnaby Kabeel Tanta University Paul Durai Leon Dharmadurai SNS College of Technology Sathiyaseelan Vasanthaseelan KPR Institute of Engineering and Technology Prof. Dr. Ravishankar Sathyamurthy (  raviannauniv23@gmail.com ) KPR Institute of Engineering and Technology https://orcid.org/0000-0002-2881-3455 Bharathwaaj Ramani KPR Institute of Engineering and Technology Athikesavan Muthu Manokar BS Abdur Rahman Crescent Engineering College: B S Abdur Rahman Crescent Institute of Science & Technology Ali Chamkha Kuwait College of Science and Technology Research Article Keywords: Fish, Moisture, Drying, Solar, Hygiene Posted Date: June 2nd, 2021 DOI: https://doi.org/10.21203/rs.3.rs-357329/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Environmental Science and Pollution Experimental Studies on Natural Convection Open and Closed Solar Drying Using External Re§ector Abd Elnaby Kabeel Tanta University Paul Durai Leon Dharmadurai SNS College of Technology Sathiyaseelan Vasanthaseelan KPR Institute of Engineering and Technology Prof. Dr. Ravishankar Sathyamurthy (  raviannauniv23@gmail.com ) KPR Institute of Engineering and Technology https://orcid.org/0000-0002-2881-3455 Bharathwaaj Ramani KPR Institute of Engineering and Technology Athikesavan Muthu Manokar BS Abdur Rahman Crescent Engineering College: B S Abdur Rahman Crescent Institute of Science & Technology Ali Chamkha Kuwait College of Science and Technology Research Article Keywords: Fish, Moisture, Drying, Solar, Hygiene Research Article Version of Record: A version of this preprint was published at Environmental Science and Pollution Research on August 5th, 2021. See the published version at https://doi.org/10.1007/s11356-021-15768- 4 4. 1 Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 Abd Elnaby Kabeela, Paul Durai Leon Dharmaduraib, Sathiyaseelan Vasanthaseelanc, Ravishankar 3 Sathyamurthya,c, Bharathwaaj Ramanic, Athikesavan Muthu Manokard, Ali Chamkhae,f,* 4 a Mechanical Power Engineering Department, Faculty of Engineering, Tanta university Tanta, Egypt 5 b Department of Automobile Engineering, SNS College of Technology, Coimbatore, Tamil Nadu, 6 India. 7 c Department of Mechanical Engineering, KPR Institute of Engineering and Technology, Arasur, 8 Coimbatore, Tamil Nadu, India 9 d Department of Mechanical Engineering, B.S Abdur Rahman Crescent Institute of Science and 10 Technology, Vandalur, Chennai, 600048, Tamil Nadu, India 11 e Faculty of Engineering, Kuwait College of Science and Technology, Doha District, Kuwait 12 f Institute of Research and Development, Duy Tan University, Da Nang, 550000, Viet Nam 13 Corresponding author Ali J Chamkha, Faculty of Engineering, Kuwait College of Science and 14 Technology, Doha District, Kuwait alichamkha@duytan.edu.vn 15 Abstract 16 This work presents the experimental investigation on a solar food dryer equipped with external 17 reflectors to enhance the rate of drying efficiency by removing the moisture content available in the 18 anchovy fish. A comparison on conventional open solar drying is carried out to assess the parameters 19 such as drying efficiency, moisture removal rate, and heat energy required for drying the anchovy fish 20 using the modified solar dryer using natural convection technique. The physical examination results on 21 the dried products proved that using modified solar dryer, the dried product are free from insects, dust. 22 Also, it is found that the loss of colour from the product is minimum while compared to conventional 23 open solar drying. The results showed that the thermal efficiency of dryer 1 and dryer 2 are 24 comparatively higher as compared to that of conventional open solar drying and found as 16.73 and 25 19.34 % respectively. 26 Keywords: Fish, Moisture, Drying, Solar, Hygiene 27 1. Introduction 28 The increase in global population and economic prosperity is followed by an increase in material 29 demands and energy consumption. Since the Earth's fossil fuel supplies are finite and will inevitably 30 run out, the loss of these fuels is a danger to life on the world. Research Article Furthermore, the use of fossil fuels as a 31 Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 Abd Elnaby Kabeela, Paul Durai Leon Dharmaduraib, Sathiyaseelan Vasanthaseelanc, Ravishankar 3 Sathyamurthya,c, Bharathwaaj Ramanic, Athikesavan Muthu Manokard, Ali Chamkhae,f, 4 a Mechanical Power Engineering Department, Faculty of Engineering, Tanta university Tanta, Egypt 5 b Department of Automobile Engineering, SNS College of Technology, Coimbatore, Tamil Nadu, 6 India. 7 c Department of Mechanical Engineering, KPR Institute of Engineering and Technology, Arasur, 8 Coimbatore, Tamil Nadu, India 9 d Department of Mechanical Engineering, B.S Abdur Rahman Crescent Institute of Science and 10 Technology, Vandalur, Chennai, 600048, Tamil Nadu, India 11 e Faculty of Engineering, Kuwait College of Science and Technology, Doha District, Kuwait 12 f Institute of Research and Development, Duy Tan University, Da Nang, 550000, Viet Nam 13 *Corresponding author Ali J Chamkha, Faculty of Engineering, Kuwait College of Science and 14 Technology, Doha District, Kuwait alichamkha@duytan.edu.vn 15 Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 The increase in global population and economic prosperity is followed by an increase in material 29 demands and energy consumption. Since the Earth's fossil fuel supplies are finite and will inevitably 30 run out, the loss of these fuels is a danger to life on the world. Furthermore, the use of fossil fuels as a 31 1 1 source of energy has had a variety of environmental implications. Following the heavy use of coal, oil, 32 and natural gas over the past 150 years, many countries are now invested in renewable energy supplies. 33 Globally, there is an increasing realisation that expanded renewable energy adoption is vital for 34 mitigating climate change, generating new economic prospects, and delivering energy coverage to the 35 billions of people who already lack modern energy services. Solar energy has received a lot of interest 36 among renewable energy options because of its availability, free connectivity, and commercial 37 potential. Photovoltaic and solar thermal systems are two popular methods for using solar energy in 38 use (Dharmadurai et al. (2020); Bal et al. (2010); Mustayen et al. (2014)). The fish is washed cleanly 39 with the water and made it to dry for sometimes. Then the fish is made to be cut as per the requirement, 40 then again, the fish is cleaned and made it to dry for some time. The salt is applied on the layer of the 41 fish to kill the micro-organism. Then the salted fish is exposed to sun light as shown in the Fig. 1. At 42 present there are many methods for drying the fish namely open drying method, using electric dryer, 43 and using solar dryer. The salted fish is exposed to the sunlight till it attains 10 % moisture level. The 44 advantage of this process is that the capital cost is less and the more products can be dried at the same 45 time. The disadvantages of this method are that there is no hygiene and it takes more time to dry. The 46 dry fish can’t be saved during the unexpected rain time. Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 47 48 Figure 1: Drying fish in Open Sun light in India 49 Figure 1: Drying fish in Open Sun light in India Figure 1: Drying fish in Open Sun light in India For centuries, drying of agricultural crops, vegetables, and fruits are carried out using open solar drying 50 technique (Fudholi et al. 92010; Mustayen et al. (2014); Jairaj et al. (2009)). There are several 51 drawbacks in open solar drying technique as it requires large space, labor and the time required for 52 drying is higher. The quality of products are also considered to be main influential parameter in open 53 solar drying technique as it was affected by unexpected rain, and foreign bodies left by the animals 54 For centuries, drying of agricultural crops, vegetables, and fruits are carried out using open solar drying 50 technique (Fudholi et al. 92010; Mustayen et al. (2014); Jairaj et al. (2009)). There are several 51 drawbacks in open solar drying technique as it requires large space, labor and the time required for 52 drying is higher. The quality of products are also considered to be main influential parameter in open 53 solar drying technique as it was affected by unexpected rain, and foreign bodies left by the animals 54 2 2 and birds. To overcome these issues, a solar dryer was developed to create a hygiene food processing. 55 The dryer can be classified into two major groups, namely: (a) passive solar-energy drying systems, 56 (b) active solar-energy drying systems. Either passive or active solar dryer, it is further divided into 57 (a) direct (b) indirect and (c) mixed type. Indirect solar dryers can be either be forced circulation or 58 natural circulation mode (Fig. 2) (Kumar et al. (2016); Ekechuku & Norton (1999)). Mohana et al. 59 (2020) made a comprehensive review on different concepts on solar dryer design for drying food. 60 (2020) made a comprehensive review on different concepts on solar dryer design for drying food. 60 The solar dryer is a combination of different components. The component used in the solar dryer are 61 Insulators, outer cover, inner cover, glazing. The thermal Insulator is used to prevent the transfer of 62 heat from one body to another body. The thermal Insulator prevents the conduction and convection 63 modes of heat transfer, but it is quite challenging to control the heat transfer through radiation. Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 In order to optimize the results, the 100 drying process is carried out in three different stages namely, drying the anchovy fish at 70◦C for 150 101 min followed by drying the fish at 50◦C for 100 min, and finally drying at 40◦C to attain equilibrium 102 moisture content. Koya et al. (2015) undergone research to find the most efficient angle and direction 103 for the solar radiation when the solar power is introduced by using Geographic Information System 104 (GIS) in Kitakyushu city, Fukuoka Prefecture as an example. The results revealed that the south-facing 105 angle of 25 degrees is the most efficient angle on the solar power for Kitakyushu city, Fukuoka 106 Prefecture, Japan. The results are promising and required for the best utilization of solar energy in 107 Japan. The method is applicable anywhere in the world with the availability of GIS data access. 108 Subarkah & Hendrarsakti (2013) conducted experiments to determine if the drying rate interval of 109 anchovy fish shifts from steady to dropping when the essential moisture content is between 48.8 and 110 50.3 % (wet basis). Even at temperatures of 70 °C, there is no case-hardening in these studies. To 111 obtain the best results, the drying process should be done in three steps: drying at 70 °C for 150 112 minutes, drying at 50 °C for 100 minutes, and finally drying at 40 °C before equilibrium moisture 113 content is achieved. 114 To dry agricultural products Lingayat et al. (2017) designed and experimented a indirect solar dryer. 115 Solar dryers are made up of a solar flat plate air collector with V-corrugated absorption plates, an 116 enclosed drying cavity, and an exhaust air chimney. The collectors have a gross surface area of 2 m2. 117 The qualitative study for banana drying indicated that the moisture content of banana was decreased 118 With a moisture content of 48.8%-50.3% on a wet basis Subarkah & Hendrarsakti (2013) 98 experimented the effect of drying anchovy fish using solar drying chamber technique. Even at 99 temperature of 70◦C it was found that there is no case-hardening. In order to optimize the results, the 100 drying process is carried out in three different stages namely, drying the anchovy fish at 70◦C for 150 101 min followed by drying the fish at 50◦C for 100 min, and finally drying at 40◦C to attain equilibrium 102 moisture content. Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 An 64 excellent thermal insulator should possess less thermal conductivity. An increase in the thickness of 65 the thermal Insulator decreases the heat transfer. The Insulator is placed between the outer cover and 66 the inner cover. Some of the thermal insulator materials are polystyrene, glass wool, glass fiber, 67 polyurethane, mineral wool, wood, glass, cotton, paper, cork. The outer cover is used as the protection 68 layer for the dryer. The outer cover will give the shape for the dryer. The materials commonly used 69 for outer cover are wood, mild steel, galvanized iron sheet. The inner cover or absorber is the main 70 part of the dryer, and it is mostly painted in black for arresting the heat from the incident solar radiation. 71 The material primarily used are aluminum, mild steel, galvanized iron sheet. The glazing serves as an 72 essential part of transmitting the radiation inside the dryer for drying. The most commonly used 73 material is glass. The effect of absorber plate solar air collector on indirect solar drying was 74 experimentally investigated by Essalhi et al. (2017). The absorber plate was fixed with cylindrical fins 75 kept parallel to the absorber plate while the air flows over the cylindrical absorber of the solar collector 76 to dry the food products available in the solar dryer. Results showed that the effect of cylindrical 77 corrugations improved the drying chamber efficiency. On a 24- hour operation the mass of samples 78 kept inside the dryer was reduced from 997.3 to 135.13 g. 79 Eltawil et al. (2018) studied the enhancement of a solar tunnel dryer by coupling a flat plate collector 80 and a solar photovoltaic system to remove the moisture content available in peppermint effectively. 81 The photovoltaic panel is incorporated to run the DC fan for forced convection, while the direct solar 82 radiation is avoided by providing curtains. A comparison of open solar drying was made to study the 83 effect of tunnel dryer on single, double, and three layers of mint. With an energy payback time of 2 84 years, the dryer efficiency, overall efficiency, and photovoltaic efficiency were found as 30.7, 16.3, 85 and 9.3%, respectively. Also, it was estimated that the CO2 mitigation for the entire lifetime was 31.8 86 3 tons. Nabnean et al. (2016) experimentally analyzed an innovative solar dryer to dry the osmotically 87 dehydrated cherry tomatoes. Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 A comparison was carried out to assess the performance of solar dryer 88 over natural solar drying techniques. It was reported that there was a significant reduction in the time 89 of drying while compared to that of natural drying. Similarly, the efficiency of the modified solar dryer 90 was found in the range of 21-69 %. 91 tons. Nabnean et al. (2016) experimentally analyzed an innovative solar dryer to dry the osmotically 87 dehydrated cherry tomatoes. A comparison was carried out to assess the performance of solar dryer 88 over natural solar drying techniques. It was reported that there was a significant reduction in the time 89 of drying while compared to that of natural drying. Similarly, the efficiency of the modified solar dryer 90 was found in the range of 21-69 %. 91 tons. Nabnean et al. (2016) experimentally analyzed an innovative solar dryer to dry the osmotically 87 dehydrated cherry tomatoes. A comparison was carried out to assess the performance of solar dryer 88 over natural solar drying techniques. It was reported that there was a significant reduction in the time 89 of drying while compared to that of natural drying. Similarly, the efficiency of the modified solar dryer 90 was found in the range of 21-69 %. 91 Sengar et al. (2009) used a low-cost dryer to assess the performance, which was used for drying the 92 salted and unsalted fish products. A comparison was also made between the solar dryer and open solar 93 drying techniques. It was reported that the time taken to dry the fish using the proposed method was 8 94 hours, while the open solar drying technique required 15 hours to reach the same amount of moisture 95 content. It was also found that the average drying efficiency using salted and unsalted fish was 14 and 96 11%, respectively, and the overall thermal efficiency of the system was found as 70.97 %. 97 With a moisture content of 48.8%-50.3% on a wet basis Subarkah & Hendrarsakti (2013) 98 experimented the effect of drying anchovy fish using solar drying chamber technique. Even at 99 temperature of 70◦C it was found that there is no case-hardening. Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 Koya et al. (2015) undergone research to find the most efficient angle and direction 103 for the solar radiation when the solar power is introduced by using Geographic Information System 104 (GIS) in Kitakyushu city, Fukuoka Prefecture as an example. The results revealed that the south-facing 105 angle of 25 degrees is the most efficient angle on the solar power for Kitakyushu city, Fukuoka 106 Prefecture, Japan. The results are promising and required for the best utilization of solar energy in 107 Japan. The method is applicable anywhere in the world with the availability of GIS data access. 108 To dry agricultural products Lingayat et al. (2017) designed and experimented a indirect solar dryer. 115 Solar dryers are made up of a solar flat plate air collector with V-corrugated absorption plates, an 116 enclosed drying cavity, and an exhaust air chimney. The collectors have a gross surface area of 2 m2. 117 The qualitative study for banana drying indicated that the moisture content of banana was decreased 118 4 from 356 percent (db) to 16.32 % in tray 1, whereas, from tray 2, tray 3, tray 4 and solar opendrying 119 the moisture content available was found as 19.47 %, 21.15 %, 31.15 %, and 42.37 % respectively. 120 From the results it was found that the overall thermal efficiency of collected as 31.5 % whereas, the 121 chamber drying efficiency was found as 22.38%. The most critical and effective factor during drying 122 is the temperature of the drying air. The humidity of the air, as well as the velocity of the air, are 123 essential factors in improving the drying rate. An indirect solar dryer with forced circulation for drying 124 the commodities utilizing a DC powered fan was analysed by Chandrasekar et al.(2018). The air from 125 the solar collector was integrated to the drying chamber for improved kinetics of drying. 126 Swami et al. (2018) performed a comparative analysis on solar dryers using rock bed, pebbles as heat 127 storage to increase the efficiency of solar dryer with decreased drying time. To improve the overall 128 system performance, Daghigh and Shafieian (2016) integrated a heat recovery system to drying 129 chamber. The hot water from the storage tank is integrated with the help of a heat exchanger and the 130 air is blown into the dryer for improved efficiency. Experimental studies on natural convection open and closed solar 1 drying using external reflector 2 The hot air reaches the drying main chamber, which 131 holds the drying materials to dry completely. The mean outlet air temperature of the dryer was 132 approximately 44.3°C at a volumetric flow rate of 0.0328 m3/s. At the end of the day, the system's 133 exergetic performance reaching a peak, which was about 11.7 %. Azam et al. (2020) compared the 134 drying performance of PV and solar collector integrated green house dryer. Pretreatments of the 135 tomatoes was done before drying. Udomkun et al. (2020) reviewed the performance of different solar 136 dryers for agricultural products in Asian and African regions. Murali et al. (2020) designed a dryer 137 using solar-LPG dryer in a hybrid mode for drying shrimps. The moisture content available in the fresh 138 shrimps was reduced from initial moisture content of 76.71 % to the final moisture content of 15.38 139 %. Spall and Sethi (2020) designed a multi-rack solar dryer with reflectors for concentrating the solar 140 radiation. 141 5 5 151 52 Figure 2: Types of solar dryer 53 152 Figure 2: Types of solar dryer 153 Figure 2: Types of solar dryer Figure 2: Types of solar dryer 2. Experimental setup and procedure 154 p p p The conventional open solar drying is carried out by placing the required amount of anchovy fishes in 155 the basket and kept in the outdoor condition of Coimbatore. In the modified solar drying, the dryer 156 with an effective area of 0.152 m2 with width and height of 0.48 and 0.32 m respectively. The dryer is 157 installed with a capacity of handling 3 kg of fish. Additionally, to improve the drying efficiency, top 158 reflector is used to focus the solar radiation. Two trays are placed inside the dryer namely tray 1 and 159 tray 2. Tray 1 is placed at the top and tray 2 is placed below the tray 1. The dryer is oriented east-west, 160 and the reflector is oriented east-west. Two trays are placed inside the drying chamber and the product 161 is placed to dry. The sun's rays would reach the glazing region and be reflected within the dryer. Since 162 the black absorber is less reflective, heat is lost within the dryer. The trapped heat dries the put 163 substance. The air difference will frequently expel moist air from the dryer and replace it with fresh 164 air. 165 Experiments are carried out between 9:00 and 17:00 Hrs and the ambient temperature, dryer 166 temperature (tray 1 and tray 2), basket temperature, relative humidity (ambient and dryer) are 167 measured every hour. The wind velocity and solar intensity variations are measured using anemometer 168 and solar power meter respectively. Temperature are measured using PT100 type RTD sensors. An 169 electronic balance is used measure the weight of the anchovy fish while the relative humidity is 170 measured using thermo hygrometer. The specification of the instruments used in the present study is 171 given in Table. 1. 172 6 173 Figure 3: Graphical representation of modified solar drying technique with external reflector 174 Table 1: Specification of instruments used in the present study 175 S. No Name of instrument Range Error ( % ) 1 Solar power meter 0-2500 W/m2 ±3.5 % 2 Anemometer 0-45 m/s ±4 % 3 Thermo hygrometer 0-100% ±2 % 4 RTD sensors (PT-100 type) 0-1000◦C ±1.5 % 5 Electronic balance 0-2.500 kg ±0.001 % 3. Results and discussion 176 173 Figure 3: Graphical representation of modified solar drying technique with external reflector 174 Table 1: Specification of instruments used in the present study 175 S. 2. Experimental setup and procedure 154 No Name of instrument Range Error ( % ) 1 Solar power meter 0-2500 W/m2 ±3.5 % 2 Anemometer 0-45 m/s ±4 % 3 Thermo hygrometer 0-100% ±2 % 4 RTD sensors (PT-100 type) 0-1000◦C ±1.5 % 5 Electronic balance 0-2.500 kg ±0.001 % 3 Results and discussion 176 173 173 Figure 3: Graphical representation of modified solar drying technique with external reflector 174 Table 1: Specification of instruments used in the present study 175 S. No Name of instrument Range Error ( % ) 1 Solar power meter 0-2500 W/m2 ±3.5 % 2 Anemometer 0-45 m/s ±4 % 3 Th h 0 100% 2 % 173 Figure 3: Graphical representation of modified solar drying technique with external reflector 174 Table 1: Specification of instruments used in the present study 175 S. No Name of instrument Range Error ( % ) 1 Solar power meter 0-2500 W/m2 ±3.5 % 2 Anemometer 0-45 m/s ±4 % 3 Thermo hygrometer 0-100% ±2 % 4 RTD sensors (PT-100 type) 0-1000◦C ±1.5 % 5 Electronic balance 0-2.500 kg ±0.001 % 3. Results and discussion 176 raphical representation of modified solar drying technique with external reflector Figure 3: Graphical representation of modified solar drying technique with extern Table 1: Specification of instruments used in the present study 3. Results and discussion 176 The effect of temperature and humidity has a vital part in the drying. For perfect drying, the 177 temperature should be high and the humidity should be less. It is observed that the temperature inside 178 the dryer tends to be more than the open drying temperature. The humidity is less for dryer drying 179 compared to open drying. The temperature changes from time to time according to the incident solar 180 radiation. The high temperature and low humidity inside the dryer is responsible for quick drying of 181 products. But the temperature should not be so high because it will spoil the drying of the products. 182 The hourly variation of wind velocity, ambient temperature, tray temperature, relative humidity, and 183 solar intensity during experimental days using conventional open solar dryer is shown in Fig. 4. During 184 the experimental days, it is observed that the relative humidity of open basket and the ambient is almost 185 7 equal and the temperature of basket is increasing steadily with respect to the incident rays falling on 186 the fish. 2. Experimental setup and procedure 154 On increased solar intensity the relative humidity measured from the basket and ambient 187 decreases. The relative humidity of day 2 is found to be in the ranges of 50-65%, while the relative 188 humidity of day 3 and day 4 are found in the ranges of 35-55 and 37-56% respectively. With lower 189 humidity, the temperature of basket reaches a peak value of 45 and 48◦C during day 3 and day 4 190 respectively. With higher moisture content available in the basket, the peak temperature recorded 191 during day 1 and day 2 are found as 44 and 40◦C while the relative humidity recorded during the peak 192 solar intensity for the days are found as 50 and 48% respectively. 193 equal and the temperature of basket is increasing steadily with respect to the incident rays falling on 186 the fish. On increased solar intensity the relative humidity measured from the basket and ambient 187 decreases. The relative humidity of day 2 is found to be in the ranges of 50-65%, while the relative 188 humidity of day 3 and day 4 are found in the ranges of 35-55 and 37-56% respectively. With lower 189 humidity, the temperature of basket reaches a peak value of 45 and 48◦C during day 3 and day 4 190 respectively. With higher moisture content available in the basket, the peak temperature recorded 191 during day 1 and day 2 are found as 44 and 40◦C while the relative humidity recorded during the peak 192 solar intensity for the days are found as 50 and 48% respectively. 193 194 195 194 195 195 8 196 197 196 197 196 197 9 198 Figure 4: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 199 humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional 200 open solar dryer 201 Similarly, the hourly variation of wind velocity, ambient temperature, tray temperature, relative 202 humidity, and solar intensity during the experimental days using modified solar dryer is shown in Fig. 203 5. It can be clearly seen from the experimental days that the influence of dryer configuration with 204 reflectors increased the temperature of tray to a maximum of 60, 58 and 65◦C with peak solar intensity 205 in tray 1 for day 1, day 2 and day 3 respectively. 2. Experimental setup and procedure 154 On increased drying condition of fishes in the tray, 206 the relative humidity inside the dryer reduced with increased solar intensity and tray temperature. The 207 relative humidity measured inside the dryer ranged from 35-65% during the day 1 operation while the 208 day 2 and day 3 showed 23-55 and 21-50 % respectively. 209 198 Figure 4: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 199 humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional 200 open solar dryer 201 198 Figure 4: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 199 humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional 200 open solar dryer 201 Figure 4: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 199 humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional 200 open solar dryer 201 Similarly, the hourly variation of wind velocity, ambient temperature, tray temperature, relative 202 humidity, and solar intensity during the experimental days using modified solar dryer is shown in Fig. 203 5. It can be clearly seen from the experimental days that the influence of dryer configuration with 204 reflectors increased the temperature of tray to a maximum of 60, 58 and 65◦C with peak solar intensity 205 in tray 1 for day 1, day 2 and day 3 respectively. On increased drying condition of fishes in the tray, 206 the relative humidity inside the dryer reduced with increased solar intensity and tray temperature. The 207 relative humidity measured inside the dryer ranged from 35-65% during the day 1 operation while the 208 day 2 and day 3 showed 23-55 and 21-50 % respectively. 3.1. Moisture content 225 Moisture content lost during drying process is defined as the ratio of difference in moisture content 226 before and after drying to the initial moisture content available in the product. Mathematically it is 227 expressed as (Akoy et al. (2006); Khalifa et al. (2012); Panchal et al. (2013)), 228 1 2 1 Total moisture content loss = 100 % M M M   (1) 229 Where M1 and M2 are the sample mass before and after drying (kg) 230 1 2 1 Total moisture content loss = 100 % M M M   (1) 229 1 2 1 Total moisture content loss = 100 % M M M   (1) 229 (1) 1 M Where, M1 and M2 are the sample mass before and after drying (kg) 30 For every two hours the samples are weighed in the electronic balance and mass of fish is recorded. 232 The variation in mass of fish during the drying process using conventional open solar drying and 233 modified solar drying is plotted in Fig. 8. It can be observed that the mass of fish using open solar 234 drying is more while compared to mass of fish in tray 1 and tray 2. 235 While drying with different seasoning time period, the mass of anchovy fish is reduced exponentially 236 with respect to time period as the moisture level in the fish is removed by drying with solar intensity. 237 Also, it is observed that the mass of anchovy fish inside tray 2 is higher than tray 1 as it is directly 238 exposed to solar radiation and additional solar intensity is focused to heat the product with external 239 reflector. The amount of time taken to dry the 1.6 kg of anchovy fish using open solar drying technique 240 is found as 96 hours while the solar dryer required an time of 56 hours. The variations in moisture 241 removal from the anchovy fish using conventional open solar drying and modified solar drying is 242 plotted in Fig. 9. It is observed that the initial moisture content available in the anchovy fish during 243 conventional open solar drying and modified solar dryer with reflector is same at the start of 244 experiment. On increasing the drying time, the moisture content available in the fish starts to decrease. 2. Experimental setup and procedure 154 209 10 210 211 210 0 1 211 211 11 212 Figure 5: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 213 humidity, and solar intensity during (a) Day 1 (b) Day 2 and (3) Day 3 using solar dryer 214 215 212 Figure 5: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 213 humidity, and solar intensity during (a) Day 1 (b) Day 2 and (3) Day 3 using solar dryer 214 215 Figure 5: Hourly variation of wind velocity, ambient temperature, tray temperature, relative 213 humidity, and solar intensity during (a) Day 1 (b) Day 2 and (3) Day 3 using solar dryer 214 216 Figure 6: Conventional solar drying of anchovy fish (a) before and (b) after 217 216 Figure 6: Conventional solar drying of anchovy fish (a) before and (b) after 217 218 Figure 6: Conventional solar drying of anchovy fish (a) before and (b) after Fig. 6 and 7 shows the conventional open solar drying and modified solar drying of anchovy fish 219 before and after drying respectively. It is observed that the product which is opened dried has loosen 220 more colour when compared to dryer drying. The colour of the product will retain when it has indirect 221 exposure to the sunlight for drying. 222 12 223 Figure 7: Drying of anchovy fish using modified solar dryer (a) before and (b) after 224 Figure 7: Drying of anchovy fish using modified solar dryer (a) before and (b) afte 3.1. Moisture content 225 245 While drying with different seasoning time period, the mass of anchovy fish is reduced exponentially 236 with respect to time period as the moisture level in the fish is removed by drying with solar intensity. 237 Also, it is observed that the mass of anchovy fish inside tray 2 is higher than tray 1 as it is directly 238 exposed to solar radiation and additional solar intensity is focused to heat the product with external 239 reflector. The amount of time taken to dry the 1.6 kg of anchovy fish using open solar drying technique 240 is found as 96 hours while the solar dryer required an time of 56 hours. The variations in moisture 241 removal from the anchovy fish using conventional open solar drying and modified solar drying is 242 plotted in Fig. 9. It is observed that the initial moisture content available in the anchovy fish during 243 conventional open solar drying and modified solar dryer with reflector is same at the start of 244 experiment. On increasing the drying time, the moisture content available in the fish starts to decrease. 245 13 The amount of moisture present in the anchovy fish available in tray 1 and tray 2 are found as 23.3 ad 246 27.2% respectively. The moisture content in tray 2 is higher to about 3.9% than the moisture present 247 in tray 1. This is due to the thermal barrier by the insulator placed in tray 1 and also, tray 2 is kept just 248 beneath the tray 1. Moreover tray 1 is exposed directly to the radiation from the sun and the reflected 249 radiation from the mirror reduced the moisture content present in anchovy fish. On comparing the 250 moisture content available in conventional open solar drying technique, the moisture content available 251 in the fish is found as 32.68 %. 252 The amount of moisture present in the anchovy fish available in tray 1 and tray 2 are found as 23.3 ad 246 27.2% respectively. The moisture content in tray 2 is higher to about 3.9% than the moisture present 247 in tray 1. This is due to the thermal barrier by the insulator placed in tray 1 and also, tray 2 is kept just 248 beneath the tray 1. 3.1. Moisture content 225 Moreover tray 1 is exposed directly to the radiation from the sun and the reflected 249 radiation from the mirror reduced the moisture content present in anchovy fish. On comparing the 250 moisture content available in conventional open solar drying technique, the moisture content available 251 in the fish is found as 32.68 %. 252 253 Figure 8: Variation in mass of fish during the drying process using conventional open solar drying 254 and modified solar dryer 255 253 Figure 8: Variation in mass of fish during the drying process using conventional open solar drying 254 and modified solar dryer 255 Figure 8: Variation in mass of fish during the drying process using conventional open solar drying 254 and modified solar dryer 255 3.2. Drying rate 256 The drying rate is defined as the difference in the amount of moisture content available before drying 257 and after drying to the total time taken for drying the sample. Mathematically it is expressed as (Akoy 258 et al. (2006); Khalifa et al. (2012); Panchal et al. (2013)), 259 1 2 Drying rate d M M t   (2) 260 (2) Where, td is the time for drying sample (days) 261 Where, td is the time for drying sample (days) 261 Where, td is the time for drying sample (days) 261 14 262 Figure 9: Variation in moisture content using conventional open solar drying and modified solar 263 drying 264 Figure 9: Variation in moisture content using conventional open solar drying and modified solar 263 drying 264 The drying rate of both the open drying and dryer dying is compared for fishes. The drying rate of 265 fishes at open drying and dryer dying is 0.012 kg/h and 0.031 kg/h respectively (Table. 2). On 266 comparing both the values, it is clear that the drying rate of dryer drying is much higher than the open 267 drying. The time taken for drying the product is also less for dryer drying when compared to the open 268 drying. 269 3.3. Dryer efficiency 3.3. Dryer efficiency 270 The dryer efficiency of the solar dryer using proposed modification is mathematically expressed as 271 (Akoy et al. (2006); Khalifa et al. (2012); Panchal et al. (2013)), 272 The dryer efficiency of the solar dryer using proposed modification is mathematically expressed as 271 (Akoy et al. (2006); Khalifa et al. (2012); Panchal et al. (2013)), 272   1 2 d Thermal efficiency of dryer, ( ) l d d M M h A I t t       (3) 273   1 2 d Thermal efficiency of dryer, ( ) l d d M M h A I t t       (3) 273 (3) 15 From the experiments it is found that due to direct exposure of the fish with direct solar radiation and 279 reflected radiation from the mirrors, the upper tray (dryer 2) exhibits higher thermal efficiency than 280 lower tray (dryer 1). The average daily efficiency of the modified solar dryer with external reflectors 281 were found as 19.34 and 16.73% from dryer 2 and dryer 1 respectively. 3.2. Drying rate 256 282 From the experiments it is found that due to direct exposure of the fish with direct solar radiation and 279 reflected radiation from the mirrors, the upper tray (dryer 2) exhibits higher thermal efficiency than 280 lower tray (dryer 1). The average daily efficiency of the modified solar dryer with external reflectors 281 were found as 19.34 and 16.73% from dryer 2 and dryer 1 respectively. 282 16 were found as 19.34 and 16.73% from dryer 2 and dryer 1 respectively. 282 3.4. Heat energy for drying 283 The heat energy used for drying the fish is defined as the product of dryer efficiency, latent heat of 284 vaporization and mass of water removed. Mathematically it is expressed as (Akoy et al. (2006); Khalifa 285 et al. (2012)), 286 Heat used for drying =mw × h1 × ηd (4) 287 Where, ηd – Efficiency of dryer ( s), Mw – amount of water removed ( kg ) 288 Similarly, the latent heat of vaporization is estimated using Equation (5) (Youcef-Ali et al. (2001); 289 Akoy et al. (2006); Khalifa et al. (2012)), 290     3 Latent heat of vaporization = 4.186 10 597 0.56 p T    (5) 291 Where, Tp- is the temperature of product (K) 292 Based on the amount of water removed from the dryer, dryer efficiency and latent heat, the amount of 293 heat for drying the estimated for solar dryer drying is found as 362.3 kJ. 294 Table. 2 Comparison of drying rate, drying efficiency, energy saving time, amount of heat for drying, 295 and heat conduction rate through dryer insulation 296 Date of experiment Drying rate (kg/hr) Drying efficiency (%) Energy saving time (%) Amount of heat for drying (kJ) Heat conduction rate through dryer insulation (kJ) Dryer Open solar drying Dryer Open solar drying 07.05.2019 0.031 0.012 19.37 15.73 20.05 362.3 365.09 10.05.2019 0.029 0.013 21.05 15.85 21.20 361.2 364.24 15.05.2019 0.030 0.012 20.04 14.95 21.35 362.4 365.14 20.05.2019 0.028 0.013 21.34 15.93 21.50 360.2 363.15 25.05.2019 0.027 0.011 21.39 15.51 22.14 361.2 364.28 01.06.2019 0.025 0.015 23.54 14.21 22.25 362.1 363.17 The heat energy used for drying the fish is defined as the product of dryer efficiency, latent heat of 284 vaporization and mass of water removed. Mathematically it is expressed as (Akoy et al. (2006); Khalifa 285 et al. 3.2. Drying rate 256 (2012)), 286 The heat energy used for drying the fish is defined as the product of dryer efficiency, latent heat of 284 vaporization and mass of water removed. Mathematically it is expressed as (Akoy et al. (2006); Khalifa 285 et al. (2012)), 286 Heat used for drying =mw × h1 × ηd (4) 287 Where, ηd – Efficiency of dryer ( s), Mw – amount of water removed ( kg ) 288 Similarly, the latent heat of vaporization is estimated using Equation (5) (Youcef-Ali et al. (2001); 289 Akoy et al. (2006); Khalifa et al. (2012)), 290 (4)     3 Latent heat of vaporization = 4.186 10 597 0.56 p T    (5) 291 Where T is the temperature of product (K) 292     3 Latent heat of vaporization = 4.186 10 597 0.56 p T    (5) 291 16 Where, Tp is the temperature of product (K) 292 Based on the amount of water removed from the dryer, dryer efficiency and latent heat, the amount of 293 heat for drying the estimated for solar dryer drying is found as 362.3 kJ. 294 Table. 2 Comparison of drying rate, drying efficiency, energy saving time, amount of heat for drying, 295 and heat conduction rate through dryer insulation 296 Date of experiment Drying rate (kg/hr) Drying efficiency (%) Energy saving time (%) Amount of heat for drying (kJ) Heat conduction rate through dryer insulation (kJ) Dryer Open solar drying Dryer Open solar drying 07.05.2019 0.031 0.012 19.37 15.73 20.05 362.3 365.09 10.05.2019 0.029 0.013 21.05 15.85 21.20 361.2 364.24 15.05.2019 0.030 0.012 20.04 14.95 21.35 362.4 365.14 20.05.2019 0.028 0.013 21.34 15.93 21.50 360.2 363.15 25.05.2019 0.027 0.011 21.39 15.51 22.14 361.2 364.28 01.06.2019 0.025 0.015 23.54 14.21 22.25 362.1 363.17 Based on the amount of water removed from the dryer, dryer efficiency and latent heat, the amount of 293 heat for drying the estimated for solar dryer drying is found as 362.3 kJ. 294 16 3.5. Percentage of time saving for solar drying 297 The heat energy used for drying the fish is defined as the product of dryer efficiency, latent heat of 298 vaporization and mass of water removed. Mathematically it is expressed as (Akoy et al. (2006); Khalifa 299 et al. 3.2. Drying rate 256 (2012)), 300 Energy time saving from dryer = 100 % open dryer open T T T   (5) 301 Energy time saving from dryer = 100 % open dryer open T T T   (5) 301 (5) topen – time for drying in open solar drying (hrs) 303 tdryer – time for drying in solar dryer drying (hrs) 304 topen – time for drying in open solar drying (hrs) 303 tdryer – time for drying in solar dryer drying (hrs) 304 Using the modified 305 The rate of heat conduction through the insulator plays a vital role in heat transfer between the trays 307 for efficient drying the fishes. It depends of the temperature inside the dryer, thermal conductivity of 308 the insulator, ambient temperature and area of the tray. Mathematically it is given as (Holman (2008)), 309   Heat conduction rate through Insulator = d d a kA T T t  (6) 310   Heat conduction rate through Insulator = d d a kA T T t  (6) 310 (6) (6) Td – Drying chamber temperature ( K), 312 Td – Drying chamber temperature ( K), 312 Td – Drying chamber temperature ( K), 312 Td Drying chamber temperature ( K), 312 Ta – Surrounding ambient temperature ( K), 313 t – Thermal insulator thickness ( m ) 314 The amount of heat conduction rate through the dryer insulation provided is provided in Table. 2. It is 315 seen that the rate of conduction through the dryer insulation is found as 365 kJ. 316 4. Conclusions 317 The following conclusions are derived from the experimental investigation: 318 • The drying is carried out using conventional open solar drying and modified solar dryer techniques 319 for drying anchovy fish. 320 • The product dried in dryer is free from dust and insects and also the loss of colour in the product 321 is less compared to open drying. 322 • The drying thermal efficiency of dryer 1 is 16.73% and the drying thermal efficiency of dryer 2 is 323 19.34%. The drying thermal efficiency is found to be higher in the dryer 2. 324 325 326 Ta – Surrounding ambient temperature ( K), 313 The amount of heat conduction rate through the dryer insulation provided is provided in Table. 2. 3.2. Drying rate 256 It is 315 seen that the rate of conduction through the dryer insulation is found as 365 kJ. 316 The amount of heat conduction rate through the dryer insulation provided is provided in Table. 2. It is 315 seen that the rate of conduction through the dryer insulation is found as 365 kJ. 316 The following conclusions are derived from the experimental investigation: 318 • The drying is carried out using conventional open solar drying and modified solar dryer techniques 319 for drying anchovy fish. 320 • The drying thermal efficiency of dryer 1 is 16.73% and the drying thermal efficiency of dryer 2 is 323 19.34%. The drying thermal efficiency is found to be higher in the dryer 2. 324 17 18 Ethical Approval 327 Not Applicable 328 Consent to Participate 329 Not Applicable 330 Consent to Publish 331 Not Applicable 332 Authors Contributions 333 Conceptualization, Methodology, Resources, Formal anal 334 preparation, review and editing, Supervision and investigation 335 Kabeel, Ravishankar Sathyamurthy, Ali Chamkha 336 Writing - original draft preparation, review and editing were 337 Dharmadurai, Sathiyaseelan Vasanthaseelan, Bharathwaaj 338 Manokar 339 Funding 340 There is no funding received for the research work carried out 341 Competing Interests 342 The authors declare that there is no competing interest 343 Availability of data and materials 344 Not Applicable 345 References 346 Akoy, E., Ismail, M. A., Ahmed, E. F. A., & Luecke, W. (2006). D 347 dryer for mango slices. Proceedings of International Research on 348 Management and Rural Development-Tropentag. University of B 349 Azam, M. M., Eltawil, M. A., & Amer, B. M. (2020). Thermal analys 350 integrated with greenhouse dryer for drying tomatoes. Energy, 21 351 Bal, L. M., Satya, S., & Naik, S. N. (2010). Solar dryer with thermal 352 agricultural food products: A review. Renewable and Sustainab 353 2314. 354 Chandrasekar, M., Senthilkumar, T., Kumaragurubaran, B., & Fern 355 investigation on a solar dryer integrated with condenser unit o 356 enhancing drying rate. Renewable Energy, 122, 375-381. 3.2. Drying rate 256 An experimental study of a heat pipe evacuated tube solar dryer 358 with heat recovery system. Renewable Energy, 96, 872-880. 359 with heat recovery system. Renewable Energy, 96, 872-880. 359 Dharmadurai, P. L., Vasanthaseelan, S., Bharathwaaj, R., Dharmaraj, V., Gnanasekaran, K., Balaji, 360 D & Sathyamurthy R (2020) A comparative study on solar dryer using external reflector for 361 Dharmadurai, P. L., Vasanthaseelan, S., Bharathwaaj, R., Dharmaraj, V., Gn 60 Dharmadurai, P. L., Vasanthaseelan, S., Bharathwaaj, R., Dharmaraj, V., Gnanasekaran, K., Balaji, 360 D., & Sathyamurthy, R. (2020). A comparative study on solar dryer using external reflector for 361 drying grapes. Materials Today: Proceedings. 362 D., & Sathyamurthy, R. (2020). A comparative study on solar dryer using external reflector for 361 drying grapes. Materials Today: Proceedings. 362 Ekechukwu, O. V., & Norton, B. (1999). Review of solar-energy drying systems ii: an overview of 363 solar drying technology. Energy conversion and management, 40, 615–655. 364 Eltawil, M. A., Azam, M. M., & Alghannam, A. O. (2018). Energy analysis of hybrid solar tunnel 365 dryer with pv system and solar collector for drying mint (menthaviridis). Journal of Cleaner 366 Production, 181, 352–364. 367 Essalhi, H., Tadili, R., & Bargach, M. N. (2017). Conception of a solar air collector for an indirect 368 solar dryer. pear drying test. Energy Procedia, 141, 29–33. 369 , , , , g , ( ) p solar dryer. pear drying test. Energy Procedia, 141, 29–33. 369 Fudholi, A., Sopian, K., Ruslan, M. H., Alghoul, M., & Sulaiman, M. Y. (2010). Review of solar 370 dryers for agricultural and marine products. Renewable and sustainable energy reviews, 14, 1–30. 371 Holman J P (2008) Heat Transfer (Si Units) Sie Tata McGraw Hill Education 372 Fudholi, A., Sopian, K., Ruslan, M. H., Alghoul, M., & Sulaiman, M. Y. ( 70 Fudholi, A., Sopian, K., Ruslan, M. H., Alghoul, M., & Sulaiman, M. Y. (2010). Review of solar 370 dryers for agricultural and marine products. Renewable and sustainable energy reviews, 14, 1–30. 371 dryers for agricultural and marine products. Renewable and sustainable energy reviews, 14, 1–30. 371 Holman, J. P. (2008). Heat Transfer (Si Units) Sie. Tata McGraw-Hill Education. Jairaj, K., Singh, S., & Srikant, K. (2009). A review of solar dryers developed for grape drying. Solar 373 energy, 83, 1698–1712. 374 Khalifa, A. J. N., Al-Dabagh, A. M., & Al-Mehemdi, W. M. (2012). 3.2. Drying rate 256 357 Ethical Approval 327 Not Applicable 328 Consent to Participate 329 Not Applicable 330 Consent to Publish 331 Not Applicable 332 Authors Contributions 333 Conceptualization, Methodology, Resources, Formal analysis, Writing - original draft 334 preparation, review and editing, Supervision and investigation were carried out by Abd Elnaby 335 Kabeel, Ravishankar Sathyamurthy, Ali Chamkha 336 Writing - original draft preparation, review and editing were carried out by Paul Durai Leon 337 Dharmadurai, Sathiyaseelan Vasanthaseelan, Bharathwaaj Ramani, Athikesavan Muthu 338 Manokar 339 Funding 340 There is no funding received for the research work carried out. 341 Competing Interests 342 The authors declare that there is no competing interest 343 Availability of data and materials 344 Not Applicable 345 18 Ethical Approval 327 Not Applicable 328 Consent to Participate 329 Not Applicable 330 Consent to Publish 331 Not Applicable 332 Authors Contributions 333 Conceptualization, Methodology, Resources, Formal analysis, Writing - original draft 334 preparation, review and editing, Supervision and investigation were carried out by Abd Elnaby 335 Kabeel, Ravishankar Sathyamurthy, Ali Chamkha 336 Writing - original draft preparation, review and editing were carried out by Paul Durai Leon 337 Dharmadurai, Sathiyaseelan Vasanthaseelan, Bharathwaaj Ramani, Athikesavan Muthu 338 Manokar 339 Funding 340 There is no funding received for the research work carried out. 341 Competing Interests 342 The authors declare that there is no competing interest 343 Availability of data and materials 344 Not Applicable 345 References 346 Akoy, E., Ismail, M. A., Ahmed, E. F. A., & Luecke, W. (2006). Design and construction of a solar 347 dryer for mango slices. Proceedings of International Research on Food Security, Natural Resource 348 Management and Rural Development-Tropentag. University of Bonn, Bonn, Germany. 349 Azam, M. M., Eltawil, M. A., & Amer, B. M. (2020). Thermal analysis of PV system and solar collector 350 integrated with greenhouse dryer for drying tomatoes. Energy, 212, 118764. 351 Bal, L. M., Satya, S., & Naik, S. N. (2010). Solar dryer with thermal energy storage systems for drying 352 agricultural food products: A review. Renewable and Sustainable Energy Reviews, 14(8), 2298- 353 2314. 354 Chandrasekar, M., Senthilkumar, T., Kumaragurubaran, B., & Fernandes, J. P. (2018). Experimental 355 investigation on a solar dryer integrated with condenser unit of split air conditioner (A/C) for 356 enhancing drying rate. Renewable Energy, 122, 375-381. 357 18 Daghigh, R., & Shafieian, A. (2016). 3.2. Drying rate 256 An experimental study of vegetable 375 solar drying systems with and without auxiliary heat. International Scholarly Research 376 Notices, 2012. 377 Khalifa, A. J. N., Al-Dabagh, A. M., & Al-Mehemdi, W. M. (2012). An experimental study of vegetable 375 solar drying systems with and without auxiliary heat. International Scholarly Research 376 Khalifa, A. J. N., Al-Dabagh, A. M., & Al-Mehemdi, W. M. (2012). An experimental study of vegetable 375 solar drying systems with and without auxiliary heat. International Scholarly Research 376 Notices, 2012. 377 Koya, Y., Shiota, A., Mitani, Y., Qudaih, Y. S., & Fuji, K. (2015). An index to evaluate the amount of 378 the solar radiation for a surface with eight directions. International Journal of Smart Grid and Clean 379 Energy, 4, 241–246. 380 Kumar, M., Sansaniwal, S. K., & Khatak, P. (2016). Progress in solar dryers for drying various 381 commodities. Renewable and Sustainable Energy Reviews, 55, 346–360. 382 Lingayat, A., Chandramohan, V. P., & Raju, V. R. K. (2017). Design, development and performance 383 of indirect type solar dryer for banana drying. Energy Procedia, 109, 409-416. 384 Mohana, Y., Mohanapriya, R., Anukiruthika, T., Yoha, K. S., Moses, J. A., & Anandharamakrishnan, 385 C. (2020). Solar dryers for food applications: Concepts, designs, and recent advances. Solar 386 Energy, 208, 321-344. 387 Kumar, M., Sansaniwal, S. K., & Khatak, P. (2016). Progress in solar dryers for drying various 381 commodities. Renewable and Sustainable Energy Reviews, 55, 346–360. 382 commodities. Renewable and Sustainable Energy Reviews, 55, 346 360. 382 Lingayat, A., Chandramohan, V. P., & Raju, V. R. K. (2017). Design, development and performance 383 of indirect type solar dryer for banana drying. Energy Procedia, 109, 409-416. 384 Mohana, Y., Mohanapriya, R., Anukiruthika, T., Yoha, K. S., Moses, J. A., & Anandharamakrishnan, 385 C. (2020). Solar dryers for food applications: Concepts, designs, and recent advances. Solar 386 Energy, 208, 321-344. 387 19 Murali, S., Amulya, P. R., Alfiya, P. V., Delfiya, D. A., & Samuel, M. P. (2020). Design and 388 performance evaluation of solar-LPG hybrid dryer for drying of shrimps. Renewable Energy, 147, 389 2417-2428. 390 Mustayen, A. G. M. B., Mekhilef, S., & Saidur, R. (2014). Performance study of different solar dryers: 391 A review. Renewable and Sustainable Energy Reviews, 34, 463-470. 392 Mustayen, A. G. M. B., Mekhilef, S., & Saidur, R. (2014). Performance study of different solar dryers: 391 A review. 3.2. Drying rate 256 Renewable and Sustainable Energy Reviews, 34, 463-470. 392 Mustayen, A., Mekhilef, S., & Saidur, R. (2014). Performance study of different solar dryers: A 393 review. Renewable and Sustainable Energy Reviews, 34, 463–470. 394 Mustayen, A., Mekhilef, S., & Saidur, R. (2014). Performance study of different solar dryers: A 393 review. Renewable and Sustainable Energy Reviews, 34, 463–470. 394 review. Renewable and Sustainable Energy Reviews, 34, 463–470. 394 review. Renewable and Sustainable Energy Reviews, 34, 463–470. 394 Nabnean, S., Janjai, S., Thepa, S., Sudaprasert, K., Songprakorp, R., & Bala, B 95 Nabnean, S., Janjai, S., Thepa, S., Sudaprasert, K., Songprakorp, R., & Bala, B. (2016). Experimental 395 performance of a new design of solar dryer for drying osmotically dehydrated cherry tomatoes. 396 Renewable energy, 94, 147–156. 397 Panchal, S., Solanki, S. K., Tilkar, A. K., & Nagaich, R. (2013). Design, construction and testing of 398 Panchal, S., Solanki, S. K., Tilkar, A. K., & Nagaich, R. (2013). Design, construction and testing of 398 solar dryer with roughened surface solar air heater. International journal of innovative Research in 399 Engineering and Science, 7(2), 7-12. 400 Panchal, S., Solanki, S. K., Tilkar, A. K., & Nagaich, R. (2013). Design, construction and testing of 398 solar dryer with roughened surface solar air heater. International journal of innovative Research in 399 Engineering and Science, 7(2), 7-12. 400 Sengar, S., Khandetod, Y., & Mohod, A. (2009). Low cost solar dryer for fish. African Journal of 401 Environmental Science and Technology, 3. 402 Sengar, S., Khandetod, Y., & Mohod, A. (2009). Low cost solar dryer for fish. African Journal of 401 Environmental Science and Technology, 3. 402 Spall, S., & Sethi, V. P. (2020). Design, modeling and analysis of efficient multi-rack tray solar cabinet 403 dryer coupled with north wall reflector. Solar Energy, 211, 908-919. 404 Spall, S., & Sethi, V. P. (2020). Design, modeling and analysis of efficient multi-rack tray solar cabinet 403 dryer coupled with north wall reflector. Solar Energy, 211, 908-919. 404 Subarkah, R., & Hendrarsakti, J. (2013). Drying characteristic of anchovy fish. Journal of Food Science 405 and Engineering, 3, 87. 406 Subarkah, R., & Hendrarsakti, J. (2013). Drying characteristic of anchovy fish. Journal of Food Science 405 and Engineering, 3, 87. 406 Subarkah, R., & Hendrarsakti, J. (2013). Drying characteristic of anchovy fish. Journal of Food Science 407 Swami, V. M., Autee, A. T., & Anil, T. R. (2018). 3.2. Drying rate 256 Experimental analysis of solar fish dryer using phase 409 change material. Journal of Energy Storage, 20, 310-315. 410 Swami, V. M., Autee, A. T., & Anil, T. R. (2018). Experimental analysis of solar fish dryer using phase 409 change material. Journal of Energy Storage, 20, 310-315. 410 Udomkun, P., Romuli, S., Schock, S., Mahayothee, B., Sartas, M., Wossen, T., Njukwe, E., Vanlauwe, 411 B. and Müller, J., 2020. Review of solar dryers for agricultural products in Asia and Africa: An 412 innovation landscape approach. Journal of Environmental Management, 268, p.110730. 413 Youcef-Ali, S., Messaoudi, H., Desmons, J. Y., Abene, A., & Le Ray, M. (2001). Determination of the 414 average coefficient of internal moisture transfer during the drying of a thin bed of potato 415 slices. Journal of food engineering, 48(2), 95-101. 416 417 20 Figures Figure 1 Drying ¦sh in Open Sun light in India Drying ¦sh in Open Sun light in India Figure 2 Figure 3 Graphical representation of modi¦ed solar drying technique with external re§ector Figure 3 Figure 3 Graphical representation of modi¦ed solar drying technique with external re§ector Graphical representation of modi¦ed solar drying technique with external re§ector Figure 4 Hourly variation of wind velocity, ambien Figure 4 Hourly variation of wind velocity, ambient temperature, tray temperature, relative humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional open solar dryer Figure 4 Figure 4 Hourly variation of wind velocity, ambient temperature, tray temperature, relative humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional open solar dryer Hourly variation of wind velocity, ambient temperature, tray temperature, relative humidity, and solar intensity during (a) Day 1 (b) Day 2 (3) Day 3 and (4) Day 4 using conventional open solar dryer Figure 5 Hourly variation of wind velocity, ambient temperature, tray temperature, relative humidity, and solar intensity during (a) Day 1 (b) Day 2 and (3) Day 3 using solar dryer Figure 5 Hourly variation of wind velocity, ambient temperature, tray temperatur intensity during (a) Day 1 (b) Day 2 and (3) Day 3 using solar dryer Figure 5 Hourly variation of wind velocity, ambient temperature, tray temperature, relative humidity, and solar intensity during (a) Day 1 (b) Day 2 and (3) Day 3 using solar dryer Figure 6 Conventional solar drying of anchovy ¦sh (a) before and (b) after Figure 7 Drying of anchovy ¦sh using modi¦ed solar dryer (a) before and (b) after Figure 6 Conventional solar drying of anchovy ¦sh (a) before and (b) after Figure 6 Figure 6 Conventional solar drying of anchovy ¦sh (a) before and (b) after Conventional solar drying of anchovy ¦sh (a) before and (b) after Figure 7 Drying of anchovy ¦sh using modi¦ed solar dryer (a) before and (b) after Figure 7 Drying of anchovy ¦sh using modi¦ed solar dryer (a) before and (b) after Drying of anchovy ¦sh using modi¦ed solar dryer (a) before and (b) after igure 8 ariation in mass of ¦sh during the drying process using conventional open solar drying and modi¦ed olar dryer Figure 8 Variation in mass of ¦sh during the drying process using conventional open solar drying and modi¦ed solar dryer Figure 9 Variation in moisture content using conventional open solar drying and modi¦ed solar drying Figure 9 Variation in moisture content using conventional open solar drying and modi¦ed solar drying Variation in moisture content using conventional open solar drying and modi¦ed solar drying Variation in moisture content using conventional open solar drying and modi¦ed solar drying
https://openalex.org/W4383961156	https://fis.uni-bamberg.de/bitstream/uniba/59940/1/fisba59940.pdf	English	null	‘Stay at home order’ : A COVID-19’s child of necessity with implications for the ‘Female’ in Yoruba (Nigeria) home settings	Bible in Africa studies	2,021	cc-by	6,282	10.20378/irb-59940 10.20378/irb-59940 10.20378/irb-59940 Abstract Studies on COVID–19 regarding its cause, effects and remedies are many. However, insignificant attention is paid to the implications of the effect of the disease on the ‘fe- male’, as noticed in some extant studies. This work examines the ‘Stay at Home Order’ as one of the mechanisms put in place to minimise the spread of COVID-19 with a view to ascertaining whether or not the act is a child of necessity in Yoruba (Nigeria) home settings. Research data were gathered through telephone interviews with 50 families, cou- pled with several sessions of observation in select streets of Ibadan. Consequently, we found out that there was division of labour between male and female children and hus- bands and wives in many homes in dealing with house chores and religious activities due to the ‘Stay at Home Order’. The social relationship, however, improved through helping one another, playing games together, respecting one another in relationships, mutual involvement in taking decisions, engaging in division of labour in handling the numerous home chores, tolerating one another, and managing and sharing peacefully the quantity of food that parents could provide. Such a new experience in male and female dynamics in Yoruba (Nigeria) home settings may help redefine gendered inter- actions, now and in the future. ‘STAY AT HOME ORDER’: A COVID-19’S CHILD OF NECESSITY WITH IMPLICATIONS FOR THE ‘FEMALE’ IN YORUBA (NIGERIA) HOME SETTINGS Gloria Njideka Ayantayo Introduction The year 2019 ended with many sad stories, including with the emergence of the incurable and novel COVID-19 disease, which is said to have begun in Wuhan, China, emerging in December 2019 and since spreading to many countries globally. The disease has remarkably thrown the world off balance because of its ravaging nature. Consequently, it has affected many people including the health workers who are managing victims. One of the universal effects of the disease is the declaration of lockdowns other- wise known as ‘Stay at Home Orders’. The Stay at Home Order, as it shall 99 BiAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 2021 \| UBP 2021 be described in the course of this discussion has effect on many things but significantly, on the female in Nigerian Yoruba home settings. Given the tremendous impact of the COVID-19 phenomenon and its global reach, the attention of scholars and non-scholars has been consid- erable. Many of the scholarly works, especially from the scientific perspec- tive, regarding its causes, remedies, measurement, effects, and sugges- tions-cum-solutions in addressing the pandemic are widely consulted and dispersed. But, all in all, we are of the view that too little attention has been given to its sociological dimension, particularly, as it relates to the Stay at Home Order (given by the government). It is on this premise that the current study examines the Stay at Home Order, its forms and condi- tions, with a view to underscoring its effects on the ‘female’ in Yoruba (Nigeria) home settings. The work is divided into sub-sections and dis- cussed as follows: a brief literature review, the ‘female’ in Yoruba (Nigeria) in pre-COVID-19 home settings, COVID-19, Stay at Home/lockdown, im- plications of the Stay at Home Order for Yoruba home settings in Ibadan, effects of the Stay at Home Order on the ‘female’ in Yoruba homes and concluding remarks. be described in the course of this discussion has effect on many things but significantly, on the female in Nigerian Yoruba home settings. COVID-19 Pandemic in Perspective has led many governments to institute a variety of control measures in- cluding a Stay at Home Order. But our point of departure into the study is that it has not looked at the aftermath of the order and its sociological implications (Harapan et al. 2020:667-673). Still from the scientific point of view, Ahmet et al. (2019:1-7) also in a collaborative work shed significant light on COVID-19 from the medical and oncological perspectives. These authors examined the meaning, his- tory, outbreak, treatment, and prevention of COVID-19 in its early stage. These authors also concluded instructively that there is an urgent need to impose some restrictions on people’s movement on the account that coro- navirus has the capacity to spread widely in the coming years. These au- thors are of the opinion that there is also a need for increased scientific collaboration to fight the disease now and in the future. A cursory look at the position of the authors indicate the justification for the Stay at Home Order, which some people criticize on the grounds that it has affected na- tional/global economies. The current study does not underrate the eco- nomic matter that may arise from the lockdown order but also considers the phenomenon and its sociological impacts. From the above, we are of the position that the books reviewed have pro- vided some insight into the problem of COVID-19 and the possible justi- fications for the declaration of the Stay at Home Order. This however calls for moving a step further, which includes a probe into the gender dimen- sion of the Order by looking at the fate of the female in Yoruba home settings in Nigeria. We shall expound on the meaning of ‘the female’ and the context in which it is used as we commence our discussion. Writing from the view of sociology of religion, data were gathered through telephone interviews with 50 families in Ibadan, regarding their disposi- tions to the fate of the female during the lockdown. We must state that the families interviewed are categorised into two. Thus, the elite family, that is, the ones whose father, mother and children are educated to Uni- versity first degree level. The majority of them are lecturers and the senior cadre of civil servants. The second is non-elite with the educational quali- fication of the members of the family between primary and secondary schools. COVID-19 Pandemic in Perspective As earlier indicated, much of the literature on COVID-19 is science-based and for this reason, the researcher intends to review two. Harapan et al (2020) in collaborative research reviewed COVID-19 in the context of its causative agents, pathogenesis and immune responses, epidemiology, di- agnosis, treatment and management of the disease, control and preven- tion strategies. Backing the above scientific findings with necessary data, these authors recommend that due to the rapid transmission in countries around the world, there should be an increase in attention to disease sur- veillance systems to scale up the country’s readiness to respond to the challenges posed by COVID-19. Such efforts should include the establish- ment of rapid response teams and, improving the capacity of the national laboratory systems to carry out tests on COVID-19 patients appropriately and in timely fashion around the globe (Akanti 2020). No doubt, the work provides us with some basic background information about the pandemic from which we can draw some inference, particularly, on the attention paid to the perceived risk associated with the spread of the disease, which 100 ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … has led many governments to institute a variety of control measures in- cluding a Stay at Home Order. But our point of departure into the study is that it has not looked at the aftermath of the order and its sociological implications (Harapan et al. 2020:667-673). Still from the scientific point of view, Ahmet et al. (2019:1-7) also in a collaborative work shed significant light on COVID-19 from the medical and oncological perspectives. These authors examined the meaning, his- tory, outbreak, treatment, and prevention of COVID-19 in its early stage. These authors also concluded instructively that there is an urgent need to impose some restrictions on people’s movement on the account that coro- navirus has the capacity to spread widely in the coming years. These au- thors are of the opinion that there is also a need for increased scientific collaboration to fight the disease now and in the future. A cursory look at the position of the authors indicate the justification for the Stay at Home Order, which some people criticize on the grounds that it has affected na- tional/global economies. The current study does not underrate the eco- nomic matter that may arise from the lockdown order but also considers the phenomenon and its sociological impacts. COVID-19 Pandemic in Perspective The two categories comprise both literates and semi-literates. We weighed the responses of the two sets in terms of whether they still retain their traditions, or accept also (some) modern ideas held about the ‘fe- male’ in their Yoruba settings. Furthermore, we also had casual interviews with randomly selected ten females and ten male adults in five streets in 101 BiAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 2021 the Ajibode area of Ibadan, regarding their disposition and experience with regards to gender differentiation in the sharing of house chores among children during the on-going ‘Stay at Home Order’. This was com- plemented with a non-participant observation regarding the sex of chil- dren and adults fetching water on the five streets in Ibadan. It is important to note, that Ibadan, though a large city in Nigeria, has problems with access to potable water and this is one of the reasons why people have to go about in the morning and evening looking for houses with boreholes from which they can fetch water. We also gathered information from fifteen selected YouTube video clips about husband and wife with regards to their handling of house chores during the lockdown. These included: “How to share household respon- sibilities”, “Splitting house tasks in marriage”,” Let’s cook dinner to- gether”, “Should a man do house chores with the wife?”, “Husband and wife divide household”, “Talking about household chores in English”, “3 Tips to avoid fighting with your partner over house chore”, “How to divide house chores”, “How to get husbands to do chores without nagging”, “Staying home is actually fun”, “Are chores hurting your relationship? The remaining four were not labelled, they were posted via WhatsApp messages. The video clips were subjected to content analysis while the other sources of information earlier mentioned were subjected to descrip- tive and sociological analysis. The Challenges of the ‘Female’ in the Pre-COVID 19 Home Settings The Yoruba is a group of Nigerians found in the South-western states of Nigeria. The Yoruba are made up of Oyo, Ogun, Osun, Ondo, Ekiti, and Lagos States. A few of them are also found in Kogi and Kwara states in Nigeria. There are three major religions practised by the people. These are Traditional Religions, Christianity and Islam. For emphasis and clarity, the word ‘female ‘is used in the context of fem- ininity to refer to a girl, woman, wife and mother. It is germane to men- tion that the Yoruba people, like many other African ethnic groups, are patriarchal in nature. That is, the group and her society hold the belief that the male, comprising boy, man and husband, is superior to the fe- male counterpart. Thus, based on the patriarchal tradition, the girls, at the domestic level, are saddled with more duties, including washing of plates, 102 ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … grinding of pepper, beans, maize, and melon, fetching of water for cook- ing and bathing, fetching of firewood, sweeping of the house and its en- virons, cooking and serving of food. It is also a common practice that while girls are doing the jobs mentioned above among others, their coun- terparts, that is, the boys, are playing around or sitting somewhere, wait- ing for the food to be served and the plates to be carried and washed by the girls. In the same vein, the men dictate what the women should do regarding taking care of the compound like sweeping, general cleaning, and super- vision of the girls in doing their jobs. At this point, the man is also con- sidered superior to the woman. The same scene plays out regarding the relationship between husband and wife. The husband is the overlord who determines, designs both the economic, domestic and political agenda for the running of their homes. The wife is expected to listen attentively to the rules of family engagement designed by the husband such as what food and soup to be cooked, who takes what portion of meat (whenever it is available), when to sleep and when to wake up and who handles which aspects of the house chores. The Challenges of the ‘Female’ in the Pre-COVID 19 Home Settings On childcare, the wife is expected to bathe the children, guide them to do their school assignments where applicable, carry the baby on her back while working in the house, or do house chores, with the husband relax- ing by reading newspapers or watching movies. When the child is playing around but needs attention, the husband calls the wife with strict instruc- tions – ‘So, So and So, come and carry your child’. Or, ‘take this child away from the sitting room because he/she is disturbing our peace’ At times, he shouts to the child: ‘go and meet your mummy in the kitchen’, ‘Don’t disturb me in my library’. On religious matters, particularly at the level of conducting the family re- ligious service in the morning or at the night (which some Christians call the family altar), it is seen as primarily a male affair. The husband con- ducts the whole service starting with singing of choruses, reading of Bible passages, delivery of sermon, and benediction. In all these, both the girl child and the mother or wife play subsidiary roles under the patriarchal canopy. Ayantola (2001:79-88) has discussed this tradition, pinpointing its social, political, moral and economic implications both in positive and negative terms, but we are not going into a full debate about it, rather making allu- sion to some of the reasons as is apposite. A lot of reasons are adduced to 103 BiAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 2021 \| UBP 2021 the marginalisation of girls, women or mothers in the home setting. For instance, there is a cultural belief that the male is superior to the female and should, on that account, take the lead in everything that concerns them over their female counterparts. Those belonging to this school of thought lean on an oral tradition which remarks that the males are more rational and less emotional than women and by so doing, women can hardly handle things emotion-free, hence the saying, Oro-Obinrin ko se te le, which means “it is not wise for a man to follow a woman’s advice or act on a woman’s suggestion”. The Challenges of the ‘Female’ in the Pre-COVID 19 Home Settings Beyond the shore of Yoruba state, this stance is also upheld by a sociologist – Steven Goldberg (1987:148), who wrote in his work: “The inevitability of patriarchy’ that males have a hormonal ag- gressive advantage over females and therefore, are better in leadership than women.” None of the above arguments is absolute because both male and female are making significant contributions to development in the field of sci- ence, technology, business, education and politics among others but the cultural belief about gender disparity still holds water in some quarters. Practically, we observe that among the Yoruba people, there is male child preference over female children. This has provoked some mothers whose children are all girls to pray to have male children at all cost. Evidently, some women, among whom are Christians, go as far as having many chil- dren all in the name of looking for a male child (Ojebode 2016). Some of the Christians who prefer male children to female children are probably influenced by the traditions of Judeo-Christian religions, which gave man superior right of leadership over woman, as epitomized in the work of Josephus (1962:373), who argues that women are inferior to men in every way as strengthened in the biblical injunction that wives are subject to their husbands with their husbands being their heads. COVID-19 and Stay at Home /Lockdown According to the World Health Organisation (WHO), the Coronavirus dis- ease (COVID-19) is an infectious disease caused by a newly discovered coronavirus said to have originated in Wuhan, China, in late 2019. Iden- tified in December, it spread from there to many countries globally (who.int/health-topic). The disease spreads primarily from person to per- son through small droplets from the nose or mouth, which are expelled when a person with COVID-19 coughs, sneezes, or speaks. These droplets 104 ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … can land on objects and surfaces around the person such as tables, door- knobs, and handrails. On this note, it is proved that people can become infected by touching these objects or surfaces, then touching their eyes, nose or mouth (who.int/emergencies). The disease manifests with some symptoms such as fever, dry cough, tiredness, aches and pains, sore throat, and sometimes also diarrhoea, conjunctivitis, headache, loss of taste or smell, difficulty breathing or shortness of breath, chest pain, or pressure, loss of speech or movement and a rash on skin, or discoloura- tion of fingers or toes. Since its emergence, the COVID-19 virus as at 9th May 2020 is reported to have infected 3.94 million, out of which 1.32 million have recovered and 275 thousand have died worldwide. The rapid spread of the disease provokes global concern about how to stop the spread, and get vaccines to protect against infection. This led to setting out some regulations spear- headed by the World Health Organisation and reinforced by national gov- ernments in countries that are affected by the pandemic. Substantial por- tions of the regulations are regularly and thoroughly cleaning of hands with an alcohol-based hand rub, washing of hands with soap and water, in order to kill viruses that may be on the hands, maintaining at least 1 meter (3 feet) distance between two people because when someone coughs, sneezes, or speaks they spray small liquid droplets from their nose or mouth, which may contain the virus. If one is too close, one can breathe in the droplets, and get infected with COVID-19 if the one who coughed has the virus. On account of this, crowded places are to be avoided. This is because when people come together in crowds, one is more likely to come into close contact with someone with COVID-19. COVID-19 and Stay at Home /Lockdown Other strategies include avoiding touching of eyes, nose and mouth be- cause hands touch many surfaces and can pick up viruses. Once contam- inated, hands can transfer the virus to our eyes, nose or mouth. Also im- portant is following good respiratory hygiene. This involves covering of one’s mouth and nose with the bend of one’s elbow or a tissue when one coughs or sneezes, disposing of used tissue immediately and washing one’s hands, because droplets spread the virus, wearing of masks covering both the nose and the mouth, staying at home to avoid contact with an unsuspecting carrier of the virus and, self-isolation for someone who has even minor symptoms of COVID-19 such as a cough, headache, or mild fever, until you recover, seeking of medical attention for anyone who has fever, cough and difficulty breathing by calling the National Centre for 105 BiAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 2021 \| UBP 2021 Disease Control, or visiting the centre personally and finally, keeping up to date on the latest information about the disease from trusted sources, such as WHO, local and national health authorities. Disease Control, or visiting the centre personally and finally, keeping up to date on the latest information about the disease from trusted sources, such as WHO, local and national health authorities. Of the above regulations, we shall dwell on the ‘Stay at home Order’. The Stay at Home Order is also synonymous with Lockdown in Nigeria and Movement Control in South East Asia. The term stands for an order from a government authority to restrict movements of a population as a mass quarantine strategy for suppressing, or mitigating an epidemic, or pan- demic, by ordering residents to stay home except for essential tasks or to work in essential businesses. Different nations redefine the scope and contents of the ‘Stay at Home Order’ which generally include: i. The stay at home order was characterised by lockdown which involved restriction of movement of human beings from their homes to their places of work or places of their choice. ii. It goes with closing of business centres, companies, ware- houses, and religious centres such as the Mosque and the Church. COVID-19 and Stay at Home /Lockdown On religious matters, all religious activities such as congregational pray- ers, vigils, Friday Jumat services for the pilgrimage exercise for Muslims, Sabbath services for Seventh Day Adventist Church members, and Sun- day services for other Christians, as well as conventions, camp meetings and many other liturgical services, were banned. Religious centres in sup- port of the government issued a series of bulletins and letters readable in public domains to ensure that people stayed at home. Other rules include allowance for minimum hours of outdoor activities, attending to non-es- sential business activities to be carried out at home, the imposition of cur- few for a number of hours – that is 7.00 pm to 6.00 am as in the case of Oyo State, banning of social and religious engagements except where the ones to attend are not more than 20 people, banning of commercial activ- ities except those that are essential to living and management of COVID- 19 (e.g. selling and distribution of food stuff, health and security), curtail- ing of non-essential journeys but permitting what are described as essen- tial journeys (which include shopping for food, travelling to work, access- ing healthcare, and engaging in physical exercise within 1 km of the home for up to 1 hour). Others also include the closure of schools at all levels, suspension of parliamentary sessions, prohibition of all gatherings of more than 10 people, and permission to operate essential services such as 106 ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … supermarkets, petrol stations, and health services. Also enforced were for- mal and informal closure of bars and the prohibition of the sale of alcohol, cancelation of sporting activities as well as non-essential services such as pools, bars, cafes, restaurants, playgrounds and closing of permitted busi- ness activities in specific hours of the day at the commencement of the curfew. To make this Order effective, security agents were drafted to major roads, interstate roads and some hotspots in the main cities and towns, as a strat- egy for preventing unauthorised journeys. This led to the emergence of road blocks and police or security checkpoints to check people who were out and about had good reason to be and that their exemption declarations were in tandem with the existing stay at home order. Implications of Stay at home Order for Yoruba Home settings in Ibadan In most cases, there was an allowance for online worship particularly for those who had access to some social media applications like WhatsApp, and Zoom applications, or their church television stations. In some cases, the order of service including the sermon was sent online to church members for them to follow during their family Sunday service. On the whole, there are many needs to attend to, many assignments to carry out in the running of the home and more expenses to be accommo- dated with little or no resources particularly for families whose livelihood depends on daily income. All these put together call for a new order, a re- orientation, and absolute tolerance as parameters to manage the home. On the whole, there are many needs to attend to, many assignments to carry out in the running of the home and more expenses to be accommo- dated with little or no resources particularly for families whose livelihood depends on daily income. All these put together call for a new order, a re- orientation, and absolute tolerance as parameters to manage the home. Implications of Stay at home Order for Yoruba Home settings in Ibadan The ‘Stay at Home Order’ with its attendant instructions has implications for all Yoruba homes but I limit my research area to Ibadan, the capital of Oyo State, Nigeria where I have gathered first-hand information. Based on research findings, the Stay at Home Order means the following: hus- band and wife have to be indoors and by implication have to relate with one another more than during the pre-COVID period when they usually met at home possibly late evening after each of them might have returned from work as the case may be. Children too are fully home because their schools have been closed. It has given them ample opportunity to enjoy more intimacy with their parents. Children being at home calls for full feeding, i.e. three meals per day, seven days a week. It implies that their parents would need to spend more money on food, compared with pre- COVID times. It also has implications on whoever is the bread winner among the husband and wife. There will be more cooking of food, fetch- ing of water where there is no flowing water at home, washing of plates and grinding of pepper and beans among others. This is followed by more cleaning of the house, mopping of the floor and sweeping of the house and more of the general house chores than before. Where electricity sup- ply is stable, the ‘Stay at Home Order’ calls for more consumption of en- ergy, which leads to purchasing more energy power points. In essence, the ‘Stay at Home Order’ led to the closure of religious centres and re- 107 BiAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 2021 \| UBP 2021 sulted in the emergence of home cells and family altars for Sunday wor- ship. In most cases, there was an allowance for online worship particularly for those who had access to some social media applications like WhatsApp, and Zoom applications, or their church television stations. In some cases, the order of service including the sermon was sent online to church members for them to follow during their family Sunday service. sulted in the emergence of home cells and family altars for Sunday wor- ship. Effects of the Stay at Home Order on the ‘Female’ in Yoruba homes Leaning on our research tools as mentioned earlier, we discovered that the Stay at Home Order has some negative and positive effects on the ‘female’ in Yoruba homes selected for our study. The findings are dis- cussed thematically as seen below: 2. Division of Labour Far away from the first research findings, we gathered that some homes, especially the elites, subscribed to what can be described as division of labour among boys and girls, and husband and wives in handling the de- mands in their home occasioned by the stay at home order. The division of labour has an outstanding history in terms of its meaning, theory and applications in economic history and it positively correlates with societal advancement because it increases as society progresses (Smith 1970). The thesis of the division of labour as mentioned elsewhere is characterised by the division of a large task, contract, or project into a smaller task so that each of the people involved in the task or labour would have a separate schedule of duty based on their specialty within the overall project sched- ule; allocation of tasks to individuals or organizations according to their capacities (Ayantoyo 2019:173-185). This is exemplified in some homes where boys were asked to wash plates and fetch water; a job that was typ- ically reserved for males. In fact, according to Prof B (anonymous name), a timetable was made for children in the house (including a child of a family friend who came on visit but could not travel back before the lock- down) on who should cook on each day, wash the toilets and cars, sweep and mop the floor without reference to gender factor. The same sharing of work is displayed among some husbands and wives on driving, bathing of children (particularly those who have small children) feeding the chil- dren, monitoring them while doing their online lessons and assignments among others. In fact, it is interesting to note that there was a WhatsApp video clip showing a man putting one of their twin babies on the back and at the same time working with the wife in the kitchen. The emerging development in the home as characterised by sharing of responsibilities has sociological dimensions because the practice is a form of role responsibility. Role responsibility is a sociological term denoting the specific tasks or duties that members are expected to complete accord- ing to their roles. In other words, role responsibility in the context of our discussion is the specific home chore, specifying what each member of the family should do at a specific time and for which individual members of the family are held accountable. ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … G. AYANTAYO \| 1. Maintenance of status quo We gathered that a majority of the families belonging to the non-elite still maintain the status quo; all the house chores such as sweeping, cooking, fetching of water and wood among others are still reserved for the female members of the family. This information is further strengthened by the researcher’s observation during one of their morning exercises where a conversation ensued among some girls fetching water by the borehole about whether or not they have male siblings at home who could have joined them in the task. The majority of them said yes, and that their male counterparts were still sleeping. They added that they are the ones who have been doing the major work in the house while boys are just playing one game or the other, sleeping or pressing their phones while they are saddled with so much of the house chores. What this revelation translated to is that some homes still do not approve of gender equality or equity even in a time like this. This is suggestive that the female under these settings has many painful stories of being overwhelmed with work asso- ciated with the lockdown. It is a trend that needs further attention as the global world strives towards gender respect and tolerance. 108 2. Division of Labour The responsibility is not limited to sharing home chores and other domes- tic activities, it is also extended to the handling of religious programmes which various churches in most cases have sent to their church members 109 BiAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 2021 for the purpose of Sunday service in replacement of congregational ser- vice they would have held on Saturday or Sundays. For example, a family notified me how the division of labour is applied in handling the pro- gramme sent to them by their church – Christ Apostolic Church, (Oke Ife) Agbowo, Ibadan in the order listed below: Pre-service Choruses: Female Children , Hymn 1 (GHB 70: The God of Abraham praise): All of us with our eldest daughter taking the lead, Processional Hymn: (101 70: Thank God of Abraham) All of us with our eldest son taking the lead in turns, Opening Prayer: Daddy , Lesson/Bible Reading: I Pet. 4:7-: Mummy, Hymn 2 (GHB 823: God moves in a mysterious way): All of us with our eldest daughter taking the lead in turns, Testimonies/Offerings/ Thanks- giving: All of us comes out one after the other to give testimony; Male Child takes the offering and Mummy does the Thanksgiving prayers, Hymn 3 (GHB 454: Break Thou the bread of life): Thank God of Abra- ham). All of us with our eldest son taking the lead in turns, with the Ser- mon by Mummy, and the General Prayers by Daddy, The Fourth Hymn 4 (GHB 723: How, my comrades! see the signal): All of us with our eldest daughter taking the lead in turns, and the Benediction by Daddy. From the above, we can see that both male and female are involved in the conduct of the service; both the female and male children were given the task to take the lead in handling one part of the programme or the other, while both Daddy and Mummy functioned in equal capacities. 3. Improved social relations We also gathered information that there is a considerable improvement in social relationship between male and female, boys and girls and hus- bands and wives in the majority of the homes of both the elites and non- elites. The social relationship is manifested in all of them helping one an- other, playing games together, respecting one another in the relationship, We also gathered information that there is a considerable improvement in social relationship between male and female, boys and girls and hus- bands and wives in the majority of the homes of both the elites and non- elites. The social relationship is manifested in all of them helping one an- other, playing games together, respecting one another in the relationship, mutual involvement in taking decisions, engaging in division of labour in handling the numerous home chores, tolerating one another, managing and sharing peacefully the quantity of food that parents could provide. In addition, it is noted that husband and wife have much more time to inter- act, interrelate, communicate, share ideas of their experiences in their places of work, discuss issues bothering their family prospects and coping strategies for the aftermath of COVID-19. The improvement in social relation is of sociological significance because the act propels cooperation, which implies colleagueship, co-partnership, 110 ‘Stay at Home Order‘: A COVID-19’s Child of Necessity with Implications … and tie-up relationship. It is a good development because as Ayantayo puts it, ‘where there is cooperative spirit, there is always the growth of frater- nity, solidarity, fellow-feeling, voluntary association, coalition, federation, united fronts, common front, mutual assistance, reciprocity, spirit of give- take mutual concession, compromise and comradeship’ (Ayantoyo 2009:78). 4. Domestic Violence The researcher discovered that a few women in Ibadan suffered domestic violence from their husbands during the lock down. The violence mani- fested in physical, verbal, emotional, and sexual abuse. Three cases of such were reported to the Women and Child Rights wing of the Justice, Development and Protection Commission (JDPC) in Ibadan. The violence erupted as a result of the inability of both the wives and husbands to re- solve some issues related to finance. Further investigation shows that the husbands and wives that were involved had been avoiding each other be- fore the enforced stay at home. Now that they were forced to stay together, their disagreements apparently and consequently developed into domes- tic violence. Conclusion It is pertinent to remark that most ‘females’ in our research locations en- joyed some level of respect, appreciation, cooperation, and justice from their male counterparts during the Stay at Home Order. This is a good departure point from the pre-existing order, in which the female was made a second-class citizen. We are of the opinion that the gain made by the ‘female’ during the period of Stay at Home Order will endure for a long period. The new experience may help define a lot of things regarding male and female interactions, now and in the future. Hopefully, many conferences, talk shows, workshops, and various fora for discussions will emerge after COVID-19, the result of which could help raise the dignity of the female even more. 111 iAS 31 / ERA 8 \| COVID-19 – AFRICAN WOMEN AND THE WILL TO SURVIVE \| UBP 202 References Akanti, Bindu (2020), ‘Cerebrovascular events in COVID -19 patients,’ Journal of Monaldi Archives for chest disease, 90/2. Ayantayo, G.N. (2019), ‘Division of Labour among Church Workers of the Living Faith Church Worldwide and Its Sociological Import,’ in C.O. Oshun, and J.K. Ayantayo, (eds.), Ministerial Ethics and Ecclesiological Challenges in Nigeria, A Festschrift in honour of Pastor Stephen Olurotimi Adeola Afolabi, Ibadan: Hope Publications. Ayantayo, J. K. (2001), ‘Revisiting Gender Difference from the Context of Biblical Ethics,’ Ibadan Journal of Multicultural/Multidisciplinary Studies, 8/1. Ayantayo, J. K. (2009), Fundamentals of Religious Ethics, Ibadan: Samprints and Graphic and Co. Goldberg, S. (1987), ‘The inevitability of Patriarchy,’ in Thomas, Mappers, and Jane Zembath (eds.), Social Ethics, New York: McGraw Hill. Harapan, H., N. Itoh, A. Yufika, W. Wirandi, M. Mudatsir (May 2020), ‘Corona- virus disease 2019 (COVID-19): A Literature review,’ Journal of Infection and Public Health, 13/5: 667-673. Josephus, A. (1962), Against Anion or On the Antiquity of the Jews Book, 11, papa 201, trans. H. St. J. Thackeray, (Leus Classical Library), London: Heinemann. Ojebode, Peter Olayiwola (2016), ‘Biblical Ethics and Gender Preference of a Male Child among Christian Couples in Oyo Baptist Conference, Oyo State, Nige- ria,’ (Ph. D thesis Submitted to Nigerian Baptist Seminary, Ogbomoso, Nige- ria). Sahin, Ahmet Riza, Aysegul Erdogan, Pelin Mutlu Agaoglu, Yeliz Dineri, Ah- meengage, Yusuf Cakirci, Mahmut Egemen Senel, Ramazan Azim Okyay, Ali Muhittin Tasdogan (2020), ‘Novel Coronavirus (COVID-19) Outbreak: A Re- view of the Current Literature,’ Eurasian Journal of Medicine and Oncology, 4/1. Smith, Adam (1970), The Wealth of Nations, Baltimore: Penguin Books. 112
https://openalex.org/W3119989816	https://europepmc.org/articles/pmc7822689?pdf=render	English	null	Ehretiquinone from Onosma bracteatum Wall Exhibits Antiaging Effect on Yeasts and Mammals through Antioxidative Stress and Autophagy Induction	Oxidative medicine and cellular longevity	2,021	cc-by	13,840	Hindawi Oxidative Medicine and Cellular Longevity Volume 2021, Article ID 5469849, 15 pages https://doi.org/10.1155/2021/5469849 Hindawi Oxidative Medicine and Cellular Longevity Volume 2021, Article ID 5469849, 15 pages https://doi.org/10.1155/2021/5469849 Hindawi Research Article Ehretiquinone from Onosma bracteatum Wall Exhibits Antiaging Effect on Yeasts and Mammals through Antioxidative Stress and Autophagy Induction Yanjun Pan ,1 Yanan Liu ,1 Rui Fujii ,2 Umer Farooq ,1,3 Lihong Cheng ,1 Akira Matsuura ,2 Jianhua Qi ,1 and Lan Xiang 1 1College of Pharmaceutical Sciences, Zhejiang University, Yu Hang Tang Road 866, Hangzhou 310058, China 2Department of Biology, Graduate School of Science, Chiba University, Chiba 263-8522, Japan 3Faculty of Pharmacy, University of Central Punjab, Lahore 54660, Pakistan Correspondence should be addressed to Jianhua Qi; qijianhua@zju.edu.cn and Lan Xiang; lxiang@zju.edu.cn Received 1 August 2020; Revised 19 December 2020; Accepted 29 December 2020; Published 15 January 2021 Academic Editor: Liang Jun Yan Copyright © 2021 Yanjun Pan et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The antiaging benzoquinone-type molecule ehretiquinone was isolated in a previous study as a leading compound from the herbal medicine Onosma bracteatum wall. This paper reports the antiaging eﬀect and mechanism of ehretiquinone by using yeasts, mammal cells, and mice. Ehretiquinone extends not only the replicative lifespan but also the chronological lifespan of yeast and the yeast-like chronological lifespan of mammal cells. Moreover, ehretiquinone increases glutathione peroxidase, catalase, and superoxide dismutase activity and reduces reactive oxygen species and malondialdehyde (MDA) levels, contributing to the lifespan extension of the yeasts. Furthermore, ehretiquinone does not extend the replicative lifespan of Δsod1, Δsod2, Δuth1, Δskn7, Δgpx, Δcat, Δatg2, and Δatg32 mutants of yeast. Crucially, ehretiquinone induces autophagy in yeasts and mice, thereby providing signiﬁcant evidence on the antiaging eﬀects of the molecule in the mammalian level. Concomitantly, the silent information regulator 2 gene, which is known for its contributions in prolonging replicative lifespan, was conﬁrmed to be involved in the chronological lifespan of yeasts and participates in the antiaging activity of ehretiquinone. These ﬁndings suggest that ehretiquinone shows an antiaging eﬀect through antioxidative stress, autophagy, and histone deacetylase Sir2 regulation. Therefore, ehretiquinone is a promising molecule that could be developed as an antiaging drug or healthcare product. 1. Introduction Seven fractions were then obtained, and fraction 4 (400 mg), which was eluted with n-hexane/CH2Cl2 (0 : 100), was puriﬁed by the ODS open column with CH3OH/H2O (40 : 60, 50 : 50, 55: 45, 57 : 43, 60: 40, and 100 : 0). Among the ﬁve fractions obtained, fraction 3 (37.2 mg) was subjected to HPLC puriﬁ- cation (C30-UG-5 (Φ 10 × 250 mm, Nomura Chemical), 68% methanol, ﬂow rate: 3 mL/min, and detector: 210 nm) to yield EHR (18.7 mg, tR = 22:1 min). The chemical structure of EHR was identiﬁed by comparing the MS, 1H NMR, and 13C NMR spectra and speciﬁc rotation data with the litera- ture [24]: 1H NMR (500 MHz, CDCl3): δ H = 6:84 (1H, d, J = 10:5 Hz, H-5′), 6.62 (1H, d, J = 8:7 Hz, H-6), 6.58 (1H, d, J = 2:95 Hz, H-3), 6.53 (1H, m, H-5), 6.52 (1H, d, J = 10:6 Hz, H-6′), 6.25 (1H, d, J = 16:1 Hz, H-8′), 5.65 (1H, d, J = 6:3 Hz, H-8), 5.49 (1H, d, J = 16:1 Hz, H-7′), 5.04 (1H, s, H-10′b), 4.97 (1H, s, H-10′a), 3.83 (1H, d, J = 6:3 Hz, H-7), 2.78 (1H, d, J = 19:3 Hz, H-10b), 2.52 (1H, d, J = 19:2 Hz, H-10a), 1.73 (3H, s, H-11′), 1.69 (3H, s, H-11); 13C NMR (125 MHz, CDCl3): δC = 195:3 ðC −1′Þ, 193.2 (C-4′), 149.9 (C-4), 145.0 (C-1), 140.8 (C-9′), 139.1 (C-5′), 138.7 (C-6′), 137.4 (C-8′), 131.8 (C-9), 127.7 (C-2), 124.1 (C-7′), 122.6 (C-8), 119.3 (C-10′), 117.6 (C-6), 114.9 (C-5), 114.3 (C-3), 80.7 (C-3′), 55.3 (C-2′), 39.4 (C-7), 36.2 (C-10), 22.7 (C- 11), 18.2 (C-11′). ½α16 D +1.01(c 0.12, MeOH); high- p p [ ] Onosma bracteatum is a medicinal herb used in Asian countries due to its antileprotic, antibacterial, and anti- inﬂammatory functions [21] and its role in enhancing mem- ory and immunity [22]. As an important part of the Unani system, O. bracteatum has been used to treat Alzheimer’s dis- ease, arrhythmias, and hypertension [22]. The K6001 bioas- say screening system also highlights the antiaging activity of this herb. A series of antiaging molecules are isolated on the basis of this system, and their chemical structure and antiag- ing activity are described [23]. Moreover, the structure–activ- ity relationship study suggests that ehretiquinone (EHR) shows the best performance in extending the replicative life- span of K6001. 1. Introduction Studying the antiaging mechanism of EHR is imperative because it is a benzoquinone-type molecule with a signiﬁcant antiaging activity. The mechanism behind the antiaging eﬀects of EHR is reported in this paper. 1. Introduction to oxidative stress and improve aging [3, 4]. Oxidative stress is caused by the disequilibrium between the production of reactive oxygen species (ROS) and antioxidant systems. Although a low level of ROS is necessary to conduct a normal physiological function, the superﬂuous level of ROS can impair cellular lipids, proteins, or DNA and consequently produce harmful materials [5]. ROS can also be scavenged by catalase (CAT) and glutathione peroxidase (GPx), which are antioxidant enzymes that contribute to either extending or shortening the lifespan of an organism [5, 6]. Moreover, if the damaged substance is eliminated in time, then its neg- ative eﬀects on physiological functions can be minimized. As a highly conserved metabolic process, autophagy in organ- isms signiﬁcantly contributes to the removal of damaged Human beings demand a healthy life but are vulnerable to age-related diseases, including Alzheimer’s disease, diabetes, cancer, and cardiovascular diseases [1]. With the continuous aging of the global population, the medical expenditures for the elderly increase the burden for their families and society. Therefore, eﬀective therapy for controlling and preventing age-related diseases must be developed. As a vital risk factor of age-related diseases, aging refers to the degradation of an organism as characterized by the gradual accumulation of damaged substances [2]. These substances are triggered by oxidative stress and eliminated by autophagy; some genes, speciﬁcally the silent information regulator 2 (SIR2), respond Oxidative Medicine and Cellular Longevity 2 China) or reversed-phase C18 (Octadecylsilyl, ODS) silica gel (Nacalai Tesque, Kyoto, Japan). molecules. The damaged organelles or proteins are delivered to the lysosome after inducing autophagy and then degraded and recycled to maintain cellular homeostasis, which con- tributes to the longevity of organisms [7]. Previous studies showed that inducing autophagy alleviated the toxicity caused by oxidative stress while the defects of this process promoted oxidative stress, thereby highlighting the relation- ship and potential eﬀect of autophagy on the aging mecha- nism of organisms [8]. The Sir2, which is an evolutionarily highly conserved NAD+-dependent deacetylase, can prevent senescence by protecting ribosomal DNA stability [9]. The deﬁciency and overexpression of the SIR2 gene can shorten and extend the replicative lifespan, respectively [10]. How- ever, whether SIR2 also shows the same eﬀect on the chrono- logical lifespan remains debated [11, 12]. 2.2. Plant Material and Reagents. The plant material was pur- chased in Mansehra, Khyber Pakhtunkhwa, Pakistan, and identiﬁed by the associate professor Dr. 1. Introduction Zafar Ullah Zafar from the Institute of Pure and Applied Biology of the Bahauddin Zakariya University in Multan, Pakistan. The voucher specimen (20170220) of the plant was preserved at the School of Pharmaceutics of Zhejiang University. The fol- lowing reagents were purchased from the indicated suppliers: chemical reagents of HPLC (TEDIA, Rhode Island, USA) and analytical grades (Sinopharm Chemical Reagent Co. Ltd., Shanghai, China), resveratrol (RES) (J&K Scientiﬁc Ltd., Beijing, China), rapamycin (Solarbio, Beijing, China), DAPI dihydrochloride (DAPI) (Macklin, Shanghai, China), leupeptin (Macklin, Shanghai, China), and dimethyl sulfox- ide (DMSO) (Sigma, Saint Louis, USA). Yeast models have been extensively used in aging studies due to their relatively simple and short life cycle; a small genome comprising 6000 genes has been completely sequenced and mapped [13], and the ortholog of approxi- mately 30% of yeast genes relates to human diseases [14]. Among the available yeast models, the K6001 yeast strain is uniquely characterized by the growth of only mother cells in the glucose medium [15], thereby facilitating the replica- tive lifespan evaluation of yeast while screening active com- pounds. Therefore, the K6001 yeast mutant has been utilized as an antiaging screening bioassay system, and many antiaging compounds, including cucurbitacin B, parishin, cholesterol, nolinospiroside F, and phloridzin, have been iso- lated from natural products in previous studies [16–20]. 2.3. Isolation and Structural Elucidation of EHR. Dried plant material (1.5 kg) was crushed to powder and soaked in 100% methanol (CH3OH) for three days with constant shaking at room temperature. The obtained crude extract (120 g) was partitioned between ethyl acetate (EtOAc) and water. The active EtOAc layer sample (30 g) was concentrated before its separation by a silica gel open column eluted with n-hex- ane/CH2Cl2 (100 : 0, 80 : 20, 50: 50, and 0 : 100) and CH2Cl2/CH3OH (98 : 2, 95 : 5, 90: 10, 80 : 20, and 0 : 100). A total of 11 fractions were obtained. Fraction 3 (5.2 g), which was eluted with CH2Cl2/CH3OH (95 : 5), was further sepa- rated by a silica open column with n-hexane/CH2Cl2 (100 : 0, 50: 50, 30: 70, 20 : 80, 10 : 90, and 0 : 100) and CH2Cl2/CH3OH (95 : 5 and 0 : 100). resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. The chronological lifespan assay was conducted by fol- lowing the described methodology [25]. The YOM36 or Δsir2, Δuth1 of YOM36, was incubated in YPD by shaking for 24h, transferred to a 20 mL SD medium, and incubated for another 24h. On day 0, the yeast cell suspension was inoculated in a new 100 mL SD medium with an initial OD600 value of 0.01 and treated with EHR at concentrations of 0, 3, and 10μM with shaking at 28°C. On the third day, nearly 200 yeast cells were spread on the glucose medium agar plate with the stabilization of yeast growth, and the colony-forming units (CFUs) on the plates were counted after a two-day incubation. This process was undertaken every two days, and the EHR in diﬀerent concentrations was added to the corresponding medium on day 7. The CFUs at day 3 were ﬁxed as a 100% survival. 2.4. Yeast Strains, Cell Lines, and Culture Medium. The K6001 derived from W303 was used along with Δsod1, Δsod2, Δskn7, Δuth1, Δgpx, Δcat, Δatg2, and Δagt32 yeast strains with a K6001 background in the replicative lifespan assay. By contrast, the YOM36 derived from BY4742 was used along with the Δsir2 and Δuth1 yeasts of a YOM36 background in the chronological lifespan assay. Wild-type BY4741 and YOM38 containing plasmid pRS316-GFP- ATG8 were used to investigate antioxidative stress and detect autophagy levels. All yeast strains were stored in a freezer at −30°C and validated by Professors Breitenbach (Salzburg University, Austria) and Matsuura (Chiba University, Japan). The genotypes of yeast strains and mutants in this study were described in the previous study [16]. 2.6. Yeast-Like Chronological Lifespan Assay. The yeast-like chronological lifespan assay was performed following the methods described in reference [26]. First, 80,000 PC12 cells were seeded in each well of a 96-well plate and cultured for 24 h. Then, the medium was replaced with 1 mL serum- free DMEM containing a test sample at diﬀerent concen- trations or DMSO of 0.5%. The cells were continually incubated for six days, and the medium was replaced every two days by a new serum-free DMEM containing doses of samples or DMSO (0.5%). Rapamycin inhibits the mTOR signaling pathway and has beneﬁcial eﬀects on the health span and lifespan of all cellular and organism systems [27]. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. Afterward, approximately 4000 cells calculated by a hemacytometer were evenly spread on the surface of the glucose medium agar plate containing RES at 10 μM or EHR at doses of 0, 0.1, 1, 3, 10, and 30 μM. The microcolony formation on the agar plates after cultivation at 28°C for 48 h was observed with a microscope. Forty microcolonies from each group were randomly selected, and the number of daughter cells produced by a mother cell in microcolonies was counted. The performance of the repli- cative lifespan assays of K6001 mutants (Δsod1, Δsod2, Δuth1, Δskn7, Δgpx, Δcat, Δatg2, and Δatg32) was similar to that of the K6001 strain. RES, as a well-known antiaging substance, was used as a positive control. Ethanol was used in the yeast experiment to dissolve compounds and as the vehicle for the control group. EHR did not aﬀect the cell growth at doses of 1 to 30 μM. 2.5. Replicative and Chronological Lifespan Assays. In the rep- licative lifespan assay, a K6001 yeast strain, which was stored in a freezer at −30°C, was taken out, washed thrice with phos- phate buﬀer solution (PBS), inoculated in a liquid galactose medium, and incubated in a shaker for 24 h. The K6001 cells were collected by centrifugation at 1800×g for 10min and washed thrice with PBS. Afterward, approximately 4000 cells calculated by a hemacytometer were evenly spread on the surface of the glucose medium agar plate containing RES at 10 μM or EHR at doses of 0, 0.1, 1, 3, 10, and 30 μM. The microcolony formation on the agar plates after cultivation at 28°C for 48 h was observed with a microscope. Forty microcolonies from each group were randomly selected, and the number of daughter cells produced by a mother cell in microcolonies was counted. The performance of the repli- cative lifespan assays of K6001 mutants (Δsod1, Δsod2, Δuth1, Δskn7, Δgpx, Δcat, Δatg2, and Δatg32) was similar to that of the K6001 strain. RES, as a well-known antiaging substance, was used as a positive control. Ethanol was used in the yeast experiment to dissolve compounds and as the vehicle for the control group. EHR did not aﬀect the cell growth at doses of 1 to 30 μM. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. Therefore, rapamycin was used as a positive control in the experiment. The cells treated in the 96-well plate were subsequently trypsinized, and a 5% aliquot was plated on fresh medium-ﬁlled six-well plates. Finally, the colony formation on the plate after 15 days was stained with crystal violet and photographed. The yeast culture media used in this work included galac- tose liquid medium (3% galactose, 2% peptone, and 1% yeast extract), glucose liquid medium (YPD) (2% glucose, 2% pep- tone, and 1% yeast extract), glucose medium agar plate (2% glucose, 2% peptone, 2% agar, and 1% yeast extract), and SD (2% glucose, 0.17% yeast nitrogen base without amino acids, and 0.5% ammonium sulfate (BD Ditco)). The liquid culture was performed by using a shaker at 28°C, and a solid culture was obtained in an incubator at 28°C. The PC12 cell line, which was derived from rat pheochro- mocytoma cells, was purchased from the Type Culture Col- lection of the Chinese Academy of Sciences (Shanghai, People’s Republic of China). PC12 cells were cultured in Dul- becco’s Modiﬁed Eagle’s Medium (DMEM) with high glu- cose (Cellmax, Beijing, China) supplemented with 10% horse serum (Cellmax, Beijing, China) and 7.5% fetal bovine serum (Solarbio, Beijing, China) in a 5% CO2 incubator at 37°C. 2.7. Antioxidative Stress Assay. The BY4741 yeast strain stored in the freezer at −30°C was taken out, washed thrice with PBS, inoculated in YPD, and incubated for 24 h. The BY4741 cells with an initial optical density at 600 nm (OD600) of 0.1 were inoculated in a 20 mL YPD containing the positive control, RES at 10 μM, or EHR at 0, 3, and 10 μM. After a 12 h cultivation, the yeast suspension of each group with an OD600 of 1.5 was dropped on the glucose medium agar plate containing 9.0 mM of H2O2 and continu- ally incubated at 28°C for three days. The qualitative growth of yeast was also photographed. 2.5. Replicative and Chronological Lifespan Assays. In the rep- licative lifespan assay, a K6001 yeast strain, which was stored in a freezer at −30°C, was taken out, washed thrice with phos- phate buﬀer solution (PBS), inoculated in a liquid galactose medium, and incubated in a shaker for 24 h. The K6001 cells were collected by centrifugation at 1800×g for 10min and washed thrice with PBS. Oxidative Medicine and Cellular Longevity Oxidative Medicine and Cellular Longevity 3 2. Materials and Methods 2.1. General. A preparative high-performance liquid chroma- tography (HPLC) system was equipped with two ELITE P- 230 pumps and a UV detector. Optical rotations were per- formed by using a JASCO P-1030 digital polarimeter, and nuclear magnetic resonance (NMR) spectra were detected by a Bruker AV III-500 spectrometer (Bruker, Billerica, USA). The Agilent Technologies 6224A Accurate Mass TOF LC/MS system (Santa Clara, CA, USA) was used for the high-resolution ESI-TOF-MS analyses. Column chroma- tography was conducted by using silica gel (with 200 to 300 meshes, Yantai Chemical Industry Research Institute, Yantai, resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. In the MDA assay, after culturing BY4741 in a 20mL YPD for 24 or 48 h, all yeasts were harvested and washed with PBS thrice for 10min at 1800×g and suspended in 500 μL PBS. The yeast cells were ultrasonicated for 5 min, followed by ﬁve freezing cycles in liquid nitrogen for 5 min, soaking in a 37°C water bath for 2 min and another 5 min of ultraso- nication. The cell lysates were centrifuged at 4°C for 10 min at 12,000×g to obtain the supernatant, which was then used to evaluate the MDA level by using the MDA assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). 2.9. CAT, Total GPx, and SOD Activity Assays. Fig3 a, b, c, d, e and f.BY4741 yeast cells were cultured as described in the MDA assay for 24 h. These cells were then washed with PBS thrice, suspended in 250 μL of PBS, and ultrasonicated on ice for 5 min. Afterward, these cells were centrifuged at 4°C for 10 min at 12,000×g to obtain a supernatant. Conse- quently, this supernatant measured the CAT, GPx, and superoxide dismutase (SOD) activities by, respectively, using CAT, GPx (Beyotime Biotechnology Limited Company, Shanghai, China), and SOD (Nanjing Jiancheng Bioengineer- ing Institute, Nanjing, China) assay kits following the instructions of the manufacturer. For the CAT enzyme activ- ity assay, diﬀerent concentrations of hydrogen peroxide solu- tion were ﬁrst taken. Then, a color working solution was added to the CAT enzyme activity assay kit and reacted at 25°C for 15 min. The absorption value of A520 was mea- sured, and the standard curve of hydrogen peroxide concen- tration was determined. Afterward, the catalase buﬀer and 250 mM hydrogen peroxide were added to each sample (5 μg protein). The enzyme reaction termination solution was added after reaction at 25°C for 1–5 min to terminate the reaction. The color working solution was also added and reacted at 25°C for 15 min, and the absorption value of A520 was measured. For the GPx enzyme activity assay, 2 μg of protein of each sample was taken, and the total gluta- thione enzyme assay kit was used to determine the glutathi- one enzyme activity. The general protocol is as follows: the GPx detection solution, samples, GPx detection working solution, and peroxide reagent were added in a 96-well plate in proper order. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. A total of 200 cells from the treated or untreated groups were evenly spread on the glucose medium agar plate con- taining H2O2 at 0 or 5 mM and cultured for 48 h to obtain quantitative results. The CFUs were then counted to calculate the survival rate as follows: the number of cells survived in the 5 mM H2O2 divided by that survived in the H2O2 untreated group. 2.8. ROS and MDA Level Assays. For the ROS assay, BY4741 was cultured in 20 mL YPD for 23 or 47 h. Approximately, 5 × 107 cells from each group were harvested and washed with PBS thrice and suspended in 1 mL of PBS. Afterward, 2′,7′-dichlorodihydroﬂuorescein diacetate (DCFH-DA) was 4 Oxidative Medicine and Cellular Longevity was measured after adding color reagents and standing at room temperature for 10min. added and mixed under dark conditions to reach a ﬁnal con- centration of 10 μM. The suspended mixture was incubated by shaking for 1 h at 28°C, and yeast cells were collected and washed with PBS thrice. By using a SpectraMax M3 mul- timode microplate reader (Molecular Devices Corporation, California, USA), the DCF (2′,7′-dichloroﬂuorescein) ﬂuo- rescence intensity of 1 × 107 cells was recorded with excita- tion and emission wavelengths of 488 and 525 nm, respectively. 2.10. Fluorescence Image of Yeast Autophagy. The YOM38 yeast strain containing the pRS316-GFP-ATG8 plasmid was primarily cultured in 20 mL YPD medium for 24 h. The yeast cells were then harvested and washed with PBS before inocu- lation in an synthetic deﬁned (SD) medium with an initial OD600 of 0.1 and treated with a positive control RES at 300 μM and EHR at 0, 3, and 10 μM. After cultivating for 22 h, the yeast cells were collected, washed, and suspended in 245 μL PBS and stained with 5 μL DAPI (1 mg/mL) in the dark for 10min. These cells were then washed with PBS thrice and suspended in 10 μL PBS to observe the diﬀerential interference contrast and the green and blue ﬂuorescence images under two-photon confocal ﬂuorescence microscopy (Olympus FV1000BX-51, Tokyo, Japan). The images were acquired and analyzed by using computer software (Olym- pus Fluoview Ver.4.1 Viewer). The light was avoided during the experiment. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. The absorbance value of A340 at every 4 min was measured seven times after mixing. For the SOD enzyme activity assay, 25 μg of each sample protein was ﬁrst mixed with reagent VII and reacted for 1 min to remove the Mn- SOD enzyme activity in the samples. The supernatant was then obtained as samples after centrifugation. The reagent I, blank control, samples, and the samples treated by reagent VII were added to the 96-well plate. Afterward, reagents II, III, and IV were added, mixed well, and incubated at 37°C for 40min. Finally, the A550 absorbance value of samples 2.11. Mice Autophagy Measurement. C57BL/6 mice were purchased from the Zhejiang Academy of Medical Sciences in Hangzhou, China. The autophagy assay was conducted by following the procedures of reference [26]. Six-week- old male C57BL/6 wild-type mice were injected intraperito- neally with 3 μM EHR. The EHR was dissolved in 50 μL DMSO before dilution with 150 μL saline. The mice were treated with leupeptin in 200 μL saline or vehicle saline after 4 h. The mice were then sacriﬁced after 2 h, and their liver and heart were collected. This animal experiment was per- formed in accordance with international ethical standards and was guided and approved by the Committee of Exper- imental Animal Care of Zhejiang University (permit num- ber ZJU20190143). Approximately, 100 mg of the liver tissue or 1 whole heart was homogenized in a lysis buﬀer containing a RIPA lysis buﬀer, 1% protease inhibitor and 1% cocktail 2 (CoWin Biotech, Beijing, China), and 1% cocktail 3 (Sigma, Saint Louis, USA). The homogenization was followed by a 20 min settlement, and the samples were centrifuged at 12,000×g for 10min at 4°C. The supernatants were collected, and the protein concentration was measured by using the BCA protein assay kit (CoWin Biotech, Beijing, China). The proteins of the heart and liver tissues were used for the western blot analysis. 2.12. Real-Time Polymerase Chain Reaction (RT-PCR) Analysis. First, the wild-type BY4741 was incubated with EHR at a concentration of 0, 3, and 10 μM in a glucose medium for 12 h at 28°C with shaking at 180 rpm. Total RNA was extracted using the hot phenol method. The reverse transcription method was employed to synthesize cDNA using the HiFi-MMLV cDNA Kit (Cowin Biotech, Beijing, P. R. China) and 5 ug of RNA. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. YOM38 yeast cells containing the pRS316-GFP-ATG8 plasmid were treated with 300 μM RES and EHR at 0, 1, 3, and 10μM for 22 h or with EHR at 3 μM for 0, 8, 15, and 22h to examine the changes in free green ﬂuorescent protein (GFP) in yeast autophagy. All treated and untreated cells were harvested by centrifugation at 12,000×g for 10 min, washed thrice with PBS, and sus- pended in 150 or 250μL PBS. The yeast cells were collected, sonicated on ice for 5 min, and centrifuged to obtain the supernatant. The BCA assay kit was used to measure the con- centration of proteins in the supernatant. Brieﬂy, the 200 μL BCA working solution and 25 μL of each standard or sample were added into each well of 96-well plate. Each sample was repeated twice and mixed well. Subsequently, the 96-well plate was incubated at 37°C for 25 min, and the absorbance of the BSA standard and protein samples at 562 nm of wave- length was measured with the BioTek Microplate Reader (detail method is provided in supplementary information). Approximately, 20 and 100μg of protein, respectively, from the yeast and animal samples were separated by 12% or 13% SDS-PAGE and transferred to PVDF membranes. These membranes were then incubated with primary antibodies speciﬁc to GFP (1 : 1000) (#598, Medical & Biological Labora- tories, Nagoya, Japan), β-actin (#CW0096, CoWin Biotech, Beijing, China), microtubule-associated protein 1 light chain 3B (LC3B) (1: 1000) (#2775 s, Cell Signaling Technology, Boston, USA), or glyceraldehyde 3-phosphate dehydroge- nase (#CW0100a, Beijing ComWin Biotechnology, Beijing, China) for 1 h, followed by secondary antibodies (1 : 5000) (horseradish peroxidase-linked anti-rabbit (#CW0103) or anti-mouse IgGs (#CW0102) (CoWin Biotech, Beijing, China). The antigens were visualized by using the ECL West- ern Blot Kit (CoWin Biotech, Beijing, China), whereas the protein bands were analyzed by using the ImageJ software (National Institute of Health, Rockville, MD, United States). 3.2. Antioxidative Stress Activity of EHR. Oxidative stress, which is a crucial factor in the advancement of the aging pro- cess, is caused by excessive levels of ROS; these levels cannot be resisted by the antioxidant system. This stress can damage DNAs, proteins, and lipids and impair physiological func- tions [28]. The survival capability of yeast under H2O2 was evaluated to highlight the important role of antioxidative stress in the antiaging eﬀect of EHR. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. Quantitative RT-PCR was per- formed by using CFX96 Touch (Bio-Rad, Hercules, USA) and SYBR Premix EX Taq (Takara, Otsu, Japan) based on the previous study [16]. The thermal cycling parameters were as follows: 40 cycles, 94°C for 15 s, 60°C for 25 s, and 72°C for Oxidative Medicine and Cellular Longevity 5 ing the replicative lifespan of K6001 at 1, 3, and 10μM (Figure 1(b)). Thus, this molecule is a valuable substance for intensive studies. The chronological lifespan of YOM36 after treatment with EHR was then studied. The survival time of EHR-treated groups was 15 days, which was signiﬁ- cantly longer than that of the control group (13 days) at con- centrations of 3 and 10μM (Figure 1(c)). Although yeast has many vital features that are evolutionally conserved in mam- mal cells, the antiaging activity of EHR in mammal cells must be conﬁrmed to evaluate the antiaging activity of this molecule. Therefore, the conducted studies have examined the eﬀect of EHR on the yeast-like chronological lifespan of PC12 cells and used rapamycin as a positive control. EHR can signiﬁcantly increase the viability of PC12 cells at concentrations of 0.1 and 0.3 μM compared with the control group (Figures 1(d) and 1(e), respectively). These results generally demonstrate the antiaging eﬀects of EHR on yeast and mammals. 20 s. The sequences of the primers for RT-PCR were as fol- lows: for SIR2, sense 5′-CGT TCC CCA AGT CCT GAT TA-3′ and anti-sense 5′-CCA CAT TTT TGG GCT ACC AT-3′; for TUB1, sense 5′-CCA AGG GCT ATT TAC GTG GA-3′ and anti-sense 5′-GGT GTA ATG GCC TCT TGC AT-3′. The relative gene expression data were analyzed by the 2-ΔΔCt method. The levels of SIR2 mRNA were nor- malized to those of TUB1. 20 s. The sequences of the primers for RT-PCR were as fol- lows: for SIR2, sense 5′-CGT TCC CCA AGT CCT GAT TA-3′ and anti-sense 5′-CCA CAT TTT TGG GCT ACC AT-3′; for TUB1, sense 5′-CCA AGG GCT ATT TAC GTG GA-3′ and anti-sense 5′-GGT GTA ATG GCC TCT TGC AT-3′. The relative gene expression data were analyzed by the 2-ΔΔCt method. The levels of SIR2 mRNA were nor- malized to those of TUB1. 2.13. Western Blot Analysis. resolution ESI-TOF-MS m/z 371.1250, which is calculated for C22H20O4Na (M + Na)+371.1254. Diﬀerences were initially observed in the survivability of each group under oxidative stress (Figure 2(a)). Moreover, after a certain number of these cells were cultured with H2O2, the antioxidant activity of EHR was quantitatively identiﬁed and compared with RES as a positive control (Figure 2(b)) according to the survival rate of yeast cells. The survival capability of yeast signiﬁcantly increased and became comparable with that of the positive control after EHR treatment. In the mitochondrial electron transport chain, ROS is generated as a byproduct involved in the aging process [29]. The ROS at low concentrations can maintain a normal cellu- lar metabolism but may fatally damage cells, tissues, and organisms at abnormally high concentrations [5]. MDA, as an index of membrane lipid peroxidation, is also produced by ROS during lipid reaction. Therefore, ROS and MDA levels in yeast have been studied to determine the degree of intracellular oxidation. At 24h, EHR signiﬁcantly reduced the MDA level in yeast at concentrations of 3 and 10 μM; at 48 h, EHR signiﬁcantly reduced the MDA level at a concen- tration of 10 μM (Figure 2(c)). Figure 2(d) shows that the ﬂuorescence intensity of yeast signiﬁcantly decreased after EHR treatment at 24 and 48 h. Therefore, EHR can con- stantly abate the ROS and MDA levels of yeast, thereby dem- onstrating its antioxidative stress activity. 2.14. Statistical Analysis. The data were expressed as mean ± SEM values of three independent experiments and sub- jected to one-way ANOVA and Dunnett’s multiple compar- ison tests or log-rank (Mantel-Cox) test by using the GraphPad Prism software. Statistical signiﬁcance was estab- lished at p < 0:05. g y Antioxidant enzymes, including CAT, GPx, and SOD, are important components in the antioxidant system of organisms. CAT catalyzes the decomposition of hydrogen peroxide, GPx breaks down the peroxide in organisms to pre- vent oxidative stress, and SOD catalyzes O2 −to less toxic hydrogen peroxide [28, 30]. Therefore, the enzymatic activi- ties of CAT, GPx, and SOD were measured in this work. A signiﬁcant increase in the total GPx and CAT activities was 3. Results 3.1. EHR Extension of the Lifespan. A series of benzoquinone- type molecules with antiaging activity were isolated from O. bracteatum in a previous study based on the K6001 yeast bioassay system [23]. Among these molecules, EHR (Figure 1(a)) demonstrated the best performance in extend- Oxidative Medicine and Cellular Longevity 6 O O HO (a) 0 4 8 12 16 20 0 20 40 60 80 100 Control RES EHR (0.1 𝜇M) EHR (1 𝜇M)⁎ EHR (3 𝜇M)⁎⁎ EHR (10 𝜇M)⁎⁎⁎ EHR (30 𝜇M) Generations Viability (%) (b) 0 3 6 9 12 15 0 20 40 60 80 100 Control EHR (3 𝜇M)⁎⁎⁎ EHR (10 𝜇M)⁎⁎⁎ Survival rate (%) Days (c) EHR (0.1 𝜇M) EHR (0.3 𝜇M) Rapamycin (0.5 𝜇M) C EHR (1 𝜇M) (d) C 0.1 0.3 1 0.5 0 50 100 150 Rapamycin (𝜇M) EHR (𝜇M) Number of cell colonies ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ (e) The chemical structure and eﬀect of EHR on the lifespan of yeasts and mammals. (a) The chemical structure of EHR. ( e lifespan of K6001 yeast after treating EHR. The mean lifespan of each treatment group is as follows: control (7:42 ± 0:5 (10:20 ± 0:77∗∗), EHR at 0.1 μM (7:30 ± 0:48), EHR at 1 μM (9:63 ± 0:64∗), EHR at 3 μM (10:33 ± 0:70∗∗), EHR at 0:57∗∗∗), and EHR at 30 μM (7:78 ± 0:54). (c) The chronological lifespan of YOM36 after treating EHR. The mean lifes tment group is as follows: control (6:99 ± 0:15), EHR at 3 μM (8:57 ± 0:22∗∗∗), and EHR at 10 μM (7:95 ± 0:20∗∗∗). ( olony-forming units (CFUs) of PC12 cells. (e) Corresponding number of cell colonies in (d). Each assay was performed ample numbers of each group were six. In the present study, one-way ANOVA and Dunnett’s multiple comparison tes replicative lifespan and yeast-like chorological lifespan assay. The log-rank (Mantel-Cox) test was used for the yeast choro ssay. ∗, ∗∗, and ∗∗∗represent signiﬁcant diﬀerences from the control group at p < 0:05, p < 0:01, and p < 0:001, respective e lifespan assay needed 4 days; the chronological lifespan assay needed 15-20 days as well as the yeast-like chronological l ded 22 days. 3. Results 0 3 6 9 12 15 0 20 40 60 80 100 Survival rate (%) Days Survival rate (%) C C 0.1 0.3 1 0.5 0 50 100 150 Rapamycin (𝜇M) EHR (𝜇M) Number of cell colonies ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ (e) EHR (0.3 𝜇M) EHR (1 𝜇M) (d) (d) (e) Figure 1: The chemical structure and eﬀect of EHR on the lifespan of yeasts and mammals. (a) The chemical structure of EHR. (b) The replicative lifespan of K6001 yeast after treating EHR. The mean lifespan of each treatment group is as follows: control (7:42 ± 0:54), RES at 10 μM (10:20 ± 0:77∗∗), EHR at 0.1 μM (7:30 ± 0:48), EHR at 1 μM (9:63 ± 0:64∗), EHR at 3 μM (10:33 ± 0:70∗∗), EHR at 10 μM (10:58 ± 0:57∗∗∗), and EHR at 30 μM (7:78 ± 0:54). (c) The chronological lifespan of YOM36 after treating EHR. The mean lifespan of each treatment group is as follows: control (6:99 ± 0:15), EHR at 3 μM (8:57 ± 0:22∗∗∗), and EHR at 10 μM (7:95 ± 0:20∗∗∗). (d) The survival colony-forming units (CFUs) of PC12 cells. (e) Corresponding number of cell colonies in (d). Each assay was performed thrice, and the sample numbers of each group were six. In the present study, one-way ANOVA and Dunnett’s multiple comparison tests were used for replicative lifespan and yeast-like chorological lifespan assay. The log-rank (Mantel-Cox) test was used for the yeast chorological lifespan assay. ∗, ∗∗, and ∗∗∗represent signiﬁcant diﬀerences from the control group at p < 0:05, p < 0:01, and p < 0:001, respectively. The replicative lifespan assay needed 4 days; the chronological lifespan assay needed 15-20 days as well as the yeast-like chronological lifespan assay needed 22 days. Figure 1: The chemical structure and eﬀect of EHR on the lifespan of yeasts and mammals. (a) The chemical structure of EHR. (b) The replicative lifespan of K6001 yeast after treating EHR. The mean lifespan of each treatment group is as follows: control (7:42 ± 0:54), RES at 10 μM (10:20 ± 0:77∗∗), EHR at 0.1 μM (7:30 ± 0:48), EHR at 1 μM (9:63 ± 0:64∗), EHR at 3 μM (10:33 ± 0:70∗∗), EHR at 10 μM (10:58 ± 0:57∗∗∗), and EHR at 30 μM (7:78 ± 0:54). (c) The chronological lifespan of YOM36 after treating EHR. 3. Results (c) The MDA level of yeast in 24 h and 48 h (∗p < 0:05, ∗∗p < 0:01, and ∗∗∗p 7 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 Survival rate (%) ⁎ ⁎⁎ ⁎⁎ 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 Survival rate (%) ⁎ ⁎⁎ ⁎⁎ (b) 0 1 2 3 4 5 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 Fluorescence/107 cells ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎ Control RES EHR (3 𝜇M) EHR (10 𝜇M) (a) Control RES EHR (3 𝜇M) EHR (10 𝜇M) (a) 0 20 40 C RES EHR (𝜇M) 3 10 Survival (b) 0.0 0.2 0.4 0.6 0.8 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 MDA (nmol/mg protein) ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎ ⁎ (c) 0 1 2 3 4 5 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 Fluorescence/107 cells ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎ (d) 0 50 100 150 C 3 10 EHR (𝜇M) RES GPx activity (mU/mg) ⁎ ⁎ (e) 0 2 4 6 8 10 C 3 10 EHR (𝜇M) RES CAT activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎ (f) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 T-SOD activity (units/mg protein) ⁎⁎⁎ ⁎ ⁎ (g) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 CuZn-SOD activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎⁎⁎ (h) antioxidant activity of EHR in yeast. (a) Qualitative survival condition of yeast under H2O2. (b) Quantitative survival rate of O2 (∗p < 0:05 and ∗∗p < 0:01). (c) The MDA level of yeast in 24 h and 48 h (∗p < 0:05, ∗∗p < 0:01, and ∗∗∗p < 0:001). (d) ROS as represented by the ﬂuorescence intensity of 1 × 107 yeast cells in 24 h and 48 h (∗p < 0:05, ∗∗p < 0:01, and ∗∗∗p < 0:001). (e) yeast (∗p < 0:05). (f) CAT activity of yeast (∗p < 0:05 and ∗∗p < 0:01). (g, h) Total SOD activity and CuZn-SOD activity after HR at 24 h (∗p < 0:05, ∗∗p < 0:01, ∗∗∗p < 0:001). Each assay was performed thrice, and the sample numbers of each group were ANOVA and Dunnett’s multiple comparison tests were used to analyze the results of each experiment. 3. Results The mean lifespan of each treatment group is as follows: control (6:99 ± 0:15), EHR at 3 μM (8:57 ± 0:22∗∗∗), and EHR at 10 μM (7:95 ± 0:20∗∗∗). (d) The survival colony-forming units (CFUs) of PC12 cells. (e) Corresponding number of cell colonies in (d). Each assay was performed thrice, and the sample numbers of each group were six. In the present study, one-way ANOVA and Dunnett’s multiple comparison tests were used for replicative lifespan and yeast-like chorological lifespan assay. The log-rank (Mantel-Cox) test was used for the yeast chorological lifespan assay. ∗, ∗∗, and ∗∗∗represent signiﬁcant diﬀerences from the control group at p < 0:05, p < 0:01, and p < 0:001, respectively. The replicative lifespan assay needed 4 days; the chronological lifespan assay needed 15-20 days as well as the yeast-like chronological lifespan assay needed 22 days. 7 Oxidative Medicine and Cellular Longevity Control RES EHR (3 𝜇M) EHR (10 𝜇M) (a) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 Survival rate (%) ⁎ ⁎⁎ ⁎⁎ (b) 0.0 0.2 0.4 0.6 0.8 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 MDA (nmol/mg protein) ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎ ⁎ (c) 0 1 2 3 4 5 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 Fluorescence/107 cells ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎ (d) 0 50 100 150 C 3 10 EHR (𝜇M) RES GPx activity (mU/mg) ⁎ ⁎ (e) 0 2 4 6 8 10 C 3 10 EHR (𝜇M) RES CAT activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎ (f) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 T-SOD activity (units/mg protein) ⁎⁎⁎ ⁎ ⁎ (g) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 CuZn-SOD activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎⁎⁎ (h) antioxidant activity of EHR in yeast. (a) Qualitative survival condition of yeast under H2O2. (b) Quantitati 2O2 (∗p < 0:05 and ∗∗p < 0:01). 3. Results The average lifespan of K6001 in the control group was 6:58 ± 0:38, whereas those of RES and EHR at 10 μM were 9:1 ± 0:67 (p < 0:01) and 9:2 ± 0:64 (p < 0:001), respectively. (a) The average lifespan of Δsod1 in the control group was 5:78 ± 0:28, whereas that of EHR at 10 μM was 6:43 ± 0:43. (b) The average lifespan of Δsod2 in the control group was 6:45 ± 0:48, whereas that of EHR at 10 μM was 6:43 ± 0:45. (c) The average lifespan of Δuth1 in the control group was 10:58 ± 0:66 (p < 0:001), whereas that of EHR at 10 μM was 9:15 ± 0:63. (d) The average lifespan of Δskn7 in the control group was 7:65 ± 0:57, whereas that of EHR at 10 μM was 7:6 ± 0:51. In the lifespan assay of Δgpx and Δcat, the average lifespan of K6001 in the control group was 7:98 ± 0:58, whereas those of RES and EHR at 10 μM were 11:50 ± 0:69 (p < 0:001) and 11:48 ± 0:59 (p < 0:001), respectively. (e) The average lifespan of Δgpx in the control group was 7:28 ± 0:56, whereas those of RES and EHR at 10 μM were 8:10 ± 0:61 and 6:65 ± 0:41, respectively. (f) The average lifespan of Δcat in the control group was 6:73 ± 0:46, whereas those of RES and EHR at 10 μM were 6:85 ± 0:50 and 6:73 ± 0:42. Each experiment was repeated thrice. Viability (%) Viability (%) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥skn7) EHR 10 𝜇M (𝛥skn7) (d) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations Viability (%) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥cat) RES 10 𝜇M (𝛥cat) EHR 10 𝜇M (𝛥cat) (f) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥gpx) RES 10 𝜇M (𝛥gpx) EHR 10 𝜇M (𝛥gpx) (e) (f) (e) Figure 3: The eﬀects of EHR on the replicative lifespan of Δsod1 (a), Δsod2 (b), Δuth1 (c), Δskn7 (d), Δgpx (e), and Δcat (f) with a K6001 background. The average lifespan of K6001 in the control group was 6:58 ± 0:38, whereas those of RES and EHR at 10 μM were 9:1 ± 0:67 (p < 0:01) and 9:2 ± 0:64 (p < 0:001), respectively. 3. Results (e) The average lifespan of Δgpx in the control group of RES and EHR at 10 μM were 8:10 ± 0:61 and 6:65 ± 0:41, respectively. (f) The average lifespan of Δca ± 0:46, whereas those of RES and EHR at 10 μM were 6:85 ± 0:50 and 6:73 ± 0:42. Each experiment was re 0 4 8 12 16 20 0 20 40 60 80 100 Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥sod1) EHR 10 𝜇M (𝛥sod1) Generations Viability (%) (a) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥sod2) EHR 10 𝜇M (𝛥sod2) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations (b) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations Viability (%) Viability (%) 0 4 8 12 16 20 0 20 40 Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥sod1) EHR 10 𝜇M (𝛥sod1) Generations Viabili (a) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥sod2) EHR 10 𝜇M (𝛥sod2) Viabili 0 4 8 12 16 20 0 20 40 Generations (b) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥uth1) EHR 10 𝜇M (𝛥uth1) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations (c) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥skn7) EHR 10 𝜇M (𝛥skn7) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations (d) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥gpx) RES 10 𝜇M (𝛥gpx) EHR 10 𝜇M (𝛥gpx) Generations (e) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥cat) RES 10 𝜇M (𝛥cat) EHR 10 𝜇M (𝛥cat) (f) Figure 3: The eﬀects of EHR on the replicative lifespan of Δsod1 (a), Δsod2 (b), Δuth1 (c), Δskn7 (d), Δgpx (e), and Δcat (f) with a K6001 background. 3. Results (b) 0 1 2 3 4 5 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 Fluorescence/107 cells ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎ (a) 0.0 0.2 0.4 0.6 0.8 24 h 48 h C RES EHR (𝜇M) 3 10 C RES EHR (𝜇M) 3 10 MDA (nmol/mg protein) ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎ ⁎ (c) (d) 0 50 100 150 C 3 10 EHR (𝜇M) RES GPx activity (mU/mg) ⁎ ⁎ 0 2 4 6 8 10 C 3 10 EHR (𝜇M) RES CAT activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎ (d) (c) (𝜇 ) (e) (𝜇 ) (f) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 T-SOD activity (units/mg protein) ⁎⁎⁎ ⁎ ⁎ (g) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 CuZn-SOD activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎⁎⁎ (h) (f) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 CuZn-SOD activity (units/mg protein) ⁎⁎ ⁎⁎ ⁎⁎⁎ (h) (e) 0 20 40 60 80 100 C RES EHR (𝜇M) 3 10 T-SOD activity (units/mg protein) ⁎⁎⁎ ⁎ ⁎ (g) (f) (h) (g) Figure 2: The antioxidant activity of EHR in yeast. (a) Qualitative survival condition of yeast under H2O2. (b) Quantitative survival rate of yeast under H2O2 (∗p < 0:05 and ∗∗p < 0:01). (c) The MDA level of yeast in 24 h and 48 h (∗p < 0:05, ∗∗p < 0:01, and ∗∗∗p < 0:001). (d) ROS level of yeast was represented by the ﬂuorescence intensity of 1 × 107 yeast cells in 24 h and 48 h (∗p < 0:05, ∗∗p < 0:01, and ∗∗∗p < 0:001). (e) GPx activity of yeast (∗p < 0:05). (f) CAT activity of yeast (∗p < 0:05 and ∗∗p < 0:01). (g, h) Total SOD activity and CuZn-SOD activity after treatment of EHR at 24 h (∗p < 0:05, ∗∗p < 0:01, ∗∗∗p < 0:001). Each assay was performed thrice, and the sample numbers of each group were ﬁve. One-way ANOVA and Dunnett’s multiple comparison tests were used to analyze the results of each experiment. 3. Results Oxidative Medicine and Cellular Longevity 8 0 4 8 12 16 20 0 20 40 60 80 100 Generations Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥sod1) EHR 10 𝜇M (𝛥sod1) Generations Viability (%) (a) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥sod2) EHR 10 𝜇M (𝛥sod2) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations (b) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥uth1) EHR 10 𝜇M (𝛥uth1) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations (c) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥skn7) EHR 10 𝜇M (𝛥skn7) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations (d) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥gpx) RES 10 𝜇M (𝛥gpx) EHR 10 𝜇M (𝛥gpx) Generations (e) Viability (%) 0 4 8 12 16 20 0 20 40 60 80 100 Generations Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥cat) RES 10 𝜇M (𝛥cat) EHR 10 𝜇M (𝛥cat) (f) eﬀects of EHR on the replicative lifespan of Δsod1 (a), Δsod2 (b), Δuth1 (c), Δskn7 (d), Δgpx (e), and Δcat he average lifespan of K6001 in the control group was 6:58 ± 0:38, whereas those of RES and EHR at 10 μM 9:2 ± 0:64 (p < 0:001), respectively. (a) The average lifespan of Δsod1 in the control group was 5:78 ± 0:28, was 6:43 ± 0:43. (b) The average lifespan of Δsod2 in the control group was 6:45 ± 0:48, whereas that of EH The average lifespan of Δuth1 in the control group was 10:58 ± 0:66 (p < 0:001), whereas that of EHR at 10 verage lifespan of Δskn7 in the control group was 7:65 ± 0:57, whereas that of EHR at 10 μM was 7:6 ± 0:51 nd Δcat, the average lifespan of K6001 in the control group was 7:98 ± 0:58, whereas those of RES and EH < 0:001) and 11:48 ± 0:59 (p < 0:001), respectively. 3. Results (a) The average lifespan of Δsod1 in the control group was 5:78 ± 0:28, whereas that of EHR at 10 μM was 6:43 ± 0:43. (b) The average lifespan of Δsod2 in the control group was 6:45 ± 0:48, whereas that of EHR at 10 μM was 6:43 ± 0:45. (c) The average lifespan of Δuth1 in the control group was 10:58 ± 0:66 (p < 0:001), whereas that of EHR at 10 μM was 9:15 ± 0:63. (d) The average lifespan of Δskn7 in the control group was 7:65 ± 0:57, whereas that of EHR at 10 μM was 7:6 ± 0:51. In the lifespan assay of Δgpx and Δcat, the average lifespan of K6001 in the control group was 7:98 ± 0:58, whereas those of RES and EHR at 10 μM were 11:50 ± 0:69 (p < 0:001) and 11:48 ± 0:59 (p < 0:001), respectively. (e) The average lifespan of Δgpx in the control group was 7:28 ± 0:56, whereas those of RES and EHR at 10 μM were 8:10 ± 0:61 and 6:65 ± 0:41, respectively. (f) The average lifespan of Δcat in the control group was 6:73 ± 0:46, whereas those of RES and EHR at 10 μM were 6:85 ± 0:50 and 6:73 ± 0:42. Each experiment was repeated thrice. 9 Oxidative Medicine and Cellular Longevity 9 Control TG8 DIC RES (300 𝜇M) EHR (10 𝜇M) EHR (3 𝜇M) (a) 0 20 40 60 Control RES (300 𝜇M) 3 10 EHR (𝜇M) Percentage (%) ⁎⁎ ⁎⁎ ⁎ (b) Control RES 1 3 10 (kDa) 41 27 43 𝛽-Actin EHR (𝜇M) GFP-ATG8 GFP (c) 0 1 2 3 1 3 10 Control Relative protein level of free GFP RES (300 𝜇M) EHR (𝜇M) ⁎⁎⁎ ⁎ ⁎ ⁎ (d) 0 h 8 h 15 h 22 h 22 h (kDa) 41 27 43 𝛽-Actin 3 𝜇M Control GFP-ATG8 GFP (e) 0 1 2 3 0 h 8 h 15 h 22 h Control (22 h) 3 𝜇M EHR Relative protein level of free GFP ⁎⁎⁎ ⁎ (f) Figure 4: Continued. 3. Results observed in EHR-treated groups as shown in Figures 2(e) and 2(f), while total SOD and CuZn-SOD activities were signiﬁ- cantly improved after treatment of EHR at doses of 3 and 10 μM for 24 h, respectively (Figures 2(g) and 2(h)). This ﬁnding suggests that EHR counters oxidative stress by increasing the activities of CAT, GPx, and SOD. Overall, EHR signiﬁcantly reduces ROS and MDA levels while increasing CAT, total GPx, and SOD activities to protect cells from oxidative stress, thereby highlighting its antioxidative stress activity. observed in EHR-treated groups as shown in Figures 2(e) and 2(f), while total SOD and CuZn-SOD activities were signiﬁ- cantly improved after treatment of EHR at doses of 3 and 10 μM for 24 h, respectively (Figures 2(g) and 2(h)). This ﬁnding suggests that EHR counters oxidative stress by increasing the activities of CAT, GPx, and SOD. Overall, EHR signiﬁcantly reduces ROS and MDA levels while increasing CAT, total GPx, and SOD activities to protect cells from oxidative stress, thereby highlighting its antioxidative stress activity. 3.4. EHR Induces Autophagy in Yeast. The autophagy level decreases with aging; speciﬁcally, inducing autophagy shows a potent antiaging eﬀect, whereas decreasing autophagy can accelerate aging [36]. In yeast and mammalian cells, Atg8 and LC3B are molecular markers for monitoring the autoph- agic process. The Atg8 conjugated to phosphatidylethanol- amine is localized to a preautophagosomal structure, which is crucial in autophagosome formation. Therefore, Atg8 plays an important role in measuring the autophagy level [37]. The YOM38 yeast strain, which expresses GFP-Atg8, was used in this study to monitor the autophagy level via ﬂuorescence microscopy. Figure 4(a) presents the ﬂuorescent images of GFP in autophagosome in yeast, while Figure 4(b) presents the statistical results. EHR signiﬁcantly increased the per- centage of cells with green ﬂuorescence at concentrations of 3 and 10 μM. The EHR concentration at 3 μM showed a bet- ter eﬀect than that of 10μM EHR. The autophagy ﬂux repre- sents the entire autophagy process, and free GFP was eventually released in the vacuole. Therefore, the expression of free GFP was analyzed using western blot to further con- ﬁrm the autophagy level. As shown in Figures 4(c) and 4(d) and S1a, EHR signiﬁcantly enhanced the level of free GFP at 1, 3, and 10 μM, respectively, among which the EHR at 3 μM exhibited the best eﬀect. 3. Results 0 20 40 60 Control RES (300 𝜇M) 3 10 EHR (𝜇M) Percentage (%) ⁎⁎ ⁎⁎ ⁎ (b) RES (300 𝜇M) EHR (3 𝜇M) EHR (b) 0 1 2 3 1 3 10 Control Relative protein level of free GFP RES (300 𝜇M) EHR (𝜇M) ⁎⁎⁎ ⁎ ⁎ ⁎ (d) Control RES 1 3 10 (kDa) 41 27 43 𝛽-Actin EHR (𝜇M) GFP-ATG8 GFP (c) (d) 0 1 2 3 0 h 8 h 15 h 22 h Control (22 h) 3 𝜇M EHR Relative protein level of free GFP ⁎⁎⁎ ⁎ (f) (c) (d) 0 h 8 h 15 h 22 h 22 h (kDa) 41 27 43 𝛽-Actin 3 𝜇M Control GFP-AT GFP (e) (e) (f) Figure 4: Continued. Oxidative Medicine and Cellular Longevity 10 0 2 4 6 8 10 12 14 16 18 20 0 20 40 60 80 100 Generations Viability (%) 0 2 4 6 8 10 12 14 16 18 20 0 20 40 60 80 100 Generations Viability (%) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥atg2) RES 10 𝜇M (𝛥atg2) EHR 10 𝜇M (𝛥atg2) (g) 0 2 4 6 8 10 12 14 16 18 20 0 20 40 60 80 100 Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥atg32) RES 10 𝜇M (𝛥atg32) EHR 10 𝜇M (𝛥atg32) Generations Viability (%) (h) Figure 4: Eﬀects of EHR on the autophagy of yeast. (a) The merge ﬂuorescence images of DAPI and GFP and the image of YOM38-GFP-atg8 yeast after EHR treatment. (b) The percentage of yeast cells containing autophagosome in the observation cells (∗p < 0:05 and ∗∗p < 0:01). (c, e) Western blot analysis results for the free GFP of yeast in dose and time courses. (d, f) The digital results of (c) and (e), respectively (∗p < 0:05 and ∗∗∗p < 0:001). (g, h) The changes on replicative lifespan of Δatg2 and Δatg32 of yeast. The average lifespan of K6001 in the control group was 8:05 ± 0:59, whereas those of RES and EHR at 10 μM were 11:8 ± 0:73 (∗∗∗p < 0:001) and 11:6 ± 0:60 (∗∗∗p < 0:001), respectively. (g) The average lifespan of Δatg2 in the control group was 6:78 ± 0:45, whereas those of RES and EHR at 10 μM were 7:53 ± 0:56 and 7:78 ± 0:60, respectively. 3. Results (h) The average lifespan of Δatg32 in the control group was 8:40 ± 0:64, whereas those of RES and EHR at 10 μM were 7:75 ± 0:50 and 8:48 ± 0:58, respectively. Each experiment was repeated thrice, and the sample numbers of each group were four. One-way ANOVA and Dunnett’s multiple comparison tests were used to analyze the results of each experiment. g y 0 2 4 6 8 10 12 14 16 18 20 0 20 40 60 80 100 Generations Viability (%) Generations Generations Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥atg2) RES 10 𝜇M (𝛥atg2) EHR 10 𝜇M (𝛥atg2) (g) Control (K6001) RES 10 𝜇M (K6001) EHR 10 𝜇M (K6001) Control (𝛥atg32) RES 10 𝜇M (𝛥atg32) EHR 10 𝜇M (𝛥atg32) (h) (h) (g) Figure 4: Eﬀects of EHR on the autophagy of yeast. (a) The merge ﬂuorescence images of DAPI and GFP and the image of YOM38-GFP-atg8 yeast after EHR treatment. (b) The percentage of yeast cells containing autophagosome in the observation cells (∗p < 0:05 and ∗∗p < 0:01). (c, e) Western blot analysis results for the free GFP of yeast in dose and time courses. (d, f) The digital results of (c) and (e), respectively (∗p < 0:05 and ∗∗∗p < 0:001). (g, h) The changes on replicative lifespan of Δatg2 and Δatg32 of yeast. The average lifespan of K6001 in the control group was 8:05 ± 0:59, whereas those of RES and EHR at 10 μM were 11:8 ± 0:73 (∗∗∗p < 0:001) and 11:6 ± 0:60 (∗∗∗p < 0:001), respectively. (g) The average lifespan of Δatg2 in the control group was 6:78 ± 0:45, whereas those of RES and EHR at 10 μM were 7:53 ± 0:56 and 7:78 ± 0:60, respectively. (h) The average lifespan of Δatg32 in the control group was 8:40 ± 0:64, whereas those of RES and EHR at 10 μM were 7:75 ± 0:50 and 8:48 ± 0:58, respectively. Each experiment was repeated thrice, and the sample numbers of each group were four. One-way ANOVA and Dunnett’s multiple comparison tests were used to analyze the results of each experiment. background (Figure 3). In other words, SOD1, SOD2, GPx, CAT, UTH1, and SKN7 genes are involved in the antiaging eﬀect of EHR. Oxidative Medicine and Cellular Longevity 11 Control Liver (vehicle) (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (a) Liver (leupeptin) Control (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (b) Control Heart (vehicle) (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (c) Heart (leupeptin) Control (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (d) 0 1 2 3 4 5 C EHR C EHR Vehicle Leupeptin Liver Relative protein level of LC3B-II ⁎⁎⁎ ⁎ (e) 0 1 2 3 ns C EHR C EHR Vehicle Leupeptin Heart Relative protein level of LC3B-II ⁎⁎⁎ (f) gure 5: The eﬀects of EHR on the autophagy in mice in vivo. (a, b) The LC3B protein level in the mice liver after administrating EHR and HR plus leupeptin, respectively. (c, d) The LC3B protein level in the mice heart after administrating EHR and EHR plus leupeptin, spectively. (e) The digital results of western blotting analysis in (a) and (b). (f) The digital results of western blotting analysis in (c) and d). Each experiment was repeated thrice, and the sample numbers of each group were three. ∗and ∗∗∗represent signiﬁcant diﬀerences om the control group at p < 0:05 and p < 0:001, respectively. One-way ANOVA and Dunnett’s multiple comparison tests were used to nalyze the results of each experiment. Liver (leupeptin) Control (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (b) Heart (leupeptin) Control (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (d) H (b) Heart (leupeptin) Control (kDa) 16 14 38 LC3B-I LC3B-II GAPDH EHR (3 𝜇M) (d) Heart (d) 0 1 2 3 ns C EHR C EHR Vehicle Leupeptin Heart Relative protein level of LC3B-II ⁎⁎⁎ (f) (f) (e) Figure 5: The eﬀects of EHR on the autophagy in mice in vivo. (a, b) The LC3B protein level in the mice liver after administrating EHR and EHR plus leupeptin, respectively. (c, d) The LC3B protein level in the mice heart after administrating EHR and EHR plus leupeptin, respectively. (e) The digital results of western blotting analysis in (a) and (b). (f) The digital results of western blotting analysis in (c) and (d). Each experiment was repeated thrice, and the sample numbers of each group were three. ∗and ∗∗∗represent signiﬁcant diﬀerences from the control group at p < 0:05 and p < 0:001, respectively. 3. Results Therefore, the EHR at a con- centration of 3 μM was used to examine the time- dependent behavior of GFP after treatment. The released 3.3. EHR cannot Extend the Replicative Lifespan of Oxidative Stress-Related Yeast Mutants. As an antioxidant enzyme, SOD can eﬀectively remove superoxide anion radicals to counteract oxidative stress, and the overexpression of SOD encoded by SOD1 and SOD2 genes can extend cellular lon- gevity [31, 32]. UTH1 is a gene that can respond to oxidative stress, and its deletion extends the lifespan of yeast [33]. Skn7 is a major transcription factor that controls the oxidative stress response of Saccharomyces cerevisiae [34]. Replicative lifespan assays of these mutants were performed to under- stand the relationship between the antiaging eﬀects of EHR and those of SOD1, SOD2, GPx, CAT, UTH1, and SKN7 genes. Consistent with previous reports, Figure 3 shows that the replicative lifespan of Δsod1, Δsod2, Δgpx, and Δcat with a K6001 background is similar to that of wild-type K6001, whereas the replicative lifespan of the Δuth1 with a K6001 background is longer than that of wild-type K6001 [33–35]. Furthermore, EHR cannot extend the replicative lifespan of Δsod1, Δsod2, Δgpx, Δcat, Δuth1, and Δskn7 with a K6001 Oxidative Medicine and Cellular Longevity 11 Oxidative Medicine and Cellular Longevity One-way ANOVA and Dunnett’s multiple comparison tests were used to analyze the results of each experiment. GFP is time-dependent within the tested time range (Figures 4(e), 4(f), and S1b), and ATG2 and ATG32 are two of the most important genes that mediate autophagy in yeast. Therefore, the replicative lifespan assay of Δatg2 and Δatg32 of yeast with a K6001 background was further performed. EHR did not aﬀect the replicative lifespan of these mutants (Figures 4(g) and 4(h)), thereby conﬁrming the involvement of autophagy in the antiaging activity of EHR in yeast. inhibitor is an eﬀective tool for blocking the autophagy ﬂux when determining the actual autophagy level. Therefore, leu- peptin was used to further investigate the change in the autophagy level after EHR treatment. In the experiment, intraperitoneal injection of EHR was administered among mice at 3 μM for 6 h, and the mice were treated with or with- out leupeptin after injection of EHR for 4 h. EHR signiﬁ- cantly promoted the expression of LC3B-II in the liver and heart of mice without leupeptin (Figures 5(a), (c), (e), and (f) and S2). Interestingly, the expression of LC3B-II in the liver after treatment with leupeptin signiﬁcantly increased with EHR treatment (Figures 5(b), (e), and S2). However, such an expression did not change in the heart (Figures 5(d), (f), and S2). Therefore, EHR signiﬁcantly induces autophagy in the liver of mice. 3.5. EHR Induces Autophagy in Mice. Autophagy is also important in determining the lifespan of some organisms [38]. Compared with yeast, which is a unicellular eukaryote, mice have 30,000 protein-coding genes similar to humans, and approximately 80% of the human genome has a single identiﬁable ortholog in mouse genes [39]. Hence, using mice can provide highly reliable evidence. Microtubule-associated protein 1 LC3B in mammal cells is a homologous protein of Atg8 to form autophagosome [40]. This protein started to appear in the cytoplasm as LC3B-I and transformed into LC3B-II during autophagy. Leupeptin, one kind of autoph- agy inhibitor, can impede autophagosome degradation. This 3.6. Eﬀect of EHR on the SIR2 Gene Expression and the Chronological Lifespan of Δsir2 and Δuth1 of Yeast with YOM36 Background. Oxidative Medicine and Cellular Longevity The increase in SIR2 and the reduction of the UTH1 gene expression can extend replicative lifespan because Sir2 is considered to be a prolongevity factor, while 12 Oxidative Medicine and Cellular Longevity 12 0.0 0.5 1.0 1.5 2.0 Control 3 10 EHR (𝜇M) Relative level of SIR2 mRNA ⁎ ⁎⁎⁎ (a) 0 3 6 9 12 15 0 20 40 60 80 100 Control (YOM36) EHR (3 𝜇M, YOM36)⁎⁎⁎ EHR (3 𝜇M, 𝛥sir2) Control (𝛥sir2)### Days Survival rate (%) (b) Survival rate (%) 0 2 4 6 8 10 12 14 0 20 40 60 80 100 Control (YOM36) EHR (3 𝜇M, YOM36)⁎⁎⁎ EHR (3 𝜇M, 𝛥uth1) Control (𝛥uth1) Days (c) Figure 6: The eﬀects of EHR on the SIR2 gene expression of yeast (a), the chronological lifespan of Δsir2 (b) and Δuth1 (c) of yeast with the YOM36 background. This assay was taken thrice, and the sample numbers of each group were ﬁve or three. The log-rank (Mantel-Cox) test was used for the yeast chorological lifespan assay. ∗and ∗∗∗indicate a signiﬁcant diﬀerence between the control group and EHR-treated group at p < 0:05 and p < 0:001, whereas ### indicates a signiﬁcant diﬀerence between the Δsir2 control group and YOM36 control group at p < 0:001. 0.0 0.5 1.0 1.5 2.0 Control 3 10 EHR (𝜇M) Relative level of SIR2 mRNA ⁎ ⁎⁎⁎ (a) 0 3 6 9 12 15 0 20 40 60 80 100 Control (YOM36) EHR (3 𝜇M, YOM36)⁎⁎⁎ EHR (3 𝜇M, 𝛥sir2) Control (𝛥sir2)### Days Survival rate (%) (b) Survival rate (%) 0 2 4 6 8 10 12 14 0 20 40 60 80 100 Days 0.0 0.5 1.0 1.5 2.0 Control 3 10 EHR (𝜇M) Relative level of SIR2 mRNA ⁎ ⁎⁎⁎ 0 3 6 9 12 15 0 20 40 60 80 100 Days Survival rate (%) (a) (b) Survival rate (%) 0 2 4 6 8 10 12 14 0 20 40 60 80 100 Days Control (YOM36) EHR (3 𝜇M, YOM36)⁎⁎⁎ EHR (3 𝜇M, 𝛥uth1) Control (𝛥uth1) (c) (c) Figure 6: The eﬀects of EHR on the SIR2 gene expression of yeast (a), the chronological lifespan of Δsir2 (b) and Δuth1 (c) of yeast with the YOM36 background. This assay was taken thrice, and the sample numbers of each group were ﬁve or three. The log-rank (Mantel-Cox) test was used for the yeast chorological lifespan assay. Oxidative Medicine and Cellular Longevity This ﬁnding highlights the importance of the SIR2 gene in extending the chronological lifespan of yeast. The EHR treatment at 3μM signiﬁcantly extended the chronological lifespan of YOM36 from 11 days to 13 days. However, EHR did not increase the chronological lifespan of Δsir2 and Δuth1 with the YOM36 background (Figures 6(b) and 6(c), respectively). However, the extension Oxidative Medicine and Cellular Longevity ∗and ∗∗∗indicate a signiﬁcant diﬀerence between the control group and EHR-treated group at p < 0:05 and p < 0:001, whereas ### indicates a signiﬁcant diﬀerence between the Δsir2 control group and YOM36 control group at p < 0:001. the deﬁciency of the UTH1 gene can prolong the replicative lifespan in yeast [10, 33]. However, the similarities in the roles of these genes in the chronological lifespan in yeast remain uncertain. The SIR2 gene expression analysis and chronological lifespan assay of Δsir2 and Δuth1 with a YOM36 background were performed to understand the eﬀect of EHR on the SIR2 gene and the relationship between SIR2 and UTH1 genes. A signiﬁcant increase in the SIR2 gene expression in the EHR treatment group was observed com- pared with the control group at doses of 3 and 10 μM (Figure 6(a)). The chronological lifespan of Δsir2 with a YOM36 background was signiﬁcantly shorter than that of the YOM36 yeast (Figure 6(b)). This ﬁnding highlights the importance of the SIR2 gene in extending the chronological lifespan of yeast. The EHR treatment at 3μM signiﬁcantly extended the chronological lifespan of YOM36 from 11 days to 13 days. However, EHR did not increase the chronological lifespan of Δsir2 and Δuth1 with the YOM36 background (Figures 6(b) and 6(c), respectively). However, the extension of the chronological lifespan of Δuth1 with the YOM36 back- ground was not observed in Figure 6(c). This ﬁnding suggests the involvement of SIR2 and UTH1 genes in the antiaging eﬀect of EHR for the chronological lifespan of yeast. UTH1 genes did not take important roles in the chronological life- span of YOM36 yeast but were required to participate in the antiaging eﬀect of EHR. the deﬁciency of the UTH1 gene can prolong the replicative lifespan in yeast [10, 33]. However, the similarities in the roles of these genes in the chronological lifespan in yeast remain uncertain. The SIR2 gene expression analysis and chronological lifespan assay of Δsir2 and Δuth1 with a YOM36 background were performed to understand the eﬀect of EHR on the SIR2 gene and the relationship between SIR2 and UTH1 genes. A signiﬁcant increase in the SIR2 gene expression in the EHR treatment group was observed com- pared with the control group at doses of 3 and 10 μM (Figure 6(a)). The chronological lifespan of Δsir2 with a YOM36 background was signiﬁcantly shorter than that of the YOM36 yeast (Figure 6(b)). 5. Conclusion Overall, EHR produced antiaging eﬀects on yeast and mam- mal cells by inducing autophagy and antioxidative stress. Several well-known antiaging molecules, such as RES, nico- tinamide mononucleotide, and rapamycin, have been recently reported to ease the symptoms of type 2 diabetes, Alzheimer’s disease, cancer, cardiovascular diseases, and obe- sity [41]. Therefore, antiaging substances may have potential applications in the treatment of age-related illnesses. The new function and mechanism of EHR in various age-related pathological models and the safety evaluation of EHR must be examined in the future to develop EHR as a drug or health food. Supplementary Materials p SIR2 is a longevity gene in which deletion or overexpres- sion can shorten or prolong the replicative lifespan of yeast [10]. A previous study found that SIR2 and UTH1 genes interacted with each other, and deletion of the latter extends the replicative lifespan of yeast [17]. The SIR2 gene expres- sion was investigated to determine the involvement of these genes in the chronological lifespan of yeast and their contri- butions to the antiaging eﬀect of EHR. Δsir2 and Δuth1 with a YOM36 background were also constructed for a chronolog- ical lifespan assay. The changes in the SIR2 gene expression (Figure 6(a)) and the chronological lifespan of Δsir2 with the YOM36 background (Figure 6(b)) suggest the involve- ment of the SIR2 gene in the chronological lifespan of yeast and the antiaging eﬀect of EHR. However, some reports showed that Δsir2 can extend the chronological lifespan [11], but other studies did not ﬁnd any diﬀerences [12]. Such discrepancies can be ascribed to the usage of diﬀerent yeast Figure S1: origin data of western blotting analysis of free GFP in yeast in Figures 4(c) and 4(e). Figure S2: origin data of western blotting analysis of LC3B-II in the liver and heart of mice in Figures 5(a), 5(b), 5(c) and 5(d). (Supplementary Materials) 4. Discussion The previous study conducted by the authors found a series of antiaging benzoquinone-type molecules from O. Bractea- tum and determined the chemical structure of these com- pounds [23]. Among these molecules, EHR not only has signiﬁcant antiaging activities but also emerges as the best molecule. Therefore, further studies must focus on the anti- aging eﬀects and mechanism of the EHR action. The lifespan assay results (Figures 1(b) and 1(c)) of yeast and the yeast- like chronological lifespan in PC12 cells (Figures 1(d) and Oxidative Medicine and Cellular Longevity 13 1(e)), respectively, suggest that EHR has an antiaging eﬀect for yeast and mammal cells. strains and media in these studies. Meanwhile, the deletion of the UTH1 gene can extend the replicative lifespan of K6001 but does not aﬀect the chronological lifespan of YOM36 (Figures 3(c) and 6(c)). This ﬁnding suggests that the UTH1 gene can be ignored in the chronological lifespan of YOM36. However, the UTH1 gene is required to partici- pate in the antiaging eﬀect of EHR. Oxidative stress plays a crucial role in the aging process [5]. Previous studies revealed the isolation of many natural products with antioxidative stress properties from food and herb medicines [16, 17, 35]. Therefore, the current study focused on the antioxidative stress of EHR. The signiﬁcant increase in survival rates under oxidative conditions (Figures 2(a) and 2(b)), the GPx, CAT, and SOD enzyme activities (Figures 2(e), 2(f), 2(g), and 2(h), respectively) and the reduction in MDA and ROS levels (Figures 2(c) and 2(d), respectively) indicate the involvement of antioxida- tive stress in the antiaging eﬀect of EHR. The lifespan assays of Δsod1, Δsod2, Δgpx, Δcat, Δuth1, and Δskn7 with a K6001 background, which all had related eﬀects on antioxidative stress, were performed to obtain additional evidence to sup- port the present hypothesis. The changes in the lifespan of these mutants in Figure 3 suggest that SOD1, SOD2, GPx, CAT, UTH1, and SKN7 contribute to the antiaging eﬀect of EHR. Interestingly, all the natural products from previous studies have the function of antioxidative stress despite dif- ferences in structural features [16–20, 35]. Therefore, the antioxidant properties of these molecules will be examined in the future. Conflicts of Interest The authors declare no ﬁnancial or commercial conﬂict of interest. Data Availability Autophagy takes an important role in removing damaged molecules, and enhanced autophagy can extend the lifespan of yeasts, worms, ﬂies, and mice [38]. A previous study indicated that autophagy was involved in the antiaging activity of cucur- bitacin B [16]. Therefore, the eﬀects of EHR on autophagy in yeasts and mammals were also investigated. The increase in free GFP in YOM38-GFP-ATG8 yeast (Figures 4(a)–(f)), the absence of changes in the replicative lifespan of Δatg2 and Δagt32 of yeast (Figures 4(g)–(h)), and the increase in LC3B- II of the liver and heart in mice (Figure 5) after treatment of EHR conﬁrmed that autophagy is crucial in the antiaging activity of EHR. However, opposite ﬁndings were obtained for the autophagy of the liver and heart of mice. Speciﬁcally, the EHR-induced LC3B-II protein level in the heart was higher than that in the liver. The signiﬁcant changes in the EHR-induced LC3B-II in the heart were diminished after the leupeptin administration, but the opposite was observed for the expression of LC3B-II in the liver. All the ﬁgures and table used to support the ﬁndings of this study are included within the article. Acknowledgments The authors are grateful to Michael Breitenbach (Salzburg University, Austria) for the gifts of the K6001 yeast strain. This work was ﬁnancially supported by the National Key R&D Program of China (Grant No. 2019YFE0100700) and the National Natural Science Foundation of China (Grant Nos. 21661140001, 21877098, and 21572204). Oxidative Medicine and Cellular Longevity Mostoslavsky, “A sirtuin's role in pre- venting senescence by protecting ribosomal DNA,” Journal of Biological Chemistry, vol. 293, no. 28, pp. 11251-11252, 2018. [25] Y. Maruyama, T. Ito, H. Kodama, and A. Matsuura, “Availabil- ity of amino acids extends chronological lifespan by suppress- ing hyper-acidiﬁcation of the environment in Saccharomyces cerevisiae,” PLoS One, vol. 11, no. 3, p. e0151894, 2016. [10] M. Kaeberlein, M. McVey, and L. Guarente, “The SIR2/3/4 complex and SIR2 alone promote longevity in Saccharomyces cerevisiae by two diﬀerent mechanisms,” Genes & Develop- ment, vol. 13, no. 19, pp. 2570–2580, 1999. [26] D. Carmona-Gutierrez, A. Zimmermann, K. Kainz et al., “The ﬂavonoid 4,4′-dimethoxychalcone promotes autophagy- dependent longevity across species,” Nature communications, vol. 10, no. 1, p. 651, 2019. [11] P. Fabrizio, C. Gattazzo, L. Battistella et al., “Sir2 blocks extreme life-span extension,” Cell, vol. 123, no. 4, pp. 655– 667, 2005. [27] I. Bjedov and C. Rallis, “The target of rapamycin signalling pathway in ageing and lifespan regulation,” Genes, vol. 11, no. 9, p. 1043, 2020. [12] D. L. Smith, J. M. McClure, M. Matecic, and J. S. Smith, “Cal- orie restriction extends the chronological lifespan ofSaccharo- myces cerevisiaeindependently of the Sirtuins,” Aging Cell, vol. 6, no. 5, pp. 649–662, 2007. [28] B. G. Trist, D. J. Hare, and K. L. Double, “Oxidative stress in the aging substantia nigra and the etiology of Parkinson's dis- ease,” Aging Cell, vol. 18, no. 6, 2019. [13] H. Gershon and D. Gershon, “The budding yeast, _Saccharo- myces cerevisiae_ , as a model for aging research: a critical review,” Mechanisms of Ageing and Development, vol. 120, no. 1-3, pp. 1–22, 2000. [29] R. Mittler, “ROS are good,” Trends in Plant Science, vol. 22, no. 1, pp. 11–19, 2017. [30] O. M. Ighodaro and O. A. Akinloye, “First line defence antioxidants-superoxide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GPX): their fundamental role in the entire antioxidant defence grid,” Alexandrea Journal of Medicine, vol. 54, no. 4, pp. 287–293, 2018. [14] H. Karathia, E. Vilaprinyo, A. Sorribas, and R. Alves, “Saccha- romyces cerevisiae as a model organism: a comparative study,” PLoS One, vol. 6, no. 2, 2011. [15] S. Jarolim, J. Millen, G. Heeren, P. Laun, D. S. Goldfarb, and M. Breitenbach, “A novel assay for replicative lifespan in,” FEMS Yeast Research, vol. 5, no. 2, pp. 169–177, 2004. [31] N. D. Bonawitz, M. S. Rodeheﬀer, and G. S. Oxidative Medicine and Cellular Longevity 14 stress by deacetylating SigH ( σH) and FOXO3a in Biﬁdobac- terium longum and HEK293T cell respectively,” Free Radical Biology and Medicine, vol. 108, pp. 929–939, 2017. lifespan of yeast via the pathway involved in SOD and UTH1,” International Journal of Molecular Sciences, vol. 14, no. 3, pp. 4461–4475, 2013. [20] [4] H. R. Tai, Z. Wang, H. Gong et al., “Autophagy impairment with lysosomal and mitochondrial dysfunction is an important characteristic of oxidative stress-induced senescence,” Autoph- agy, vol. 13, no. 1, pp. 99–113, 2017. [20] L. Xiang, K. Y. Sun, J. Lu et al., “Anti-aging eﬀects of phlorid- zin, an apple polyphenol, on yeast via the SOD and Sir2 genes,” Bioscience, Biotechnology, and Biochemistry, vol. 75, no. 5, pp. 854–858, 2011. [5] J. Luo, K. Mills, S. Cessie, R. Noordam, and D. Heemst, “Age- ing, age-related diseases and oxidative stress: what to do next?,” Ageing Research Reviews, vol. 57, 2020. [21] J. S. Ali, I. Haq, A. Ali, M. Ahmed, and M. Zia, “Onosma brac- teatumWall andCommiphora stocksianaEngl extracts gener- ate oxidative stress inBrassica napus: an allelopathic perspective,” Cogent Biology, vol. 3, no. 1, 2017. [6] J. Q. Fang, J. P. Yang, X. Wu et al., “Metformin alleviates human cellular aging by upregulating the endoplasmic reticu- lum glutathione peroxidase 7,” Aging Cell, vol. 17, no. 4, p. e12765, 2018. [22] S. Badruddeen, H. Fareed, H. Siddiqui, S. E. Haque, M. Khalid, and J. Akhtar, “Psychoimmunomodulatory eﬀects of Onosma bracteatum wall. (Gaozaban) on stress model in Sprague daw- ley rats,” Journal of Clinical and Diagnostic Research, vol. 6, no. 7, pp. 1356–1360, 2012. [7] L. R. Lapierre, C. Kumsta, M. Sandri, A. Ballabio, and M. Hansen, “Transcriptional and epigenetic regulation of autophagy in aging,” Autophagy, vol. 11, no. 6, pp. 867–880, 2015. [23] U. Farooq, Y. J. Pan, D. Disasa, and J. H. Qi, “Novel anti-aging benzoquinone derivatives from Onosma bracteatum Wall,” Molecules, vol. 24, no. 7, p. 1428, 2019. [8] R. Malaviya, J. D. Laskin, and D. L. Laskin, “Oxidative stress- induced autophagy: role in pulmonary toxicity,” Toxicology and Applied Pharmacology, vol. 275, no. 2, pp. 145–151, 2014. [24] Y. C. Chien, C. H. Lin, M. Y. Chiang et al., “Secondary metab- olites from the root of Ehretia longiﬂora and their biological activities,” Phytochemistry, vol. 80, pp. 50–57, 2012. [9] J. P. Etchegaray and R. References [1] V. Atella, A. P. Mortari, J. Kopinska et al., “Trends in age- related disease burden and healthcare utilization,” Aging Cell, vol. 18, no. 1, p. e12861, 2019. [2] C. López-Otín, M. A. Blasco, L. Partridge, M. Serrano, and G. Kroemer, “The hallmarks of aging,” Cell, vol. 153, no. 6, pp. 1194–1217, 2013. [3] Q. Guo, S. Y. Li, Y. J. Xie et al., “The NAD+-dependent deace- tylase, Biﬁdobacterium longum Sir2 in response to oxidative Oxidative Medicine and Cellular Longevity Oxidative Medicine and Cellular Longevity Shadel, “Defective mitochondrial gene expression results in reactive oxygen species-mediated inhibition of respiration and reduction of yeast life span,” Molecular and Cellular Biology, vol. 26, no. 13, pp. 4818–4829, 2006. [16] Y. F. Lin, Y. Kotakeyama, J. Li et al., “Cucurbitacin B exerts antiaging eﬀects in yeast by regulating autophagy and oxida- tive stress,” Oxidative Medicine and Cellular Longevity, vol. 2019, Article ID 4517091, 15 pages, 2019. [32] V. D. Longo, E. B. Gralla, and J. S. Valentine, “Superoxide dis- mutase activity is essential for stationary phase survival inSac- charomyces cerevisiae,” Journal of Biological Chemistry, vol. 271, no. 21, pp. 12275–12280, 1996. [17] Y. F. Lin, Y. J. Sun, Y. F. Weng, A. Matsuura, L. Xiang, and J. H. Qi, “Parishin from Gastrodia elata extends the lifespan of yeast via regulation of Sir2/Uth1/TOR signaling pathway,” Oxida- tive Medicine and Cellular Longevity, vol. 2016, Article ID 4074690, 11 pages, 2016. [33] N. M. Camougrand, M. Mouassite, G. M. Velours, and M. G. Guerin, “The "SUN" Family: _UTH1,_ an Ageing Gene, Is Also Involved in the Regulation of Mitochondria Biogenesis in _Saccharomyces cerevisiae_,” Archives of Biochemistry and Biophysics, vol. 375, no. 1, pp. 154–160, 2000. [18] Y. J. Sun, Y. F. Lin, X. L. Cao, L. Xiang, and J. H. Qi, “Sterols from Mytilidae show anti-aging and neuroprotective eﬀects via anti-oxidative activity,” International Journal of Molecular Sciences, vol. 15, no. 12, pp. 21660–21673, 2014. [34] J. Lee, C. Godon, G. Lagniel et al., “Yap1 and Skn7 control two specialized oxidative stress response regulons in yeast,” Journal of Biological Chemistry, vol. 274, no. 23, pp. 16040–16046, 1999. [19] K. Y. Sun, S. N. Cao, L. Pei, A. Matsuura, L. Xiang, and J. H. Qi, “A steroidal saponin from Ophiopogon japonicus extends the 15 Oxidative Medicine and Cellular Longevity Oxidative Medicine and Cellular Longevity 15 [35] U. Farooq, Y. Pan, Y. Lin et al., “Structure characterization and action mechanism of an antiaging new compound from Gas- trodia elata Blume,” Oxidative Medicine and Cellular Longev- ity, vol. 2019, Article ID 5459862, 10 pages, 2019. [36] F. Madeo, N. Tavernarakis, and G. Kroemer, “Can autophagy promote longevity?,” Nature Cell Biology, vol. 12, no. 9, pp. 842–846, 2010. [37] R. Kellner, J. C. De-la-Concepcion, A. Maqbool, S. Kamoun, and Y. F. Dagdas, “ATG8 expansion: a driver of selective autophagy diversiﬁcation?,” Trends in Plant Science, vol. 22, no. 3, pp. 204–214, 2017. [38] B. Levine and G. Kroemer, “Biological functions of autophagy genes: a disease perspective,” Cell, vol. 176, no. 1-2, pp. 11–42, 2019. [39] R. H. Waterston, K. Lindblad-Toh, E. Birney et al., “Initial sequencing and comparative analysis of the mouse genome,” Nature, vol. 420, no. 6915, pp. 520–562, 2002. [40] A. Ashkenazi, C. F. Bento, T. Ricketts et al., “Polyglutamine tracts regulate beclin 1-dependent autophagy,” Nature, vol. 545, no. 7652, pp. 108–111, 2017. [41] F. Madeo, D. Carmona-Gutierrez, S. J. Hofer, and G. Kroemer, “Caloric restriction mimetics against age-associated disease: targets, mechanisms, and therapeutic potential,” Cell Metabo- lism, vol. 29, no. 3, pp. 592–610, 2019.
https://openalex.org/W2128722301	https://academicjournals.org/journal/JPHE/article-full-text-pdf/566D83242932.pdf	English	null	Assessment of long-lasting insecticidal net coverage, use and physical integrity one year after universal distribution campaign in Plateau department in South-East Benin	Journal of Public Health and Epidemiology	2,014	cc-by	4,721	Accepted 20 January, 2014 Long-lasting insecticidal nets (LLINs) have proved to be an important tool for the malaria control and other vector-borne diseases. Benin, by its National Malaria Control Program, conducted in July, 2011 a universal distribution campaign where approximately 5 million of nets were distributed. But after this mass-distribution, questions arise: Do people effectively use or not use the mosquito nets freely distributed? To clarify these questions, this study was conducted on LLINs coverage, use and physical condition in Benin one year after their distribution. The households were randomly selected from 32 clusters. Data on bed net ownership and usage, physical condition of the nets, other characters and issues related to sourcing were asked of all targets to assess the origins of LLIN found at the household level. Of the total surveyed, 88.96% had at least one LLIN. 87.19% of these nets come from the last campaign, 9.1% were from pregnant women voucher clinic delivery systems and 3.20% were purchased full price. 84% of the nets were reportedly used the night preceding the survey. Around 89% of the total of LLINs observed was in good condition. The universal distribution campaign in Benin has significantly increased LLIN coverage and ownership policy in the community. Key words: Long-lasting insecticidal nets (LLINs), coverage, use, physical integrity, Benin Vol. 6(2), pp. 76-84, February 2014 DOI: 10.5897/JPHE2013.0592 ISSN 2006-9723 ©2014 Academic Journals http://www.academicjournals.org/JPHE Journal of Public Health and Epidemiology Journal of Public Health and Epidemiology Full Length Research Paper Assessment of long-lasting insecticidal net coverage, use and physical integrity one year after universal distribution campaign in Plateau department in South- East Benin Full Length Research Paper Full Length Research Paper Filémon T. Tokponnon1, 2,,5,6, Yolande Sissinto3, Aurore Hounto Ogouyémi3, Arthur Sovi5,6, Virgile Gnanguenon5,6 , Adicath Adéola Adéothy3, Abel Wapo3, Bruno Aholoukpe1,2, Patrick Makoutodé3, Alioun Adéchoubou1, Ossè Razaki5,6, Raphael N’guessan7, Dina Gbénou4, Mariam Oke1,2, Dorothée Kinde Gazard2, Achille Massougbodji3 and Martin C. Akogbéto5 1National Malaria Control Program, Cotonou, Benin. 1National Malaria Control Program, Cotonou, Benin. 2Ministry of Health, Benin. 3Faculté des Sciences de la Santé de l’Université d’Abomey Calavi, Benin. 4World Health Organisation, Benin. 5Faculte des Sciences et Techniques de l’Université d’Abomey-Calavi, Benin. 6Centre de Recherche Entomologique de Cotonou (CREC), Cotonou, Benin. 7London School of Hygiene and Tropical Medicine, Kepel Street, London, UK. y , aculté des Sciences de la Santé de l’Université d’Abomey Calavi, Benin. 4 g q ( ), , 7London School of Hygiene and Tropical Medicine, Kepel Street, London, UK. Study area Plateau is a county of Benin in West Africa. The selection of this county was based on its geographic accessibility, the high use of mosquito nets by children under 5. Entomological surveys conducted in the plateau have shown that there are both high and low pyrethroids resistance areas (Yadouleton et al., 2010). According to the general report of the distribution campaign, 85.5% of the households received a LLIN with an average of 2.7 LLINs/Household. Ketou, Pobe, Adja-Ouere, Ifangni and Sakete are the top 5 cities that compose it. But our study was focused on only four districts (Ketou, Pobe, Ifangni and Sakete). Ifangni district is located at 2° 43' 14"E and 6° 38' 56"N; its area is 242 km² representing 7.28% of Plateau territory. Sakete district is located at 2° 39' 7"E and 6° 46' 3"N; covering an area of 432 km², it represents 13.29% of plateau territory. Ketou is located is at 2° 36' 4"E and 7° 27' 21"N; it has an area of 1775 km², representing 54.38% of Plateau county's area. Pobè district is located at 2° 41' 51"E and 7° 5' 12"N; it has an area of 400 km², representing 11% of the county's area. 32 rural villages were selected through all four districts (Figure 1). ) Benin, by its National Malaria Control Program, supported since 2000 the initiative RBM and conducted in July, 2011 a universal distribution campaign where approximately 5 million of nets were distributed. This distribution is intended to significantly increase the national coverage of insecticide-treated nets. Before the distribution, the proportion of households owning at least one net at national level was estimated at 52% (PNLP- AFRICARE BENIN-CRS BENIN, 2010). After the distribution of July, 2011, this proportion was around 86.4% (PNLP, 2011). In the net level analysis, factors independently associated in both surveys with reduced likelihood that a net would be used were: increasing net age, increasing damage of nets, increasing household net density (nets/person). Some studies have shown that several factors are associated with LLIN ownership and effective use (Graves et al., 2011). Factors associated with ownership were assumed to be: cultural beliefs and practices mechanisms of LLIN distribution and distance to LLIN suppliers, rumours about LLINs and social support and pressure (Wiseman et al., 2007; Baume and Marin, 2007). Study area But factors associated with net used were reported to be: perceived benefits and disadvantages of nets, trust in health workers providing health education and LLINs, knowledge of appropriate net use/care practices, and net-hanging skills, household size and composition, the number of children under five years of age and use of other vector control measures (Toé et al., 2009; Edelu et al., 2010). Other studies conducted in Benin showed that several mosquito nets after 12 month of use had already holes (Gnanguenon, personnal communication). From these observations, questions arise: Do people effectively use the mosquito nets freely distributed? Or do people not use LLINs to their insufficient number or their poor physical condition? To clarify these questions, we conducted a study on insecticide-treated nets coverage, use and physical condition in plateau department in southern Benin one year after the distribution. This study aimed to provide useful data on the effectiveness of the distribution Tokponnon et al. vector-borne diseases (World Health Organisation (WHO), 2005). Several studies in malaria endemic coun- tries have shown the usefulness of LLINs in reducing man-vector contact from malaria (Greenwood et al., 2005; Lengeler, 2004; Eisele et al., 2006; Thwing et al., 2008). It is a technology based on the slow release of pyrethroid insecticides, rendering it wash-resistant and extending insecticide residual effectiveness to at least three years without the need of re-treatment. This is why the recent objective of roll back malaria (RBM) focuses duplication of efforts including maintaining universal coverage of nets to achieve a 75% reduction compared to 2000, the number of reported malaria cases in endemic countries in the African region, 2015 (OMS, 2011; AFR/RC50/12 – WHO, 2012) . campaign. campaign. INTRODUCTION an important tool for the control of malaria and other an important tool for the control of malaria and other Long-lasting insecticidal nets (LLINs) have proved to be Corresponding author. E-mail: filemont@yahoo.fr. Tel: (229) 97272631. Fax: (229) 21333930. Tokponnon et al. 77 Study design Larvae prospection’s for insecticides susceptibility tests on Anopheles gambiae, main malaria vector in Benin have been conducted in many villages of the targeted districts. This baseline study on the resistance of malaria vectors to deltamethrin in the department of Plateau helped to make the choice of sentinel villages where various activities were held in our work. A total of 32 clusters were selected including 17 clusters at Ifangni, 6 at Sakété, 2 at Pobè and 8 at Kétou. Each cluster (village) was composed of several hamlets and comprised a minimum of 100 Children under five years old. Household cross-sectional surveys were undertaken in each cluster in May and August, 2012, conducting during high malaria transmission period. The survey covered the targeted groups in different villages. The choice of the targeted population has followed the standards set in the collection of basic data on morbidity and mortality due to malaria in the monitoring/evaluation RBM/RBM in Benin in 2004 (Kinde-Gazard et al., 2004). The targeted persons by cluster were: 30 mothers of children or care for children less than 5 years to evaluate fever or malaria during the last two weeks, 25 Pregnant women in the third trimester of pregnancy and women who gave birth in the last 6 months to assess malaria prevention, 30 householders for the availability and use of LLINs. The households were randomly selected from each cluster. Data on bed net ownership and usage, physical condition of the nets, demographics of household members, other characters and issues related to sourcing were asked of all targets to assess the origins of LLIN found at the household level. Data were gathered using an adapted version of the standard Malaria Indicator Survey (RBM, 2013). Specific questions relating to the Universal Campaign Coverage process were asked to the householder and to the mothers of children or care for children less than 5 years. The physical condition of the nets was estimated using one of the two indicators recommended by WHO: the proportion of LLINs with any hole(s) (WHO, 2011). The main hole category in the LLINs was recorded as follows: J. Public Health Epidemiol. 78 Figure 1. Map showing the study villages in Plateau department (Benin). < thumb (0.5 to 2 cm); > thumb < fist (2 to 10 cm); fi h d (10 2 ) Statistical analysis Interviews were conducted using questions At the Figure 1. Study design Map showing the study villages in Plateau department (Benin). T1: holes size < thumb (0.5 to 2 cm); T2: holes size > thumb < fist (2 to 10 cm); T3: holes size > fist < head (10 to 25 cm); T4: holes size > head (> 25 cm). RESULTS The sex of the heads of households (male and female) as well as their education level were not associated with ITN ownership (p > 0.05). Also, the composition of the household size in children under five, pregnant women and persons over five years old was not associated with ITN ownership (p > 0.05) (Table 3). Statistical analysis Interviews were conducted using questions. At the end of the survey, data were recorded with Epi-Info and data were transferred 79 Tokponnon et al. Table 1. Background characteristics of the households surveyed. Table 1. Background characteristics of the households surveyed. Table 1. Background characteristics of the households surveyed. Characteristic Frequency (%) Sex of the heads of households (n = 960) Male 386 (40.20) Female 574 (59.80) Educational level of the heads of households Illiterate 655 (68.23) Elementary 188 (19.58) Post elementary 117 (12.19) Household size (n = 4688) Children under 5 years 1339 (28.56) Pregnants women 800 (17.06) Person over 5 years 2549 (54.37) The investigators had cleaned and and 2549 (54.37%) were people over five years old Characteristic Frequency (%) Sex of the heads of households (n = 960) Male 386 (40.20) Female 574 (59.80) Educational level of the heads of households Illiterate 655 (68.23) Elementary 188 (19.58) Post elementary 117 (12.19) Household size (n = 4688) Children under 5 years 1339 (28.56) Pregnants women 800 (17.06) Person over 5 years 2549 (54.37) and 2549 (54.37%) were people over five years old (without pregnant women). into SPSS 16.0 software. The investigators had cleaned and analysed the data using the same software program. Household ownership of LLIN was calculated as a proportion of households having at least one LLIN among the total surveyed households. LLIN use was estimated as the proportion of households using at least one LLIN in the LLIN owning households. Once the LLIN ownership and use were determined; the data were filtered into a separate file of LLIN owned households. Then, households using LLINs were compared to those who did not use any to identify the factors associated with LLINs non-use. Barriers of LLIN use were determined by using enter method multivariate logistic regression model. Data on the physical integrity of LLINs and those relating to the origins of LLINs were extracted. The various holes observed were divided into four and frequencies were estimated to assess the general condition of operation of LLINs (WHO, 2011; Kilian, 2012). Study households characteristics 960 households and 4688 people were included in the survey and the response rate was 100% (Table 1). 59.80% of the heads of households interviewed were females versus 40.20% of males. 68.23% of the heads of households were illiterate, 19.58% have elementary school degree and 12.19% have an educational level over elementary school degree. Of the 4688 people included (Table 1), 1339 (28.56%) were under five years old; 800 (17.06%) were reported to be pregnant women Ethics approval This study was planned and approved by the Ministry of Health. The protocol was also reviewed and approved by National Ethics Committee for Health Research at the Ministry of Health. A briefing note indicating the objectives of the study, the advantages and disadvantages was given to the respondents in order to obtain consent. Confidentiality was respected and questionnaires were anonymous. LLINs ownership Of the total surveyed households, 88.96% had at least one LLIN while 11.04% did not have any type of mosquito nets. The majority of households that owned LLIN had either one or two LLINs irrespective of their household size. 28.22% (25.47 to 31.16) had one LLIN, 31.85% (29.00 to 34.89) had two LLINs, 16.76% (14.54 to 19.27) had three LLINs and 12.18% (10.27 to 14.41) had four LLINs. The number of ITNs did not vary significantly between cluster (Table 2) and the average LLIN ownership among LLIN owned household was 1.82 (Table 2). 87.19% (84.76 to 89.27) of these nets come from the National distribution campaign of 2011, 9.1% (7.79 to 11.78) were from pregnant women voucher clinic delivery systems and 3.20% (2.21 to 4.61) were purchased full price (Figure 2). ITNs utilization Of the total of 1746 reported LLINs, 42.96% (n = 750) LLINs were observed by the surveyors. 84% of the nets were reportedly used the night preceding the survey. The 80 J. Public Health Epidemiol. Table 2. LLINs ownership by cluster. Table 2. LLINs ownership by cluster. Parameter Cluster LLINs No. of households Mean CI 95% Ketou Okpometa 52 30 1.73 [1.40-2.05] Omou 69 30 2.3 [2.03-2.56] Adjozounme 82 30 2.73 [2.54-2.92] Kpankoun 59 30 1.97 [1.69-2.25] Oke Ola 38 30 1.27 [1.02-1.52] Mowodani 74 30 2.47 [2.24-2.70] Idena2 45 30 1.5 [1.11-1.88] Idena3 49 30 1.63 [1.24-2.01] Pobe Okoffi 2 73 30 2.43 [2.20-2.75] Agbarou 44 30 1.47 [1.11-1.82] Sakete Igboabikou 39 30 1.3 [1.04-1.55] Igbola 54 30 1.8 [1.44-2.15] Adjegounle/Alabansa 48 30 1.6 [1.22-1.97] Iwai 61 30 2.03 [1.68-2.37] Ikemon 52 30 1.73 [1.51-1.94] Djohounkolle 53 30 1.77 [1.51-2.02] Ifangni Akadja 69 30 2.3 [2.08-2.51] Araromi 40 30 1.33 [0.92-1.73] Banigbe 48 30 1.6 [1.28-1.91] Daagbe 49 30 1.63 [1.23-2.02] Djegou Djedji 65 30 2.17 [1.93-2.40] Gblo Gblo 23 30 0.77 [0.43-1.10] Ita Kpako 59 30 1.97 [1.62-2.32] Itassumba 56 30 1.87 [1.54-2.19] Ketougbekon 47 30 1.57 [1.24-1.90] Ko Dogba 56 30 1.87 [1.54-2.19] Ko-Aîdjedo 70 30 2.33 [2.12-2.54] Kokoumolou 51 30 1.7 [1.42-1.97] Lokossa 42 30 1.4 [1.00-1.79] Tchaada 64 30 2.13 [1.85-2.40] Zian 70 30 2.33 [1.98-2.67] Zougoudo 45 30 1.5 [1.09-1.90] Total 1746 960 1.82 [1.76-1.87] CI = confidence interval. found with hole(s) (Table 5). No significant difference was observed between categories of holes (T1 to T4) and while comparing the physical condition of the nets by the level of instruction of their owners (p < 0.05). proportion of nets in use varies from 47 to 100%, but not significantly different from one cluster to another (Table 4). It was only at Araromi that the level of net use was significantly low compared to other clusters. Among the target groups, 76.4% (73.46 to 79.34) of pregnant women really used LLINs, 82.88% (80.87 to 84.91) children under 5 years used nets while 69.78% (68.22 to 71.34) people over 5 years were reported as using LLINs (Figure 3). On the 807 LLINs observed, around 16% was proportion of nets in use varies from 47 to 100%, but not significantly different from one cluster to another (Table 4). It was only at Araromi that the level of net use was significantly low compared to other clusters. ITNs utilization Among the target groups, 76.4% (73.46 to 79.34) of pregnant women really used LLINs, 82.88% (80.87 to 84.91) children under 5 years used nets while 69.78% (68.22 to 71.34) people over 5 years were reported as using LLINs (Figure 3). On the 807 LLINs observed, around 16% was DISCUSSION ITNs coverage was increased by the universal distribution Tokponnon et al. Figure 2. Source of ITN ownership. Table 3. Multivariate regression on ITN ownership. Factors % ITN ownership (n) CI = 95% p-value Heads of households 88.96 (854) Sex of the heads of households Male 43.85 (421) [40.72-46.99] 0.680 Female 45.10 (433) [41.96-48.25] Educational level of the heads of households Illiterate 60.31 (579) [57.22-63.41] 0.332 Elementary 17.5 (168) [15.10-19.90] Post elementary 11.15 (107) [9.16-13.14] n = number CI = Confidence Interval. n campaign with a proportion of 88.96% of s that owned at least one ITN. This proportion when compared to the preceding distribution evaluation (Tokponnon et al., 2013), and the umber of ITNs per household was around 2. campaign is to increase the distribution level coverage (PNLP, 2012) at least 80% and have a two people in the general population. In the protocol of the LLIN distribution camp July, 2011, households of two people were cover 81 Tokponnon et al. Tokponnon Figure 2. Source of ITN ownership. Figure 2. Source of ITN ownership. Table 3. Multivariate regression on ITN ownership. Factors % ITN ownership (n) CI = 95% p-value Heads of households 88.96 (854) Sex of the heads of households Male 43.85 (421) [40.72-46.99] 0.680 Female 45.10 (433) [41.96-48.25] Educational level of the heads of households Illiterate 60.31 (579) [57.22-63.41] 0.332 Elementary 17.5 (168) [15.10-19.90] Post elementary 11.15 (107) [9.16-13.14] n = number CI = Confidence Interval. Table 3. Multivariate regression on ITN ownership. campaign is to increase the distribution level of 56% coverage (PNLP, 2012) at least 80% and have a LLIN for two people in the general population. distribution campaign with a proportion of 88.96% of households that owned at least one ITN. This proportion is higher when compared to the preceding distribution campaign evaluation (Tokponnon et al., 2013), and the average number of ITNs per household was around 2. Mosquito nets (LLINs) that were from the distribution campaign were significantly higher (87%) comparatively to those that were purchased full price (3%) or received from pregnant women antenatal consultation (9%). In our study, the average number of LLINs per household visited was 1.82 considering the 960 households. The average household size was 4.88 people visited. Thus we can say that about two nets are available for 5 people at the household level. DISCUSSION Djegou - Djegi 6 24 30 80 4.5714 0.0094 Gblo-Gblo 14 16 30 53 1.3 0.6058 Ita - Kpako 0 30 30 100 69.41 0.0039 Itassumba 2 28 30 93 16 0.0007 Ketougbekon 11 19 30 63 1.97 0.1967 Ko Dogba 9 21 30 70 2.66 0.0698 Ko Aïdjedo 0 30 30 100 69.41 0.0039 Kokoumolou 9 21 30 70 2.66 0.0698 Lokossa 10 20 30 67 2.28 0.1208 Tchaada 9 21 30 70 2.66 0.0698 Zian 2 28 30 93 16 0.0007 Zougoudo 9 21 30 70 2.66 0.0698 Total 208 752 960 78.3 M+: household who’s everybody use LLIN the days before survey. M-: household who’s not everybody use LLIN the days before survey. OR: odd ratio; Significant at p < 0.05 level. M+: household who’s everybody use LLIN the days before survey. M-: household who’s not everybody use LLIN the days before survey. OR: odd ratio; Significant at p < 0.05 level. Table 5. Physical condition of LLINs. Level of instruction LLINs without hole LLINs with holes Total LLINs with hole(s) % LLINs with holes T1 T2 T3 T4 Illiteracy 542 28 52 15 11 106 16,36 Primary school 163 10 12 3 3 28 14,66 Post primary school 102 7 3 4 1 15 12,82 Total 807 45 67 22 15 149 15,58 Table 5. Physical condition of LLINs. proportion was higher compared to those with three or four due to the fact that LLIN distribution campaign has limited the number of nets distributed for two people and a maximum of 8 LLINs for large households at the same size. The high ITNs usage could be a consequence of the increased availability of ITNs at the household level due to the universal distribution campaign. This observation has already been noticed in Tanzania by West and colleagues (West et al., 2012). The level of ITN use could also be influenced by high temperature and mosquito density (Graves et al., 2011). But comparatively to other studies, ITN usage observed in target group in this study was higher than those observed in Tanzania and Nigeria (Ye et al., 2012; West et al., 2012). ITN ownership and use seems to reflect the general coverage and ITNs ownership in the whole population. The proportion of LLINs from the distribution campaign found in good condition was significantly higher than those observed in serviceable condition or torn out. DISCUSSION However, the objective of this In the protocol of the LLIN distribution campaign of July, 2011, households of two people were covered with two LLINs and this rule has been well respected in rural areas in which the distribution has been fewer problems (Tokponnon et al., 2013). Our observations were consistent with the same requirements a year after the campaign. A disadvantage of the distribution was that 11% of households did not receive LLINs (PNLP, 2012). There were no significant differences between the availability of LLINs in different villages when considering the proportion of households with at least two LLINs. This 82 J. Public Health Epidemiol. p Figure 3. Net usages in target and non target group. Figure 3. Net usages in target and non target group. Figure 3. Net usages in target and non target group. Figure 3. Net usages in target and non target group. Table 4. Multivariate regression on ITN use. Districts Clusters LLIN utilization Proportion M+ (%) OR p M- M+ Total Ketou Okpometa 7 23 30 77 3.75 0.0194 Omou 1 29 30 97 33.14 0.0012 Adjozounme 0 30 30 100 69.41 0.0039 Kpankoun 3 27 30 90 10.26 0.001 Oke Ola 7 23 30 77 3.75 0.0194 Mowodani 9 21 30 70 2.66 0.0698 Idena 2 10 20 30 67 2.28 0.1208 Idena 3 8 22 30 73 3.14 0.0379 Pobe Okoffi 2 4 26 30 87 7.42 0.002 Agbarou 13 17 30 57 1.49 0.4391 Sakete Igboabikou 3 27 30 90 10.26 0.001 Igbola 4 26 30 87 7.42 0.002 Adjegounle/Alabansa 11 19 30 63 1.97 0.1967 Iwai 4 26 30 87 7.42 0.002 Ikemon 1 29 30 97 33.14 0.0012 Djohounkolle 5 25 30 83 5.71 0.0044 Ifangni Akadja 6 24 30 80 4.5714 0.0094 Araromi 16 14 30 47 1 - Banigbe 6 24 30 80 4.5714 0.0094 Daagbe 9 21 30 70 2.66 0.0698 83 Tokponnon et al. Table 4. Contd. Table 4. Contd. DISCUSSION This observation is a good performance and the coverage attained was similar to what has been achieved in other countries (Tanzania, Nigeria and Togo) (West et al., 2012; Stevens et al., 2013). Continuous distribution of LLINs to pregnant women and children under five is an important way to increase nets coverage and replace torn nets (RBM, 2011). An additional continuous distribution system, via NGOs for example, can also be planned for other households. 78% of the net observed were reported to be used the previous night, and the proportion of nets use did not vary significantly between clusters, sug- gesting that awareness for net use was a success. But full net used by target group was not achieved. REFERENCES AFR/RC50/12 - WHO (2012). Regional Office for Africa [http://www.afro.who.int/fr/downloads/doc_details/1106- afrrc5012.html?tmpl=component]. West PA, Protopopoff N, Rowland MW, Kirby MJ, Oxborough RM, Mosha FW, Malima R, Kleinschmidt I (2012). Evaluation of a national universal coverage campaign of long-lasting insecticidal nets in a rural district in North-west Tanzania. Malar. J. 11:273. Baume CA, Marin MC (2007). Intra-household mosquito net use in Ethiopia, Ghana, Mali, Nigeria, Senegal, and Zambia: are nets being used? Who in the household uses them? Am. J. Trop. Med. Hyg. 77:963–971. WHO (2005). World Malaria Report 2005. [http://www.who.int/malaria/publications/atoz/9241593199/en/] Edelu B, Ikefuna A, Emodi J, Adimora G (2010). Awareness and use of insecticide-treated bed nets among children attending outpatient clinic at UNTH, Enugu - the need for an effective mobilization process. Afr. Health Sci. 10:117–119. WHO (2011). Guidelines for monitoring the durability of long-lasting insecticidal mosquito nets under operational conditions. [http://apps.who.int/iris/handle/10665/44610] Eisele TP, Macintyre K, Yukich J, Ghebremeskel T (2006). Interpreting household survey data intended to measure insecticide-treated bednet coverage: results from two surveys in Eritrea. Malar. J. 5:36. [ p pp ] Wiseman V, Scott A, McElroy B, Conteh L, Stevens W (2007). Determinants of bed net use in the Gambia: implications for malaria control. Am. J. Trop. Med. Hyg. 76:830–836. Graves PM, Ngondi JM, Hwang J, Getachew A, Gebre T, Mosher AW, Patterson AE, Shargie EB, Tadesse Z, Wolkon A, Reithinger R, Emerson PM, Richards FO (2011). Factors associated with mosquito net use by individuals in households owning nets in Ethiopia. Malar. J. 10:354. Yadouleton AW, Padonou G, Asidi A, Moiroux N, Bio-Banganna S, Corbel V, N’guessan R, Gbenou D, Yacoubou I, Gazard K, Akogbeto MC (2010). Insecticide resistance status in Anopheles gambiae in Southern Benin. Malar. J. 9:83. Ye Y, Patton E, Kilian A, Dovey S, Eckert E (2012). Can universal insecticide-treated net campaigns achieve equity in coverage and use? the case of northern Nigeria. Malar. J. 11:32. Greenwood BM, Bojang K, Whitty CJM, Targett GAT (2005). Malaria. Lancet 365:1487–1498. Kilian A (2012). Textile Testing Methods for Nets. Ki d G d D Gb D T h S D Sil C Kinde-Gazard D, Gbenou D, Tohon S, Da Silva C, Nahum A, Quenum A, Houndigande E, Houndekon R, Ekoue S, Massougbodji A (2004). Indicateurs de suivi et d’évaluation en 2001 de l’initiative Faire reculer le paludisme au Bénin. Bull. Soc. Pathol. Exot. 97:349–352. Lengeler C (2004). Insecticide-treated bed nets and curtains for preventing malaria. Cochrane Database Syst. Rev. ACKNOWLEDGEMENTS We are grateful to the WHO and Bill & Melinda Gates Foundation which supported financially this study. The authors would also like to thank the populations of 32 clusters, the team of health system in the Plateau department, and the investigators for their collaboration. Thwing J, Hochberg N, Vanden Eng J, Issifi S, Eliades MJ, Minkoulou E, Wolkon A, Gado H, Ibrahim O, Newman RD, Lama M (2008). Insecticide-treated net ownership and usage in Niger after a nationwide integrated campaign. Trop. Med. Int. Health 13:827–834. Toé LP, Skovmand O, Dabiré KR, Diabaté A, Diallo Y, Guiguemdé TR, Doannio JM, Akogbeto M, Baldet T, Gruénais M-E (2009). Decreased motivation in the use of insecticide-treated nets in a malaria endemic area in Burkina Faso. Malar. J. 8:175. Tokponnon F, Aholoukpe B, Denon E, Gnanguenon V, Bokossa A, N’guessan R, Oke M, Kinde-Gazard D, Akogbeto M (2013). Evaluation of the coverage and effective use rate of Long-Lasting Insecticidal Nets after Nation-wide scale up of their distribution in Benin. Parasit. Vectors., 6:265. Conclusion 76.4% of pregnants women were using ITNs while only 82.88% of children under five were using ITNs. This observation suggests that net usage could be improved in target group. The universal distribution campaign in Benin has significantly increased ITN coverage and ownership policy in the community. Non target persons and target J. Public Health Epidemiol. 84 PNLP (2012). Campagne de Distribution Gratuite Des Moustiquaires Imprégnées à Longue Durée D’action Aux Ménages Du Bénin. Cotonou: Ministère de la Santé.11p. persons (pregnant women and children under five) had similar level in ITN ownership and usage. The level of ITN ownership and usage were also similar between villages. This is an important indicator of universal coverage goal. But additional effort must be done to fully achieve universal coverage goal, and routine distribution must be used to maintain ITN coverage. p RBM (2011). Continuous LLIN Distributions: A Guide to Concepts and Planning/ K4Health [http://www.k4health.org/toolkits/continuous- distribution-malaria/continuous-llin-distributions-guide-concepts-and- planning]. p g] RBM (2013). Malaria Indicator Surveys - Access to Reports, MIS Datasets, Survey Information [http://www.malariasurveys.org/surveys.cfm]. Stevens ER, Aldridge A, Degbey Y, Pignandi A, Dorkenoo MA, Hugelen-Padin J (2013). Evaluation of the 2011 long-lasting, insecticide-treated net distribution for universal coverage in Togo. Malar. J. 12:162. REFERENCES CD000363. OMS (2011). Rapport 2011 sur le paludisme dans le monde [http://www.who.int/malaria/world_malaria_report_2011/fr/] PNLP-AFRICARE BENIN-CRS BENINRS (2010). Evaluation Finale Du Projet D’appui à La Lutte Contre Le Paludisme. Cotonou, 42p. PNLP (2011). Evaluation post campagne de la distribution gratuite de MIILD en 2011 au Bénin. Cotonou, 58p.
https://openalex.org/W2336847247	https://lipidworld.biomedcentral.com/track/pdf/10.1186/s12944-016-0241-4	English	null	Effect of n-3 fatty acids on the expression of inflammatory genes in THP-1 macrophages	Lipids in health and disease	2,016	cc-by	5,495	© 2016 Allam-Ndoul et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Effect of n-3 fatty acids on the expression of inflammatory genes in THP-1 macrophages Bénédicte Allam-Ndoul1, Frédéric Guénard1, Olivier Barbier2 and Marie-Claude Vohl1* Bénédicte Allam-Ndoul1, Frédéric Guénard1, Olivier Barbier2 and Marie-Claude Vohl1* Abstract Background: Uncontrolled inflammation participates in the development of inflammatory diseases. Beneficial effects of polyunsaturated fatty acids belonging to the n-3 family such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) on inflammation have been reported. ( ) p The present study investigates the basal effects of EPA, DHA and a mixture EPA + DHA on the expression of 10 genes (AKT1, MAPK, NFKB, TNFA, IL1Β, MCP1, ALOX5, PTGS2, MGST1and NOS2) related to inflammation in unstimulated cultured THP1 macrophages. Cells were incubated for 24 h with n-3 PUFAs (50 μM and 10 μM EPA, DHA, EPA + DHA). Expression levels of inflammatory genes were analyzed by real-time PCR. Results: 50 μM, 10 μM EPA and 50 μM EPA + DHA decreased the expression of genes involved in the NF-κB pathway (MAPK, AKT1, and NFKB). Treatment with 50 μM, 10 μM EPA, 50 μM DHA and EPA + DHA decreased expression levels of cytokines genes IL1Β and MCP1. TNFA expression was decreased by 50 μM, 10 μM of EPA, DHA and with 50 μM EPA + DHA. Two genes involved in the fatty acid metabolism (PTGS2 and ALOX5) were also modulated by the n-3 FAs. 50 μM of DHA and EPA + DHA inhibited PTGS2 expression when the two concentrations of EPA, 50 μM DHA and EPA + DHA inhibited ALOX5 expression. Finally, the effects of n-3 FAs were studied among genes involved in the oxidative stress. 50 μM of each fatty acid increased MGST1 expression. Both concentration of EPA and 50 μM DHA decreased NOS2 expression. Conclusion: EPA seems to be more effective than DHA and EPA + DHA in modulating expression l nflammatory genes. The concentration of 50 μM was globally more effective than 10 μM. Conclusion: EPA seems to be more effective than DHA and EPA + DHA in modulating expression levels of selected inflammatory genes. The concentration of 50 μM was globally more effective than 10 μM. Keywords: Omega-3, Eicosapentaenoic acid, Docosahexaenoic acid, Inflammation, THP-1, Genes Keywords: Omega-3, Eicosapentaenoic acid, Docosahexaenoic acid, Inflammation, THP-1, Genes peripheral blood mononuclear cells (PBMCs) triggering in- flammatory diseases [8, 20, 32]. Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 DOI 10.1186/s12944-016-0241-4 Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 DOI 10.1186/s12944-016-0241-4 Open Access * Correspondence: marie-claude.vohl@fsaa.ulaval.ca 1Institute of Nutrition and Functional Foods (INAF), Laval University, Pavillon des Services, 2440 Hochelaga Blvd, Québec, Québec G1V 0A6, Canada Full list of author information is available at the end of the article Background fl Inflammation is part of the body’s immediate response to injury. It starts the immunologic elimination of invading pathogens and toxins in order to repair damaged tissues. During this stage, immune cells produce inflammatory molecules participating in the destruction of foreign agents [6, 31]. Although inflammation is a normal response, it can harm host tissues when uncontrolled. Such inappropriate inflammatory responses are in part characterized by an abnormal production of inflammatory mediators by Polyunsaturated fatty acids (PUFAs) are natural con- stituents of the diet, having a wide spectrum of physio- logical roles [5, 8, 28]. Omega-6 (n-6) and omega-3 (n-3) fatty acids (FAs) are two classes of PUFAs. Linoleic acid (n-6 fatty acids) and alpha linolenic acid ((ALA) n-3 FAs) are essential FAs. They cannot be synthetized by the body and must be provided through the diet. The major end product of n-6 FAs is arachidonic acid (AA). AA generally leads to the production of pro-inflammatory eicosanoids. ALA is the precursor of eicosapentaenoic (EPA, 20:5n-3) and docosahexaenoic (DHA, 22:6n-3) acids producing anti-inflammatory eicosanoids involved in the resolution of inflammation [4, 29]. Unfortunately, cellular Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Page 2 of 7 Page 2 of 7 mechanisms by which n-3 FAs act on immune cells are not completely understood. Cell culture and Fatty acids treatment The human THP-1 cell line, an acute monocytic leukemia cell line (American Type Culture Collection (ATCC), Rockville, MD, USA), was cultured in RPMI- 1640 media supplemented with 10 % FBS, penicillin (100 U/ml) and streptomycin (100 μM/ml) at 37 °C in a 5 % CO2 incubator. To induce differentiation of monocytes into macrophages, 106 cells per ml were seeded into 12 well plates, with 200 nM of PMA for 72 h. After incuba- tion, non-attached cells were removed by aspiration, and the adherent cells were washed with PBS three times. Preliminary tests were done to determine the optimum conditions for this study which were an incubation of 24 h in 50 and 10 μM of n-3 FAs. Stock solutions of FAs (EPA-DMSO 330 nM and DHA-DMSO 760nM) pre- pared in serum-free RPMI 1640 medium were diluted in culture medium to obtain 50 and 10 μM concentrations. Reagents and cell lines Phosphate-buffered saline (PBS) solution was obtained from Life Technologies (Burlington, Canada). Cell cul- ture media, Roswell Park Memorial Institute medium 1640 (RPMI), Fetal Bovine Serum (FBS), penicillin and streptomycin media supplements, phorbol 12-myristate 13-acetate (PMA), dimethyl sulfoxide (DMSO) were pur- chased from Thermo Scientific (Walthman, USA). EPA, DHA and reagents for reverse transcription were ob- tained from Applied biosystems (Oakville, Canada). Statistical analyses Gene expression levels were analyzed using SAS soft- ware (version 9.12). Experimental results are reported as means ± S.E. from at least three independent experi- ences performed each in triplicate. Treatments were compared to control and to different concentrations using the least square means. The level of significance was defined as P < 0.05 for all tests. RNA isolation and quantitative real-time PCR RNA isolation and quantitative real-time PCR After 24 h, total RNA was extracted using RNeasy Mini Kit (Qiagen) following the protocol provided. Then cDNA was produced from RNA using High Capacity Transcription Kit (Applied biosystems). Real-time PCR was used to determine the expression of several inflam- matory genes (Table 1) related to FA metabolism, FA re- ceptors, oxidative stress, NF-κB signalling or toll-like receptor signaling. PCR samples were normalized against 18S gene expression. Primers and TaqMan® probes were obtained from Applied Biosystems (Table 1). Finally, each experimental treatment was performed three times and replicates were analyzed. It is important to further understand the molecular mechanisms underlying beneficial effects of PUFAs, such as EPA and DHA, in non-inflamed human immune cells. Therefore, the aim of this study was to investigate the ef- fect of EPA, DHA and a mixture of EPA and DHA on expression levels of genes involved in inflammation in THP-1 macrophages were the inflammation was not in- duced. This study was realized on THP-1 macrophages because of the similarity of their transcriptomic profile with human primary cells [22]. Background fl Cells were thereafter incubated in 50 and 10 μM of EPA, DHA and in 50 μM (25 μM EPA + 25 μM DHA) or 10 μM (5 μM EPA + 5 μM DHA) EPA + DHA for 24 h. The controls in this experiment were THP-1 cells incu- bated with the vehicle, DMSO. In vitro studies revealed that both EPA and DHA may influence the functional responses of immune cells to stimulation by reducing the production of pro- inflammatory mediators. Early studies in experimental models of autoimmunity [15] and clinical trials on fish oil in patients with rheumatoid arthritis [13] demon- strated significant anti-inflammatory activity of the com- bination of EPA and DHA. The beneficial effects of n-3 FAs are thought to result from their anti-inflammatory effects, in part through changes in gene expression levels. In women with type 2 diabetes and without hypertiglyceridemia Kabir et al. showed that expression levels of inflammation-related genes were reduced in adipose tissue after n-3 FA supplementation [10]. In the same manner, a transcriptomic study in elderly subjects by Bowens et al. showed that a supplementation with 1.8 g of EPA + DHA per day could alter gene expression toward a more anti-inflammatory profile among participants with- out inflammatory disease [3]. Finally, our team demon- strated that n-3 FAs supplementation (1.9 g EPA+ 1.1 g DHA/day during 6 weeks) in healthy men and women could alter gene expression levels of inflammation-related genes [24]. A decrease of the expression of TNFA (37 %) and MGST1 (17 %) was seen following this supplementa- tion, reinforcing the notion that EPA and DHA have an anti-inflammatory effect. This study also showed an alter- ation of pathways biologically related to inflammation and influenced by n-3 FAs supplementation. In fact, NFKB and oxidative stress signalling pathways were modulated after this supplementation [24]. Effects of n-3 FAs on cytokine gene expression Effects of n-3 FAs on cytokine gene expression EPA significantly reduced the expression of genes coding for cytokines IL1B, MCP1 and TNFA, Fig. 1. EPA con- centrations of 10 and 50 μM equally reduced the expres- sion of MCP1 and IL1B. As far as TNFA was concerned 50 μM of omega-3 FA reduced its expression more ef- fectively than 10 μM. When cells were incubated with Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Page 3 of 7 Table 1 Genes and primers sets for real-time PCR 5’ Forward primer 3’ Reverse primer NFKB pathway : AKT1 CTCCTGAGGAGCGGGAGGAGTGG GTCCACTCCTCCCGCTCCTCAGGA MAPK TATGGCTCTGTGTGTGCTGCTTTTG CAAAAGCAGCACACACAGAGCCAT NFKB ACAACTATGAGGTCTCTGGGGGTA GTACCCCCAGAGACCTCATAGTTGT Cytokines production : TNFA CCATGTTGTAGCAAACCCTCAAGCT AGCTTGAGGGTTTGCTACAACATGG IL1Β CAGATGAAGTGCTCCTTCCAGGACC GGTCCTGGAAGGAGCACTTCATCTG MCP1 CGCTCAGCCAGATGCAATCAATGCC GGCATTGATTGCATCTGGCTGAGCG Fatty acids metabolism : ALOX5 GGCATTGATTGCATCTGGCTGAGCG CGCTCAGCCAGATGCAATCAATGCC PTGS2 CTGGGCCATGGGGTGGACTTAAATC GATTTAAGTCCACCCCATGGCCCAG Oxydation system : MGST1 TTGACAAGAAAGGTTTTTGCCAATC GATTGGCAAAAACCTTTCTTGTCAA NOS2 CCATGGAACATCCCAAATACGAGT CACTCGTATTTGGGATGTTCCATG HO-1 AAGATTGCCCAGAAAGCCCTGGAC AACTGTCGCCACCAGAAAGCTGAG Housekeeping genes : RPL37A GGTGCCTGGACGTACAATACCACTT AAGTGGTATTGTACGTCCAGGCACC 18S CCATTGGAGGGCAAGTCTGGTGCC TGGCACCAGACTTGCCCTCCAATGG Effects of n-3 FAs on the expression of genes involved in FA The effect of EPA, DHA and EPA + DHA on the expres- sion of two genes involved in the FA metabolism (ALOX5 and PTGS2) is presented in Fig. 2. Relative to vehicle control-related cells, 10 and 50 μM of EPA equally decreased the expression of ALOX5. DHA and EPA + DHA also decreased its expression, but only when the concentration of 50 μM was used. The expres- sion of PTGS2 was also modulated by the omega-3 FAs. A treatment of THP-1 cells with 50 μM of DHA and EPA + DHA reduced its expression when EPA did not have any effect on PTGS2 gene expression levels. CCATGTTGTAGCAAACCCTCAAGCT AGCTTGAGGGTTTGCTACAACATGG CAGATGAAGTGCTCCTTCCAGGACC GGTCCTGGAAGGAGCACTTCATCTG CGCTCAGCCAGATGCAATCAATGCC GGCATTGATTGCATCTGGCTGAGCG Effects of n-3 FAs on NF-κB pathway gene expression Figure 3 shows the effect of EPA, DHA and EPA + DHA on three genes (AKT1, MAPK and NFKB) involved in the NFKB pathway. 50 and 10 μM of EPA equally inhib- ited the expression of each of these genes. No significant changes in the expression of NFKB, MAPK and AKT1 were seen when cells were incubated in DHA. The mix- ture EPA + DHA at a concentration of 50 μM decreased the expression of NFKB and AKT1 but had no effect on MAPK. The concentration of 10 μM did not change the expression of these 3 genes. Effects of n-3 FAs on cytokine gene expression EPA seemed to be more ef- fective on inhibiting NFKB, MAPK and AKT1 than DHA or the mixture EPA + DHA. GGCATTGATTGCATCTGGCTGAGCG CGCTCAGCCAGATGCAATCAATGCC CTGGGCCATGGGGTGGACTTAAATC GATTTAAGTCCACCCCATGGCCCAG Effects of n-3 FAs on the expression of oxidative stress genes Finally, we studied changes in the expression of two genes involved in the oxidative stress (MGST1 and NOS2-Fig. 4). At a concentration of 50 μM EPA and DHA up-regulated the expression of MGST1 whereas 10 μM did not have any effect. For NOS2, 50 μM of EPA and DHA decreased its expression more efficiently than 10 μM. The mixture EPA + DHA did not have any effect on the expression of these genes. EPA had a more powerful effect on the expression of MGST1 and NOS2 than DHA or the mixture EPA + DHA. Discussion Inflammation is a condition which contributes to a range of human diseases [7, 27]. EPA and DHA are the major n-3 FAs found in oily fish and fish oil supplements. There is substantial evidence that these FAs are able to partly inhibit several aspects of inflammation, including production of inflammatory cells [27] and eicosanoids from n-6 [31]. The literature had shown that the influ- ence of EPA and DHA on immune cell functions is me- diated by several mechanisms [4]. In this study, we focused on the action of EPA and DHA on the regula- tion of inflammatory gene expression levels on unstimu- lated THP-1 cells. Literature showed that an increased intake of EPA and DHA resulted in an increased 50 μM of DHA, the down-regulation of TNFA and IL1B was more intense than with10 μM. The expression of MCP1 was equally decreased with 50 or 10 μM of both EPA and DHA. The mixture EPA + DHA down regulated the expression of MCP1, TNFA and IL1B but only with the concentration of 50 μM while 10 μM of EPA + DHA did not affect their expression. Except for TNFA on which EPA and DHA had the same effect; EPA seemed to have a more powerful action on the expression of MCP1 and IL1B than DHA and EPA + DHA. Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Page 4 of 7 Fig. 1 Effects of n-3 FAs on cytokine gene expression after a 24 h incubation of THP-1 macrophages in 10 μM and 50 μM of EPA, DHA and EPA + DHA. a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration Fig. 1 Effects of n-3 FAs on cytokine gene expression after a 24 h incubation of THP-1 macrophages in 10 μM and 50 μM of EPA, DHA and EPA + DHA. a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration incorporation of n-3 FAs in membrane phospholipids of immune cells [6]. Discussion a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration Fig. 3 Effects of omega-3 FAs on the expression of genes involved in NF-κB pathway. a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration It is important to highlight the fact that the aim of the present study was to investigate the effect of EPA, DHA and EPA + DHA on the expression of inflammatory genes, in cells under baseline inflammatory conditions. Studies investigating the effects of n-3 FAs on inflamma- tion without induction of inflammation are limited. In most in vitro studies, induction of inflammation with different chemicals (LPS, TNFA, and interferon gamma) is triggered to evaluate the effects of n-3 FAs [19]. Changes in gene expression levels are thus more import- ant in these models. It might be a reason why the effects of n-3 FAs are not as pronounced here as in studies with induced inflammation. In vitro and clinical [2] studies showed that EPA, DHA and fish oil supplements can de- crease production of pro-inflammatory cytokines. In addition, a study by Wang and coll., on macrophages treated with EPA and DHA showed reduced expression of TNFA, IL6 and MCP1 compared to controls [30]. These results highlight the ability of EPA, DHA and EPA + DHA to down-regulate pro-inflammatory markers, even in the absence of pro-inflammatory stimuli thus suggesting a protective effect of n-3 FAs on THP-1 macrophages in- flammatory profile concordant with reported health bene- fits (7). Combined with data from the present study, these findings suggest that the inverse relationship between n-3 FAs and inflammation may be partly due to their ability to modulate macrophage cytokine secretion via NF-κB pathway. Eicosanoids are biologically active lipid mediators pro- duced from PUFA, mainly AA. They are important media- tors of inflammation [16]. The prostaglandin-endoperoxide synthase (PTGS; also called cyclooxygenase (COX)) and Fig. 4 Effects of n-3 FAs on the expression of genes involved in oxidative stress. Discussion N-3 FAs can then modulate the ex- pression of several inflammatory genes by interacting with various nuclear receptor and transcription factors [4]. NFKB is a key transcription factor involved in the up-regulation of cyclooxygenase (COX2) gene, adhesion molecules and inflammatory cytokines [14, 26]. Previous studies reported that fish oil decrease LPS-induced acti- vation of NFKB in human monocytes [17, 32]. Singer et al., suggested that EPA and DHA could inhibit NFKB ac- tivity at several levels during NF-κB pathway [32]. Fol- lowing this lead, we studied in unstimulated macrophages, the expression levels of genes involved in this pathway. We demonstrate that EPA and EPA + DHA, but not DHA, slightly depressed the expression of genes involved in NF-κB pathway (AKT1, MAPK and NFKB). The modulation of these genes expression being very low, we can hypothesise that in unstimulated mac- rophages the action of EPA and DHA does not seem to go through NF-κB pathway alone. Other mechanisms might be involved. In addition, EPA suppressed the mRNA expression of pro-inflammatory cytokines (IL1B, MCP1 and TNFA) in THP-1 monocyte-derived macrophages. As mentioned previously, in inflammatory conditions, NFKB plays key roles in up-regulating pro-inflammatory cytokine gene transcription [9]. The results of this study suggest that the effects observed on cytokines can reflect upstream actions on genes from the NF-κB pathway. Effects on gene expression being relatively modest, we can also hy- pothesise that n-3 FAs may act on expression levels of cytokine genes by other mechanisms. Fig. 2 Effects of n-3 FAs on the expression of genes involved in FA metabolism. a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration Fig. 2 Effects of n-3 FAs on the expression of genes involved in FA metabolism. a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Page 5 of 7 Fig. 3 Effects of omega-3 FAs on the expression of genes involved in NF-κB pathway. Discussion The present study shows a dose and n-3 FAs dependent reduction of the expression of ALOX5 and PTGS2 when cells are ex- posed to these FAs. As shown in the literature, the incorp- oration of EPA and DHA into inflammatory cells occurs at the expense of AA [6]. This might explain the decrease in gene expression levels of PTGS2 and ALOX5. g p We also studied the effect of n-3 FAs on the expression of genes involved in oxidative stress. Reactive oxygen spe- cies (ROS) are produced by the body and are important in normal cellular functioning. Neutralising ectopic oxidative stress is very important to maintain the body’s integrity. The potential role of oxidative stress as an early event in inflammatory diseases has been widely noted [23]. To prevent abnormal oxidative stress, several physiological anti-oxidant systems act as protective mechanisms. Microsomal glutathione S- transferase 1 (MGST1) and heme oxygenase (HO-1) are enzymes which detoxify re- active intermediates. Nitric Oxide Synthase (NOS) cata- lyzes the production of nitric oxide (NO), a free radical, as a defense mechanism. During pathological inflammation NO is produced in an excessive manner leading to cell damages. Figure 4 suggests that even when inflammation is not induced, n-3 FAs act on oxidative stress by reducing the expression of a pro-oxidative gene (NOS2) and en- hancing the expression of MGST1, an anti-oxidative gene. In the previous study led by our laboratory [24], the 30 participants had a body mass index between 25 and 40 kg/m2. It means that they were overweight to obese participants. Thus the results of this study were from individuals with a potential chronic inflammatory state whereas our in vitro study was performed on un- stimulated macrophages. In the clinical trial, each par- ticipant received 5 capsules containing 1 g oil each per day (1.9 g EPA and 1.1 g DHA). It is of note that the re- sponses observed in this human clinical study may be modulated by genetic variations that influenced the in- dividual response to n-3 FAs supplementation. These facts might be partly explaining the difference regarding MGST1 expression between the human and the in vitro studies. It is also important to consider the fact that the in vitro study considerably simplifies what is observed in the human body. In fact it doesn’t really take into ac- count different interactions that can be observed in the human body. Authors’ contributions BAN i d h BAN carried out the experiments, interpreted the data and drafted the manuscript. BAN and FG performed statistical analysis. BO participated in the design of the study. MCV supervised the entire project. All authors read and approved the final manuscript. Discussion a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration Fig. 4 Effects of n-3 FAs on the expression of genes involved in oxidative stress. a,b Represents the differences (P ≤0.05) between the 2 concentrations for each omega-3 FAs and for the mixture EPA + DHA. P ≤0.05 relative to DMSO; 1P ≤0.05 relative to EPA within the same concentration; 2P ≤0.05 relative to DHA within the same concentration Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Page 6 of 7 Page 6 of 7 N-3 FAs have beneficial effects on inflammation [4]. It is important to know whether effects of each o n-3 FAs on inflammation are similar. This study shows that in unstimulated macrophages, depending on the metabolic pathway of interest, the n-3 FA having the most import- ant effect is different. Globally, EPA has the most potent effect on gene expression comparatively to DHA and EPA + DHA. Interestingly, in most studies on the effect of EPA and DHA on inflammatory genes, DHA was a more potent inhibitor of inflammatory genes than EPA [11, 12, 19, 21]. However, inflammation was induced in these studies while in the present study effects of n-3 FAs on macrophages were studied without induced inflamma- tion. The greater effect of DHA seems to come from its secretion of more anti-inflammatory lipid mediators (Resolvin D, maresins, protectins) than EPA (Resolvin E) [6, 25]. It could explain that in studies where inflammation is induced, the effects of DHA on expression levels of in- flammatory genes were more intense than EPA. Indeed, the induction of inflammation triggers the synthesis of resolvins and other lipid mediators derived from n-3 FAs not seen in cells when inflammation is not induced. Stimulation of a macrophage cell culture with an inflam- matory agent may lead to different results but this need to be confirmed. the lipoxygenase (ALOX) are two enzymes contributing to the synthesis of eicosanoids. AA is abundant in membrane phospholipids. It produces pro-inflammatory eicosanoids through these enzymes [5]. EPA and DHA are weaker sub- strates for ALOX5 and PTGS2 than AA. They produce less potent eicosanoids and lipid mediators (resolvins, maresins, protectins), with an anti-inflammatory profile. Discussion The role of n-3 FAs on the restoration of free radical homeostasis is not fully understood al- though studies suggest that EPA and DHA reduce oxi- dative damages in human and animals [1, 18]. A dose effect of the omega-3 FA is also seen. Globally incubating the cells with 50 μM of each omega-3 FAs seems to be more efficient then with 10 μM. Thus, when studying the effects of omega-3 FAs, the concentrations must be carefully set in order to truly observe their action. Conclusion N-3 FAs may modulate the expression of genes related to inflammation. In fact these FAs seem to down- regulate the expression of genes promoting inflamma- tion (cytokines, NF-κB pathway) and up-regulate the ex- pression of MGST1 involved in the prevention of inflammation through detoxification of ROS. The n-3 FAs used in this study have distinct effects on gene ex- pression, EPA having a more potent effect than DHA and or EPA + DHA combined. A dose effect was also seen, in fact higher concentrations (50 μM) being more potent than lower ones (10 μM). Competing interests The authors declare that they have no competing interest. References 1. An WS, Kim HJ, Cho KH, Vaziri ND. Omega-3 fatty acid supplementation attenuates oxidative stress, inflammation, and tubulointerstitial fibrosis in the remnant kidney. Am J Physiol Renal Physiol. 2009;297:F895–903. 1. An WS, Kim HJ, Cho KH, Vaziri ND. Omega-3 fatty acid supplementation attenuates oxidative stress, inflammation, and tubulointerstitial fibrosis in the remnant kidney. Am J Physiol Renal Physiol. 2009;297:F895–903. 25. Serhan CN, Chiang N, Van Dyke TE. Resolving inflammation: dual anti- inflammatory and pro-resolution lipid mediators. Nat Rev Immunol. 2008;8:349–61. 2. Baumann KH, Hessel F, Larass I, Muller T, Angerer P, Kiefl R, von, SC. Dietary omega-3, omega-6, and omega-9 unsaturated fatty acids and growth factor and cytokine gene expression in unstimulated and stimulated monocytes. A randomized volunteer study. Arterioscler Thromb Vasc Biol. 1999;19:59–66. 26. Sigal LH. Basic science for the clinician 39: NF-kappaB-function, activation, control, and consequences. J Clin Rheumatol. 2006;12:207–11. 3. Bouwens M, van de Rest O, Dellschaft N, Bromhaar MG, de Groot LC, Geleijnse JM, Muller M, Afman LA. Fish-oil supplementation induces antiinflammatory gene expression profiles in human blood mononuclear cells. Am J Clin Nutr. 2009;90:415–24. 27. Simopoulos AP. Omega-3 fatty acids in inflammation and autoimmune diseases. J Am Coll Nutr. 2002;21:495–505. 28. Siriwardhana N, Kalupahana NS, Moustaid-Moussa N. Health benefits of n-3 polyunsaturated fatty acids: eicosapentaenoic acid and docosahexaenoic acid. Adv Food Nutr Res. 2012;65:211–22. 4. Calder PC. Polyunsaturated fatty acids, inflammatory processes and inflammatory bowel diseases. Mol Nutr Food Res. 2008;52:885–97. 29. Turk HF, Chapkin RS. Membrane lipid raft organization is uniquely modified by n-3 polyunsaturated fatty acids. Prostaglandins Leukot Essent Fatty Acids. 2013;88:43–7. 5. Calder PC. Mechanisms of action of (n-3) fatty acids. J Nutr. 2012;142:592S–9. 6. Calder PC. N-3 fatty acids, inflammation and immunity: new mechanisms to explain old actions. Proc Nutr Soc. 2013;72:326–36. 30. Wang S, Wu D, Lamon-Fava S, Matthan NR, Honda KL, Lichtenstein AH. In vitro fatty acid enrichment of macrophages alters inflammatory response and net cholesterol accumulation. Br J Nutr. 2009;102:497–501. 7. Calder PC, Albers R, Antoine JM, Blum S, Bourdet-Sicard R, Ferns GA, Folkerts G, Friedmann PS, Frost GS, Guarner F, Løvik M, Macfarlane S, Meyer PD, M'Rabet L, Serafini M, van Eden W, van Loo J, Vas Dias W, Vidry S, Winklhofer-Roob BM, Zhao J. Inflammatory disease processes and interactions with nutrition. Br J Nutr. 2009;101 Suppl 1:S1–45. 31. Yaqoob P, Calder P. Effects of dietary lipid manipulation upon inflammatory mediator production by murine macrophages. Received: 21 January 2016 Accepted: 30 March 2016 Received: 21 January 2016 Accepted: 30 March 2016 23. Roberts CK, Sindhu KK. Oxidative stress and metabolic syndrome. Life Sci. 2009;84:705–12. 24. Rudkowska I, Paradis AM, Thifault E, Julien P, Tchernof A, Couture P, Lemieux S, Barbier O, Vohl MC. Transcriptomic and metabolomic signatures of an n-3 polyunsaturated fatty acids supplementation in a normolipidemic/ normocholesterolemic Caucasian population. J Nutr Biochem. 2013;24:54–61. Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Page 7 of 7 Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 Vohl is tier 1 Canada Research Chair in Genomics Applied to Nutrition and Health. We thank Alain Houde for contributing to the laboratory work. 19. Mullen A, Loscher CE, Roche HM. Anti-inflammatory effects of EPA and DHA are dependent upon time and dose–response elements associated with LPS stimulation in THP-1-derived macrophages. J Nutr Biochem. 2010;21:444–50. Author details 1 f g 20. Nathan C, Ding A. Nonresolving inflammation. Cell. 2010;140:871–82. 1Institute of Nutrition and Functional Foods (INAF), Laval University, Pavillon des Services, 2440 Hochelaga Blvd, Québec, Québec G1V 0A6, Canada. 2 1Institute of Nutrition and Functional Foods (INAF), Laval University, Pavillon des Services, 2440 Hochelaga Blvd, Québec, Québec G1V 0A6, Canada. 2Laboratory of Molecular Pharmacology, CHU de Quebec Research Center, 2705, boulevard Laurier R-4720, Québec, Québec G1V 4G2, Canada. 21. Oliver E, McGillicuddy FC, Harford KA, Reynolds CM, Phillips CM, Ferguson JF, Roche HM. Docosahexaenoic acid attenuates macrophage-induced inflammation and improves insulin sensitivity in adipocytes-specific differential effects between LC n-3 PUFA. J Nutr Biochem. 2012;23:1192–200 1Institute of Nutrition and Functional Foods (INAF), Laval University, Pavillon des Services, 2440 Hochelaga Blvd, Québec, Québec G1V 0A6, Canada. 2 2Laboratory of Molecular Pharmacology, CHU de Quebec Research Center, 2705, boulevard Laurier R-4720, Québec, Québec G1V 4G2, Canada. 22. Qin Z. The use of THP-1 cells as a model for mimicking the function and regulation of monocytes and macrophages in the vasculature. Atherosclerosis. 2012;221:2–11. Acknowledgments h k This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (grant number: 48031–2012). MC Page 7 of 7 Page 7 of 7 Allam-Ndoul et al. Lipids in Health and Disease (2016) 15:69 References Cell Immunol. 1995;163:120–8. 32. Zhao Y, Joshi-Barve S, Barve S, Chen LH. Eicosapentaenoic acid prevents LPS-induced TNF-alpha expression by preventing NF-kappaB activation. J Am Coll Nutr. 2004;23:71–8. 8. Dinarello CA. Anti-inflammatory Agents: Present and Future. Cell. 2010;140: 935–50. 9. Ghosh S, May MJ, Kopp EB. NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu Rev Immunol. 1998;16:225–60. 10. Kabir M, Skurnik G, Naour N, Pechtner V, Meugnier E, Rome S, Quignard- Boulange A, Vidal H, Slama G, Clement K, Guerre-Millo M, Rizkalla SW Treatment for 2 mo with n 3 polyunsaturated fatty acids reduces adiposity and some atherogenic factors but does not improve insulin sensitivity in women with type 2 diabetes: a randomized controlled study. Am J Clin Nutr. 2007;86:1670–9. 11. Kishida E, Tajiri M, Masuzawa Y. Docosahexaenoic acid enrichment can reduce L929 cell necrosis induced by tumor necrosis factor. Biochim Biophys Acta. 2006;1761:454–62. 12. Komatsu W, Ishihara K, Murata M, Saito H, Shinohara K. Docosahexaenoic acid suppresses nitric oxide production and inducible nitric oxide synthase expression in interferon-gamma plus lipopolysaccharide-stimulated murine macrophages by inhibiting the oxidative stress. Free Radic Biol Med. 2003; 34:1006–16. 13. Kremer JM, Bigauoette J, Michalek AV, Timchalk MA, Lininger L, Rynes RI, Huyck C, Zieminski J, Bartholomew, LE,. Effects of manipulation of dietary fatty acids on clinical manifestations of rheumatoid arthritis. Lancet. 1985;1: 184–7. Submit your next manuscript to BioMed Central and we will help you at every step: Submit your next manuscript to BioMed Central and we will help you at every step: 14. Kumar A, Takada Y, Boriek AM, Aggarwal BB. Nuclear factor-kappaB: its role in health and disease. J Mol Med. 2004;82:434–48. • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit and we will help you at every step: 15. Leslie CA, Gonnerman WA, Ullman MD, Hayes KC, Franzblau C, Cathcart ES. Dietary fish oil modulates macrophage fatty acids and decreases arthritis susceptibility in mice. J Exp Med. 1985;162:1336–49. • We accept pre-submission inquiries 16. Lewis RA, Austen KF, Soberman RJ. Leukotrienes and other products of the 5-lipoxygenase pathway. Biochemistry and relation to pathobiology in human diseases. N Engl J Med. 1990;323:645–55. 17. Lo CJ, Chiu KC, Fu M, Lo R, Helton S. Fish oil decreases macrophage tumor necrosis factor gene transcription by altering the NF kappa B activity. J Surg Res. 1999;82:216–21. 17. Lo CJ, Chiu KC, Fu M, Lo R, Helton S. Fish oil decreases macrophage tumor necrosis factor gene transcription by altering the NF kappa B activity. J Surg Res. 1999;82:216–21. 18. McAnulty SR, Nieman DC, Fox-Rabinovich M, Duran V, McAnulty LS, Henson DA, Jin F, Landram, MJ. Effect of n-3 fatty acids and antioxidants on oxidative stress after exercise. Med Sci Sports Exerc. 2010;42:1704–11.
https://openalex.org/W2808745154	https://biotechnologyforbiofuels.biomedcentral.com/track/pdf/10.1186/s13068-018-1166-0	English	null	The impact of silicon on cell wall composition and enzymatic saccharification of Brachypodium distachyon	Biotechnology for biofuels	2,018	cc-by	16,190	© The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Biotechnology for Biofuels Biotechnology for Biofuels Głazowska et al. Biotechnol Biofuels (2018) 11:171 https://doi.org/10.1186/s13068-018-1166-0 Open Access Abstract Background: Plants and in particular grasses benefit from a high uptake of silicon (Si) which improves their growth and productivity by alleviating adverse effects of biotic and abiotic stress. However, the silicon present in plant tissues may have a negative impact on the processing and degradation of lignocellulosic biomass. Solutions to reduce the silicon content either by biomass engineering or development of downstream separation methods are therefore targeted. Different cell wall components have been proposed to interact with the silica pool in plant shoots, but the understanding of the underlying processes is still limited. Results: In the present study, we have characterized silicon deposition and cell wall composition in Brachypodium distachyon wild-type and low-silicon 1 (Bdlsi1-1) mutant plants. Our analyses included different organs and plant developmental stages. In the mutant defective in silicon uptake, low silicon availability favoured deposition of this element in the amorphous form or bound to cell wall polymers rather than as silicified structures. Several alterations in non-cellulosic polysaccharides and lignin were recorded in the mutant plants, indicating differences in the types of linkages and in the three-dimensional organization of the cell wall network. Enzymatic saccharification assays showed that straw from mutant plants was marginally more degradable following a 190 °C hydrothermal pretreatment, while there were no differences without or after a 120 °C hydrothermal pretreatment. Conclusions: We conclude that silicon affects the composition of plant cell walls, mostly by altering linkages of non-cellulosic polymers and lignin. The modifications of the cell wall network and the reduced silicon concentration appear to have little or no implications on biomass recalcitrance to enzymatic saccharification. Keywords: Brachypodium distachyon, Silicon, Cell wall composition, CoMPP, Recalcitrance, Hydrothermal pretreatment, Bioenergy The impact of silicon on cell wall composition and enzymatic saccharification of Brachypodium distachyon Sylwia Głazowska1, Laetitia Baldwin1, Jozef Mravec1, Christian Bukh1, Thomas Hesselhøj Hansen1, Mads Mørk Jensen2, Jonatan U. Fangel1, William G. T. Willats1, Marianne Glasius2, Claus Felby3 and Jan Kofod Schjoerring1* Backgroundh Despite multiple sources of evidence supporting the presence of Si within the cell wall matrix, the nature of interactions with cell wall polymers and the exact role of Si in the process of cell wall assembly remain poorly understood.i genetic resources and molecular tools have strengthened the position of B. distachyon among other model organ- isms [23, 24]. The cell wall composition of B. distachyon is comparable with several grass species important for agriculture and bioenergy production [25–27]. In addi- tion, the Si uptake system and the levels of Si accumula- tion in shoot tissues are similar to those in barley, maize and wheat. Together these properties make B. distachyon a valuable tool for studies of biomass conversion [28, 29]. In the present work, we have used wild-type and low- silicon 1 mutant (Bdlsi1-1) plants of B. distachyon to unravel the implications of Si on the cell wall composi- tion and the enzymatic digestibility of the biomass. We have studied the organ-specific deposition of Si in the cell walls and its effect upon cell wall polymer profiles at the ripening stage. Compositional changes associated with low Si levels in the mature plants were characterized with the overall intention to link them with the efficiency of enzymatic digestibility. We demonstrate that cell walls of low-Si plants exhibit multiple compositional altera- tions, most likely affecting the linkage of the individual polymers within the cell wall network. However, as will be shown, these changes have minor or no effects on the saccharification potential of the biomass. genetic resources and molecular tools have strengthened the position of B. distachyon among other model organ- isms [23, 24]. The cell wall composition of B. distachyon is comparable with several grass species important for agriculture and bioenergy production [25–27]. In addi- tion, the Si uptake system and the levels of Si accumula- tion in shoot tissues are similar to those in barley, maize and wheat. Together these properties make B. distachyon a valuable tool for studies of biomass conversion [28, 29]. In the present work, we have used wild-type and low- silicon 1 mutant (Bdlsi1-1) plants of B. distachyon to unravel the implications of Si on the cell wall composi- tion and the enzymatic digestibility of the biomass. We have studied the organ-specific deposition of Si in the cell walls and its effect upon cell wall polymer profiles at the ripening stage. Backgroundh Compositional changes associated with low Si levels in the mature plants were characterized with the overall intention to link them with the efficiency of enzymatic digestibility. We demonstrate that cell walls of low-Si plants exhibit multiple compositional altera- tions, most likely affecting the linkage of the individual polymers within the cell wall network. However, as will be shown, these changes have minor or no effects on the saccharification potential of the biomass. Plants benefit from Si uptake and Si fertilization of many crop species is carried out as a standard procedure in agriculture and horticulture [15, 16]. Silicon amend- ment substantially improves plant growth and vigour upon exposure to stress conditions [17]. On the other hand, the presence of Si in plant material is problematic in industrial utilization of biomass for energy, as it may form insoluble precipitates and cause increased wear of equipment [18]. In addition, cell wall-bound Si may con- tribute to the recalcitrance of biomass [18]. However, the mechanisms triggering Si deposition within cell walls remain unexplained, and the fact that Si interacts with multiple polymers inside cell wall network makes it very difficult to remove and study [19]. Backgroundh studies suggest that Si deposition is correlated with the presence of Si transport proteins (LSI) mediating uptake of orthosilicic acid [Si(OH)4] and Si translocation to shoots [2, 3]. Silicon deposited in spines, trichomes and beneath the cuticle in the leaf blades increases the physi- cal resistance of plants towards harmful environmental factors [4, 5]. Based on observations in ex vivo experi- ments showing that certain cell wall polymers added to silicic acid solution favoured precipitation of Si aggre- gates, cell wall polysaccharides have also been suggested to interact with Si [6, 7]. The most recent indications of Si The grass family (Poaceae) includes numerous plants species important worldwide as sources of food, feed and bioenergy. One of the distinctive features of grasses is their ability to accumulate silicon (Si) in high quanti- ties, usually over 2% of the shoot dry weight [1]. Multiple Correspondence: jks@plen.ku.dk 1 Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 2 of 18 Page 2 of 18 involvement in cell wall structures in planta come from studies of two Si accumulating species, rice and Equise- tum [8, 9]. Since plants belonging to both the Poales and Equisetales order contain mixed-linkage glucans (MLG), it was suggested that MLG might serve as a template for Si polymerization in plants [10]. Studies on rice cell wall mutants showed positive correlation between the Si content in biomass and cellulose, hemicelluloses and lignin [11]. In addition, Si supply stimulated deposi- tion of cellulose and hemicelluloses in rice, suggesting further involvement of Si in the formation of cell walls [11]. Hemicelluloses were identified to be ligands bind- ing Si in cell walls of rice cell cultures [9, 12, 13]. In cell walls of fern (Adiantum raddianum), Si co-localized with epitopes of homogalacturonans (HG) and 1,5-arabinans [14]. Plant material and growth conditions Plant material and growth conditions Seeds of Brachypodium distachyon (accession Bd21-3) wild-type plants and the low-silicon 1 (Bdlsi1-1) mutant were surface sterilized as previously described [30, 31]. Briefly, the seeds were soaked in MilliQ water for 2 h then washed with 70% (v/v) ethanol for 30 s, 1.3% sodium hypochlorite (v/v) for 4 min and rinsed with sterile Mil- liQ water. The seeds were vernalized in the dark at 5 °C in MilliQ water for 4 days. Thereafter, the seeds were placed in pots containing soil (Pindstrup Substrate NO. 2; Pindstrup Mosebrug A/S, Ryomgaard Denmark) and transferred to a greenhouse with a temperature around 19 °C and a 16 h photoperiod. The plants were harvested at two developmental stages corresponding to the Biolo- gische Bundesanstalt, Bundessortenamt und CHemische Industrie (BBCH) scale developmental stages of ripen- ing (BBCH 85–89) and senescence (BBCH 92–99) [32]. A total of 9 plants, sampled as 3 pools of 3 plants, were harvested at each stage, the material was frozen in liq- uid nitrogen and stored at − 80 °C. At the ripening stage, the sampled plants were subdivided into: (i) leaf blades and sheaths; (ii) stems; and (iii) spikelets (heads), which included all reproductive organs, meaning flowers, seeds and bracts (seed covers). At senescence (maturity), entire shoots, including all the above-ground organs, i.e. both the vegetative and generative parts (straw material and fi Cell walls are dynamic structures of which the compo- sition and spatial arrangement vary not only among plant species or tissue types, but also changes during plant devel- opment and in response to external factors [20]. Second- ary cell walls of grasses are formed by cellulose microfibrils embedded in a network of hemicelluloses, hydroxycin- namic acids and lignin with minor contribution of pectins and structural proteins [20]. In addition, mineral elements such as boron (B) and calcium (Ca) play crucial structural roles in the assembly of the cell wall network by cross-link- ing pectin polymers [21, 22]. Grasses are widely used feedstocks in production of second generation biofuels. To speed up progress in developing bioenergy crops, the model plant Brachypo- dium distachyon has been introduced as a tool for rapid testing of traits and strategies directed towards the use of grasses in bioenergy applications [23]. Features such as short generation time, small genome size, and multiple Głazowska et al. Cell wall polysaccharide compositional analysis Non-cellulosic polysaccharide composition and cellulose content were determined following [36] with modifica- tions. Briefly, de-starched AIR material was hydrolyzed in 2 M trifluoroacetic acid (TFA) for 90 min at 121 °C. TFA releases primarily non-cellulosic polysaccharides; however, also parts of cellulose from regions containing kinks or dislocations will be hydrolyzed. TFA was evapo- rated and pellet resuspended in 500 µL MilliQ water. The supernatant was filtered with 0.45 µm nylon filters for quantification of monosaccharides in the non-cellulosic fraction, whereas the pellet was collected for analysis of the cellulose content (mostly intact microfibrils). The monosaccharide composition was quantified by high- performance anion-exchange chromatography (HPAEC) on Dionex ICS5000 coupled with pulsed amperometric detection (PAD) using CarboPac PA1 column and guard (2 × 250 mm/2 × 30 mm) (Thermo Fisher Scientific) according to the manufacturer’s instructions (TN280) with modifications. For neutral sugars, samples were eluted with MilliQ water at a column temperature of 17 °C in 75 min at a flow rate of 0.250 mL min−1, whereas for uronic acids, a multistep gradient of 30 mM NaOH and 1 M NaOAc in 30 mM NaOH was used for elution at a flow rate of 0.250 mL min−1 over 17 min. A solution of 300 mM NaOH was applied post-column to improve the sensitivity of detection during both analyses. The separated monosaccharides were quantified using cali- bration with monosaccharide standards (l-arabinose, l-rhamnose, d-xylose, d-galactose, d-glucose, d-glucu- ronic and d-galacturonic acid) (Sigma). Plant material and growth conditions Biotechnol Biofuels (2018) 11:171 Page 3 of 18 (pH 7) containing 0.125 µM 2-(4-pyridyl)-5-((4-(2- dimethyaminoethylaminocarbamoyl)-methoxy) phenyl) oxazole (PDMPO, LysoSensor Yellow/Blue DND-160, 1 mM in DMSO, Thermo Fisher). Samples were left for 24 h to allow the labelling. PDMPO is used to measure the pH of acidic organelles; however, it has been shown to produce green fluorescence upon interaction with silica surfaces, but not silicic acid [6, 35]. Each sample was examined under a fluorescence microscope (Leica DM5000B). all the spikelets containing mature seeds) were harvested and analysed as a whole or as the straw part separately, the latter consisting of leaves with sheaths and stems. Silicon, boron and calcium concentrations in cell wall materialh The analyses were performed on samples harvested at the ripening growth stage and at maturity as described below. AIR material (20–30 mg) was weighed into Tef- lon microwave digestion tubes followed by the addition of 1.25 mL of 70% (v/v) nitric acid and 500 µL of 15% H2O2 (v/v). The tubes were then capped and the sam- ples digested in a microwave oven at 242 °C for 25 min (UltraWAVE single reaction chamber microwave diges- tion system, Milestone Inc., Shelton, CT Multiwave 3000, software version 1.24, Anton Paar GmbH, Graz, Austria). The resulting solution was filtered through a membrane filter having a pore size of 0.22 µm (Q-Max RR 25 mm CA, Frisenette ApS, Denmark) in order to withhold phy- toliths. After addition of 0.1 mL of 49% (v/v) hydrofluoric acid and 1 mL of 36% (v/v) hydrochloric acid, the filtrate was incubated overnight at room temperature, followed by dilution with MilliQ water to a final volume of 50 mL. The elemental composition of the samples was measured by inductively coupled plasma optical emission spec- trometry (ICP-OES; Agilent 5100, Agilent Technologies, Manchester, UK). Reference material (spinach leaf, NCS ZC73013, China National Analysis Center for Iron and Steel, Beijing, China) was included in the analysis to vali- date analytical precision and accuracy. The content of intact cellulose microfibrils in the TFA-resistant pellet was quantified based on the method reported by [37]. The TFA-resistant material was treated with Updegraff reagent (acetic acid: nitric acid: water in the ratio 8:1:2, v/v) at 100 °C for 30 min, which removes residual hemicelluloses and regions of cellulose with dislocations that were not removed during hydrolysis with TFA. The remaining pellet was washed one time with water and 3 times with 100% acetone and air dried overnight. Next, the pellet was Cell wall material preparation Cell wall material (alcohol insoluble residue, AIR) was prepared in biological triplicates as described by [33] with adaptations. Plant material was ground using a mortar and pestle and washed 6 times with 70% (v/v) ethanol for 10 min followed by washing with 100% ace- tone for 10 min. The pellet was left to dry and AIR was obtained. Starch was removed by enzymatic digestion according to [34]. Briefly, 100 mg AIR was resuspended in 100 mM potassium phosphate, pH 6.8 and digested with α-amylase (0.5 U mg−1 AIR, α-amylase from Bacil- lus subtilis Type II A, SIGMA A6814-1MU) for 24 h at room temperature. The pellet was washed 3 times with water and 2 times with 100% acetone and freeze-dried. Labelling of silica surfaces in phytoliths Silica residues (phytoliths) were recovered from the filters, washed several times with water followed by acetone, and left to air dry. The phytoliths were subsequently suspended in 20 mM PIPES buffer Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 4 of 18 Page 4 of 18 (4 mg) was saponified for 1 h at room temperature by suspending in 180 µL of water and 60 µL of 1 M NaOH with shaking. Afterwards, the solution was neutral- ized with HCl and centrifuged. The supernatant with released methanol (50 µL) and alcohol oxidase (0.05 U in 0.1 M sodium phosphate, pH 7.5) (Sigma) was loaded into a microplate and incubated for 20 min at room temperature with shaking. Next, 100 µL of mix contain- ing 0.02 M 2,4-pentanedione in 2 M ammonium acetate and 0.05 M acetic acid was added to each well and incu- bated for 10 min at 68 °C. Samples were cooled on ice and absorbance at 412 nm was measured in a microplate reader (Eon™ High-Performance Microplate Spectropho- tometer, BioTek). The degree of methylesterification was calculated as the molar ratio of methanol to galacturonic acid (expressed in %). The content of galacturonic acid was quantified previously by HPAEC, as described in the “Cell wall polysaccharide compositional analysis” section. hydrolyzed with 72% (v/v) H2SO4 for 45 min at room temperature to break down intact cellulose fibrils. The released glucose was quantified using the colorimet- ric anthrone assay as follows: 5 μL of each sample was added to a 96-well polystyrene microtiter plate with 95 μL of water and 200 μL of anthrone reagent (2 mg anthrone mL−1 concentrated H2SO4). The plate was then incubated at 80 °C for 30 min, cooled down and the absorbance measured at 625 nm in a microplate reader (Eon™ High-Performance Microplate Spectro- photometer, BioTek). Cellulose content was determined in triplicate for each sample. Comprehensive microarray polymer profiling (CoMPP) The analysis was performed essentially accord- ing to the method reported by [38, 39]. Briefly, 10 mg of AIR was first treated with 50 mM trans- 1,2-diaminocyclohexane-N,N,N′,N′-tetraacetic acid (CDTA), pH 7.5 for 2 h. The supernatant was collected as a pectin-enriched fraction and the remaining pellet was treated with 4 M sodium hydroxide (NaOH) con- taining 0.1% (v/v) sodium borohydride for 2 h to obtain a hemicellulose-rich fraction. Each extraction was done in triplicates and pooled to one sample. Immunolocalization and phloroglucinol stainingi p g g Approx. 1 cm long pieces of 70% ethanol-fixed internodes harvested at the ripening stage of growth were dissected, rehydrated and embedded in 8% agarose in Eppen- dorf tubes. The blocks were glued on a metallic holder and 100 μm thick sections were generated using the vibratome Leica VT1000 s. The sections were collected in PBS (140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, and 1.7 mM KH2PO4, pH 7.5) in small sieves in 24-microw- ell plate. The sections were blocked by 5% milk in PBS for 30 min, incubated for 1 h with the primary antibody (LM12, PlantProbes) at room temperature in a solution of 5% milk in PBS at the 1:10 dilution, washed two times with PBS and incubated for 1 h in anti-rat secondary antibody conjugated with AlexaFluor488 (Invitrogen) at 1:500 dilution. The sections were then washed two times with PBS stained with Calcofluor (0.1 mg mL−1 in PBS), washed and mounted on the glass slide in CITIFLUOR Labelling of silica surfaces in phytoliths Each sample was mixed 50/50 with printing buffer (55.2% glycerol, 44% water, 0.8% Triton X-100) and spotted on a nitro- cellulose membrane, pore size of 0.45 μm (Whatman, Maidstone, UK) using an Arrayjet Sprint (Arrayjet, Ros- lin, UK) with two technical replicates and three fivefold dilutions giving 8 spots per sample. The arrays were blocked with phosphate-buffered saline (PBS) contain- ing 5% (w/v) fat-free milk powder, followed by incu- bation with primary monoclonal antibodies (mAbs) or carbohydrate binding modules (CBMs) for 2 h. The probes used in this study are specific for plant cell wall polymers and listed in Additional file 1: Table S1. After washing with PBS, the arrays were probed with a secondary antibodies conjugated with alkaline phos- phatase for 2 h. The arrays were then washed and devel- oped using a 5-bromo-4-chloro-3-indolyl-phosphate (BCIP)/nitro-blue tetrazolium chloride (NBT) sub- strate and scanned using a flatbed scanner (CanoScan 9000 Mark II, Canon, Søborg, Denmark) at 2400 dpi. The signal strength was quantified using Array-Pro Analyzer 6.3 (Media Cybernetics, Rockville, MD, USA) and converted into a heatmap [38]. Quantification of (1→3;1→4)‑β‑d‑glucan The amount of (1→3;1→4)-β-d-glucan (mixed-link- age glucan, MLG) in samples was determined by the release of (1→3;1→4)-β-d-glucan oligomers after enzy- matic digestion with lichenase, followed by hydrolysis with β-glucosidase. The d-glucose produced was then quantified using a glucose oxidase/peroxidase reagent. The β-glucan (Mixed Linkage) assay kit (Megazyme, Wicklow, Ireland) was used and the quantification of (1→3;1→4)-β-d-glucan was performed as described in the manufacturer’s instructions, except that the method was scaled down to 5 mg of dry AIR. Ground barley and oat flour, provided by the manufacturer, served as refer- ence material. Mixed-linkage glucan content was deter- mined in triplicate for each sample. Methylesterification degree of uronic acidshi The degree of methylesterification was analysed by saponification of AIR followed by enzymatic oxidation of methanol released by alcohol oxidase, as described by [40, 41] with modifications. Briefly, de-starched AIR Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 5 of 18 (Agar Scientific) mounting media. The sections were scanned using a Leica SP5 laser scanning confocal micro- scope (LSCM) equipped with Ar and HeNe lasers. All sections were scanned with the same settings. (Agar Scientific) mounting media. The sections were scanned using a Leica SP5 laser scanning confocal micro- scope (LSCM) equipped with Ar and HeNe lasers. All sections were scanned with the same settings. and co-elution was evaluated by deconvolution of the TIC in Agilent Masshunter Unknowns Analysis. All com- pounds used for calculating monolignol and hydroxy- cinnamate ratios were identified by authentic standards or published mass spectra [44]. The compounds were grouped according to methoxylation into H, G or S. Monolignol ratios were calculated as the peak area of the specific monolignol in proportion to the total peak area of the three monolignols. Analogously, the hydroxycinna- mate ratio was calculated. The saturated solution of phluoroglucinol (Sigma) in 20% HCl was used to stain lignin (Wiesner stain). The sections were placed in the drop of solution, incubated app. 5 min and directly observed using a Olympus BX41 light microscope. Measurement of monolignol ratios and hydroxycinnamates by analytical pyrolysish The straw or entire shoots of mature plants were pyro- lyzed in duplicates in random order. The samples were prepared by transferring about 100–200 µg of AIR to a solvent- and flame-cleaned pyrolysis tube. Pyrolysis was performed at 500 °C (calibrated as sample received tem- perature). The pyrolysis temperature was held for 20 s by the automated pyrolysis unit (PYRO, GERSTEL) under a He flow of 51 mL min−1. The sample and pyrolysis unit were heated from 40 to 300 °C (720 °C min−1, held 0.1 min) prior to pyrolysis and maintained at 300 °C for 1 min after pyrolysis by a thermal desorption unit (TDU, GERSTEL). The transfer line was held at 320 °C and pyro- lysates were carried onto the column with a 50:1 split in the programmed temperature vaporizer inlet (CIS4, GERSTEL) held at 300 °C. The pyrolysates were sepa- rated and detected by an Agilent GC–MS (7890B GC, 5977A MSD) equipped with an Agilent VF-5 ms (60 m, 0.25 mm, 0.25 µm) column with 5 m integrated guard column. The oven program was 40 °C (5 min), to 100 °C (10 °C min−1), then to 280 °C (4 °C min−1), and finally to 320 °C (10 °C min−1, 10 min), giving a total run time of 70 min. The MS was scanning the 35–500 m/z range. Only total ion chromatogram (TIC) peak areas of com- pounds with no or insignificant co-elution were used in calculations. The TIC peak areas were integrated using Agilent Masshunter Quantitative Analysis (ver. B.07.00) Pretreatments and enzymatic saccharificationi Pretreatments and enzymatic saccharification were per- formed on mature straw and entire shoot samples, fol- lowing the method described by [39]. Dry material was ground and distributed using an automated sample preparation robotic system (Labman Automation Ltd. at Stokesley, North Yorkshire, UK) followed by microscale pressurized heat treatment, mimicking a hydrothermal pretreatment (in-house build system). Briefly, 27 mg of the dried samples was dispensed into a 96-well alumin- ium plate and 50 mM sodium citrate buffer (pH 4.8) was added to each sample. The plates were sealed with Teflon tape with a little hole above each well, placed on a heating block and further sealed with a thin aluminium plate and a Teflon plate to ensure a gas tight enclosure. The plates were heated to 120 or 190 °C at a stable pressure and incubated for 15 min and 10 min, respectively. Thereafter, the system was cooled down to room temperature prior to enzyme addition. The pretreated material was mixed with the cellulase preparation Cellic Ctec2 (10 FPU g−1 biomass; Novozymes, Bagsværd, Denmark), placed in a plate shaker at 50 °C, 450–600 rpm, and hydrolyzed for 50 h. The hydrolyzed samples were filtered through 0.45 μm filter plates (Pall Corporation, Ann Arbor, MI) and 100 μL of each filtrate was mixed with 100 μL of 5 mM H2SO4 for sugar determination and quantifica- tion using high-performance liquid chromatography (HPLC). Separation and determination of the glucose and xylose in the filtrates were carried out by an Ultimate 3000 HPLC (Dionex, Germering, Germany) equipped with a refractive index detector (Shodex, Tokyo, Japan). The separation was performed in a Phenomenex Rezex ROA column (Phenomenex Inc., Torrance, CA, USA) at 80 °C with 5 mM H2SO4 as eluent at a flow rate of 0.6 mL min−1. Measurement of acetyl bromide ligninh The concentration of acetyl bromide lignin was deter- mined according to the method described by [42]. The AIR samples (~ 20 mg) were incubated for 4 h at 50 °C in 2.5 mL 25% acetyl bromide (v/v) (Sigma) in glacial ace- tic acid. After complete digestion, samples were cooled down and cleared by centrifugation. Then, 30 μL of supernatant, 40 μL 1.5 M NaOH, 30 μL 0.5 M hydroxy- lamine and 150 μL glacial acetic acid were loaded into a quartz microplate. The lignin concentration was calcu- lated from the UV absorbance at 280 nm using a molar extinction coefficient of 18.126 g−1 L cm−1 [43] and a 0.6345 cm path length. The lignin analysis was performed in triplicate for each sample. Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plantsh The low-silicon 1 mutant (Bdlsi1-1) plants used in this study carry a mutation in the Si influx transporter Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 6 of 18 Si in the stems was present in the non-phytolith form (Table 1). (BdLSI1), localized in the roots. The mutation severely hinders Si uptake and resulted in 93% reduction of the shoot Si content [29]. No phenotypic consequences were observed as the morphology and development of the Bdlsi1-1 plants were similar to the wild-type plants [29]. However, the average weight of the mature de-hulled seeds was reduced in the mutant plants despite the fact that the total yield of spikelets was comparable to that of the wild type [29]. (BdLSI1), localized in the roots. The mutation severely hinders Si uptake and resulted in 93% reduction of the shoot Si content [29]. No phenotypic consequences were observed as the morphology and development of the Bdlsi1-1 plants were similar to the wild-type plants [29]. However, the average weight of the mature de-hulled seeds was reduced in the mutant plants despite the fact that the total yield of spikelets was comparable to that of the wild type [29]. The concentration of Si associated with the cell walls (i.e. non-phytolith Si) in the straw and entire shoots of wild-type and mutant plants increased between the ripening growth stage and maturity (Fig. 1a, b). The increase in the straw was comparable between the wild type and the mutant (53 and 65%, respectively) (Fig. 1a, b). The concentration of Si associated with cell walls in the entire shoots of the wild type was 32% higher at maturity than at the ripening growth stage, while that in the mutant increased more than twofold between the rip- ening stage and maturity. First, we harvested wild-type and mutant plants at the ripening growth stage and at maturity to quantify non-phytolith Si. At the ripening stage, the sampled plants were subdivided into (i) leaf blades and sheaths, (ii) stems, and (iii) spikelets (flowers, bracts and seeds), whereas at maturity either entire shoots (all above- ground organs) or straw only (leaves, leaf sheaths and stems) were collected. Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants Microwave-assisted acid diges- tion of the wild-type and Bdlsi1-1 plant material resulted in insoluble phytoliths that were isolated by filtration and further labelled with a silica-specific fluorescent dye (Additional file 2: Figure S1). The phytoliths varied in degree of silicification as indicated by the intensity of fluorescence (Additional file 2: Figure S1). The wild-type material contained substantially more phytoliths in all the analysed organs than did the mutant (Additional file 2: Figure S1). The silicon present in the filtrate after acid digestion represented the pool of amorphous Si and Si chemically bound with cell wall polysaccharides, together designated as non-phytolith Si. This pool accounted in all cases for less than 0.2% of the weight of cell walls (Fig. 1a, b). At the ripening stage, the Si concentrations in the leaves and spikelets of the Bdlsi1-1 plants were two and threefold lower, respectively, than in the wild type (Fig. 1a). Despite the overall reduction of non-phytolith Si in the Bdlsi1-1 plants, the distribution pattern between the organs was comparable (Table 1). Both the wild-type and the mutant plants deposited approximately 40% of the shoot Si content in the spikelets and 50–60% in the leaves (Table 1). Significant differences between the wild type and the mutant became evident when comparing the proportion of Si not allocated to phytoliths (Table 1). In the wild-type plants, more than 90% of Si in the entire shoot was associated with phytoliths, while the mutant plants only allocated about 50% of their Si to phytoliths (Table 1). The organ having the largest proportion of Si associated with the cell walls, i.e. non-phytolith Si, in the wild-type plants was the stem. Here, 20% of the Si con- tent was associated with the cell walls against only 11 and 6% in the spikelets and leaves, respectively (Table 1). The mutant plants deviated extensively from this pattern by containing more than 50% of the Si in leaves and spike- lets as amorphous or cell wall-bound, while none of the Apart from Si, boron (B) and calcium (Ca) were also quantified as these elements play important structural roles within the cell wall by interacting with pectins. Boron links two molecules of rhamnogalacturonans II (RG-II), whereas Ca forms bridges between non-meth- ylesterified regions of homogalacturonans (HG) [21, 22]. Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants Differences in B concentration between the wild type and the mutant were observed only at the ripening stage, where the concentrations in leaves and spikelets of the Bdlsi1-1 plants were 70 and 60% lower, respectively (Fig. 1c). Calcium concentrations in ripening plants were comparable between the wild type and the mutant in all tested samples, except for the stems, where the mutant had 17% lower Ca concentration than the wild type (Fig. 1e). At maturity, the mutant had 5 and 67% higher Ca concentration in the straw and entire shoots, respec- tively, than the wild type (Fig. 1f). The Bdlsi1‑1 mutant exhibits a range of alterations in the composition of non‑cellulosic polysaccharides at both the ripening stage of growth and at maturity Hemicelluloses and pectins constitute a large part of the cell wall structure and form together a fraction of non- cellulosic polysaccharides [45, 46]. Here, we character- ized the composition of monosaccharides associated with pectins (rhamnose, galacturonic acid and galactose) and hemicelluloses (arabinose, glucose, xylose and glucuronic acid) that were released from the cell walls following hydrolysis with trifluoroacetic acid (TFA). The analy- sis revealed compositional variations among the organs, growth stages as well as alterations between the wild type and the mutant (Fig. 2).h The concentrations of arabinose and galacturonic acid in cell walls of the leaves were significantly lower in the Bdlsi1-1 mutant plants than in the wild type (Fig. 2a). The same was the case for galactose and glucose in the spike- lets (Fig. 2c). Rhamnose was the least abundant monosac- charide detected in the cell walls of B. distachyon (below Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 7 of 18 Fig. 1 Concentrations of structural mineral elements in cell wall material of wild type (WT) and Bdlsi1-1 mutant plants. Concentrations of non-phytolith silicon, Si (a, b); boron, B (c, d) and calcium, Ca (e, f) in cell walls isolated from different organs of wild-type and Bdlsi1-1 plants growing in soil at 19 °C and 16 h photoperiod in a greenhouse. Material was harvested at two growth stages: (i) ripening (a, c, e) and (ii) maturity (senescence) (b, d, f). At the ripening stage, plants were subdivided into leaves with sheaths, stems and spikelets (flowers, seeds and bracts); data obtained for individual organs were used to calculate element concentrations in straw and entire shoots at the ripening stage (a, c, e). Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants At maturity, either entire shoots (all vegetative and generative above-ground organs) or straw (leaves, leaf sheaths and stems) were harvested. Concentrations were measured by ICP-OES after microwave digestion with acids. Values are mean ± SD of three replicates. Asterisks indicate significant differences between the wild type and the mutant (p ≤ 0.05; p ≤ 0.01; p ≤ 0.001) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test Fig. 1 Concentrations of structural mineral elements in cell wall material of wild type (WT) and Bdlsi1-1 mutant plants. Concentrations of non-phytolith silicon, Si (a, b); boron, B (c, d) and calcium, Ca (e, f) in cell walls isolated from different organs of wild-type and Bdlsi1-1 plants growing in soil at 19 °C and 16 h photoperiod in a greenhouse. Material was harvested at two growth stages: (i) ripening (a, c, e) and (ii) maturity (senescence) (b, d, f). At the ripening stage, plants were subdivided into leaves with sheaths, stems and spikelets (flowers, seeds and bracts); data obtained for individual organs were used to calculate element concentrations in straw and entire shoots at the ripening stage (a, c, e). At maturity, either entire shoots (all vegetative and generative above-ground organs) or straw (leaves, leaf sheaths and stems) were harvested. Concentrations were measured by ICP-OES after microwave digestion with acids. Values are mean ± SD of three replicates. Asterisks indicate significant differences between the wild type and the mutant (p ≤ 0.05; p ≤ 0.01; p ≤ 0.001) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test ntrations of structural mineral elements in cell wall material of wild type (WT) and Bdlsi1-1 mutant plants. Concentrations of tions of structural mineral elements in cell wall material of wild type (WT) and Bdlsi1-1 mutant plants. Concentrations of on, Si (a, b); boron, B (c, d) and calcium, Ca (e, f) in cell walls isolated from different organs of wild-type and Bdlsi1-1 plants Fig. 1 Concentrations of structural mineral elements in cell wall material of wild type (WT) and Bdlsi1-1 mutant plants. Concentrations of non-phytolith silicon, Si (a, b); boron, B (c, d) and calcium, Ca (e, f) in cell walls isolated from different organs of wild-type and Bdlsi1-1 plants growing in soil at 19 °C and 16 h photoperiod in a greenhouse. Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants The standard deviation of the mean values (n = 3) used in the calculations did not exceed 30% a Samples in which Si was below the detection limit (0.01 mg g−1 CW) Organ Plant Total Si (mg plant−1) Si (% of total Si plant−1) Total Si in cell walls (mg plant−1) Si (% of total Si in CW plant−1) Ratio of Si in CW to total Si (%) Leaves WT 40.5 63 2.7 48 6 Bdlsi1-1 2.7 51 1.5 58 56 Stems WT 3.6 6 0.7 13 20 Bdlsi1-1 0.7 13 < 0.01a < 1 Spikelets WT 20.0 31 2.1 39 11 Bdlsi1-1 1.9 36 1.1 42 57 Entire shoot WT 64.1 100 5.5 100 9 Bdlsi1-1 5.3 100 2.6 100 49 Table 1 Silicon distribution among organs and their cell walls in wild-type and mutant plant harvested at the ripening stage O Pl t T t l Si Si (% f t t l T t l Si i ll ll Si (% f t t l Si R ti f Si ution among organs and their cell walls in wild-type and mutant plant harvested at the ripening d their cell walls in wild-type and mutant plant harvested at the ripening Data show the average Si content in different organs of wild-type and mutant plants and the proportion of this Si present in non-phytolith form in cell wall material. The standard deviation of the mean values (n = 3) used in the calculations did not exceed 30% a Samples in which Si was below the detection limit (0 01 mg g−1 CW) gf g yp p p p p p y The standard deviation of the mean values (n = 3) used in the calculations did not exceed 30% a Samples in which Si was below the detection limit (0.01 mg g−1 CW) The standard deviation of the mean values (n 3) used in the calculations did not exceed 30% a Samples in which Si was below the detection limit (0.01 mg g−1 CW) Fig. 2 Monosaccharide composition of non-cellulosic polysaccharides in cell walls of wild type (WT) and Bdlsi1-1 plants. Cell walls (CW) were isolated from leaves, stems and spikelets of B. distachyon wild-type and Bdlsi1-1 plants harvested at the ripening stage of growth (a–c) and from straw (d) and entire shoots (e) at maturity. a Samples in which Si was below the detection limit (0.01 mg g−1 CW) Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants Material was harvested at two growth stages: (i) ripening (a, c, e) and (ii) maturity (senescence) (b, d, f). At the ripening stage, plants were subdivided into leaves with sheaths, stems and spikelets (flowers, seeds and bracts); data obtained for individual organs were used to calculate element concentrations in straw and entire shoots at the ripening stage (a, c, e). At maturity, either entire shoots (all vegetative and generative above-ground organs) or straw (leaves, leaf sheaths and stems) were harvested. Concentrations were measured by ICP-OES after microwave digestion with acids. Values are mean ± SD of three replicates. Asterisks indicate significant differences between the wild type and the mutant (p ≤ 0.05; p ≤ 0.01; p ≤ 0.001) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 8 of 18 Table 1 Silicon distribution among organs and their cell walls in wild-type and mutant plant stage Table 1 Silicon distribution among organs and their cell walls in wild-type and mutant plant harvested at the ripening stage Data show the average Si content in different organs of wild-type and mutant plants and the proportion of this Si present in non-phytolith form in cell wall material. Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants Samples were obtained by hydrolyzing CW with 2 M TFA, followed by separation and quantification by HPAEC. Data represent mean values (± SD, n = 5.) asterisks indicate significant differences (p ≤ 0.05; p ≤ 0.01; p ≤ 0.001) between the wild type and the mutant as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test. Rha l-rhamnose, Ara l-arabinose, Gal d-galactose, Glu d-glucose, Xyl d-xylose, GalA d-galacturonic acid, GlcA d-glucuronic acid Fig. 2 Monosaccharide composition of non-cellulosic polysaccharides in cell walls of wild type (WT) and Bdlsi1-1 plants. Cell walls (CW) were isolated from leaves, stems and spikelets of B. distachyon wild-type and Bdlsi1-1 plants harvested at the ripening stage of growth (a–c) and from straw (d) and entire shoots (e) at maturity. Samples were obtained by hydrolyzing CW with 2 M TFA, followed by separation and quantification by HPAEC. Data represent mean values (± SD, n = 5.) asterisks indicate significant differences (p ≤ 0.05; p ≤ 0.01; p ≤ 0.001) between the wild type and the mutant as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test. Rha l-rhamnose, Ara l-arabinose, Gal d-galactose, Glu d-glucose, Xyl d-xylose, GalA d-galacturonic acid, GlcA d-glucuronic acid Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 9 of 18 1% of CW) but the concentrations measured in the stems and spikelets at the ripening stage were nevertheless sig- nificantly lower in the Bdlsi1-1 plants than in the wild type (Fig. 2b, c). At maturity, rhamnose could still be detected, but the concentrations measured in the straw and entire shoots of the Bdlsi1-1 plants were significantly higher than in the wild type, as was also the case for ara- binose and xylose (Fig. 2d, e). Expressed as arabinose to xylose ratio, the mutant had a slightly, but significantly, higher ratio of 0.27 in the straw against 0.22 in the wild type, while in the entire shoots, the values were compa- rable (0.25). The concentration of TFA-extractable glu- cose was substantially reduced in the entire shoots of the mutant plants at maturity (Fig. 2e), but not in the straw (Fig. 2d). This corresponded with a decreased concentra- tion of glucose in mutant spikelets at the ripening stage, while the concentrations in leaves and stems were similar to the wild type (Fig. 2a–c). pattern of methylesterification as well as in the nature of interactions within the cell wall network. The Bdlsi1‑1 mutant displays rearrangements in the degree and pattern of homogalacturonan methylesterification and pattern of homogalacturonan methylesterification To identify the potential sources of alterations in the individual monomers building the non-cellulosic poly- saccharides (Fig. 2), we conducted semi-quantitative comprehensive microarray polymer profiling (CoMPP). This method gives information about extractable cell wall glycans that are released in two consecutive extractions. The first fraction obtained following extraction with cyclohexane diamine tetraacetic acid (CDTA; Fig. 3a) contains predominantly soluble cell wall polymers, espe- cially pectins and proteoglycans, while the second frac- tion, obtained by extraction with sodium hydroxide (NaOH; Fig. 4a), yields less soluble cell wall components, namely hemicelluloses. Concentrations of non‑phytolith silicon and cell wall boron are altered in the Bdlsi1‑1 plants To address this, we measured the degree of methylesterification (DM) of HG using an enzymatic assay. We found that the wild- type and the Bdlsi1-1 plants had comparable DM in all tested samples, apart from the leaves and stems at the ripening stage, where the Bdlsi1-1 had 8% higher and 25% lower DM than the wild type, respectively (Fig. 3b, c). The DM in all the straw samples was comparable between the wild type and mutant and did not change from ripening to maturity (Fig. 3b, c). However, entire shoots of both the Bdlsi1-1 and wild type had about 30% lower DM values at maturity than at the ripening stage of growth (Fig. 3b, c).i We identified the presence of RG-I in B. distachyon samples by means of the mAbs directed against the RG-I backbone (INRA-RU1), 1,5-arabinan (LM6 and LM13) and 1,4-galactan side chains (LM5). The epitopes were detected in all the samples, mostly in the frac- tion extracted with NaOH (Figs. 3a and 4a). In leaves, the abundance of the epitope of the RG-I backbone was much higher in the Bdlsi1-1 samples, while mAbs recog- nizing both branched and linear arabinose chains showed lower signal relative to the wild type (Figs. 3a and 4a). Several epitopes of AGPs and extensins were detected in both the CDTA and NaOH extractable fractions, suggest- ing varying nature and strength of their bonds with the cell wall components. However, there were no differences between the Bdlsi1-1 and the wild type plants (Figs. 3a and 4a). The concentration of (1→3;1→4)‑β‑d‑glucans is strongly reduced in the spikelets of the Bdlsi1‑1 mutant 3 Characterization of the pectin-rich fractions in wild-type (WT) and Bdlsi1-1 plants. Plants were harvested at the ripening stage of growth and at maturity and subdivided as previously described. a Comprehensive microarray polymer profiling (CoMPP) showing the relative abundance of polysaccharide epitopes in the pectin-enriched fraction extracted using 1,2-diaminocyclohexanetetraacetic acid (CDTA). The average signal intensities (n ≥ 2) are presented as a heatmap in which the white colour corresponds to the lowest intensity and red to the highest. The antibody names and their corresponding epitopes are indicated on the Y-axis. Antibodies that did not show signal for all the samples are not included in the heatmap, but listed in Additional file 1: Table S1. The degree of methylesterification of GalA (%) was quantified in the wild type and the Bdlsi1-1 mutant at the ripening stage of growth (b) and at maturity (c). Data represent the mean ± SD, n ≥ 6. All statistical comparisons were made using one-way ANOVA followed by Tukey’s HDS post hoc test, asterisks indicate statistical differences between the wild type and the mutant (p ≤ 0.01; p ≤ 0.001). HG homogalacturonan, RG-I rhamnogalacturonan-I, MLG mixed-linkage glucan, AX arabinoxylan, AGP arabinogalactan protein, GlcA glucuronic acid, FA ferulic acid, HC hemicelluloses plants (Fig. 4c), the latter reflecting the higher MLG concentration in the spikelets of the wild type. in stems and spikelets of the Bdlsi1-1 mutant at the rip- ening stage were 30 and 34% lower than those in the wild type (Fig. 4b). From ripening to maturity, the MLG concentrations in all the samples increased up to sev- eral-fold (Fig. 4b, c). Moreover, at maturity, the MLG concentration was 42% higher in the Bdlsi1-1 straw than in the wild type, while in the entire shoots, the wild type contained 70% more MLG than the Bdlsi1-1 plants (Fig. 4c), the latter reflecting the higher MLG concentration in the spikelets of the wild type. In grasses, arabinose residues of arabinoxylans (AX) can be modified via binding of ferulic acid (FA) that may bridge two adjacent AX molecules, but also cross-link hemicelluloses and lignin [48]. The presence of FA in the samples was detected by the LM12 antibody recogniz- ing epitopes of polymers modified with FA. The signal In grasses, arabinose residues of arabinoxylans (AX) can be modified via binding of ferulic acid (FA) that may bridge two adjacent AX molecules, but also cross-link hemicelluloses and lignin [48]. The concentration of (1→3;1→4)‑β‑d‑glucans is strongly reduced in the spikelets of the Bdlsi1‑1 mutant a Comprehensive microarray polymer profiling (CoMPP) showing the relative abundance of polysaccharide epitopes in the pectin-enriched fraction extracted using 1,2-diaminocyclohexanetetraacetic acid (CDTA). The average signal intensities (n ≥ 2) are presented as a heatmap in which the white colour corresponds to the lowest intensity and red to the highest. The antibody names and their corresponding epitopes are indicated on the Y-axis. Antibodies that did not show signal for all the samples are not included in the heatmap, but listed in Additional file 1: Table S1. The degree of methylesterification of GalA (%) was quantified in the wild type and the Bdlsi1-1 mutant at the ripening stage of growth (b) and at maturity (c). Data represent the mean ± SD, n ≥ 6. All statistical comparisons were made using one-way ANOVA followed by Tukey’s HDS post hoc test, asterisks indicate statistical differences between the wild type and the mutant (p ≤ 0.01; *p ≤ 0.001). HG homogalacturonan, RG-I rhamnogalacturonan-I, MLG mixed-linkage glucan, AX arabinoxylan, AGP arabinogalactan protein, GlcA glucuronic acid, FA ferulic acid, HC hemicelluloses Fig. 3 Characterization of the pectin-rich fractions in wild-type (WT) and Bdlsi1-1 plants. Plants were harvested at the ripening stage of growth and at maturity and subdivided as previously described. a Comprehensive microarray polymer profiling (CoMPP) showing the relative abundance of polysaccharide epitopes in the pectin-enriched fraction extracted using 1,2-diaminocyclohexanetetraacetic acid (CDTA). The average signal intensities (n ≥ 2) are presented as a heatmap in which the white colour corresponds to the lowest intensity and red to the highest. The antibody names and their corresponding epitopes are indicated on the Y-axis. Antibodies that did not show signal for all the samples are not included in the heatmap, but listed in Additional file 1: Table S1. The degree of methylesterification of GalA (%) was quantified in the wild type and the Bdlsi1-1 mutant at the ripening stage of growth (b) and at maturity (c). Data represent the mean ± SD, n ≥ 6. All statistical comparisons were made using one-way ANOVA followed by Tukey’s HDS post hoc test, asterisks indicate statistical differences between the wild type and the mutant (p ≤ 0.01; *p ≤ 0.001). HG homogalacturonan, RG-I rhamnogalacturonan-I, MLG mixed-linkage glucan, AX arabinoxylan, AGP arabinogalactan protein, GlcA glucuronic acid, FA ferulic acid, HC hemicelluloses h f h h f ld d dl l l h d h Fig. The concentration of (1→3;1→4)‑β‑d‑glucans is strongly reduced in the spikelets of the Bdlsi1‑1 mutant As expected, the extraction of B. distachyon cell walls with NaOH after CDTA released primarily hemicellu- loses, namely (1→3;1→4)-β-d-glucans (mixed-linkage glucans, MLG), xylans and xyloglucans (Fig. 4a). The antibody recognizing unsubstituted (1→4)-β-xylan (LM10) showed stronger signal in the Bdlsi1-1 spike- lets at the ripening stage as compared to the wild type, and a similar trend was observed in the entire mature shoots of Bdlsi1-1 plants (Fig. 4a). At the ripening stage, the signal for the MLG epitope was higher in the stems and lower in the spikelets of the Bdlsi1-1 mutant compared to the wild type, whereas at maturity, the abundance was relatively lower in both the straw and entire shoots of Bdlsi1-1 (Fig. 4a). Using a specific assay to quantify MLG, we observed that at the ripening stage MLG was present in much higher concentrations in the spikelets than in the leaves and stems (Fig. 4b) In the latter two organs, the MLG concentration was simi- lar and amounted to only approx. 9–14% of that in the spikelets (Fig. 4b). The MLG concentrations measured In grasses, pectins account for a relatively low propor- tion of the cell walls, but their main constituents, rham- nose and galacturonic acid, were nevertheless detected in all of the analysed B. distachyon organs (Fig. 2). Several pectin epitopes, mainly structures of homogalacturonans (HG) and rhamnogalacturonan-I (RG-I), were also found in material harvested at both growth stages (Figs. 3a and 4a). Homogalacturonan (HG) is a linear homopolymer of α-(1–4)-linked d-galacturonic acid (d-GalA), which can be further methylesterified at the carbon C6 [47]. In our study, HG was detected with a range of monoclonal anti- bodies (mAbs) that recognize subtle differences in meth- ylesterification patterns. At the ripening stage, leaves had the highest relative abundance of HG among the organs, and the intensity of several mAbs recognizing HG was lower in the Bdlsi1-1 leaves compared to the wild type (Fig. 3a). The alterations in the relative abundance of the non- and methylesterified HG epitopes detected with CoMPP might reflect changes in the degree and/or Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 10 of 18 Fig. 3 Characterization of the pectin-rich fractions in wild-type (WT) and Bdlsi1-1 plants. Plants were harvested at the ripening stage of growth and at maturity and subdivided as previously described. The concentration of (1→3;1→4)‑β‑d‑glucans is strongly reduced in the spikelets of the Bdlsi1‑1 mutant Concentrations of mixed-linkage glucans (MLG) were determined in the wild-type and the Bdlsi1-1 plants at the ripening growth stage (b) and at maturity (c) using a lichenase-based assay. Data represent the mean (± SD, n ≥ 4) and asterisks indicate significant differences (p ≤ 0.01; *p ≤ 0.001) between the wild type and the mutant as assessed by one-way ANOVA followed by the Tukey HDS post hoc test. d Immunostaining of the feruloylated polysaccharides with LM12 mAb and secondary anti-rat antibody (conjugated to Alexa Fluor 488; green channel) was performed on sections of internodes collected at the ripening stage. Images are overlaid with signal of Calcofluor counterstain (blue) outlining the cell walls. Arrowhead indicates stronger LM12 signal in parenchymatic cells of Bdlsi1-1 stems. Scale bars: 50 μm. HG homogalacturonan, RG-I rhamnogalacturonan-I, MLG mixed-linkage glucan, XG xyloglucan, AX arabinoxylan, AGP arabinogalactan protein, GlcA glucuronic acid, FA ferulic acid, HC hemicellulose Fig. 4 Characterization of polysaccharides in hemicellulose in wild-type (WT) and Bdlsi1-1 plants. a Comprehensive microarray polymer profiling (CoMPP) showing the relative abundance of polysaccharides in a hemicellulose-enriched fraction extracted with sodium hydroxide (NaOH) from the wild-type and the Bdlsi1-1 plants harvested at the ripening stage and at maturity. Results are presented as a heatmap of the average antibody signal intensity, where the white colour corresponds to the lowest intensity and red to the highest. The antibodies that did not show signal for all the samples tested are not included in the heatmap (Additional file 1: Table S1). Concentrations of mixed-linkage glucans (MLG) were determined in the wild-type and the Bdlsi1-1 plants at the ripening growth stage (b) and at maturity (c) using a lichenase-based assay. Data represent the mean (± SD, n ≥ 4) and asterisks indicate significant differences (p ≤ 0.01; *p ≤ 0.001) between the wild type and the mutant as assessed by one-way ANOVA followed by the Tukey HDS post hoc test. d Immunostaining of the feruloylated polysaccharides with LM12 mAb and secondary anti-rat antibody (conjugated to Alexa Fluor 488; green channel) was performed on sections of internodes collected at the ripening stage. Images are overlaid with signal of Calcofluor counterstain (blue) outlining the cell walls. Arrowhead indicates stronger LM12 signal in parenchymatic cells of Bdlsi1-1 stems. Scale bars: 50 μm. The concentration of (1→3;1→4)‑β‑d‑glucans is strongly reduced in the spikelets of the Bdlsi1‑1 mutant The presence of FA in the samples was detected by the LM12 antibody recogniz- ing epitopes of polymers modified with FA. The signal Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 11 of 18 Fig. 4 Characterization of polysaccharides in hemicellulose in wild-type (WT) and Bdlsi1-1 plants. a Comprehensive microarray polymer profiling (CoMPP) showing the relative abundance of polysaccharides in a hemicellulose-enriched fraction extracted with sodium hydroxide (NaOH) from the wild-type and the Bdlsi1-1 plants harvested at the ripening stage and at maturity. Results are presented as a heatmap of the average antibody signal intensity, where the white colour corresponds to the lowest intensity and red to the highest. The antibodies that did not show signal for all the samples tested are not included in the heatmap (Additional file 1: Table S1). Concentrations of mixed-linkage glucans (MLG) were determined in the wild-type and the Bdlsi1-1 plants at the ripening growth stage (b) and at maturity (c) using a lichenase-based assay. Data represent the mean (± SD, n ≥ 4) and asterisks indicate significant differences (p ≤ 0.01; **p ≤ 0.001) between the wild type and the mutant as assessed by one-way ANOVA followed by the Tukey HDS post hoc test. d Immunostaining of the feruloylated polysaccharides with LM12 mAb and secondary anti-rat antibody (conjugated to Alexa Fluor 488; green channel) was performed on sections of internodes collected at the ripening stage. Images are overlaid with signal of Calcofluor counterstain (blue) outlining the cell walls. Arrowhead indicates stronger LM12 signal in parenchymatic cells of Bdlsi1-1 stems. Scale bars: 50 μm. HG homogalacturonan, RG-I rhamnogalacturonan-I, MLG mixed-linkage glucan, XG xyloglucan, AX arabinoxylan, AGP arabinogalactan protein, GlcA glucuronic acid, FA ferulic acid, HC hemicellulose Fig. 4 Characterization of polysaccharides in hemicellulose in wild-type (WT) and Bdlsi1-1 plants. a Comprehensive microarray polymer profiling (CoMPP) showing the relative abundance of polysaccharides in a hemicellulose-enriched fraction extracted with sodium hydroxide (NaOH) from the wild-type and the Bdlsi1-1 plants harvested at the ripening stage and at maturity. Results are presented as a heatmap of the average antibody signal intensity, where the white colour corresponds to the lowest intensity and red to the highest. The antibodies that did not show signal for all the samples tested are not included in the heatmap (Additional file 1: Table S1). Cellulose level in the straw is slightly elevated in the Bdlsi1‑1 mutant compared to the wild type y Lignin quantification using the acetyl bromide method showed that stems were highly lignified organs in B. distachyon. To compare the distribution of lignin in stems of the Bdlsi1-1 and the wild-type plants, we used histochemical staining with phloroglucinol (Wiesner method) (Fig. 6c). The phloroglucinol stain reacts with the p-hydroxycinnamaldehyde end-groups occurring in native lignin (mainly as coniferyl aldehyde end-groups) to give a reddish-purple colour, resulting in pale yellow staining of regions with low lignification (low quantity of p-hydroxycinnamaldehyde end-groups) and increas- ing orange/red staining as lignification proceeds (high proportion of p-hydroxycinnamaldehyde end-groups) [49]. We detected the highest level of lignification in the xylem cell walls, bundle sheaths and interfascicu- lar fibres (Fig. 6c). Stems of the Bdlsi1-1 mutant, col- lected at the ripening stage displayed increased red-hued staining relative to the wild type (Fig. 6c), thus indicat- ing higher lignin level and/or higher abundance of coniferyl aldehyde end-groups in the lignin. This colour change appeared particularly in the interfascicular region (Fig. 6c). The residual pellet after TFA hydrolysis of cell wall mate- rial consists mostly of intact cellulose fibrils. At the rip- ening stage, the wild type and the mutant plants had comparable concentrations of TFA-resistant cellulose in leaves and stems, while the concentration measured in the Bdlsi1-1 spikelets was 6% higher than that in the wild type (Fig. 5a). The cellulose concentration in the straw at both the ripening growth stage and at maturity was slightly, but significantly higher in the Bdlsi1-1 mutant compared to the wild type (Fig. 5a, b). The same was the case for entire shoots at the ripening stage (Fig. 5a), but not at maturity (Fig. 5b). The concentration of (1→3;1→4)‑β‑d‑glucans is strongly reduced in the spikelets of the Bdlsi1‑1 mutant HG homogalacturonan, RG-I rhamnogalacturonan-I, MLG mixed-linkage glucan, XG xyloglucan, AX arabinoxylan, AGP arabinogalactan protein, GlcA glucuronic acid, FA ferulic acid, HC hemicellulose for the LM12 antibody was detected only in the CDTA fraction, and the relative intensity observed in the mutant leaves and stems sampled at the ripening stage was lower compared to the wild type (Fig. 3a). The analysis of the mature material showed no signal for the LM12 antibody in the straw samples (Fig. 3a). However, an LM12 signal was obtained in the entire shoots, where the wild type showed much higher relative abundance than Bdlsi1-1 (Fig. 3a). To further explore this difference, we performed immunolocalization using the LM12 antibody. While we did not succeed to obtain images for leaves, immunolabe- ling of the stem sections in the wild type revealed a strong fluorescence signal in the lignified sclerenchyma cells and in the xylem vessels (Fig. 4d). The corresponding tissues in the Bdlsi1-1 mutant showed an overall reduction of the LM12 signal although with a relative intensification Page 12 of 18 Page 12 of 18 Głazowska et al. Biotechnol Biofuels (2018) 11:171 of the signal appearing in less lignified parenchymatic cells (Fig. 4d). stage (Fig. 6a, b). On the other hand, the entire shoots of Bdlsi1-1 plants sampled at maturity contained sig- nificantly more lignin than the wild-type plants (Fig. 6b). Thus, the mature plants of the mutant contained relatively more lignin in the spikelets compared with the wild type. The Bdlsi1‑1 mutant exhibits minor changes in lignin composition To study lignin accumulation in the tissues of B. dis- tachyon, we first quantified acetyl bromide soluble lignin. At the ripening stage, the lignin concentrations varied between the organs (Fig. 6a). The highest concentra- tion was measured in the stems, where lignin constituted around 22% of the cell walls (Fig. 6a). The lignin concen- trations in all the organs sampled at the ripening stage of growth were comparable between the wild type and the mutant (Fig. 6a). At maturity, the lignin concentration in the straw was similar for the wild type and mutant and similar to the values in the straw at the ripening growth To measure the lignin composition, we applied ana- lytical pyrolysis (Py-GC/MS) that enables evaluation of pyrolytically labile monolignols and hydroxycinnamic acids. Only the samples harvested at maturity were ana- lysed as plants here have undergone full lignification. G Fig. 5 Concentrations of cellulose across organs and developmental stages in the wild-type and Bdlsi1-1 mutant. The concentration of intact cellulose microfibrils in the TFA-resistant pellet was quantified in the wild-type (WT) and the Bdlsi1-1 plants sampled at the ripening stage (a) and maturity (senescence) (b). Mean values (± SD) of three replicates are presented, asterisks indicate significant differences between the wild type and the mutant (p ≤ 0.05; *p ≤ 0.01) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test Fig. 5 Concentrations of cellulose across organs and developmental stages in the wild-type and Bdlsi1-1 mutant. The concentration of intact cellulose microfibrils in the TFA-resistant pellet was quantified in the wild-type (WT) and the Bdlsi1-1 plants sampled at the ripening stage (a) and maturity (senescence) (b). Mean values (± SD) of three replicates are presented, asterisks indicate significant differences between the wild type and the mutant (p ≤ 0.05; **p ≤ 0.01) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 13 of 18 Fig. 6 Characterization of lignin and hydroxycinnamates in wild type and Bdlsi1-1 mutant plants. The concentrations of acetyl bromide soluble lignin were quantified in the wild-type and the Bdlsi1-1 sampled at the ripening growth stage (a) and maturity (b). Results are presented as mean values (± SD, n ≥ 3), asterisk indicates statistically significant difference between the wild-type and the mutant plants (p ≤ 0.05), as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test. The Bdlsi1‑1 mutant exhibits minor changes in lignin composition c Lignin distribution in transverse sections from wild-type and Bdlsi1-1 internodes sampled at the ripening stage was visualized with phloroglucinol staining (Wiesner method) (colour from light orange to red). Scale bars: 50 μm. Composition of monolignols was characterized in mature straw (d) and entire shoots (e) using analytical pyrolysis GC/MS. Data are presented as monolignol frequency (%), which is interpreted as an abundance of given monolignols in relation to all released monolignols. The composition of hydroxycinnamic acids in mature straw (f) and entire shoots (g) was quantified analogously to the lignin composition. Mean values (± SD, n ≥ 2) are presented. ovb outer vascular bundles, ivb inner vascular bundles, if interfascicular region, bs bundle sheaths, S syringyl unit, G guaiacyl unit, p-hydroxyphenyl unit, FA ferulic acid, p-CA p-coumaric acid, SA sinapic acid Fig. 6 Characterization of lignin and hydroxycinnamates in wild type and Bdlsi1-1 mutant plants. The concentrations of acetyl bromide soluble lignin were quantified in the wild-type and the Bdlsi1-1 sampled at the ripening growth stage (a) and maturity (b). Results are presented as mean values (± SD, n ≥ 3), asterisk indicates statistically significant difference between the wild-type and the mutant plants (p ≤ 0.05), as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test. c Lignin distribution in transverse sections from wild-type and Bdlsi1-1 internodes sampled at the ripening stage was visualized with phloroglucinol staining (Wiesner method) (colour from light orange to red). Scale bars: 50 μm. Composition of monolignols was characterized in mature straw (d) and entire shoots (e) using analytical pyrolysis GC/MS. Data are presented as monolignol frequency (%), which is interpreted as an abundance of given monolignols in relation to all released monolignols. The composition of hydroxycinnamic acids in mature straw (f) and entire shoots (g) was quantified analogously to the lignin composition. Mean values (± SD, n ≥ 2) are presented. ovb outer vascular bundles, ivb inner vascular bundles, if interfascicular region, bs bundle sheaths, S syringyl unit, G guaiacyl unit, p-hydroxyphenyl unit, FA ferulic acid, p-CA p-coumaric acid, SA sinapic acid (FA) and p-coumaric acid (p-CA) monomers account- ing for almost 90% of total hydroxycinnamates released (Fig. 6f, g). units were the most abundant in all the samples, fol- lowed by S units, accounting together for over 90% of all released monolignols (Fig. 6d, e). Small amounts of H units were also detected (Fig. 6d, e). The Bdlsi1‑1 mutant exhibits minor changes in lignin composition Only the straw fraction differed in monolignol composition between the mutant and the wild type (Fig. 6d, e). The wild- type plants had comparable amounts of G and S units in the straw, whereas the Bdlsi1-1 mutant had slightly higher content of S units and lower amount of G units than the wild type (Fig. 6d). These differences were not apparent in the entire shoots due to the influence of spikelets (Fig. 6d, e). All samples had similar pro- files of released hydroxycinnamates with ferulic acid The saccharification efficiency of Bdlsi1‑1 mutant straw is roughly similar to that of the wild type We tested the sugar release following enzymatic hydrolysis of straw and entire shoots of the wild-type and the Bdlsi1-1 plants sampled at maturity (Fig. 7). Three biomass pretreatment conditions were included: (i) no pretreatment, (ii) low severity (120 °C) and (iii) medium severity (190 °C) pretreatment, the latter Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 14 of 18 Page 14 of 18 proportion of Si absorbed by plants ends up within the cell walls. This is, e.g. shown by the fact that cell walls isolated from rice cell suspension culture contained approximately 3 times higher Si concentration than in the whole cells [9]. In the shoots of wild-type B. dis- tachyon harvested at the ripening stage, the majority of Si was present as phytoliths and only 9% of the total Si was non-phytolith Si present as amorphous Si and cell wall- bound-Si (Table 1 and Additional file 2: Figure S1). In contrast, the Bdlsi1-1 mutant plants contained half of the total Si in non-phytolith form and the overall amount of phytoliths was substantially reduced in all organs (Table 1 and Additional file 2: Figure S1). This suggests that dur- ing limited Si availability, Si present in the amorphous form or directly bound with cell walls polysaccharides has priority over the formation of silicified structures. However, it remains unknown if this is mainly a result of a passive co-aggregation or also involves active regu- lation of transport and translocation pathways [52]. Cell wall B status was also affected by mutation in the Bdlsi1- 1 plants; the decrease, however, was less severe than for Si, and the difference disappeared at maturity (Fig. 1c, d). Both elements play an important role during growth and development and can affect the seed quality [29, 53, 54]. Consequently, low B and Si status in plant tissues may affect their allocation to favour deposition in the spikelets upon seed filling, as observed in our study (Table 1 and Additional file 3: Table S2). being typical for industrial hydrothermal pretreatment of straw. As expected, the low severity pretreatment (120 °C) did not improve sugar yield, whereas the more severe pretreatment (190 °C) resulted in approximately twofold higher sugar yields, compared to no pretreat- ment (Fig. 7). Low‑Si status favours Si deposition in cell wall rather than in silicified structures Comprehensive studies of B. distachyon demonstrate its utility in different areas of grass research, among other cell walls and saccharification potential [25, 50, 51]. Cell wall composition and the interactions of Si with cell wall components may play an important role for the value of lignocellulosic biomass for biofuel and biore- finery purposes. The occurrence of cell wall-bound Si is well documented in the literature, although the mecha- nism of deposition remains elusive [8, 9]. A substantial The saccharification efficiency of Bdlsi1‑1 mutant straw is roughly similar to that of the wild type The entire shoots showed a higher release of glucose, but not xylose, as compared to the straw, most likely due to the contribution from starch and MLG in seeds that are more easily hydrolyzed than cel- lulose (Fig. 7). The increase in glucose yield between the straw and the entire shoots was more pronounced in the wild type, particularly when no pretreatment or mild pretreatment was applied (Fig. 7a, b). Pretreat- ment with 190 °C slightly, but significantly, improved both the glucose and xylose release from the Bdlsi1- 1 straw compared to that of the wild type (Fig. 7a, c). Finally, when the entire shoots were hydrolyzed, the release of glucose from the Bdlsi1-1 plants was signifi- cantly reduced at all pretreatment conditions, which corresponds with the decreased concentration of glu- cose in the non-cellulosic fraction originating from breakdown of MLG (Figs. 2e, 4c and 7b). Silicon and cell wall components in straw 5, 6a, b) was not as prominent as in rice (50% in leaves). Since Si constituted < 0.1% of the mass of the cell walls (Fig. 1a, b), the increase in cellu- lose in the cell walls of the mutant was much greater than what could be explained by the relative decrease in Si. composition of the expanding cells via influencing the organization of cellulose microfibrils and interactions with AX [62]. In the present study, the B. distachyon plants generally deposited lower amounts of MLG in the vegetative tissues compared to the seeds, but con- tinued to accumulate MLG in all the organs between the ripening growth stage and maturity (Fig. 4b, c). At the ripening stage, the mutant had a relatively low amount of MLG in the straw but as the plants matured, the concentration exceeded that measured in the wild type. These alterations may affect cell wall properties and arrangement of polymers in the cell wall matrix.h The pattern and degree of substitutions of the xylan backbone may have a great impact on the structure of cell walls as they determine interactions with cellu- lose and other xylans [63]. Our observations suggest increased accumulation of AX in the mature mutant plants, but also an increase in the degree of substitu- tion with Araf (Figs. 2 and 3a). In addition, the lower extractability of FA from the Bdlsi1-1 mutant (Fig. 3a) and the weaker immunolabeling of the mutant stem sections (Fig. 4d) indicate alterations in polymer cross- linking. Dimers of FA intra- and interlink molecules of AX and FA also form cross-links between AX and lignin [48].h In grasses, pectins account for a minor fraction of cell walls, but they greatly affect the integrity and properties of the cell wall network [20, 56]. Our observations sug- gest distinct extractability and organ-specific character- istics of HG and RG-I in the B. distachyon plants (Figs. 3a and 4a). Three different observations, viz. (i) the reduced concentration of GalA (Fig. 2a), (ii) the overall higher total degree of methylesterification (Fig. 3b), and (iii) the lower relative abundance of HG epitopes in the Bdlsi1- 1 leaves (Fig. 3a), suggest that there was a reduction in the level of HG in the mutant leaves and that some HG regions were more methylesterified. These changes (high methylesterification) would favour the hydration and flexibility of cell walls. Silicon and cell wall components in straw Silicon has been suggested to play a structural role in cell walls. In rice cell suspension cultures, Si was detected within cell walls, where it improved the mechanical prop- erties and regeneration of cell walls in protoplasts [9]. Despite numerous evidences for Si involvement in the Fig. 7 Glucose and xylose release from mature straw and entire shoots after enzymatic saccharification. The experiments were based on mature straw (a, c) or entire shoots (b, d) of the wild type (WT) and the Bdlsi1-1 mutant. Three different pretreatment conditions were used, viz. no pretreatment, low severity (120 °C) and medium severity (190 °C) pretreatment. The saccharification efficiency was determined by release of glucose (a, b) and xylose (c, d) from pretreated and hydrolyzed material. Data presented are mean (± SD, n = 3); different letters indicate significant differences between the samples (p ≤ 0.05) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test Fig. 7 Glucose and xylose release from mature straw and entire shoots after enzymatic saccharification. The experiments were based on mature straw (a, c) or entire shoots (b, d) of the wild type (WT) and the Bdlsi1-1 mutant. Three different pretreatment conditions were used, viz. no pretreatment, low severity (120 °C) and medium severity (190 °C) pretreatment. The saccharification efficiency was determined by release of glucose (a, b) and xylose (c, d) from pretreated and hydrolyzed material. Data presented are mean (± SD, n = 3); different letters indicate significant differences between the samples (p ≤ 0.05) as assessed by one-way ANOVA followed by Tukey’s HDS post hoc test Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 15 of 18 Page 15 of 18 cell wall structure, most studies have focused on the func- tional roles of Si in plants exposed to stresses, while only few have looked into the role of Si under optimal growth conditions. Yamamoto et al. [55] reported enhanced sec- ondary cell wall synthesis resulting in increased accu- mulation of cellulose and lignin in rice grown without Si supply. They suggested that the alterations compensated for reduced Si-based rigidity. Conversely, Zhang et al. [11] showed lower content of cellulose, hemicelluloses and lignin in rice grown without Si supply. Elevated cel- lulose concentrations, but not lignin, were observed in the straw of the Bdlsi1-1 mutant grown in soil, but the increase (3–5%; Figs. Silicon and cell wall components in straw On the other hand, the low extractability of HG and the decline in methylesterifica- tion in Bdlsi1-1 stems (Fig. 3a, b) might indicate local cell wall stiffening via the formation of Ca-bridges between de-esterified HG residues, as also observed in response to reduced cellulose concentration in tobacco and Arabi- dopsis thaliana [57, 58]. The distinct changes observed in the pectin structure of the mutant might be a response to the lower Si concentration with different consequences for leaves and stems as part of an organ-specific remodel- ling mechanism compensating for the Si decline. The lignin units are connected through a range of C–O–C or C–C bonds; the composition and type of the linkages may vary greatly and affect the structure and properties of lignin [64]. In our study, however, the lignin fraction of the Bdlsi1-1 mutant showed only minor changes compared to the wild type (Fig. 6). It appeared that the lignin content was not altered, but minor changes in the composition of pyrolytically labile monolignols as well as differences in lignin stain- ing, nevertheless, indicate alterations of the polymer structure. Taken together our observations suggest that a low concentration of Si in the straw of B. distachyon mostly triggers compositional changes of the cell wall network, such as types of linkages and structural rearrangements of polymers, rather than affecting cell wall thickness. Mixed-linkage glucans have been proposed to inter- act with Si in plant cell walls, and possibly act as tem- plate for Si deposition [10, 59]. In rice plants, MLG participated in Si-based cell wall strengthening. Trans- genic rice lines with reduced MLG content did not show differences in the total Si concentration but exhibited reduced mechanical strength and altered Si deposition pattern [10]. In grasses, high levels of MLG have been detected in primary cell walls at the very early stages of development and in the storage endosperm, but also in mature tissues (e.g. rice stems) [27, 60, 61]. In vegeta- tive tissues, MLG were suggested to affect the cell wall Additional files Additional file 1: Table S1. Monoclonal antibodies and carbohydrate binding modules used in the study. Additional file 1: Table S1. Monoclonal antibodies and carbohydrate binding modules used in the study. Additional file 1: Table S1. Monoclonal antibodies and carbohydrate binding modules used in the study. Additional file 2: Figure S1. Phytoliths in Brachypodium distachyon wild type and Bdlsi1-1 mutant. Phytoliths remaining after microwave-assisted acid digestion of plant material (AIR) were recovered from filters, labelled with silica-specific PDMPO dye and observed by light and fluorescence microscopy. Leaves (A, B, C), stems (D, E) and spikelets (F, G) of wild type plants showed high density of silicified structures that varied in the intensity of PDMPO-labelling. In contrast, the leaves (H, I, J), stems (K, L, M) and spikelets (N, O, P) of the mutant plants contained a substantially lower amount of phytoliths. Scale bars: 100 µm. Additional file 3: Table S2. Boron distribution among organs and their cell walls in wild type and mutant plant harvested at the ripening stage. Data show the average B content in organs of dry wild type and mutant plants and the proportion of this B present in cell walls (CW). Boron concentrations were measured by inductively coupled plasma optical emission spectrometry (ICP-OES) (n = 3). The standard deviation of means used in the calculations did not exceed 20% of the mean value. Conclusion In this study, we have characterized the deposition of Si and the cell wall composition in wild-type and low-silicon 1 (Bdlsi1-1) mutant plants of Brachypodium distachyon. We conclude that low Si availability leads to remodelling of the structure and linkages of cell wall components. The most affected cell wall components are pectins, hemicelluloses and lignin. However, the rearrangements of the cell wall network seem not to significantly alter the innate recalcitrance of cell walls towards enzymatic saccharification. Relative to the wild type, the mutant spikelets showed a low Ara/Xyl ratio and an increase in the signal intensity of the antibody LM10 recognizing unsubstituted xylans (Figs. 2c and 4a). These alterations may decrease the solu- bility of AX and favour the formation of hydrogen bonds between xylan chains. At the same time, the concentra- tion of cellulose in the Bdlsi1-1 spikelets at the ripen- ing stage was elevated (Fig. 5a). The B. distachyon seed contains about 6% cellulose in the cell walls distributed across the whole seed although most abundantly in the cell walls of the outer seed layers [60]. In our study, cellu- lose accounted for approx. 25% of the spikelet cell walls, thus, indicating that the majority derived from the bracts surrounding the seed (Fig. 5a). Increased cellulose in the Bdlsi1-1 spikelets might be a result of seed cell wall thick- ening and/or fortification of the bracts in the absence of Si. Abbreviations l h l AIR: alcohol insoluble residue; AX: arabinoxylan; AGP: arabinogalactan protein; B: boron; Ca: calcium; CDTA: 1,2-diaminocyclohexanetetraacetic acid; CoMPP: comprehensive microarray polymer profiling; CW: cell wall; DM: methylesterification degree; DW: dry weight; FA: ferulic acid; GalA: galacturonic acid; GlcA: glucuronic acid; HC: hemicelluloses; HG: homoga- lacturonan; ICP-OES: inductively coupled plasma optical emission spec- trometry; LSI: silicon transport proteins; MLG: mixed-linkage glucans; pCA: p-coumaric acid; PBS: phosphate buffered saline; PDMPO: 2-(4-piridyl)-5-((4- (2-dimethylaminoethylaminocarbamoyl)-methoxyl)phenyl)oxazole; RG-I: rhamnogalacturonan-I; Si: silicon; TFA: trifluoroacetic acid. Silicon, cell wall composition and saccharification efficiency Cell wall polysaccharides are substrates for enzymatic conversion into a range of monomers and oligomers. The heterogeneity of the macromolecules and lignin com- prising the cell walls, their complex organization into a 3D network and the various types of interactions among individual building blocks impose recalcitrance to enzy- matic hydrolysis [64, 67]. The presence of Si is another crucial factor to take into account when considering the use of lignocellulosic biomass for bioenergy and biorefin- ing purposes. Silicon may not only have a negative effect on the durability of equipment used during conversion, but may also decrease the energy output during conver- sion as well as increase the recalcitrance of the biomass [18, 68]. Despite the contrasting Si concentrations and the marked differences in the cell wall network between the wild-type and the Bdlsi1-1 plants, there was only a small improvement of the glucose and xylose yield fol- lowing standard pretreatment and saccharification of the mature straw (Fig. 7a, c). Thus, a reduction of the Si Silicon and cell wall components in seeds Silicon fertilization is commonly used in rice produc- tion as it positively affects seed quality and yield [65]. We have also shown that the availability of exogenously supplied Si affected the formation of seeds in B. distach- yon and that Si defiency caused lower seed weights [29]. Compositional analysis of the mutant cell walls revealed a significant decrease in the MLG concentration in the cell walls of the spikelets (Fig. 4b). The cell walls of the cereal endosperm are characterized by a high abundance of MLG and AX substituted with FA, but the relative Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 16 of 18 supply and the Si concentration within the biomass of B. distachyon seems to have a limited effect on the cell wall recalcitrance. This suggests that loss of Si-mediated cell wall strengthening is compensated for by rearrangements in the cell walls maintaining the inherent recalcitrance of biomass. proportion of AX to MLG varies between plant species [66]. B. distachyon is a wild grass; therefore, it accumu- lates high amounts of cell wall polysaccharides, mainly MLG, in the endosperm and, as opposed to domesticated cereals, only small quantities of starch [60]. In that con- text, MLG were proposed to function as storage poly- saccharides to be remobilized during seed germination. Arabinoxylans were also detected in the seeds of B. dis- tachyon, accounting for approximately 4.7% of the weight of de-hulled seeds [60]. The high arabinose and xylose concentration in the spikelets observed in the present work (Fig. 2c) thus likely represent a contribution from the bracts. References 26. Rancour DM, Marita JM, Hatfield RD. Cell wall composition throughout development for the model grass Brachypodium distachyon. Front Plant Sci. 2012;3:266. 1. Hodson MJ, White PJ, Mead A, Broadley MR. Phylogenetic variation in the silicon composition of plants. Ann Bot. 2005;96:1027–46. 1. Hodson MJ, White PJ, Mead A, Broadley MR. Phylogenetic variation in the silicon composition of plants. Ann Bot. 2005;96:1027–46. 2. Deshmukh R, Bélanger RR, Hartley S. Molecular evolution of aquaporins and silicon influx in plants. Funct Ecol. 2016;30:1277–85. 2. Deshmukh R, Bélanger RR, Hartley S. Molecular evolution of aquaporins and silicon influx in plants. Funct Ecol. 2016;30:1277–85. 27. Christensen U, Alonso-Simon A, Scheller HV, Willats WG, Harholt J. Characterization of the primary cell walls of seedlings of Brachypodium distachyon—a potential model plant for temperate grasses. Phytochem- istry. 2010;71:62–9. and silicon influx in plants. Funct Ecol. 2016;30:1277–85. 3. Montpetit J, Vivancos J, Mitani-Ueno N, Yamaji N, Remus-Borel W, Belzile F, Ma JF, Belanger RR. Cloning, functional characterization and heterologous expression of TaLsi1, a wheat silicon transporter gene. Plant Mol Biol. 2012;79:35–46. 28. Deshmukh RK, Vivancos J, Ramakrishnan G, Guerin V, Carpentier G, Sonah H, Labbe C, Isenring P, Belzile FJ, Belanger RR. A precise spacing between the NPA domains of aquaporins is essential for silicon permeability in plants. Plant J. 2015;83:489–500. 4. Kim SG, Kim KW, Park EW, Choi D. Silicon-induced cell wall fortification of rice leaves: a possible cellular mechanism of enhanced host resistance to blast. Genet Resist. 2002;92:1095–103. 29. Głazowska S, Murozuka E, Persson DP, Castro PH, Schjoerring JK. Silicon affects seed development and leaf macrohair formation in Brachypodium distachyon. Physiol Plant. 2018;163:231–46. https://doi.org/10.1111/ ppl.12675. 5. Kumar S, Soukup M, Elbaum R. Silicification in grasses: variation between different cell types. Front Plant Sci. 2017;8:438. f 6. Law C, Exley C. New insight into silica deposition in horsetail (Equisetum arvense). BMC Plant Biol. 2011;11:112. 30. Vain P, Worland B, Thole V, McKenzie N, Alves SC, Opanowicz M, Fish LJ, Bevan MW, Snape JW. Agrobacterium-mediated transformation of the temperate grass Brachypodium distachyon (genotype Bd21) for T-DNA insertional mutagenesis. Plant Biotechnol J. 2008;6:236–45. 7. Perry CC, Lu Y. Preparation of silicas from silicon complexes: role of cel- lulose in polymerisation and aggregation control. J Chem Soc Faraday Trans. 1992;88:2915–21. 8. Currie HA, Perry CC. Chemical evidence for intrinsic ‘Si’ within Equisetum cell walls. Phytochemistry. 2009;70:2089–95. 31. Alves SC, Worland B, Thole V, Snape JW, Bevan MW, Vain P. Consent for publication Not applicable. 19. Guerriero G, Hausman JF, Legay S. Silicon and the plant extracellular matrix. Front Plant Sci. 2016;7:463. Authors’ contributions SG, LB and JKS conceived the study. SG prepared plant material, carried out the experiments, analysed and interpreted data, produced all figures, and drafted the manuscript. JM performed lignin staining, immunofluorescence microscopy experiments and revised the manuscript. CB developed the HPAEC analysis of sugars. THH carried out quantification of elements in the cell wall using ICP-OES. MMJ performed analytical pyrolysis. JUF processed the CoMPP data. SG and JKS revised and finalized the article with contributions from CF, WGTW and MG. All authors read and approved the final manuscript. Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 17 of 18 Page 17 of 18 Page 17 of 18 Funding g This work was supported by the Innovation Fund Denmark—Grant Num- ber 001-2011-4 (B21st: Biomass for the 21st century) and Grant Number 12-132625 (Bio-Value). 22. Kobayashi M, Matoh T, Azuma J. Two chains of rhamnogalacturonan II are cross-linked by borate-diol ester bonds in higher plant cell walls. Plant Physiol. 1996;110:1017–20. 23. Brkljacic J, Grotewold E, Scholl R, Mockler T, Garvin DF, Vain P, Brutnell T, Sibout R, Bevan M, Budak H, et al. Brachypodium as a model for the grasses: today and the future. Plant Physiol. 2011;157:3–13. Competing interests 16. Guntzer F. Benefits of plant silicon for crops: a review. Agron Sustain Dev. 2011;32:201–13. Author details 1 11. Zhang J, Zou W, Li Y, Feng Y, Zhang H, Wu Z, Tu Y, Wang Y, Cai X, Peng L. Silica distinctively affects cell wall features and lignocellulosic saccharifi- cation with large enhancement on biomass production in rice. Plant Sci. 2015;239:84–91. 1 Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark. 2 Department of Chem- istry and INANO, Aarhus University, Langelandsgade 140, 8000 Aarhus C, Denmark. 3 Department of Geosciences and Natural Resource Management, University of Copenhagen, Rolighedsvej 23, 1958 Frederiksberg, Denmark. 12. He C, Ma J, Wang L. A hemicellulose-bound form of silicon with potential to improve the mechanical properties and regeneration of the cell wall of rice. New Phytol. 2015;206:1051–62. Received: 25 January 2018 Accepted: 8 June 2018 25. Meineke T, Manisseri C, Voigt CA. Phylogeny in defining model plants for lignocellulosic ethanol production: a comparative study of Brachypodium distachyon, wheat, maize, and Miscanthus x giganteus leaf and stem biomass. PLoS ONE. 2014;9:e103580. Ethics approval and consent to participate Not applicable. 20. Vogel J. Unique aspects of the grass cell wall. Curr Opin Plant Biol. 2008;11:301–7. 21. Morris ER, Powell DA, Gidley MJ, Rees DA. Conformations and interac- tions of pectins I. Polymorphism between gel and solid states of calcium polygalacturonate. J Mol Biol. 1982;155:507–16. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. 24. Draper J, Mur LAJ, Jenkins G, Ghosh-Biswas GC, Bablak P, Hasterok R, Rout- ledge APM. Brachypodium distachyon. A new model system for functional genomics in grasses. Plant Physiol. 2001;127:1539–55. Received: 25 January 2018 Accepted: 8 June 2018 33. Fry SC. The growing plant cell wall: chemical and metabolic analysis. Essex: Longman Scientific and Technical; 1988. Availability of data and materials 17. Debona D, Rodrigues FA, Datnoff LE. Silicon’s role in abiotic and biotic plant stresses. Annu Rev Phytopathol. 2017;55:85–107. All data generated and/or analysed during the current study are available from the corresponding author on reasonable request. 18. Le DM, Sørensen HR, Knudsen NO, Meyer AS. Implications of silica on biorefineries—interactions with organic material and mineral elements in grasses. Biofuels Bioprod Biorefin. 2015;9:109–21. Acknowledgements 13. Ma J, Cai H, He C, Zhang W, Wang L. A hemicellulose-bound form of sili- con inhibits cadmium ion uptake in rice (Oryza sativa) cells. New Phytol. 2015;206:1063–74. We would like to thank Søren K. Rasmussen for Brachypodium mutant seeds, Emiko Murozuka for obtaining homozygous mutant plants, Lena Byrgesen, Louise Nancke and Britta Skov for their excellent technical assistance. The authors acknowledge the Center for Advanced Bioimaging (CAB) (University of Copenhagen, Denmark), for providing facility and equipment for sample preparation and collecting imaging data in this study. 14. Leroux O, Leroux F, Mastroberti AA, Santos-Silva F, Van Loo D, Bagniewska- Zadworna A, Van Hoorebeke L, Bals S, Popper ZA, de Araujo Mariath JE. Heterogeneity of silica and glycan-epitope distribution in epidermal idio- blast cell walls in Adiantum raddianum laminae. Planta. 2013;237:1453–64. 15. Ma JF, Takahashi E. Soil, fertilizer, and plant silicon research in Japan. Amsterdam: Elsevier; 2002. Competing interests The authors declare that they have no competing interests. References A protocol for Agrobacterium-mediated transformation of Brachypodium distachyon community standard line Bd21. Nat Protoc. 2009;4:638–49. 9. He C, Wang L, Liu J, Liu X, Li X, Ma J, Lin Y, Xu F. Evidence for ‘silicon’ within the cell walls of suspension-cultured rice cells. New Phytol. 2013;200:700–9. 32. Hong SY, Park JH, Cho SH, Yang MS, Park CM. Phenological growth stages of Brachypodium distachyon: codification and description. Weed Res. 2011;51:612–20. 10. Kido N, Yokoyama R, Yamamoto T, Furukawa J, Iwai H, Satoh S, Nishi- tani K. The matrix polysaccharide (1;3,1;4)-beta-d-glucan is involved in silicon-dependent strengthening of rice cell wall. Plant Cell Physiol. 2015;56:268–76. 33. Fry SC. The growing plant cell wall: chemical and metabolic analysis. Essex: Longman Scientific and Technical; 1988. Głazowska et al. Biotechnol Biofuels (2018) 11:171 Page 18 of 18 Page 18 of 18 34. Fleischer A, O’Neill MA, Ehwald R. The pore size of non-graminaceous plant cell walls is rapidly decreased by borate ester cross-linking of the pectic polysaccharide rhamnogalacturonan II. Plant Physiol. 1999;121:829–38. wall architecture of isoxaben-habituated Arabidopsis suspension- cultured cells: global transcript profiling and cellular analysis. Plant J. 2004;40:260–75. 59. Fry SC, Nesselrode BH, Miller JG, Mewburn BR. Mixed-linkage (1→3,1→4)-beta-d-glucan is a major hemicellulose of Equisetum (horse- tail) cell walls. New Phytol. 2008;179:104–15. 35. Shimizu K, Del Amo Y, Brzezinski MA, Stucky GD, Morse DE. A novel fluorescent silica tracer for biological silicification studies. Chem Biol. 2001;8:1051–60. 60. Guillon F, Bouchet B, Jamme F, Robert P, Quemener B, Barron C, Larre C, Dumas P, Saulnier L. Brachypodium distachyon grain: characterization of endosperm cell walls. J Exp Bot. 2011;62:1001–15. 36. Foster CE, Martin TM, Pauly M. Comprehensive compositional analysis of plant cell walls (lignocellulosic biomass) part II: carbohydrates. J Vis Exp. 2010. https://doi.org/10.3791/1837. 61. Vega-Sanchez ME, Verhertbruggen Y, Christensen U, Chen X, Sharma V, Varanasi P, Jobling SA, Talbot M, White RG, Joo M, et al. Loss of cellulose synthase-like F6 function affects mixed-linkage glucan deposition, cell wall mechanical properties, and defense responses in vegetative tissues of rice. Plant Physiol. 2012;159:56–69. p g 37. Updegraff DM. Semi-micro determination of cellulose in biological mate- rials. Anal Biochem. 1969;32:420–4. 38. Moller I, Sorensen I, Bernal AJ, Blaukopf C, Lee K, Obro J, Pettolino F, Rob- erts A, Mikkelsen JD, Knox JP, et al. High-throughput mapping of cell-wall polymers within and between plants using novel microarrays. Plant J. 2007;50:1118–28. 62. References Smith-Moritz AM, Hao Z, Fernandez-Nino SG, Fangel JU, Verhertbruggen Y, Holman HY, Willats WG, Ronald PC, Scheller HV, Heazlewood JL, et al. Structural characterization of a mixed-linkage glucan deficient mutant reveals alteration in cellulose microfibril orientation in rice coleoptile mesophyll cell walls. Front Plant Sci. 2015;6:628. 39. Zhang H, Fangel JU, Willats WGT, Selig MJ, Lindedam J, Jørgensen H, Felby C. Assessment of leaf/stem ratio in wheat straw feedstock and impact on enzymatic conversion. GCB Bioenergy. 2014;6:90–6. 40. Klavons JA, Bennett RD. Determination of methanol using alcohol oxi- dase and its application to methyl ester content of pectins. J Agric Food Chem. 1986;34:597–9. 63. Ebringerová A, Heinze T. Xylan and xylan derivatives—biopolymers with valuable properties, 1—naturally occurring xylans structures, procedures and properties. Macromol Rapid Commun. 2000;21:542–56. 41. Lionetti V, Giancaspro A, Fabri E, Giove SL, Reem N, Zabotina OA, Blanco A, Gadaleta A, Bellincampi D. Cell wall traits as potential resources to improve resistance of durum wheat against Fusarium graminearum. BMC Plant Biol. 2015;15:6. 64. Li M, Pu Y, Ragauskas AJ. Current understanding of the correlation of lignin structure with biomass recalcitrance. Front Chem. 2016;4:45. 65. Lavinsky AO, Detmann KC, Reis JV, Avila RT, Sanglard ML, Pereira LF, Sanglard LM, Rodrigues FA, Araujo WL, DaMatta FM. Silicon improves rice grain yield and photosynthesis specifically when supplied during the reproductive growth stage. J Plant Physiol. 2016;206:125–32. 42. Hatfield RD, Grabber J, Brei K. Using the acetyl bromide assay to deter- mine lignin concentrations in herbaceous plants: some cautionary notes. J Agric Food Chem. 1999;47:628–32. 66. Fincher GB, Stone BA. Cell walls and their components in cereal grain technology. Adv Cereal Sci Technol. 1986;8:207–95. 66. Fincher GB, Stone BA. Cell walls and their component 43. Fukushima RS, Hatfield RD. Comparison of the acetyl bromide spectro- photometric method with other analytical lignin methods for deter- mining lignin concentration in forage samples. J Agric Food Chem. 2004;52:3713–20. technology. Adv Cereal Sci Technol. 1986;8:207–95. 67. Xu N, Zhang W, Ren S, Liu F, Zhao C, Liao H, Xu Z, Huang J, Li Q, Tu Y, et al. Hemicelluloses negatively affect lignocellulose crystallinity for high bio- mass digestibility under NaOH and H2SO4 pretreatments in Miscanthus. Biotechnol Biofuels. 2012;5:58. 44. Ralph J, Hatfield RD. Pyrolysis-GC–MS characterization of forage materials. J Agric Food Chem. 1991;39:1426–37. 68. Tanger P, Field JL, Jahn CE, Defoort MW, Leach JE. Biomass for thermo- chemical conversion: targets and challenges. Front Plant Sci. 2013;4:218. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research ? Choose BMC and benefit from: References 45. Somerville C, Bauer S, Brininstool G, Facette M, Hamann T, Milne J, Osborne E, Paredez A, Persson S, Raab T, et al. Toward a systems approach to understanding plant cell walls. Science. 2004;306:2206–11. 46. Burton RA, Gidley MJ, Fincher GB. Heterogeneity in the chemistry, struc- ture and function of plant cell walls. Nat Chem Biol. 2010;6:724–32. p 47. Wolf S, Mouille G, Pelloux J. Homogalacturonan methyl-esterification and plant development. Mol Plant. 2009;2:851–60. 48. Buanafina MMD. Feruloylation in grasses: current and future perspectives. Mol Plant. 2009;2:861–72. 49. Pomar F, Merino F, Barceló AR. O-4-Linked coniferyl and sinapyl aldehydes in lignifying cell walls are the main targets of the Wiesner (phloroglu- cinol-HCl) reaction. Protoplasma. 2002;220:0017–28. 50. Gomez LD, Bristow JK, Statham ER, McQueen-Mason SJ. Analysis of sac- charification in Brachypodium distachyon stems under mild conditions of hydrolysis. Biotechnol Biofuels. 2008;1:15. 50. Gomez LD, Bristow JK, Statham ER, McQueen-Mason SJ. Analysis of sac- charification in Brachypodium distachyon stems under mild conditions of hydrolysis. Biotechnol Biofuels. 2008;1:15. 51. Kellogg EA. Brachypodium distachyon as a genetic model system. Annu Rev Genet. 2015;49:1–20. 51. Kellogg EA. Brachypodium distachyon as a genetic model system. Annu Rev Genet. 2015;49:1–20. 52. Yamaji N, Ma JF. Node-controlled allocation of mineral elements in Poaceae. Curr Opin Plant Biol. 2017;39:18–24. 52. Yamaji N, Ma JF. Node-controlled allocation of mineral elements in Poaceae. Curr Opin Plant Biol. 2017;39:18–24. 53. Bokor B, Ondos S, Vaculik M, Bokorova S, Weidinger M, Lichtscheidl I, Turna J, Lux A. Expression of genes for Si uptake, accumulation, and cor- relation of Si with other elements in ionome of maize kernel. Front Plant Sci. 2017;8:1063. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research ? References Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research ? Choose BMC and benefit from: Ready to submit your research ? Choose BMC and benefit from: 54. Lordkaew S, Konsaeng S, Jongjaidee J, Dell B, Rerkasem B, Jamjod S. Varia- tion in responses to boron in rice. Plant Soil. 2012;363:287–95. 55. Yamamoto T, Nakamura A, Iwai H, Ishii T, Ma JF, Yokoyama R, Nishitani K, Satoh S, Furukawa J. Effect of silicon deficiency on secondary cell wall synthesis in rice leaf. J Plant Res. 2012;125:771–9. 56. Mohnen D. Pectin structure and biosynthesis. Curr Opin Plant Biol. 2008;11:266–77. 57. Burton RA, Gibeaut DM, Bacic A, Findlay K, Roberts K, Hamilton A, Baulcombe BC, Fincher GB. Virus-induced silencing of a plant cellulose synthase gene. Plant Cell. 2000;12:691–705. 58. Manfield IW, Orfila C, McCartney L, Harholt J, Bernal AJ, Scheller HV, Gilmartin PM, Mikkelsen JD, Paul Knox J, Willats WG. Novel cell
https://openalex.org/W2884404516	https://europepmc.org/articles/pmc6051598?pdf=render	English	null	Predictive gravity models of livestock mobility in Mauritania: The effects of supply, demand and cultural factors	PloS one	2,018	cc-by	12,366	RESEARCH ARTICLE Predictive gravity models of livestock mobility in Mauritania: The effects of supply, demand and cultural factors Gae¨lle Nicolas1, Andrea Apolloni2, Caroline Coste3, G. R. William Wint4, Renaud Lancelot2, Marius Gilbert1,3 1 Spatial Epidemiology Lab (SpELL), Universite´ Libre de Bruxelles, Brussels, Belgium, 2 International Center for Agronomic Research and Development, CIRAD, Montpellier, France, 3 Fonds National de la Recherche Scientifique, Brussels, Belgium, 4 Environmental Research Group Oxford (ERGO)—Department of Zoology, University of Oxford, Oxford, United Kingdom gaelle.nicolas6@gmail.com * gaelle.nicolas6@gmail.com * gaelle.nicolas6@gmail.com Editor: Rachata Muneepeerakul, University of Florida, UNITED STATES Editor: Rachata Muneepeerakul, University of Florida, UNITED STATES Received: December 20, 2017 Accepted: June 8, 2018 Published: July 18, 2018 Copyright: © 2018 Nicolas et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Received: December 20, 2017 Accepted: June 8, 2018 Published: July 18, 2018 Copyright: © 2018 Nicolas et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Data are part of a previously published study referred as Apolloni et al. (2018) and available on: https://doi.org/10. 5061/dryad.v4m56. The full citation of the previously published work is: Apolloni A, Nicolas G, Coste C, EL Mamy AB, Yahya B, EL Arbi AS, Baba Gueya M, Baba D, Gilbert M, Lancelot R (2018). Towards the description of livestock mobility in Sahelian Africa: Some results from a survey in Mauritania. PLOS ONE 13(1): e0191565. Abstract Animal movements are typically driven by areas of supply and demand for animal products and by the seasonality of production and demand. As animals can potentially spread infec- tious diseases, disease prevention can benefit from a better understanding of the factors influencing movements patterns in space and time. In Mauritania, an important cultural event, called the Tabaski (Aïd el Kebir) strongly affects timing and structure of movements, and due to the arid and semi-arid climatic conditions, the season can also influence move- ment patterns. In order to better characterize the animal movements patterns, a survey was carried out in 2014, and those data were analysed here using social network analysis (SNA) metrics and used to train predictive gravity models. More specifically, we aimed to contrast the movements structure by ruminant species, season (Tabaski vs. Non-Tabaski) and mode of transport (truck vs. foot). The networks differed according to the species, and to the season, with a changed proportion of truck vs. foot movements. The gravity models were able to predict the probability of a movement link between two locations with moderate to good accuracy (AUC ranging from 0.76 to 0.97), according to species, seasons, and mode of transport, but we failed to predict the traded volume of those trade links. The significant predictor variables of a movement link were the human and sheep population at the source and origin, and the distance separating the locations. Though some improvements would be needed to predict traded volumes and better account for the barriers to mobility, the results provide useful predictions to inform epidemiological models in space and time, and, upon external validation, could be useful to predict movements at a larger regional scale. OPEN ACCESS Citation: Nicolas G, Apolloni A, Coste C, Wint GRW, Lancelot R, Gilbert M (2018) Predictive gravity models of livestock mobility in Mauritania: The effects of supply, demand and cultural factors. PLoS ONE 13(7): e0199547. https://doi.org/ 10.1371/journal.pone.0199547 On this occasion, young rams are slaughtered in most families on the 10th day of the month "dhou al-hija", a religious holiday during the last month of the Muslim (lunar) calendar. The date of this festival changes each year according to the Gregorian (solar) calendar and strongly structures the volume of traded sheep during the year. Annual and seasonal differences are thus observed in the sheep trade flows. The country, and the whole Sahel, was hit by severe droughts in the 1960’s, 70’s and 80’s, and more limited droughts later on till 2017 [15]. The drought of 1970 was the main climatic event for the area. Since then, the area has been mostly in deficit of rainfall [16]. The series of droughts had a profound impact on the livestock (affecting 2/3 of the production) and human population. Cattle population dropped, whilst the population of small ruminants increased, the latter being more robust than the former to harsh climatic conditions. Apolloni et al. [17] described the livestock trading mobility for the year 2014 in Mauritania, highlighting that the main trading peak related to the Tabaski took place between August and December. During this period, the price of male lambs sharply increased, and the high demand strongly affected the trading network structure. In this paper, we aimed to understand how the Tabaski festivity changed the trade network i M it i d t th t f th A ll i t l [17] id d h i Funding: This study was funded by EU grant FP7- 613996 VMERGE and is catalogued by the VMERGE Steering Committee as VMERGE000 (http://www.vmerge.eu). The contents of this publication are the sole responsibility of the authors and don’t necessarily reflect the views of the European Commission. Competing interests: The authors have declared that no competing interests exist. In this paper, we aimed to understand how the Tabaski festivity changed the trade network in Mauritania compared to the rest of the year. Apolloni et al. [17] provided a comprehensive study of the Mauritania survey data, characterizing the seasonal trade network and ruminant flows within Mauritania, and between Mauritania and the neighboring countries. Here, we first provided a complementary description of inner flows using social network analysis indi- cators, by contrasting the Tabaski and non-Tabaski periods, the different ruminant species and modes of transport. Livestock mobility in Mauritania distribution, growth and connectivity of human populations have also been changing rapidly as result of demographic transitions, conflicts or migrations. Similarly, the distribution and connectivity of traded animal populations are strongly influenced by agricultural intensifica- tion and changes in trade patterns. The combined effect of these societal and environmental changes taking place simultaneously is difficult to assess, but some, particularly mobility of livestock and traditional trading practices, have been associated with the emergence and the spread of infectious diseases [1–4], and can have strong socio-economic impacts [5,6]. In addi- tion to these long-term trends, culture and tradition strongly shape societies at national, regional and global scale. For example, human population movement during specific periods such as Chinese Spring Festival [7], annual holidays [8], or religious feast around Christmas, Ramadan, Thanksgiving or Hindu Holy feast are known to cause substantially affect global mobility [9] with significant economic and epidemiological implications [10–12]. Large move- ments of animal populations are also linked to changes in the spatial pattern of food demand, which is anticipated by the market. In low-income countries, such as Sahelian African coun- tries, rapid changes in demand for animal products linked to cultural and traditional events therefore leads to a large number of animals–notably sheep, being slaughtered to meet the seasonal food demand. As a consequence, in the months and weeks preceding such events, trading of live animals is particularly intense. Due to the dry ecoclimate of the Sahelian area, agriculture and breeding strongly depend on the amount of rainfall and the availability of pasture. As a consequence, successive droughts also dramatically affect the Livestock flow. Another example is Madagascar, where kapsile is a traditional practice consisting in an animal barters linked to the labour need in a period when breeders are cash-poor and which strongly affects the patterns of trade flows [3,13]. In West Africa, similar practices, called loans, have been described as a large number of short term animal exchanges for reproduction, food sup- ply through milk production, animal traction, etc [14]. In Mauritania, the Muslim feast of Tabaski or Aïd el-Kebir, is a major cultural and traditional event that strongly influence trad- ing patterns and could have major impact on the spread of diseases. Introduction Many factors that may influence the dynamics and transmission of infectious diseases have been rapidly changing over the last decades. Alongside climate and land use changes, often considered in emerging infectious diseases literature as main drivers, other factors such as the PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 1 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Data collection A survey among veterinary officers was conducted in June 2015 by the National Office for Live- stock Research and Development (ONARDEL) to collect their knowledge of ruminant trade flows as reported in Apolloni et al. [17]. The survey aimed to monitor the movement patterns during the year 2014. Its results were recorded as a series of trade flow events, with the following information: i) the origin, ii) the destination, iii) the type of trade movement (farm, transit, mar- ket), iv) the frequency (annual, monthly, weekly), v) the species (cattle–Bos indicus, sheep, goat or camel—Camelus dromedarius), vi) the number of heads, vii) the date of the starting of the event, viii) the transportation mode (by truck or on foot) and ix) the latitude/longitude coordi- nates of the origin and destination. Three types of movements were recorded: i) between pasto- ral areas for grazing and/or reproduction, ii) from farm to market, iii) or from market to market. Transhumance movements aiming to gradually move herds for suitable pasture areas were not included. The database was cross-checked against sanitary certificates, the scientific documents describing transhumance patterns, and the knowledge of veterinarian staff [17]. Both national and international trade-flows were recorded and the transboundary movements were double-checked through surveys on transit sites between Senegal and Mauritania. Study area Mauritania is situated in the hyper-arid (Sahara) and arid (Sahel) ecozones [18], with low annual rainfall (0–400 mm) concentrated in a short rainy season (June-September). In the northern part of the country, the driest one, only short-cycle plants grow. Livestock, mainly camels and small ruminants, are reared moving around available water points and grazing areas. The southern area, more humid and greener is mostly exploited by transhumant herds. Most of cattle population, being less mobile and demanding more water and nutrients, is con- centrated in the southern area, mainly in the region around the river Senegal. Because of the harsh conditions, mobility is a key aspect of animal rearing in Mauritania. Animals are moved almost continuously among grazing areas to optimize the consumption of good quality nutri- ents. In the absence of slaughterhouses, stocking facilities and road infrastructures, animals are traded alive and butchered at consumption markets. Past droughts indirectly contributed to the growth of cities, in particular Nouakchott, due to the migration of previous farmers and herders from the countryside to urban areas in search of jobs. Because of this, Nouakchott, the capital city has seen its population dramatically exploding during and following the drought years. As of today, almost one quarter of the total population lives in the capital city (National Bureau of statistics http://www.ons.mr/) Maurita- nia is still recovering from the latest food crisis in 2011, affecting almost 1 million of its habi- tants. In 2018, the drought indicators were at the same levels as those of 2012, indicating that the food emergency is not completely over. The inadequate levels of rainfall and the continu- ous threat of the droughts force herders to sell their livestock, in particular small ruminants, due to shortages of suitable feeding areas. Second, we developed predictive models to estimate the probability of a trade connection between two spatial units (areas around markets and/or farms, etc.) based PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 2 / 21 Livestock mobility in Mauritania on their potential production and demand characteristics and different measures of the cost distance between them. on their potential production and demand characteristics and different measures of the cost distance between them. Additional data In 2013, the Ministry for Rural Development and Environment reported a population of 16.8 million sheep and goats, 1.8 million cattle, and 1.4 million camels for Mauritania to FAOSTAT 3 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania Fig 1. Population of human and ruminants in Mauritania. The size of the circle are proportional to the number of head located within the Voronoï polygon related to the spatial location of the trading network. https://doi.org/10.1371/journal.pone.0199547.g001 Fig 1. Population of human and ruminants in Mauritania. The size of the circle are proportional to the number of head located within the Voronoï polygon related to the spatial location of the trading network. https://doi.org/10.1371/journal.pone.0199547.g001 https://doi.org/10.1371/journal.pone.0199547.g001 [19]. Accordingly, and because no finer data was available at national level, we used the most recent version of the Gridded Livestock of the World database (GLW), where the subnational livestock statistics for Mauritania dates back to 2007, and were adjusted to match the FAO- STAT 2010 national totals [20,21]. The WorldPop database was used for the human popula- tion [22]. Both databases were aggregated at a spatial resolution of 0.083333 decimal degrees (i.e. approximately 10 km at the equator) (Fig 1). To estimate the cost paths potentially affect- ing mobility between different localities, we considered two main data sources: the friction layer of Nelson accessibility map, which quantifies the time needed to travel through each pixel [23], and the elevation from the GTOPO30 database (https://lta.cr.usgs.gov/GTOPO30). PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Analysis In this study, we only considered inner movements of cattle, camels, sheep and goats within Mauritania. Social network analysis (SNA) [24] have proved to be of significant interest in animals movements analyses in the past decade [25–28]. Here, it was first used to describe the trade networks according to the species, the transport modality (using truck vs. walking), and the season (Tabaski vs. non-Tabaski). These mentioned periods were defined as strongly influenc- ing the livestock flow within the country by Apolloni et al. [17] who investigated the dataset regarding the occurrence of the Muslim festival for the year 2014. A set of network parameters were estimated for the different networks using the simplified definition provided in Wasser- man and Faust, (1994) [24]: PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 4 / 21 Livestock mobility in Mauritania • Diameter: a network-level parameter representing the greatest number of links in the short- est path between two nodes. • Average path length: a network-level parameter measuring the average number of steps along the shortest paths of all possible nodes pairs, i.e. the average number of nodes an actor has to trade through to connect to any other node. • The clustering coefficient: a node-level parameter of the density of local ties. It measures the probability that neighboring nodes of a node are connected. • The density: a network-level parameter measuring the proportion of observed links among the possible links between nodes, and indicates how strongly a network is connected. • Average degree: a network-level parameter quantifying the average number of links con- nected with a node in a network. Besides these global measures, other centrality measures highlighting the prominent role of nodes in the network were considered: node’s in- and out-degree (the number of connection towards and from each node), node’s in and out- weight (the volume of animal towards and from each node); node’s betweenness (the num- ber of shortest path passing through the node) and node’s eigenvector centrality (scoring the importance of nodes). Network’s vulnerability to target removal of nodes based on centrality measures and esti- mates of the size of the largest connected component were tested. The removal of specific nodes, and their links, cause the network to fragment in a set of smaller subnetworks. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Analysis Eq (2) can be rewritten as a model (3) with an intercept (a flow that would still be pres- ent when populations are equal to zero), and a set of predictor variables xk characterizing the origin or destination, with their associated coefficients βk. MIGij ¼ pa i pb j dg ij ð1Þ ð1Þ logðMIGijÞ ¼ api þ bpj gdij ð2Þ GijÞ ¼ b0 þ b1pi þ b2pj þ b3dij þ PK k 4bkxk ð3Þ logðMIGijÞ ¼ api þ bpj gdij ð2Þ ð2Þ logðMIGijÞ ¼ b0 þ b1pi þ b2pj þ b3dij þ PK k¼4bkxk ð3Þ ð3Þ In this study, we first aimed to estimate the probability of a trade connection between two nodes. Therefore, log(MIGij) was replaced with the logit of this probability and logistic regres- sion was used to estimate the coefficients. As the response, all pairs of connected nodes were coded with 1, and all other pairs of nodes were coded with 0. The analysis was split according to the main structuring factors of the networks, i.e. species, transport modality and season. For each sub-model, we tested seven combinations of predictors, considering them both at the ori- gin and destination with inclusion of a distance estimator ((i) great-circle distance, (ii) cost- path distance based on accessibility friction surface or (iii) cost-path distance based on eleva- tion friction surface). The different combinations of predictors that were tested in the models are shown in Table 1. The human and animal populations at the origin and destination were extracted from the Worldpop and GLW raster layers within the Thiessens’ polygons around each node (Fig 1). Thiessen’s polygons represent areas consisting of all points closer to the node than to any other node. These were used because the nodes did not correspond to any particular administrative division that could have been used (i.e. many nodes per admin unit). Although Thiessen’s polygon can produce somewhat misleading long shapes in desertic areas, this is not necessarily a problem as the livestock and human population demographics would be low in these areas anyway. In each sub-model, we first included the extracted animal popu- lation of each species at both the source and destination (cattle, sheep, goats, and camels). We used stepwise regression based on Akaike information criterion (AIC) to select a more parsi- monious model, with the lowest AIC. Analysis The size of the largest component can be thought as the maximum extent a disease can spread after the implementation of the control measure (vaccination of animal in the areas surrounding the nodes, market closure, etc.) [29]. Volkova et al. (2010) [30] introduced the notion of epidemic threshold (q) in veterinarian epidemiology. This parameter estimates the (minimum) proba- bility for a disease to be transmitted from one node to another to trigger an epidemic. The low- est the epidemic threshold the higher is the risk of an epidemics. This quantity depends on the heterogeneity of the network and the weight’s distribution. In the case of a weighted network the epidemic threshold can be estimated as: q ¼ hwouti hwin wouti Where h i indicates the average value and win, wout indicate node’s in- and out-weight, respectively. Following the same procedure as in Lancelot et al. (2017) [31], the invasion threshold for each month were estimated. Highlight on the role of occasional links were given (connections appearing just once per year). We plotted path intersections between different species, transports modalities and seasons, to highlight possible common or specific links for different combinations. In addition, metrics quantifying these intersections were estimated, such as the pairwise percentage of common and specific paths between two networks, respectively. Gravity models were used to estimate the probability of a link between two distinct nodes according to their features. These models were developed in the field of socio-economics and human migration studies [32,33]. They provide estimates for the flows of goods or people between two nodes, as a function of node-level variables (e.g. population size, socio-economic factors, etc.), and of the distance or movement cost between these nodes. Its most general for- mulation (Eq 1) shows the analogy with Newton’s gravity law. MIGij is the flow between the origin i and destination j, pi and pj are the population at the origin and destination, dij is the distance between i and j, and α, β and γ are model parameters. The equation is linearized into PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 5 / 21 Livestock mobility in Mauritania Eq (2) by a log-transformation, and its coefficients can be estimated with generalized linear models. Analysis In a second step, this analysis was repeated using the number of animals traded between these two locations as the response variable. To better differentiate the factors influencing the trade probability from those influencing its volume, the latter analysis was restricted to pairs of Table 1. General equation of the tested models. MODEL GENERAL EQUATION E1 Hpi + Hpj + dijg E2 Lpi + Lpj + dijg E3 Hpi + Hpj + Lpi + Lpj + dijg E4 Hpi + Hpj + Lpi + Lpj + dija E5 Hpi + Hpj + Lpi + Lpj + dijel E6 Δ(Hpi, Hpj) + Δ(Lpi, Lpj) + dijg E7 E3 + mode + mode: pi + mode: pj + mode: dijg pi: population at the origin (L: livestock, H: human); pj: population at destination (L: livestock, H: human); dij: distance (g: great circle distance, a: costhpath distance based on accessibility friction surface, el: costhpath distance based on elevation friction surface), mode: transport modality (by truck or on foot). htt //d i /10 1371/j l 0199547 t001 Table 1. General equation of the tested models. MODEL GENERAL EQUATION E1 Hpi + Hpj + dijg E2 Lpi + Lpj + dijg E3 Hpi + Hpj + Lpi + Lpj + dijg E4 Hpi + Hpj + Lpi + Lpj + dija E5 Hpi + Hpj + Lpi + Lpj + dijel E6 Δ(Hpi, Hpj) + Δ(Lpi, Lpj) + dijg E7 E3 + mode + mode: pi + mode: pj + mode: dijg pi: population at the origin (L: livestock, H: human); pj: population at destination (L: livestock, H: human); dij: distance (g: great circle distance, a: costhpath distance based on accessibility friction surface, el: costhpath distance based on elevation friction surface), mode: transport modality (by truck or on foot). https://doi.org/10.1371/journal.pone.0199547.t001 Table 1. General equation of the tested models. https://doi.org/10.1371/journal.pone.0199547.t001 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Analysis MODEL GENERAL EQUATION E1 Hpi + Hpj + dijg E2 Lpi + Lpj + dijg E3 Hpi + Hpj + Lpi + Lpj + dijg E4 Hpi + Hpj + Lpi + Lpj + dija E5 Hpi + Hpj + Lpi + Lpj + dijel E6 Δ(Hpi, Hpj) + Δ(Lpi, Lpj) + dijg E7 E3 + mode + mode: pi + mode: pj + mode: dijg pi: population at the origin (L: livestock, H: human); pj: population at destination (L: livestock, H: human); dij: distance (g: great circle distance, a: costhpath distance based on accessibility friction surface, el: costhpath distance based on elevation friction surface), mode: transport modality (by truck or on foot). pi: population at the origin (L: livestock, H: human); pj: population at destination (L: livestock, H: human); dij: distance (g: great circle distance, a: costhpath distance based on accessibility friction surface, el: costhpath distance based on elevation friction surface), mode: transport modality (by truck or on foot). PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 6 / 21 Livestock mobility in Mauritania locations with an existing trade link. All analyses were coded and carried out using R [34]. The “sna” package was used to describe and analyse the trade network [35]. Table 2. Number of national and international traded animals in Mauritania. Values represent the number of animals moved for each given period. Movements are defined as national if both the origin and destination are located within the Mauritanian border. Small ruminants (SR) include sheep and goats. Some of the records did not differentiate the species individually and are counted as “Cattle and SR” or “Mixed”. Cattle Cattle and SR Camels SR Mixed Total Truck 67 240 43 760 1 105 870 1 216 870 Aug-Dec 34 398 19 119 843 047 896 565 international 2 400 2 120 410 500 415 020 national 31 998 16 999 432 547 481 545 Jan-July 32 842 24 641 262 822 320 305 international 9 600 3 280 980 13 860 national 23 242 21 361 261 842 306 445 Foot 1 801 846 1 660 598 034 3 334 755 96 563 5 832 859 Aug-Dec 342 444 277 112 740 716 645 19 740 1 191 847 international 283 403 28 011 518 420 829 835 national 59 041 277 84 729 198 225 19 740 362 012 Jan-July 1 459 402 1 383 485 294 2 618 110 76 823 4 641 012 international 119 1123 271 755 2 286 295 3 749 172 national 268 279 1 383 213 539 331 815 76 823 891 839 Total 1 869 086 1 660 641 794 4 440 625 96 563 7 049 729 https://doi.org/10.1371/journal.pone.0199547.t002 Table 2. Number of national and international traded animals in Mauritania. Values represent the number of animals moved for each given period. Movements are defined as national if both the origin and destination are located within the Mauritanian border. Small ruminants (SR) include sheep and goats. Some of the records did not differentiate the species individually and are counted as “Cattle and SR” or “Mixed”. locations with an existing trade link. All analyses were coded and carried out using R [34]. The “sna” package was used to describe and analyse the trade network [35]. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Results The dataset consisted in 2,219 trade movements involving 7.1 million head. The subset of national movements, the focus of this analysis, included 1,178 movement events involving 2.1 million head (Table 2). International movements involved around 5 million head, sold or bought to or from Senegal, Mali, Ivory Coast, Guinea Bissau and Morocco. As the destination of these international movements within these countries was unknown, they could not be included in the network analysis or gravity models. Within Mauritania, the trading network was composed of 65 nodes and 84 unique paths. Transport by foot was the most represented mode, corresponding to 83% of the animal flow (Tables 2 and 3). The foot and truck transport modalities presented contrasted patterns. The largest share of foot movements being short to medium distance (0–200 km), whereas the opposite was observed for movements by truck, where the largest share was represented by movements > 500 km (Fig 2). Similarly, those con- trasting patterns somewhat matched the seasonal pattern. The Tabaski period represented 41% of the inner animal movements, with 57% of them by truck. In contrast, only 25% of the head were moved by truck outside the Tabaski period (Table 2). The Mauritanian network was weakly connected. The density value indicated that 2% of possible node pairs were actually connected, and the network level centrality parameters and clustering coefficient were low (Table 3). However, each of the exchange networks (full net- work, or species-specific sub-networks) contained a single component in which the average length of the shortest path between node pairs was lower than 2 links, the maximum value (diameter) being 5 links. The goat and camel trading networks were smaller with a diameter of 2 and 3 links. On average, in the full species network, a given node was directly connected with approximately 2 other nodes on average (average degree). Both the in-degree and in-weight distributions are right-skewed. Only 5 locations from the livestock trading network attracted more than 50% of the connections, and are also the destination for more than 60% of the PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 7 / 21 Livestock mobility in Mauritania Fig 3. Network’s centrality measures of the Mauritanian’s livestock trade. Node size show the importance of the measured centrality values. in and out-weight measures were scaled on the total volume of traded livestock; eigenvector centrality (centrality measure) were scored from 0 to 1, betweenness was considered for the fraction of paths passing through the node. h //d i /10 13 1/j l 0199 4 003 Fig 3. Network’s centrality measures of the Mauritanian’s livestock trade. Node size show the importance of the measured centrality values. in and out-weight Fig 3. Network’s centrality measures of the Mauritanian’s livestock trade. Node size show the importance of the measured centrality values. in and out-weight measures were scaled on the total volume of traded livestock; eigenvector centrality (centrality measure) were scored from 0 to 1, betweenness was considered for the fraction of paths passing through the node. https://doi.org/10.1371/journal.pone.0199547.g003 traded volume. Nouakchott, the main urban consumption market of the country, acts as hub for livestock mobility with 18 links which concentrate around 1/3 of the traded animals’ total volume (Fig 3, Table 4). With the highest centrality score, the capital city is the most important node of the network. Almost 2/3 of the nodes has at most one outgoing link, whilst Aleg and traded volume. Nouakchott, the main urban consumption market of the country, acts as hub for livestock mobility with 18 links which concentrate around 1/3 of the traded animals’ total volume (Fig 3, Table 4). With the highest centrality score, the capital city is the most important node of the network. Almost 2/3 of the nodes has at most one outgoing link, whilst Aleg and Table 4. List of nodes with highest values for centrality measures. Each column corresponds to a specific network (all species or by single species) Each line relates to a specific centrality measure. Only the name of the node corresponding with the largest value for each measure is reported. In the case of multiple nodes with same value of the measure, all the names are reported. Table 4. List of nodes with highest values for centrality measures. Each column corresponds to a specific network (all species or by single species) Each line relates to a specific centrality measure. Only the name of the node corresponding with the largest value for each measure is reported. Livestock mobility in Mauritania ber of links of the national movement networks in Mauritania. Links are provided according to the transport modality and -Tabaski: Jan–July). Table 3. Network parameters and number of links of the national movement networks in Mauritania. Links are provided according to the transport modality and Tabaski periods (Tabaski: Aug–Dec; Non-Tabaski: Jan–July). ALL SPECIES CATTLE SR SHEEP GOAT CAMEL Network-level parameter Number of nodes 65 65 65 65 65 65 Number of links 84 49 56 54 32 34 Diameter 5 4 5 5 3 2 Clustering coefficient 0.19 0.15 0.18 0.18 0.15 0.03 Average path length 1.84 1.56 1.73 1.73 1.34 1.13 Density 0.020 0.012 0.013 0.013 0.0077 0.0082 Average degree 2.65 1.51 1.72 1.66 0.985 1.05 Number of link Transport modality Truck 33 13 28 28 13 14 Foot 55 37 30 28 20 22 Intersection 2 2 2 2 1 2 Trading period Non-Tabaski 73 45 46 44 29 31 Tabaski 57 31 39 39 19 21 Intersection 44 27 14 29 16 18 Fig 2. Great-circle distance of the national movements by truck (left) and on foot (right) recorded in the survey. Distance are given in kilometers for each species (thick line: small ruminants, dashed line: cattle, thick dash line: camel). https://doi org/10 1371/journal pone 0199547 g002 Fig 2. Great-circle distance of the national movements by truck (left) and on foot (right) recorded in the survey. Distance are given in kilometers for each species (thick line: small ruminants, dashed line: cattle, thick dash line: camel). https://doi.org/10.1371/journal.pone.0199547.g002 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 8 / 21 In the case of multiple nodes with same value of the measure, all the names are reported. ALL SPECIES CATTLE SR CAMEL INDEGREE Nouakchott Nouakchott Nouakchott Nouakchott INWEIGHT Nouakchott Selibabi Nouakchott Zouerate OUTDEGREE Aleg Aleg Kiffa Aleg Kiffa Kiffa Mmat Laakarich II M. Lahjar OUTWEIGHT Tintane Kaedi Tintane Nbeika BETWEENNESS Boutilimit Adel-Bagrou Boutilimit Boutilimit Mbout Nbeika EIGENVECTOR CENTRALITY Nouakchott Kaedi Nouakchott Nbeika https://doi org/10 1371/journal pone 0199547 t004 9 / 21 Livestock mobility in Mauritania Table 5. Correlation coefficients among centrality measures. Pearson correlations coefficients among centrality measure. Only significant (p-value <0.05) coefficients are reported. INDEGREE OUTDEGREE INWEIGHT OUTWEIGHT INDEGREE 0.88 OUTDEGREE 0.63 INWEIGHT 0.88 OUTWEIGHT 0.63 https://doi.org/10.1371/journal.pone.0199547.t005 Kiffa, in the southern region, are connected to other 7 localities, and Tintane has the largest out-weight. Finally, the betweenness distribution is right-skewed and Boutilimit appears to be the node with highest betweenness. Few of the centrality measures are significantly correlated (p-value<0.05) and reported in Table 5. Results of the percolation analysis on networks cohesion are shown in Fig 4. Nodes are removed based on the centrality measures (indegree, outdegree, incoming and outgoing Fig 4. Effect of targeted removal on the connected component size. Values on the x axis indicate the percentage of nodes removed (cumulative), together with their links. Values on y axis indicate the percentage of nodes in the largest connected sub-network, after the removal. Color indicates the removal procedure based on centrality measure score (starting from the highest score nodes) or randomly. Shaded areas correspond to 95% CI for the random procedure. https://doi.org/10.1371/journal.pone.0199547.g004 Fig 4. Effect of targeted removal on the connected component size. Values on the x axis indicate the percentage of nodes removed (cumulative), together with their links. Values on y axis indicate the percentage of nodes in the largest connected sub-network, after the removal. Color indicates the removal procedure based on centrality measure score (starting from the highest score nodes) or randomly. Shaded areas correspond to 95% CI for the random procedure. https://doi.org/10.1371/journal.pone.0199547.g004 Fig 4. Effect of targeted removal on the connected component size. Values on the x axis indicate the percentage of nodes removed (cumulative), together with their links. Values on y axis indicate the percentage of nodes in the largest connected sub-network, after the removal. Color indicates the removal procedure based on centrality measure score (starting from the highest score nodes) or randomly. Shaded areas correspond to 95% CI for the random procedure. https://doi.org/10.1371/journal.pone.0199547.g004 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 10 / 21 Livestock mobility in Mauritania Fig 5. Species-level movement networks. Each diagram represents the movement links that are specific to the species (red: by truck, dark grey: by foot, both truck and foot: blue). The last plot shows the entire network of all species. https://doi.org/10.1371/journal.pone.0199547.g005 Fig 5. Species-level movement networks. All species: Association of small ruminants, Cattle and Camel trading networks. Each diagram represents the movement links that are specific to the species (red: by truck, dark grey: by foot, both truck and foot: blue). The last plot shows the entire network of all species. Fig 5. Species-level movement networks. Each diagram represents the movement links that are specific to the species (red: by truck, dark grey: by foot, both truck and foot: blue). The last plot shows the entire network of all species. https://doi.org/10.1371/journal.pone.0199547.g005 https://doi.org/10.1371/journal.pone.0199547.g005 volume, betweenness, eigenvector centrality) and, for comparison purposes, randomly. volume, betweenness, eigenvector centrality) and, for comparison purposes, randomly. Removing nodes in order of their incoming connections, incoming volume and centrality result as the most effective strategies of fragmentation. In the first case, removing less than 20% of the nodes (13 nodes) results in decomposing network in a set of subnetworks whom the largest one contains less than 10% of the nodes (7 nodes). Removing nodes in order of their incoming connections, incoming volume and centrality result as the most effective strategies of fragmentation. In the first case, removing less than 20% of the nodes (13 nodes) results in decomposing network in a set of subnetworks whom the largest one contains less than 10% of the nodes (7 nodes). In total, 94% of the goats trading network paths were shared with the sheep (Fig 5, Table 6). These two networks were merged into a single small ruminant trading network in the gravity models. A high degree of overlap between pairs of species networks was highlighted (Table 6). However, overlapping only involved 20% of the full set of trading links (Fig 5, Table 6). the different species. The values correspond to the percentage (italic) and the number (bracket) of common links and of species. Table 6. Comparison of the trading networks of the different species. The values correspond to the percentage (italic) and the number (bracket) of common links and difference between the trading network of the pair of species. network SPECIES 1 network SPECIES 2 INTERSECTION % (n) DIFFERENCE SP1/SP2 % (n) DIFFERENCESP2/SP1 % (n) CATTLE/SR 49 56 65 (32) 35 (17) 43 (24) CAMEL/SR 34 56 65 (22) 35 (12) 61 (34) CAMEL/CATTLE 34 49 47 (16) 53 (18) 67 (33) GOAT/SHEEP 32 54 94 (30) 6 (2) 44 (24) ALL SPECIES 84 - 20 (17) Table 6. Comparison of the trading networks of the different species. ation of small ruminants, Cattle and Camel trading networks. https://doi.org/10.1371/journal.pone.0199547.t006 All species: Association of small ruminants, Cattle and Camel trading networks. https://doi.org/10.1371/journal.pone.0199547.t006 The values correspond to the percentage (italic) and difference between the trading network of the pair of species. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 11 / 21 Livestock mobility in Mauritania Among the remaining links, many were species-specific: 30% of the small ruminant (17/56), 33% of the cattle (16/49) and 32% of the camel (11/34) links were not included in any of the other networks. Very few links were shared between the truck and foot networks and were identical for all species (2/84, Table 6). The role of nodes could change depending on the species considered. Previous results about most central nodes hold when we consider the “small ruminants only” network. Whilst Nouakchott remains always the most connected node, most of the volume of cattle and camels are directed towards Selibabi and Zouerate, respectively. The markets of Kaedi, for cattle, and Nbeika, for camels, become more central in the network (see S1 File and Table 4). In term of cohesiveness, specie-specific network are more vulnerable to the target removal of nodes. As in the all species case target removal based on indegree and inweight are the most efficient pro- cedures. In fact, in both cases, removing a limited quantity of nodes, 8 for camel’s network 10 for cattle’s one and 12 for small ruminants one, results in decomposing the respective networks in a set of subnetworks whom the largest one contains at most 8 nodes. This is particularly rel- evant for small ruminants and the cattle networks, whose largest subnetworks consist of 7 and 10 nodes, respectively, after the removal of only 3 most connected nodes. There was a high number of shared links between the Tabaski and non-Tabaski periods (44/84, Table 6). Both Tabaski and non-Tabaski networks included truck and foot movements, but their relative proportion was different. During the Tabaski period, there were fewer foot movements and higher truck movements (Fig 6). Very few trade connections involved both truck and foot movements in the two periods (Table 6, Figs 5 and 6). Fig 6. Movement networks according to the species and seasons. Note that for each season the links present in both defined transportation mode are not plotted and are the same for both defined season (red: by truck, dark grey: by foot, nf: number of movement links by foot, Jan-July: Non-Tabaski period; Aug-Dec: Tabaski period). Fig 7. Variations of network quantities along the year. Bottom: monthly epidemic threshold variations along the year; Centre Volume of animals traded during the month; Top number of links active. For each month, we have considered 2 networks: all network including occasional links appearing on that month; backbone, excluding occasional link. The different colors correspond to the quantity evaluated for the specific network. https://doi.org/10.1371/journal.pone.0199547.g007 Fig 7. Variations of network quantities along the year. Bottom: monthly epidemic threshold variations along the year; Centre Volume of animals traded during the month; Top number of links active. For each month, we have considered 2 networks: all network including occasional links appearing on that month; backbone, excluding occasional link. The different colors correspond to the quantity evaluated for the specific network. https://doi.org/10.1371/journal.pone.0199547.g007 As the network changes along the year so its proneness to diffuse diseases. Fig 7, present the variation of the epidemic threshold (denoted q), estimated every month, along the year in comparison with the number of links active and the volume of livestock traded. Since the net- work can change along the year, particularly new active links around Tabaski, we considered the backbone network from Apolloni et al. [17] (containing links present more than 2 months) and the total network containing all the link active that month to elicit the role that Tabaski plays on the risk of transmission. The epidemic threshold is at the lowest values between the March and June, when the volume and the number of exchanges (links) is at maximum, and around Tabaski, when a second peak of movements (mainly small ruminants) whose volume is almost equal to the first peak, is observed. Occasional links, appearing only for the Tabaski reason, decreases the invasion threshold and consequently the risk of disease spreading is higher in this period. Table 7 presents the different sets of gravity models that were applied to the full network (binary outcome: 1 if two nodes were connected and 0 otherwise), and to the networks broken down by species, period and transport modality. The best results were obtained with models including human and animal populations at the origin and destination, with a great-circle distance (model E3, Tables 1 and 7). Replacing this distance with cost-distance functions of accessibility (E4) or elevation (E5) did not improve the results. Fig 3 should be considered to see the redundancy between transportation mode (blue edge). Fig 6. Movement networks according to the species and seasons. Note that for each season the links present in both defined transportation mode are not plotted and are the same for both defined season (red: by truck, dark grey: by foot, nf: number of movement links by foot, Jan-July: Non-Tabaski period; Aug-Dec: Tabaski period). Fig 3 should be considered to see the redundancy between transportation mode (blue edge). https://doi.org/10.1371/journal.pone.0199547.g006 12 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania Fig 7. Variations of network quantities along the year. Bottom: monthly epidemic threshold variations along the year; Centre Volume of animals traded during the month; Top number of links active. For each month, we have considered 2 networks: all network including occasional links appearing on that month; backbone, excluding occasional link. The different colors correspond to the quantity evaluated for the specific network. https://doi.org/10.1371/journal.pone.0199547.g007 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 AIC model n link n unique link E1 E2 E3 E4 E5 E6 E7 ALL SPECIES All 1178 140 1501.8 1604.1 1464.1 1488.9 1485.1 1461.9 1342.7 NTab 518 53 1339.8 1418.6 1303.3 1323.7 1323.6 1303.3 Tab 660 87 995.6 1133.7 979.8 983.1 982.9 976.1 Truck 735 121 427.2 601.1 405.0 408.9 407.3 416.9 Truck—NTab 443 91 315.7 470.7 302.4 309.1 305.4 307.8 Truck—Tab 257 40 175.2 280.5 172.0 174.9 172.2 175.2 Foot 129 21 899.3 885.9 877.5 951.1 934.7 885.6 Foot—NTab 478 81 843.8 832.2 821.8 886.8 874.0 832.1 Foot—Tab 182 41 460.4 454.7 455.2 498.3 484.9 455.7 SR All 620 58 589.5 639.7 572.0 578.9 576.7 570.3 NTab 374 46 482.7 530.3 470.1 474.1 476.5 470.7 Tab 246 40 412.9 473.1 404.0 404.2 403.7 403.7 Truck 324 31 210.3 290.5 204.5 205.7 205.7 205.9 Foot 296 27 315.3 308.2 304.7 331.3 327.0 307.9 CT All 341 47 494.1 526.9 480.6 474.9 487.2 489.4 NTab 214 43 465.0 492.4 451.3 444.5 455.8 459.3 Tab 127 31 342.3 384.2 329.3 322.5 328.0 340.2 Truck 109 13 107.2 146.7 105.6 105.9 106.0 106.0 Foot 232 34 305.7 295.8 296.0 304.1 307.9 307.7 CM All 217 35 411.5 425.1 399.2 397.4 399.5 395.8 NTab 147 32 390 393.9 377.0 375.9 377.6 373.4 Tab 70 20 241.7 265.8 236.1 235.7 236.1 233.4 Truck 113 13 116.6 156.3 109.9 110.8 108.9 112.3 Foot 104 22 266.6 258.7 261.3 264.9 265.9 258.6 https://doi.org/10.1371/journal.pone.0199547.t007 Table 7. AIC values of the logistic regression models. The models are broken down by species (SR: small ruminants, CT: cattle, CM: camels), period (Tab: Tabaski, NTab: Non-Tabaski) and transport modality. E1-E6 correspond to different models described in Table 1. Table 7. AIC values of the logistic regression models. The models are broken down by species (SR: small ruminants, CT: cattle, CM: camels), period (Tab: Tabaski, NTab: Non-Tabaski) and transport modality. E1-E6 correspond to different models described in Table 1. Table 7. AIC values of the logistic regression models. The models are broken down by species (SR: small ruminan NTab: Non-Tabaski) and transport modality. E1-E6 correspond to different models described in Table 1. Table 7. AIC values of the logistic regression models. The models are broken down by species (SR: small ruminants, CT: cattle, CM: camels), period (Tab: Tabaski, NTab: Non-Tabaski) and transport modality. E1-E6 correspond to different models described in Table 1. AIC model n link n unique link E1 E2 E3 E4 E5 E6 E7 ALL SPECIES All 1178 140 1501.8 1604.1 1464.1 1488.9 1485.1 1461.9 1342.7 NTab 518 53 1339.8 1418.6 1303.3 1323.7 1323.6 1303.3 Tab 660 87 995.6 1133.7 979.8 983.1 982.9 976.1 Truck 735 121 427.2 601.1 405.0 408.9 407.3 416.9 Truck—NTab 443 91 315.7 470.7 302.4 309.1 305.4 307.8 Truck—Tab 257 40 175.2 280.5 172.0 174.9 172.2 175.2 Foot 129 21 899.3 885.9 877.5 951.1 934.7 885.6 Foot—NTab 478 81 843.8 832.2 821.8 886.8 874.0 832.1 Foot—Tab 182 41 460.4 454.7 455.2 498.3 484.9 455.7 SR All 620 58 589.5 639.7 572.0 578.9 576.7 570.3 NTab 374 46 482.7 530.3 470.1 474.1 476.5 470.7 Tab 246 40 412.9 473.1 404.0 404.2 403.7 403.7 Truck 324 31 210.3 290.5 204.5 205.7 205.7 205.9 Foot 296 27 315.3 308.2 304.7 331.3 327.0 307.9 CT All 341 47 494.1 526.9 480.6 474.9 487.2 489.4 NTab 214 43 465.0 492.4 451.3 444.5 455.8 459.3 Tab 127 31 342.3 384.2 329.3 322.5 328.0 340.2 Truck 109 13 107.2 146.7 105.6 105.9 106.0 106.0 Foot 232 34 305.7 295.8 296.0 304.1 307.9 307.7 CM All 217 35 411.5 425.1 399.2 397.4 399.5 395.8 NTab 147 32 390 393.9 377.0 375.9 377.6 373.4 Tab 70 20 241.7 265.8 236.1 235.7 236.1 233.4 Truck 113 13 116.6 156.3 109.9 110.8 108.9 112.3 Foot 104 22 266.6 258.7 261.3 264.9 265.9 258.6 https://doi.org/10.1371/journal.pone.0199547.t007 rather than their absolute values (E6) did not lead to improved models for the different break- downs. In all models, and for all species networks, when the animal population was kept as a predictor, the number of sheep was the most important predictor. rather than their absolute values (E6) did not lead to improved models for the different break- downs. In all models, and for all species networks, when the animal population was kept as a predictor, the number of sheep was the most important predictor. Table 8 presents the details of the final models broken down by species, season and trans- port mode with the human and sheep population at the origin and destination and great-circle distance as predictors. The predictive power of the models were moderate to very good, according to the species, season and mode of transport, with AUC values ranging from 0.76 to 0.97. Similarly, although we noted some improvements for some combinations of species and transport modalities, the use of the population difference between origin and destination PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 13 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania rather than their absolute values (E6) did not lead to improved models for the different break- downs. In all models, and for all species networks, when the animal population was kept as a predictor, the number of sheep was the most important predictor. Table 8 presents the details of the final models broken down by species, season and trans- port mode with the human and sheep population at the origin and destination and great-circle distance as predictors. The predictive power of the models were moderate to very good, according to the species, season and mode of transport, with AUC values ranging from 0.76 to 0.97. Considering all species, seasons and transport modes, positive associations were found between the probability of a trade connection and: i) low human population at the origin, ii) high human population at the destination, iii) high sheep population at the origin, iv) low sheep population at the destination, and v) a low great-circle distance (Table 8). The results were similar for the small ruminants’ network model, whereas for the cattle model, the human population at the origin and the sheep population at the destination were not significant. For the camel network, the human population at the origin and the great-circle distance were not significant. The sign of the significant effects was coherent across the sub-models, i.e. a higher probability of trade event was always associated with low human population at the origin, or high human population at the destination, a high population of sheep at the origin, a low pop- ulation of sheep at the destination and a low great-circle distance, or a combination of these effects. In addition, meaningful differences were noticed in some sub-models. For example, in small ruminant models split by transport mode, human population at the destination, and sheep population at the origin were not significant in the foot-movement sub-model. In Table 7. AIC values of the logistic regression models. The models are broken down by species (SR: small ruminants, CT: cattle, CM: camels), period (Tab: Tabaski, NTab: Non-Tabaski) and transport modality. E1-E6 correspond to different models described in Table 1. Multivariate linear model and significance of the parameters given for the selected model (E3). g Hpop: Human population at the origin (i) and at destination (j); Spop: Sheep population at the origin (i) and at destination (j). https://doi org/10 1371/journal pone 0199547 t008 g Hpop: Human population at the origin (i) and at destination (j); Spop: Sheep population at the origin (i) and at destination (j). Hpop: Human population at the origin (i) and at destination (j); Spop: Sheep population at the origin (i) and at destination (j). https://doi.org/10.1371/journal.pone.0199547.t008 contrast, both predictors at the origin were not significant in the truck-movement sub-model. Great-circle distance and both human and sheep population were not significant for the latter, though. The observed and estimated trade links are illustrated in Fig 8 for these small rumi- nant models for the Tabaski and non-Tabaski periods, and movements by trucks or on foot. During the non-Tabaski period, the fitted values correctly captured the co-existence of short- and long-distance movements, whereas long-distance movements were prominent during the Tabaski period. Similarly, the prediction of the truck or foot movements corresponded to their respective long or short distances. None of the models correctly fitted the south/north movements. None of the gravity models with the flow of animals as the response, and the same set of predictors as in the presence/absence model detailed above, showed any significant association. Considering all species, seasons and transport modes, positive associations were found between the probability of a trade connection and: i) low human population at the origin, ii) high human population at the destination, iii) high sheep population at the origin, iv) low sheep population at the destination, and v) a low great-circle distance (Table 8). The results were similar for the small ruminants’ network model, whereas for the cattle model, the human population at the origin and the sheep population at the destination were not significant. For the camel network, the human population at the origin and the great-circle distance were not significant. The sign of the significant effects was coherent across the sub-models, i.e. a higher probability of trade event was always associated with low human population at the origin, or high human population at the destination, a high population of sheep at the origin, a low pop- ulation of sheep at the destination and a low great-circle distance, or a combination of these effects. In addition, meaningful differences were noticed in some sub-models. For example, in small ruminant models split by transport mode, human population at the destination, and sheep population at the origin were not significant in the foot-movement sub-model. In PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 14 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania Table 8. Multivariate linear model and significance of the parameters given for the selected model (E3). Intercept Hpopi Hpopj Spopi Spopj Great-circle distance AUC All mode All species -4.239 -7.622 10−6 4.667 10−6 3.905 10−6 -5.104 10−6 -2.278 10−3 0.836 Small ruminants -3.895 -1.0 10−5 4.92 10−6 2.45 10−6 -4.00 10−6 -2.129 10−3 0.840 Cattle -4.619 -8.928 10−6 5.596 10−6 2.118 10−5 7.090 10−6 -4.416 10−3 0.884 Camel -5.402 -1.966 10−6 4.189 10−6 4.699 10−6 -6.374 10−6 -5.632 10−4 0.790 All non-Tabaski All species -4.210 -1.069 10−5 4.555 10−6 4.196 10−6 -5.356 10−3 -2.415 10−3 0.840 Small ruminants -3.809 -1.751 10−5 5.08 10−6 2.634 10−6 -3.575 10−6 -2.628 10−3 0.863 Cattle -4.599 -1.239 10−5 5.417 10−6 2.244 10−5 4.982 10−6 -4.095 10−3 0.885 Camel -5.563 -1.559 10−6 3.852 10−6 5.167 10−6 -6.139 10−6 -5.802 10−4 0.781 All Tabaski All species -5.355 -3.679 10−6 5.424 10−6 3.563 10−6 -4.107 10−6 -1.595 10−3 0.869 Small ruminants -4.680 -5.409 10−6 5.408 10−6 1.985 10−6 -4.637 10−6 -1.378 10−3 0.863 Cattle -6.179 -3.851 10−6 6.665 10−6 2.269 10−5 1.598 10−5 -3.287 10−3 0.901 Camel -6.391 -4.458 10−7 5.102 10−6 4.862 10−6 -7.167 10−6 -2.454 10−4 0.837 Truck All period All species -5.841 -7.730 10−8 6.876 10−6 3.242 10−6 -2.072 10−5 1.126 10−3 0.934 Small ruminants -5.432 -9.748 10−7 6.600 10−6 1.634 10−6 -8.942 10−6 8.968 10−4 0.91 Cattle -2.574 -1.350 10−7 3.652 10−5 1.903 10−5 9.883 10−6 -1.049 10−2 0.945 Camel -6.015 8.635 10−7 7.230 10−6 4.261 10−6 -3.349 10−5 1.590 10−3 0.966 Foot All period All species -2.665 -1.396 10−5 -1.435 10−6 4.702 10−6 -3.418 10−6 -5.783 10−3 0.830 Small ruminants -2.225 -1.963 10−5 3.289 10−6 3.289 10−6 -2.996 10−6 -6.409 10−3 0.853 Cattle -6.470 9.398 10−7 6.589 10−6 1.636 10−5 -1.105 10−4 8.720 10−4 0.908 Camel -4.200 -3.373 10−6 -4.043 10−6 -4.940 10−6 -4.366 10−6 -1.982 10−3 0.759 Table 8. Discussion Though international movements were not addressed in this study, it is noteworthy that in 2014, the largest sheep exportation peak was observed during the “soudure”, i.e. the period sepa- rating the end of familial cereal reserve saved after the previous harvest (millet, sorghum. . .) from the next harvest. This was also the hot, dry season, when forage and surface water resources were finished, and ruminant livestock starved. Therefore, the most obvious option for the Mau- ritanian livestock farmers was to sell most of the offspring, only keeping the core of reproductive PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 15 / 21 Livestock mobility in Mauritania Fig 8. Observed and predicted movements links from the gravity model applied to the small ruminant networks. Models were applied to movements occurring within the non-Tabaski period (A), the Tabaski period (B), without distinction of period (C-D): movements by truck (C) and by foot (D). https://doi.org/10.1371/journal.pone.0199547.g008 Fig 8. Observed and predicted movements links from the gravity model applied to the small ruminant networks. Models were applied to movements occurring within the non-Tabaski period (A), the Tabaski period (B), without distinction of period (C-D): movements by truck (C) and by foot (D). https://doi.org/10.1371/journal.pone.0199547.g008 https://doi.org/10.1371/journal.pone.0199547.g008 https://doi.org/10.1371/journal.pone.0199547.g008 ewes and she-goats. In addition, many of the latter spent this season in the closest areas with more abundant pastoral resources, i.e. in Senegal and Mali. Short- and mid-distance move- ments–most of them by foot, thus allowed pastoralists to exploit more suitable environmental conditions and reduce the economic cost of feeding the animals. In future years, when the Tabaski feast occurs during the soudure period, the relative importance of truck vs. foot ewes and she-goats. In addition, many of the latter spent this season in the closest areas with more abundant pastoral resources, i.e. in Senegal and Mali. Short- and mid-distance move- ments–most of them by foot, thus allowed pastoralists to exploit more suitable environmental conditions and reduce the economic cost of feeding the animals. In future years, when the Tabaski feast occurs during the soudure period, the relative importance of truck vs. foot PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 16 / 21 Livestock mobility in Mauritania movements may change and add to each other. Moreover, as the Tabaski is a mobile festivity which is celebrated annually among Muslims worldwide 70 days after Ramadan, a time slip of 10 days upstream occurs each year. Discussion An overlap with specific diseases vectors may thus occurs depending on the occurrence of the feast (year-dependence of the epidemic risk). Around Sene- gal river, Aedes mosquitoes show a peak in the periods of July-August and September during the rainy season. Thanks to the happening of the Tabaski, held in the last years between September and October, the risk of infection was elevated [36]. The risk remains high every year due to the large volume of moved animals and the traders’ preference for truck transportation and its fast- ness which allow viremic animal introduction on remote locations. The low values for the national network diameter, the presence of hubs and the low values of the epidemic threshold, indicate that the network could be prone to transmission of dis- eases. This means that also a lowly transmitted disease, once introduced in the national net- work, could reach all nodes (the network’s single component) in a short amount of time (small diameter). On the other hand, independently of the species considered, the mobility network is prone to fragmentation due to targeted intervention based on nodes activity (in-degree and in-weight). After the intervention, the network is decomposed in a set of smaller subnetworks, and virus can circulate only among nodes of the same sub-network. Vaccinating animals in largest markets (nodes with largest number of incoming animals) or closing these markets, could result in a very effective way of controlling the epidemic spread. Regarding internal movements, the gravity models correctly predicted the probability of a trade connection and their interpretation was straightforward. Locations with few sheep and high human populations, i.e. urban consumption centers, acted as movement sinks. Con- versely, areas with high sheep and low human populations, i.e. rural livestock farming areas, acted as movement sources. For a similar level of production and demand, short distance movements were more likely that long distance ones. In addition, Nouakchott, the capital city located on the coast, strongly influenced the network structures. Almost a quarter of the total population of Mauritania lives in Nouakchott (according to the National Statistical Office http://www.ons.mr/) and because of this, it is the largest terminal market with an incoming volume of almost 30% of the national one. In preparation for Tabaski, almost 50% of the total traded small ruminants are sold in Nouakchott. The largest majority of livestock provisioning Nouakchott markets comes from the South-Western area of Mauritania. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Discussion The cost-distance was estimated using the accessibility friction map which is based on the road net- work and land-use data, two expected drivers of the livestock trade networks. However, the friction layer might not be adapted to the specific constrains of animal movements in arid and hyper-arid environments. For example, the presence of water points or stopover feeding sites along the roads might be more relevant factors than those included in the global friction sur- face. Therefore, further work is needed to build friction surfaces better suited to the specific constrains of animal movement in this environment. The movement survey database also included volumes of the livestock trade flows between locations. However, these quantitative data were not correctly predicted by any of the models. This failure might be related to two non-exclusive factors. First, when a link was established, the volume of traded animals did not vary much, and not proportionally to the deficit in demand or to the distance. Second, there might be noise in the data related to inaccurate replies during the interviews with field veterinary officers. However, considering the outer trade data (international movements), estimates from this survey closely matched importation estimates from the Senegalese Veterinary Services for the Tabaski period. Because these data were of crucial importance to stabilize sheep price (and thus prevent social troubles), all efforts were made for an accurate monitoring of sheep importation. Therefore, this good match pro- vided a partial validation for the quality of data produced by this survey, as well as the good predictive power obtained with the presence/absence models. An important question underlying these analyses is the role that animal mobility might play in the spread of animal and zoonotic diseases in the region. For example, animal move- ments may contribute to the spread of Rift Valley fever (RVF) in the Sahelian region of Mauri- tania, and from Mauritania to Senegal [37,38]. The network structure predicted by this model may provide input for an epidemiological model of RVF or other important diseases affecting the region, such as Peste des petits ruminants (PPR) [6]. Gravity models are an important method in economic analyses, used mainly to predict bilateral flow of population and goods between two distant locations [39]. They were recently adapted to describe the spread of biological agents [33,40–42]. To our knowledge, this study is a first attempt to predict livestock mobility patterns. Discussion A continuous flux of animals is ensured by stockists who collect animals at collection market and transfer them by truck to the capital city. Here, stockists buy imported goods, arrived at the international port, to sell at collection markets. It might look odd that the sheep population at the origin and destination of movements was the best predictor for all species-level networks. A possible explanation might be that the spatial distributions of all ruminant species were positively correlated, thus making the sheep population a confounding factor for the other ruminant populations. This assumption was corroborated by the dominance of small ruminant movements: 1.3 million vs. 0.39 million for cattle and 0.37 million for camels. Therefore, the small ruminant network probably influenced all other ruminant trade activities. The seasonal models had good predictive power for the models of both truck or foot move- ments. These models captured the dominance of long-distance movements (truck movement network), of short-distance movements (foot movement network) and different combinations of short and long-distance movements (Tabaski vs non-Tabaski periods). However, the models did not capture some of the trade links. For example, they failed to predict the link to the northern city of Zouerate, which was involved in the small ruminant and camel trading net- works. Located in the desert and with a low accessibility, this city has a large iron mining industry, and virtually no local production. The working population there may be much higher PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 17 / 21 Livestock mobility in Mauritania than the population estimates in the Worldpop population database, which maps people according to their residence, not their working place. Such high populations of workers may generate high demand for small ruminant and camel meat, and camel milk. Armed conflicts may also influence the pattern of demand with drivers not accounted for in the models. For example, an important flow of human population was reported in south-eastern Mauritania coming from Mali after the terrorist attack of January 2013. These refugees established camps close to the border with Mali: their population may have influenced the pattern of demand and production in a way that was not captured by the model. The use of cost-distance instead of great-circle distance did not improve the models. Discussion Besides obvious applications in the field of pastoral economics, it opens new perspectives for predicting the transmission of pathogens such as PPR or RVF viruses in animal meta-populations, or extending existing models of post- vaccination immunity persistence at the population level [43]. Also, similar models might be used at the regional level–e.g, between Maghreb or Sahel countries, or between Sahel and Maghreb regions, etc.–to validate and compliment (un)available information on transbound- ary animal movements. Acknowledgments This study was funded by EU grant FP7-613996 VMERGE and is catalogued by the VMERGE Steering Committee as VMERGE000 (http://www.vmerge.eu). The contents of this publication are the sole responsibility of the authors and don’t necessarily reflect the views of the European Commission. Supporting information S1 File. Small ruminant, cattle and camel trading networks and related centrality measures for Mauritania in 2014. Node size show the importance of the measured centrality values. in 18 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania and out-weight measures were scaled on the total volume of traded livestock; eigenvector cen- trality (centrality measure) were scored from 0 to 1, betweenness was considered for the frac- tion of paths passing through the node. (DOCX) and out-weight measures were scaled on the total volume of traded livestock; eigenvector cen- trality (centrality measure) were scored from 0 to 1, betweenness was considered for the frac- tion of paths passing through the node. (DOCX) Author Contributions Conceptualization: Gae¨lle Nicolas, Marius Gilbert. Data curation: Gae¨lle Nicolas. Formal analysis: Gae¨lle Nicolas. Formal analysis: Gae¨lle Nicolas. Methodology: Gae¨lle Nicolas, Marius Gilbert. Methodology: Gae¨lle Nicolas, Marius Gilbert. Supervision: Marius Gilbert. Supervision: Marius Gilbert. Validation: Gae¨lle Nicolas. Validation: Gae¨lle Nicolas. Visualization: Gae¨lle Nicolas. Visualization: Gae¨lle Nicolas. Writing – original draft: Gae¨lle Nicolas. Writing – original draft: Gae¨lle Nicolas. Writing – review & editing: Gae¨lle Nicolas, Andrea Apolloni, Caroline Coste, G. R. William Wint, Renaud Lancelot, Marius Gilbert. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 References 1. Jones KE, Patel NG, Levy MA, Storeygard A, Balk D, Gittleman JL, et al. Global trends in emerging infectious diseases. Nature. 2008; 451: 990–993. https://doi.org/10.1038/nature06536 PMID: 18288193 2. Black P, Nunn M. Impact of climate change and environmental changes on emerging and reemerging animal disease and animal production. Conf OIE. 2009; 15–25. 3. Nicolas G, Durand B, Duboz R, Rakotondravao R, Chevalier V. Description and analysis of the cattle trade network in the Madagascar highlands: Potential role in the diffusion of Rift Valley fever virus. Acta Trop. 2013; 126: 19–27. https://doi.org/10.1016/j.actatropica.2012.12.013 PMID: 23295911 4. Grace D, Mutua F, Ochungo P, Kruska R, Jones K, Brierley L, et al. Mapping of poverty and likely zoo- noses hotspots. Zoonoses Project 4. Report to the UK Department for International Development. Nai- robi, Kenya: ILRI [Internet]. ILRI; 2012 Jul. Available: https://cgspace.cgiar.org/handle/10568/21161 5. Fèvre EM, Bronsvoort BM de C, Hamilton KA, Cleaveland S. Animal movements and the spread of infectious diseases. Trends Microbiol. 2006; 14: 125–131. https://doi.org/10.1016/j.tim.2006.01.004 PMID: 16460942 6. Jones BA, Rich KM, Mariner JC, Anderson J, Jeggo M, Thevasagayam S, et al. The Economic Impact of Eradicating Peste des Petits Ruminants: A Benefit-Cost Analysis. PLOS ONE. 2016; 11: e0149982. https://doi.org/10.1371/journal.pone.0149982 PMID: 26900944 7. Wang X, Liu C, Mao W, Hu Z, Gu L. Tracing The Largest Seasonal Migration on Earth. ResearchGate. 2014; Available: https://www.researchgate.net/publication/267811689_Tracing_The_Largest_ Seasonal_Migration_on_Earth 8. Deville P, Linard C, Martin S, Gilbert M, Stevens FR, Gaughan AE, et al. Dynamic population mapping using mobile phone data. Proc Natl Acad Sci. 2014; 111: 15888–15893. https://doi.org/10.1073/pnas. 1408439111 PMID: 25349388 19 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania 9. Chapman M, Prothero RM. Circulation in Population Movement (Routledge Revivals): Substance and concepts from the Melanesian case. Reprint edition. London: Routledge; 2013. 10. Clarkson L, Crawford M. Feast and FamineFood and Nutrition in Ireland 1500–1920 [Internet]. 2001. Available: http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780198227519.001.0001/ acprof-9780198227519 11. Colizza V, Barrat A, Barthe´lemy M, Vespignani A. The role of the airline transportation network in the prediction and predictability of global epidemics. Proc Natl Acad Sci U S A. 2006; 103: 2015–2020. https://doi.org/10.1073/pnas.0510525103 PMID: 16461461 12. Colizza V, Barthe´lemy M, Barrat A, Vespignani A. Epidemic modeling in complex realities. C R Biol. 2007; 330: 364–374. https://doi.org/10.1016/j.crvi.2007.02.014 PMID: 17502293 13. Nicolas G, Chevalier V, Tantely LM, Fontenille D, Durand B. References A Spatially Explicit Metapopulation Model and Cattle Trade Analysis Suggests Key Determinants for the Recurrent Circulation of Rift Valley Fever Virus in a Pilot Area of Madagascar Highlands. PLoS Negl Trop Dis. 2014; 8: e3346. https://doi.org/10. 1371/journal.pntd.0003346 PMID: 25474116 14. Moritz M. Chapter 13—Networks and Hazard Adaptation Among West African Pastoralists. Social Net- work Analysis of Disaster Response, Recovery, and Adaptation. Butterworth-Heinemann; 2017. pp. 193–207. Available: http://www.sciencedirect.com/science/article/pii/B9780128051962000133 15. Ibra T, Ickowicz A, Wane A, Garba I, Gerber P. Eds. Atlas of Trends in pastoral systems in the Sahel 1970–2012. SIPSA. FAO-CIRAD, 32 pages(http://umr-selmet.cirad.fr/publications-et-ressources/ documents-techniques). [Internet]. 2012. Available: https://www.researchgate.net/publication/ 261842575_Atlas_of_Trends_in_pastoral_systems_in_the_Sahel_1970-2012_SIPSA_FAO-CIRAD_ 32_pages_httpumr-selmetciradfrpublications-et-ressourcesdocuments-techniques 16. L’hoˆte Y, Mahe´ G, Some´ B, Triboulet JP. Analysis of a Sahelian annual rainfall index from 1896 to 2000; the drought continues. Hydrol Sci J. 2002; 47: 563–572. https://doi.org/10.1080/02626660209492960 17. Apolloni A, Nicolas G, Coste C, El Mamy ABO, Lancelot R, Gilbert M. Towards the description of live- stock mobility in Sahelian Africa: some results from a survey in Mauritania. Transbound Emerg Dis. 2018; 13: e0191565. https://doi.org/10.1371/journal.pone.0191565 PMID: 29364989 18. Metzger MJ, Bunce RGH, Jongman RHG, Sayre R, Trabucco A, Zomer R. A high-resolution bioclimate map of the world: a unifying framework for global biodiversity research and monitoring. Glob Ecol Bio- geogr. 2013; 22: 630–638. https://doi.org/10.1111/geb.12022 19. FAOSTAT [Internet]. [cited 12 Feb 2018]. Available: http://www.fao.org/faostat/en/#data/QA 20. Nicolas G, Robinson TP, Wint GRW, Conchedda G, Cinardi G, Gilbert M. Using Random Forest to Improve the Downscaling of Global Livestock Census Data. PLOS ONE. 2016; 11: e0150424. https:// doi.org/10.1371/journal.pone.0150424 PMID: 26977807 21. Robinson TP, Wint GRW, Conchedda G, Van Boeckel TP, Ercoli V, Palamara E, et al. Mapping the Global Distribution of Livestock. PLoS ONE. 2014; 9: e96084. https://doi.org/10.1371/journal.pone. 0096084 PMID: 24875496 22. Stevens FR, Gaughan AE, Linard C, Tatem AJ. Disaggregating census data for population mapping using random forests with remotely-sensed and ancillary data. PloS One. 2015; 10: e0107042. https:// doi.org/10.1371/journal.pone.0107042 PMID: 25689585 23. Nelson A. Travel time to major cities: A global map of Accessibility [Internet]. Ispra Italy: Global Envi- ronment Monitoring Unit—Joint Research Centre of the European Commission; 2008. Available: Avail- able at http://gem.jrc.ec.europa.eu/ 24. Wasserman S, Faust K. Social Network Analysis: Methods and Applications. Cambridge University Press; 1994. 25. Beaune´e G, Vergu E, Ezanno P. Modelling of paratuberculosis spread between dairy cattle farms at a regional scale. Vet Res. 2015; 46. https://doi.org/10.1186/s13567-015-0247-3 PMID: 26407894 26. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 References Moslonka-Lefebvre M, Gilligan CA, Monod H, Belloc C, Ezanno P, Filipe JAN, et al. Market analyses of livestock trade networks to inform the prevention of joint economic and epidemiological risks. J R Soc Interface. 2016; 13. https://doi.org/10.1098/rsif.2015.1099 PMID: 26984191 27. Lichoti JK, Davies J, Kitala PM, Githigia SM, Okoth E, Maru Y, et al. Social network analysis provides insights into African swine fever epidemiology. Prev Vet Med. 2016; 126: 1–10. https://doi.org/10.1016/ j.prevetmed.2016.01.019 PMID: 26848113 28. Relun A, Grosbois V, Alexandrov T, Sa´nchez-Vizcaı´no JM, Waret-Szkuta A, Molia S, et al. Prediction of Pig Trade Movements in Different European Production Systems Using Exponential Random Graph Models. Front Vet Sci. 2017; 4. https://doi.org/10.3389/fvets.2017.00027 PMID: 28316972 29. Kiss IZ, Green DM, Kao RR. The network of sheep movements within Great Britain: network properties and their implications for infectious disease spread. J R Soc Interface. 2006; 3: 669–677. https://doi.org/ 10.1098/rsif.2006.0129 PMID: 16971335 20 / 21 PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 Livestock mobility in Mauritania 30. Volkova VV, Howey R, Savill NJ, Woolhouse MEJ. Sheep Movement Networks and the Transmission of Infectious Diseases. PLOS ONE. 2010; 5: e11185. https://doi.org/10.1371/journal.pone.0011185 PMID: 20567504 31. Lancelot R, Be´ral M, Rakotoharinome VM, Andriamandimby S-F, He´raud J-M, Coste C, et al. Drivers of Rift Valley fever epidemics in Madagascar. Proc Natl Acad Sci U S A. 2017; 114: 938–943. https://doi. org/10.1073/pnas.1607948114 PMID: 28096420 32. Simini F, Gonza´lez MC, Maritan A, Baraba´si A-L. A universal model for mobility and migration patterns. Nature. 2012; 484: 96–100. https://doi.org/10.1038/nature10856 PMID: 22367540 33. Balcan D, Colizza V, Gonc¸alves B, Hu H, Ramasco JJ, Vespignani A. Multiscale mobility networks and the spatial spreading of infectious diseases. Proc Natl Acad Sci. 2009; 106: 21484–21489. https://doi. org/10.1073/pnas.0906910106 PMID: 20018697 34. R Development Core Team. R: A language and Environment for Statistical Computing. Vienna Austria R Found Stat Comput. 2015; 35. Butts CT. sna: Tools for Social Network Analysis [Internet]. 2016. Available: https://cran.r-project.org/ web/packages/sna/index.html 36. Biteye B, Fall AG, Ciss M, Seck MT, Apolloni A, Fall M, et al. Ecological distribution and population dynamics of Rift Valley fever virus mosquito vectors (Diptera, Culicidae) in Senegal. Parasit Vectors. 2018; 11. https://doi.org/10.1186/s13071-017-2591-9 PMID: 29316967 37. El Mamy ABO, Baba MO, Barry Y, Isselmou K, Dia ML, Hampate B, et al. Unexpected Rift Valley Fever Outbreak, Northern Mauritania. Emerg Infect Dis. 2011; 17: 1894–1896. https://doi.org/10.3201/ eid1710.110397 PMID: 22000364 38. PLOS ONE \| https://doi.org/10.1371/journal.pone.0199547 July 18, 2018 References El Mamy AB, Lo MM, Thiongane Y, Diop M, Isselmou K, Doumbia B, et al. Comprehensive Phyloge- netic Reconstructions of Rift Valley Fever Virus: The 2010 Northern Mauritania Outbreak in the Came- lus dromedarius Species. Vector Borne Zoonotic Dis. 2014; 14: 856–861. https://doi.org/10.1089/vbz. 2014.1605 PMID: 25514121 39. Truscott J, Ferguson NM. Evaluating the Adequacy of Gravity Models as a Description of Human Mobil- ity for Epidemic Modelling. PLoS Comput Biol. 2012; 8. https://doi.org/10.1371/journal.pcbi.1002699 PMID: 23093917 40. Xia Y, Bjørnstad ON, Grenfell BT. Measles Metapopulation Dynamics: A Gravity Model for Epidemio- logical Coupling and Dynamics. Am Nat. 2004; 164: 267–281. https://doi.org/10.1086/422341 PMID: 15278849 41. Barrios JM, Verstraeten WW, Maes P, Aerts J-M, Farifteh J, Coppin P. Using the Gravity Model to Esti- mate the Spatial Spread of Vector-Borne Diseases. Int J Environ Res Public Health. 2012; 9: 4346– 4364. https://doi.org/10.3390/ijerph9124346 PMID: 23202882 42. Sarzynska M, Udiani O, Zhang N. A study of gravity-linked metapopulation models for the spatial spread of dengue fever. ArXiv13084589 Q-Bio Stat. 2013; Available: http://arxiv.org/abs/1308.4589 43. Hammami P, Lancelot R, Lesnoff M. Modelling the Dynamics of Post-Vaccination Immunity Rate in a Population of Sahelian Sheep after a Vaccination Campaign against Peste des Petits Ruminants Virus. PloS One. 2016; 11: e0161769. https://doi.org/10.1371/journal.pone.0161769 PMID: 27603710 21 / 21
https://openalex.org/W4379743553	https://www.ecologyandsociety.org/vol28/iss2/art24/ES-2023-14133.pdf	English	null	Toward spatial fit in the governance of global commodity flows	Ecology and society	2,023	cc-by	16,604	INTRODUCTION and non-material flows (Liu et al. 2013, Eakin et al. 2014, Friis et al. 2016). The telecoupling concept supports analysis of how social-ecological changes in one place are related to social- ecological processes elsewhere. Rather than confronting globalization as a diffuse, complex, and all-pervasive phenomenon, a focus on telecoupling helps to delineate and analyze particular connections, place-specific social and environmental impacts, and their (often remote) drivers in a globalizing world (Challies et al. 2014, Friis and Nielsen 2019, Sonderegger et al. 2020). Local sustainability problems are increasingly shaped by distal actors and processes through global flows of information, people, goods, and services. Demand for commodities such as palm oil, soy, meat, cocoa, and rubber produces negative social and environmental impacts, including deforestation, biodiversity loss, food insecurity, agri-chemical pollution, and consolidation of landholdings, in production regions that are often far removed from sites of consumption (Laroche et al. 2021, Cotta et al. 2022, Roux et al. 2022). Such sustainability problems often transcend traditional political boundaries, which makes it challenging to design governance institutions to fit the scale of the problems. Where governance institutions do not match the scale of the problems they are expected to address, scholars have diagnosed “problems of fit”, “mismatches”, or “misfits” (Young 2005, Folke et al. 2007, Galaz et al. 2008). The degree of fit may pertain to alignment between a given social-ecological problem and a governance response in spatial, temporal, or functional terms (Cumming et al. 2006, Folke et al. 2007). Issues of governance fit are well researched with regard to regionally bounded or transboundary social-ecological systems such as aquatic or riverine ecosystems (Moss 2012, Bergsten et al. 2014). However, research has not yet systematically explored solutions to spatial mismatches in social-ecological systems connected across long distances, so-called telecoupled systems (Sikor et al. 2013, Munroe et al. 2019, Newig et al. 2020). Governance in telecoupled systems is challenging because the drivers and effects of global flows often lie beyond the reach of national governments, companies, or citizens. Existing sustainability governance initiatives that govern global flows of agricultural and forestry commodities, such as corporate pledges, voluntary sustainability standards, public-private partnerships, and multistakeholder initiatives, are not necessarily effective in driving sustainable supply chains (Garrett et al. 2019, 2021, Grabs et al. 2021, Meemken et al. 2021). 1Institute of Sustainability Governance, Leuphana University Lüneburg, Germany, 2Centre for Development and Environment (CDE), University of Bern, Switzerland, 3Institute of Geography, University of Bern, Switzerland, 4Earth and Life Institute, UCLouvain, Louvain-la-Neuve, Belgium, 5Fonds de la Recherche Scientifique (F.R.S.-FNRS), Brussels, Belgium, 6School of Earth and Environment, University of Canterbury, New Zealand, 7Institute of Environmental Science and Technology (ICTA), Universitat Autònoma de Barcelona, Spain, 8Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain, 9Institute of Environmental Science and Technology & Department of Geography, Universitat Autònoma de Barcelona, Bellaterra, Spain, 10Department of Geosciences and Natural Resource Management, University of Copenhagen, Denmark, 11Geography Department, Humboldt-Universität zu Berlin, Germany, 12IRI-THESys, Humboldt-Universität zu Berlin, Germany, 13Department of Anthropology, Aarhus University, Denmark, 14Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, The Netherlands, 15Institute of Social Ecology (SEC), Department of Economics and Social Sciences, University of Natural Resources and Life Sciences, Vienna, Austria, 16Wyss Academy for Nature, University of Bern, Switzerland Synthesis Specifically, we draw on examples from land use and global agricultural commodity governance to examine two overarching types of governance mismatches: boundary mismatches and resolution mismatches. We argue that one way to address mismatches is through governance rescaling and illustrate this approach with reference to examples of three broad types of governance approaches: trade agreements, due diligence laws, and landscape approaches to supply chain governance. No single governance approach is likely to address all mismatches, highlighting the need to align multiple governance approaches to govern telecoupled systems effectively. Key Words: environmental governance; human-environment interactions; scale; spatial mismatch; supply chain; telecoupling Key Words: environmental governance; human-environment interactions; scale; spatial mismatch; supply chain; telecoupling Synthesis Toward spatial fit in the governance of global commodity flows Johanna Coenen 1 , Gabi Sonderegger 2,3 , Jens Newig 1 , Patrick Meyfroidt 4,5 , Edward Challies 1,6 , Simon L. Bager 4 , Louise M. Busck-Lumholt 7 , Esteve Corbera 8,9 , Cecilie Friis 10 , Anna Frohn Pedersen 11,12,13 , Perrine C.S.J. Laroche 14 , Claudia Parra Paitan 14 , Siyu Qin 11 , Nicolas Roux 15 and Julie G. Zaehringer 2,3,16 ABSTRACT. Global commodity flows between distally connected social-ecological systems pose important challenges to sustainability governance. These challenges are partly due to difficulties in designing and implementing governance institutions that fit or match the scale of the environmental and social problems generated in such telecoupled systems. We focus on the spatial dimension of governance fit in relation to global commodity flows and telecoupled systems. Specifically, we draw on examples from land use and global agricultural commodity governance to examine two overarching types of governance mismatches: boundary mismatches and resolution mismatches. We argue that one way to address mismatches is through governance rescaling and illustrate this approach with reference to examples of three broad types of governance approaches: trade agreements, due diligence laws, and landscape approaches to supply chain governance. No single governance approach is likely to address all mismatches, highlighting the need to align multiple governance approaches to govern telecoupled systems effectively. K W d i l h i i i l i l i h l h i l li Toward spatial fit in the governance of global commodity flows Johanna Coenen 1 , Gabi Sonderegger 2,3 , Jens Newig 1 , Patrick Meyfroidt 4,5 , Edward Challies 1,6 , Simon L. Bager 4 , Louise M. Busck-Lumholt 7 , Esteve Corbera 8,9 , Cecilie Friis 10 , Anna Frohn Pedersen 11,12,13 , Perrine C.S.J. Laroche 14 , Claudia Parra Paitan 14 , Siyu Qin 11 , Nicolas Roux 15 and Julie G. Zaehringer 2,3,16 Toward spatial fit in the governance of global commodity flow ABSTRACT. Global commodity flows between distally connected social-ecological systems pose important challenges to sustainability governance. These challenges are partly due to difficulties in designing and implementing governance institutions that fit or match the scale of the environmental and social problems generated in such telecoupled systems. We focus on the spatial dimension of governance fit in relation to global commodity flows and telecoupled systems. Copyright © 2023 by the author(s). Published here under license by the Resilience Alliance. Open Access. CC-BY 4.0 Coenen, J., G. Sonderegger, J. Newig, P. Meyfroidt, E. Challies, S. L. Bager, L. M. Busck-Lumholt, E. Corbera, C. Friis, A. Frohn Pedersen, P. C. S. J. Laroche, C. Parra Paitan, S. Qin, N. Roux, and J. G. Zaehringer. 2023. Toward spatial fit in the governance of global commodity flows. Ecology and Society 28(2):24. https://doi.org/10.5751/ES-14133-280224 Copyright © 2023 by the author(s). Published here under license by the Resilience Alliance. Open Access. CC-BY 4.0 Coenen, J., G. Sonderegger, J. Newig, P. Meyfroidt, E. Challies, S. L. Bager, L. M. Busck-Lumholt, E. Corbera, C. Friis, A. Frohn Pedersen, P. C. S. J. Laroche, C. Parra Paitan, S. Qin, N. Roux, and J. G. Zaehringer. 2023. Toward spatial fit in the governance of global commodity flows. Ecology and Society 28(2):24. https://doi.org/10.5751/ES-14133-280224 THE PROBLEM OF FIT Fundamentally, the problem of fit concerns the question of how to scale or rescale governance arrangements so that they have the best possible institutional fit with the targeted social-ecological dynamics. Establishing the most appropriate fit requires a trade- off between the advantages of better coordination at higher scales, which may reduce the risk of overlooking spatial externalities, and the risk of lacking context sensitivity and legitimacy among local actors, impeding effective implementation (Newig and Moss 2017). Importantly, problems do not occur at a single scale that is objectively given, but different actors perceive and frame problems at different scales and levels (Padt et al. 2014). For example, if state actors aim to meet forest restoration commitments made under international agreements and frame the problem solely at an ecological scale, a national afforestation program fits with the objective of forest restoration for carbon storage. However, if the problem is framed at a social-ecological scale, a single homogeneous afforestation program may suffer from a resolution mismatch and fail to address context-specific challenges related to rural livelihoods (Wiegant et al. 2020, Coleman et al. 2021). Thus, evaluations of fit depend upon how a problem is framed and by whom (Epstein et al. 2015). What is perceived as the “optimal scale” may vary among actors, and the scale at which they define a problem will influence their preferences for governance rescaling. For example, political and societal actors may strategically frame certain problems at the global scale if they perceive national governments as a possible hindrance to solving the problem, or if they want to avoid assuming responsibility and implementing domestic measures (Gupta 2014). The problem of fit has been widely researched in political science and social-ecological systems literature. Scholars have examined mismatches between the spatial, temporal, and functional scales of governance institutions and the scales of social-ecological processes (Cumming et al. 2006, Folke et al. 2007, Galaz et al. 2008, Ekstrom and Young 2009, Epstein et al. 2015). Here, scale is understood as “the various levels at which a phenomenon occurs in the dimensions of space and time” (Young 2002a:26). Because of institutional mismatches, governance responses to environmental threats often struggle to address the full extent of the problem (Ekstrom and Crona 2017). For example, drivers of land-use change operate at multiple levels and spatial scales. INTRODUCTION Research has attributed the ineffectiveness of governance interventions in part to mismatches between the scale of the governance institution and the scale of the underlying problem (Young 2005). Here, we explore the problem of spatial fit between governance arrangements and the social-ecological problems they address in relation to land use, as well as global agricultural commodity governance and telecoupled systems more broadly. We focus specifically on the question of spatial fit because telecoupled sustainability problems are inherently related to issues of spatial Telecoupling denotes long-distance connections between two or more social-ecological systems that are linked through material Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety org/vol28/iss2/art24/ scale. We distinguish two overarching types of spatial mismatches: boundary mismatches and resolution mismatches, building on previous work by Cumming et al. (2006) and Bergsten et al. (2014). Whereas boundary mismatches denote situations in which social- ecological processes transcend governance boundaries, resolution mismatches refer to governance schemes designed at too coarse a spatial scale to effectively address the issue at hand (Bergsten et al. 2014).[1] We present illustrative empirical examples from land and global agricultural commodity governance to elucidate how problems of spatial fit impede the effective governance of land and land-based resources in telecoupled systems. We also examine governance approaches to address this problem. We contend that a better understanding of the types of mismatches that arise in efforts to govern global commodity flows will contribute to identification of leverage points for effective governance interventions in telecoupled systems (Carrasco et al. 2017, Munroe et al. 2019, Newig et al. 2020). to “come to terms with local variations in biogeophysical conditions and [lack] sensitivity to both the knowledge and the rights and interests of local stakeholders” (Young 2002b:283; see also Ostrom 1990). For example, much of European Union legislation has been criticized for being too insensitive to local contexts, despite the EU’s principle of subsidiarity (Article 5 Treaty on European Union), which demands that decisions should be taken at the most appropriate level of governance, and that the EU should only take action when national, regional, or local governments are unable to achieve a particular objective. The EU Water Framework Directive provides an example of governance that seeks to avoid resolution mismatches. It requires member states to develop River Basin Management Plans to guide local and context-specific implementation (Jager et al. 2016). INTRODUCTION An institutional fit emerges if the governance scale equals the social- ecological scale (Fig. 1E), as in the case of the global agreements reached in the Montreal Protocol on Substances that Deplete the Ozone Layer to address a global problem (Epstein et al. 2014). THE PROBLEM OF FIT International trade, regional development policies, national property rights regimes, and local people’s agricultural practices are among the many factors that may lead to land conversion (Geist and Lambin 2002). However, governance mechanisms typically target a single level (e.g., national forestry laws), and thus do not provide adequate solutions to the challenge of governing wider resource systems (Nagendra and Ostrom 2012). Governance arrangements that only partially cover the resource or ecosystem in question have built-in limitations that impede their ability to fulfill their goals (Young 2005). Various possible configurations of spatial mismatches exist (Fig. 1). The governance scale may be smaller than the social-ecological system scale (Fig. 1A). For example, a municipality may not be able to effectively address air pollution, which is caused by local factories but dispersed beyond municipal boundaries. Governance at larger scales, such as national regulations, may solve the problem (upscaling of governance). Similarly, the governance scale may only partially cover the social-ecological scale (Fig. 1B), as is often the case, for example, with governance of transboundary rivers. In such situations, upscaling may be more difficult in the absence of an authority at a higher governing level. Moreover, governance institutions and actors may have no jurisdiction at all over the social-ecological scale of an identified problem (Fig. 1C), such as in the case of a country lacking the authority to regulate illegal logging by a company domiciled in the country but operating in a neighboring country. Lastly, the governance scale may be greater than the social-ecological scale (Fig. 1D). In such cases, regulation at a (much) larger scale than that of the ecological problem may lack the regulatory specificity Here, we build on the concept of institutional fit, which is based on the underlying normative assumption that institutional scale can be optimized to avoid spatial externalities (Moss and Newig 2010). Thus, we focus on how individual institutions face this problem of fit. Nevertheless, we recognize that governance always involves the interplay of different institutions. Analysis of institutional fit is closely linked to the analysis of institutional interplay because social-ecological problems are typically governed by various institutions at different spatial scales (Young 2002a). Although no institution operates in a vacuum, it can be useful to assess the spatial fit of a specific institution in isolation from the broader institutional landscape. This approach simplifies Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Fig. 1. THE PROBLEM OF FIT Scale (mis-)matches between social-ecological (green) and governance (orange) scales. (A–C) Boundary mismatches. The institutional boundaries do not match with the spatial boundaries of the social-ecological problem, creating spatial spillover effects. (D) Resolution mismatch. The governance institution does not fit the specifics of the (local) social-ecological context that is to be addressed by governance and hence lacks sufficient spatial specificity. A single governance institution typically addresses a class of social-ecological problems that occurs in multiple distinct localities that have specific contextual features, to which a single governance institution cannot necessarily be adjusted. (E) Spatial fit. Illustration inspired by Newig et al. (2013:13). telecoupling is increasingly addressing this research gap by investigating the causes, drivers, and implications of globally linked social-ecological systems. Telecoupling research has referred to the problem of mismatches, but the definition and application of the concept in the context of telecoupling remains limited (Oberlack et al. 2018, Munroe et al. 2019, Zaehringer et al. 2019, Newig et al. 2020). The important question of how to align the scale of governance with the scale of the social-ecological problem at hand remains largely unaddressed in research on governing telecoupled social-ecological systems. telecoupling is increasingly addressing this research gap by investigating the causes, drivers, and implications of globally linked social-ecological systems. Telecoupling research has referred to the problem of mismatches, but the definition and application of the concept in the context of telecoupling remains limited (Oberlack et al. 2018, Munroe et al. 2019, Zaehringer et al. 2019, Newig et al. 2020). The important question of how to align the scale of governance with the scale of the social-ecological problem at hand remains largely unaddressed in research on governing telecoupled social-ecological systems. the analysis and does not consider all interdependencies, but it enhances analytical tractability and makes it easier to identify governance weaknesses and gaps (Young 2005). The analysis of institutional mismatches can be complemented with considerations of how to create linkages and facilitate interactions among various institutions. We return to considerations of the relation between institutional fit and interplay below. THE PROBLEM OF SPATIAL FIT IN TELECOUPLED SYSTEMS With transboundary water resources, for example, withdrawals in one place affect downstream availability. In telecoupled systems, however, there is usually no such direct ecological feedback. For example, tropical ecosystem degradation driven by commodity production for export to European markets causes biodiversity loss in producing regions or carbon emissions, but does not directly affect European consumers in the short term. Where feedbacks are delayed or indirect, it is also difficult to attribute specific social-ecological effects to particular activities (Carlson et al. 2018). Consequently, the actors driving telecoupled interactions do not necessarily experience the negative effects of their actions or recognize the connections between past actions and subsequent negative effects (Newig et al. 2020). They may therefore have very little incentive to formulate or adapt governance responses. Fig. 2. Ideal types of interconnected social-ecological systems and their interdependencies. Systems comprise socioeconomic building blocks (blue), ecological building blocks (green), and their interdependencies (arrows). (A) In a regionally bounded system, two socioeconomic systems share the same ecological resource base; e.g., two communities harvest wood from the same forest. (B) In a transboundary system, two socioeconomic systems rely on resources or ecosystems that are ecologically connected; e.g., pollution of a river by an upstream riparian country may affect fish populations in a downstream riparian country. (C) In telecoupled systems, the ecological systems are geographically separate but are connected through social- ecological processes such as trade in agricultural commodities. Fig. 2. Ideal types of interconnected social-ecological systems and their interdependencies. Systems comprise socioeconomic building blocks (blue), ecological building blocks (green), and their interdependencies (arrows). (A) In a regionally bounded system, two socioeconomic systems share the same ecological resource base; e.g., two communities harvest wood from the same forest. (B) In a transboundary system, two socioeconomic systems rely on resources or ecosystems that are ecologically connected; e.g., pollution of a river by an upstream riparian country may affect fish populations in a downstream riparian country. (C) In telecoupled systems, the ecological systems are geographically separate but are connected through social- ecological processes such as trade in agricultural commodities. Second, as a result of the above situation, recognition of and concern about specific problems may depend on social or political actors highlighting causal linkages between certain actions and distant outcomes. “Problem-brokers” or “political entrepreneurs” can play important roles in framing and problematizing unsustainable connections between telecoupled systems (Bastos Lima et al. 2019, Meyfroidt et al. 2022). THE PROBLEM OF SPATIAL FIT IN TELECOUPLED SYSTEMS Research on institutional fit has primarily focused on cases of natural resources in specific social-ecological systems. Studies have been conducted on forest governance (Shkaruba and Kireyeu 2013, Bodin et al. 2014, Melnykovych et al. 2018), water governance (Lebel et al. 2005, 2013, Moss 2012, Enqvist et al. 2020), and land and wildlife management (Bergsten et al. 2014, Dressel et al. 2018). Most research has focused on mismatches between local, regional, and national governance institutions and the social-ecological systems they target, but a small and growing pool of literature investigates transboundary and larger scale social-ecological problems such as depletion of the ozone layer or pollution of international watersheds (Cox et al. 2014). Challies et al. (2014) observe that social-ecological systems research itself has mostly examined small, tightly coupled systems, rather than connections and interdependencies that exist between multiple social-ecological systems linked through global production networks and supply chains (Nyström et al. 2019). Research on Telecoupling is one distinct ideal-typical configuration of interdependent social-ecological systems (Fig. 2). Telecoupled systems arise when the activities of actors in one system affect a social-ecological system elsewhere (e.g., through international trade or the displacement of extractive activities from one place to another), thereby creating social-ecological interdependencies. Consequently, feedbacks can develop, for example, when actors in one location become aware of the displaced effects of their actions and seek to mitigate them through measures such as increased conservation funding. Telecoupled systems are characterized by geographical distance between the place where the social or environmental impacts occur and the places where underlying causes are found. The geographical distance is often associated with social and institutional distances between the socioeconomic systems (Eakin Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety org/vol28/iss2/art24/ et al. 2014, Niewöhner et al. 2016, Friis and Nielsen 2017) because they tend to be governed by different, functionally independent institutional arrangements, social networks, and actors (Eakin et al. 2017). Even when distant actors are willing to work together, transaction costs of cooperating on sustainability issues are often much higher than in local or transboundary settings (Newig et al. 2020). Geographical, social, and institutional distances thus hinder the creation of appropriately scaled governance institutions in telecoupled systems in at least four ways. of resource users will have direct effects on other users (Lebel et al. 2005, Bergsten et al. 2014, Kininmonth et al. 2015). THE PROBLEM OF SPATIAL FIT IN TELECOUPLED SYSTEMS However, the EU’s ability to govern these issues has clear limitations given the national sovereignty of external countries and World Trade Organization rules. their suppliers and therefore lack the ability to control the environmental and socioeconomic effects of production. For example, approximately one-quarter of the solid wood furniture that IKEA sells is manufactured in Chinese factories that source their timber from other countries, in particular Russia (Newell and Simeone 2014). IKEA attempted to control the timber sourcing of its Chinese subcontractors to “green” its supply chain but was unsuccessful because of the geographical distance to upstream activities, the large number of intermediaries between timber extraction and retail, and an inability to trace timber to a specific logging permit (Goldstein and Newell 2020). Additionally, supply chain configurations change over time (dos Reis et al. 2020). China has long depended on Russian wood for the manufacture of finished wood products for export to the United States, but the specific companies within these supply chains change regularly (Goldstein and Newell 2020). Even where large, powerful retailers dictate prices and quality standards to their suppliers, their ability to control sustainability along the value chain is often limited because of the mismatch between their governance reach and the scale of the social- ecological problem. Companies are often not able to monitor their indirect suppliers, which makes it difficult to implement chain-wide sustainability policies (zu Ermgassen et al. 2022). THE PROBLEM OF SPATIAL FIT IN TELECOUPLED SYSTEMS Once distant ecological or social conditions attract sufficient public attention and concern, a policy window opens wherein various governance interventions may become possible (Kingdon 1984, Eakin et al. 2017). Improved transparency, through the collection and dissemination of information on flows and impacts, can enable or instigate governance responses to telecoupled issues (Gardner et al. 2019). For instance, increasing media attention on environmental issues such as deforestation has put pressure on the EU to address soybean production in the Amazon region (Mempel and Corbera 2021). Several interventions have emerged to tackle deforestation embedded in international trade and to reduce “imported deforestation” from EU consumption (Bager et al. 2021). First, the absence of manifest ecological feedbacks between telecoupled systems obscures the remote causes and effects of certain decisions and actions. In many locally bounded or closely neighboring social-ecological systems, the activities of one group Third, governance mismatches arise when governance responses misdiagnose a problem or neglect its wider drivers. Interventions that target only the direct ecological effects of an activity risk merely displacing it to other social-ecological systems. For example, European demand for soy is associated with negative ecological impacts such as deforestation in producer countries (Pendrill et al. 2019, Schilling-Vacaflor et al. 2021). Addressing tropical deforestation at the scale of a single region such as the Amazon is unlikely to be effective because demand for forest-risk commodities will persist. Therefore, governance interventions such as the Brazilian Soy Moratorium, which targets the Amazon specifically, have displaced deforestation to other areas such as the Cerrado region (Dou et al. 2018). Fourth, the places and governance institutions implicated in telecoupled systems may have very little history of prior collaboration (Newig et al. 2020). The social and institutional distance between telecoupled systems may mean that separate policies, actors, and networks govern largely independently. In the absence of joint institutional structures, governing telecoupled systems is challenging because governance actors face issues that extend beyond their jurisdiction. For example, consumption in First, the absence of manifest ecological feedbacks between telecoupled systems obscures the remote causes and effects of certain decisions and actions. In many locally bounded or closely neighboring social-ecological systems, the activities of one group Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety org/vol28/iss2/art24/ the EU has social-ecological effects beyond EU borders (Kastner et al. 2015, Dorninger et al. 2021, Roux et al. 2021). Spillover In case of spillovers (Fig. 3A), part of the problem remains unaddressed because it lies outside the domain of the governance institution. The omitted part of the problem is referred to as a spillover, which is broadly understood as an indirect effect of an activity or intervention (e.g., policy, program, or new technology) that occurs outside the targeted area (Meyfroidt et al. 2020). Spillovers emerge because governance actors may not be aware of the full scale of the social-ecological problem, may be uninterested in or unable to govern what happens beyond their jurisdictional boundaries, or may intentionally neglect parts of the problem (Bastos Lima et al. 2019). For example, voluntary sustainability standards often focus on reducing harmful on-farm effects at sites of production but tend to neglect off-farm effects such as reduced downstream water availability or air pollution from pesticide use (Zaehringer et al. 2018, Parra-Paitan and Verburg 2022, Sonderegger et al. 2022). Spillovers can also cascade to further social-ecological systems (as indicated in Fig. 3A) and have cumulative effects, which makes it difficult to identify causal connections (Busck-Lumholt et al 2022a). Boundary mismatches in telecoupled systems Boundary mismatches arise in telecoupled systems when the spatial reach of governance structures is such that these structures do not internalize existing social-ecological externalities of activities (i.e., spillovers; Fig. 3A) or when public policies or transnational economic activities produce new externalities (i.e., leakages; Fig. 3B). Spillovers describe events or developments that are not targeted by a given governance intervention, whereas leakages are a form of spillover caused by a governance intervention (Meyfroidt et al. 2020). Second, companies may govern particular segments of their supply chain but neglect others, which constitutes a boundary mismatch if the goal is to create sustainable supply chains that encompass the full value chain. For example, textile certifications generally focus on either the upstream end of the supply chain (i.e., organic and fair cotton production) or the midstream section (i.e., working conditions of garment workers; Partzsch 2020), but seldom cover all segments of the supply chain. DIFFERENT TYPES OF MISMATCHES IN TELECOUPLED SYSTEMS We apply the concepts of boundary and resolution mismatches to telecoupled systems. We identify the underlying governance problem associated with each type of mismatch, outline two particular mechanisms of boundary mismatches and illustrate with examples from both public and private governance perspectives (Table 1). Our distinction between ideal-typical configurations of mismatches helps in elaborating how the scale of governance institutions often does not align with the scale of social-ecological problems.[2] Leakage A leakage may emerge when a governance intervention induces externalities (Fig. 3B). The governance intervention produces effects that contradict its objectives and reduce the overall benefit of the interventions, which constitutes a leakage effect (Meyfroidt et al. 2018, Bastos Lima et al. 2019). For example, the EU’s Renewable Energy Directive created additional demand for biofuel crops produced outside of the EU and thereby fuelled land-use change and deforestation in tropical countries, counteracting the goal of reducing greenhouse gas emissions (Bastos Lima 2021). This process has also been described as “governance inducing telecoupling” (Newig et al. 2019), i.e., situations in which governance initiatives themselves create new distal interactions with positive or negative outcomes. Recognition of the negative distal effects led to revision of the Renewable Energy Directive to mitigate indirect land-use change (Bastos Lima 2021). In other instances, the leakage effect does not occur across a great distance but can be in proximity to the target area. For instance, if a forest moratorium prohibits deforestation within designated areas, the activity may simply shift to nearby areas not covered by the moratorium (Meyfroidt et al. 2010, Leijten et al. 2021). The transnational operations of companies make it challenging to achieve institutional fit and to internalize the extra- jurisdictional social and environmental effects of global supply chains. Because multinational enterprises operate beyond the jurisdictional reach of individual states, the externalities of their activities are often not addressed by existing governance institutions. These actors are not accountable to any single authority that matches their scope of operation (Kobrin 2009). Just like public governance, private governance can have spillover effects and leakages. For instance, if private conservation actors focus their efforts on specific regions such as the Brazilian Amazon, that leaves other regions such as the Cerrado and Gran Chaco comparatively less well protected, and land conversion Private actors may encounter boundary mismatches in their efforts to govern supply chains for two reasons. First, individual companies may lack oversight and influence over some or all of Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Fig. 3. Boundary mismatches. Governance institutions neglect social-ecological problems that transcend established jurisdictional boundaries due to spillovers (A) or leakages (B). Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety.org/vol28/iss2/art24/ Fig. 3. Boundary mismatches. Governance institutions neglect social-ecological problems that transcend established jurisdictional boundaries due to spillovers (A) or leakages (B). Leakage institutions are not supranational, meaning that they do not have authority beyond that of their respective members, they rely on lower-level institutions for implementation, which, however, have limited abilities to govern the causes or effects of cross-border flows beyond their jurisdictional boundaries. If the implementation pathway is not defined and lower-level institutions have neither the capacity nor the experience to implement higher-level governance objectives, a spatial scale challenge emerges (Wiegant et al. 2020). Global environmental governance is often directed toward reaching global targets (e.g., Paris Agreement, Aichi Biodiversity Targets, Bonn Challenge). However, target-based governance has been criticized for the gap between international policy and national implementation, the missing linkages between national governments and on-the- ground actions, and the unclear definitions of some wording of the targets (Hagerman et al. 2021, Perino et al. 2022). may be displaced to those regions (Soterroni et al. 2019, Qin et al. 2022). In short, leakage occurs when the side effects of an intervention escape the scope of governance. †Adapted from Bergsten et al. (2014). ‡ In the case of a World Bank conservation project in Argentina, project concepts and ideas were decided by external actors, rather than in partnership with local beneficiaries (Busck-Lumholt et al. 2022b). Sustainability issues prioritized at the global scale may not match with local people’s understanding of and aspirations for sustainability. Self-governance and local rule development have been found to be highly important for effective natural resource management (Ostrom 1990). Otherwise, there is a high risk that international or transnational governance schemes are insufficiently adapted to local contexts. If governance actors perceive that transnational institutions do not fit the local contexts (i.e., social mismatch as result of a resolution mismatch), they may create their own institutions. This situation occurred with the establishment of the Icelandic Responsible Fisheries certification program as an alternative to the transnational Marine Stewardship Council certification scheme (Foley 2017), and with the introduction of Indonesian and Malaysian Sustainable Palm Oil schemes as alternatives to the Roundtable on Sustainable Palm Oil (Higgins and Richards 2019). †Adapted from Bergsten et al. (2014). ‡ Adapted from Bergsten et al. (2014). ‡We present the different types of mismatches from both public policy and private governance perspectives because their analytical focus differs. From a public policy perspective, the focus is on the jurisdictional scale, defined as clearly bounded political units (e.g., towns, provinces, states, or countries; Cash et al. 2006). In contrast, the private governance perspective puts more emphasis on the scale of the supply chain or associated flows. §The newly adopted EU Regulation on deforestation-free supply chains addresses this mismatch (European Commission 2022). It is expected to enter into force in summer 2023. Once it is in force, operators and traders will have 18 months to implement the new rules. As a result of resolution mismatches, new kinds of mismatches may emerge when governing institutions do not reflect the values, interests, and beliefs of different social groups. What Epstein et al. (2015) have termed “social mismatches” points to the spatial scalar challenge of matching governance objectives and rules with social customs and patterns of resource use, stakeholder expectations and needs, and social organization scales (Epstein et al. 2015). In telecoupled systems, international governance based on global goals carries a clear risk of diverging from issues that are seen as most important by local stakeholders. Global initiatives such as the Kimberley Process, for example, promote transparency in supply chains, but in so doing, they risk favoring global ideals (e.g., of traceability and accountability) over the day- to-day needs and concerns of local communities (Pedersen et al. 2021a). Research on gold mining in Tanzania, for instance, found that a centrally imposed transparency initiative had not addressed inequalities, informal structures, and power asymmetries in the mining sector (Pedersen et al. 2021b). Likewise, conservation projects that are governed by external actors (such as states, international nongovernmental organizations, or private firms) tend to subordinate local institutions, customary practices, and traditional ecological knowledge, resulting in relatively ineffective conservation management (Dawson et al. 2021). International conservation initiatives may overlook social and political complexities in local systems and create unintended and undesirable effects, including restricted access to land and natural resources and the erosion of customary natural resource governance institutions (Persson and Mertz 2019, Persson et al. 2021). If local people are merely seen as recipients of services and are not involved in the design of sustainability interventions, a mismatch between local goals and strategies and those of the wider project can emerge. Resolution mismatches in telecoupled systems Boundary mismatch Resolution mismatch Definition † Governance institutions neglect social-ecological problems that transcend established administrative or jurisdictional boundaries Governance institutions have too coarse a spatial resolution than is suitable to address the social-ecological bl t h d Governance institutions have too coarse a spatial resolution than is suitable to address the social-ecological problems at hand Lack of governance precision Panacea trap Governance institutions are not specific enough to be effectively implemented and enforced Leakage g Governance institutions address a social-ecological problem but create leakage(s), i.e., counterproductive effects outside the targeted area or domain of the intervention A forest moratorium shifts deforestation to neighboring areas or other countries, producing negative externalities in distant jurisdictions A Multilateral Environmental Agreement that is too broad in scope to govern particular telecoupled flows Supply chain actors set broad sustainability goals that are insufficiently operationalized and lack specific and measurable targets, unambiguous definitions, and exact coverage Supply chain actors implement zero-deforestation policies that target only one region, allowing actors in other regions or neighboring countries to deforest Example from a private governance perspective A Voluntary Sustainability Standard focuses on reducing harmful on-farm impacts at sites of production but neglects sustainability issues outside the farm such as air pollution from pesticide use †Adapted from Bergsten et al. (2014). ‡ Resolution mismatches in telecoupled systems Resolution mismatches represent a second problem of governance fit in telecoupled systems (Fig. 4). Because international or transnational governance institutions usually aim to address a social-ecological problem that occurs in more than one place, they are not specific to the social and ecological attributes of a particular social-ecological system or a particular telecoupling. If governance occurs at too coarse a scale, meaning that governance instruments are not context sensitive or flow specific, they are unlikely to be successful because “one-size-fits- all” panaceas do not exist (Ostrom et al. 2007, Meyfroidt et al. 2022). For example, international governance schemes such as Multilateral Environmental Agreements tend to be too general to govern specific telecoupled systems because international conventions, agreements, and commitments typically involve a large number of signatories, have a general thematic scope, and are not specific to any particular flow.[3] Of approximately 250 Multilateral Environmental Agreements worldwide, only 15 explicitly include trade-related provisions for environmental protection (World Trade Organization 2021). International governance schemes cover a large spatial scale and require a broad institutional outlook that can be implemented in heterogeneous national and local contexts. Because most international In the context of private governance, supply chain actors may set broad, blanket-coverage sustainability goals that are meant to apply across entire supply chains but are, for that reason, ambiguously defined, limited in scope, and poorly operationalized in terms of concrete and measurable targets. For example, in a sample of 513 companies in the coffee sector, only one-third reported tangible commitments to sustainability, whereas the remaining companies reported no or vague commitments (Bager and Lambin 2020). Similarly, companies may adopt zero-deforestation commitments without setting clear implementation goals, mechanisms, or deadlines, which impedes effective implementation across the contexts in which they operate (Garrett et al. 2019). Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Table 1. Resolution mismatches in telecoupled systems Boundary and resolution mismatches in Boundary Definition † Governance institutions neglect soci established administrative or jurisdic Underlying problem Lack of governance extent Mechanism Spillover Description Governance institutions do not govern a social-ecological problem that expands beyond their administrative or jurisdictional boundaries Example from a public policy perspective ‡ European countries have not (yet) implemented specific public policies to mitigate the deforestation effects of their demand for soy in remote jurisdictions § Example from a private governance perspective A Voluntary Sustainability Standard focuses on reducing harmful on-farm impacts at sites of production but neglects sustainability issues outside the farm such as air pollution from pesticide use Table 1. Boundary and resolution mismatches in the governance of telecoupled social-ecological systems. Boundary mismatch Resolution mismatch Definition † Governance institutions neglect social-ecological problems that transcend established administrative or jurisdictional boundaries Governance institutions have too coarse a spatial resolution than is suitable to address the social-ecological problems at hand Underlying problem Lack of governance extent Lack of governance precision Mechanism Spillover Leakage Panacea trap Description Governance institutions do not govern a social-ecological problem that expands beyond their administrative or jurisdictional boundaries Governance institutions address a social-ecological problem but create leakage(s), i.e., counterproductive effects outside the targeted area or domain of the intervention Governance institutions are not specific enough to be effectively implemented and enforced Example from a public policy perspective ‡ European countries have not (yet) implemented specific public policies to mitigate the deforestation effects of their demand for soy in remote jurisdictions § A forest moratorium shifts deforestation to neighboring areas or other countries, producing negative externalities in distant jurisdictions A Multilateral Environmental Agreement that is too broad in scope to govern particular telecoupled flows Example from a private governance perspective A Voluntary Sustainability Standard focuses on reducing harmful on-farm impacts at sites of production but neglects sustainability issues outside the farm such as air pollution from pesticide use Supply chain actors implement zero-deforestation policies that target only one region, allowing actors in other regions or neighboring countries to deforest Supply chain actors set broad sustainability goals that are insufficiently operationalized and lack specific and measurable targets, unambiguous definitions, and exact coverage † Table 1. Boundary and resolution mismatches in the governance of telecoupled social-ecological systems. ADDRESSING MISMATCHES IN TELECOUPLED SYSTEMS These examples suggest that global commodity flows, through boundary and resolution mismatches, pose multiple environmental governance challenges that are difficult to address through territorial or global governance approaches. Against this background, both public and private actors have attempted to rescale governance to account for social-ecological interactions Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Fig. 4. Resolution mismatches. Governance institutions have a coarser scale than is suitable to address the social-ecological problems they target. Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Fig. 4. Resolution mismatches. Governance institutions have a coarser scale than is suitable to address the social-ecological problems they target across long distances and between jurisdictions. With respect to global governance, governance rescaling has been defined as “a shift in the locus, agency, and scope of global [...] politics and governance across scales” (Andonova and Mitchell 2010:257). Scaling up governance to make it more comprehensive in terms of target area, actors, or supply chain segments can limit the risk of boundary mismatches. In contrast, scaling down governance might enhance the context sensitivity of interventions and the participation of local stakeholders, thus correcting resolution mismatches. Additionally, creating new governance scales can be another strategy to avoid mismatches. In telecoupled systems, such governance institutions comprise due diligence laws, as elucidated below. We next present three illustrative examples of public, private, and hybrid governance forms to illustrate the opportunities and challenges involved in addressing both boundary and resolution mismatches. flows may shift geographically (i.e., trade diversion), and regulated commodities may be replaced by less regulated or unregulated commodities within supply chains (i.e., substitution effect). For example, the U.S.-Peru trade agreement includes a binding Forest Annex, which details measures to strengthen forest governance in Peru, including the establishment of chain-of- custody systems to verify the legality of timber exports. However, because the Forest Annex is strongly focused on protecting CITES-listed timber species, one risk is that it increases exports of species not listed in CITES. It could also prompt U.S. importers to switch to other, less regulated markets (Del Gatto et al. 2009). Governance institutions that target specific geographic areas or commodities risk creating boundary mismatches. This situation suggests that trade agreements may be more effective at reducing leakage effects at regional scales when they contain binding, measurable, and enforceable sustainability chapters, and they involve regional blocs rather than individual countries, and commodity groups rather than single commodities. ADDRESSING MISMATCHES IN TELECOUPLED SYSTEMS However, the risk of resolution mismatches increases when the spatial scale of trade agreements increases. Landscape or jurisdictional approaches to supply chain governance Landscape approaches aim to reconcile competing social, economic, and environmental interests and objectives at the landscape scale. Landscape approaches have been widely employed in international conservation projects and are now also increasingly taken up in sustainable supply chain management (Sayer et al. 2013, Boshoven et al. 2021). They are based on multistakeholder collaboration (e.g., public authorities, producers, companies, civil society organizations), which sets them apart from purely public jurisdictional governance approaches that do not seek to involve all affected stakeholders. These relatively recent governance approaches rest on the premise that the involvement of public actors allows for the implementation and enforcement of mandatory requirements for production practices, provided that enforcement capacities exist (Bager 2021). Public actors have regulatory authority over the area covered, “allowing for better monitoring and enforcement as well as addressing the problem of institutional mismatch” (von Essen and Lambin 2021:6–7). A jurisdictional approach is a type of landscape approach that uses formal administrative boundaries to define the scope of action and involvement of stakeholders (Denier et al. 2015). Due diligence requirements often apply to specific commodities, as in the case of the EU Timber Regulation, which prohibits the sale of illegally harvested wood on the EU market, and the EU Renewable Energy Directive, under which member states can count biofuels toward the attainment of their renewable energy targets only if the biofuel production complies with certain sustainability criteria (European Union 2018), irrespective of whether the biofuel crops are produced inside or outside the EU (Scott 2020). Additionally, the EU adopted a Regulation on deforestation-free supply chains in December 2022, which prohibits the placing of palm oil, soy, wood, cattle, cocoa, coffee, rubber, and some derived products on the EU market if these commodities are linked to deforestation and forest degradation or if they are non-compliant with all relevant applicable laws in force in the country of production (European Commission 2022). These sector-specific due diligence policies use conditional market access as a mechanism to secure foreign producers’ compliance with EU rules. More recently developed, economy-wide, mandatory due diligence laws, at the national and European levels, rely on another governance mechanism, namely self-reporting and public scrutiny. The French Duty of Vigilance Law, for example, requires companies to assess and report the risks of infringing environmental and human rights in their supply chains, as well as measures to mitigate such risks. Social and environmental provisions in trade agreements The inclusion of binding, measurable, carefully monitored, and sanctionable social and environmental provisions in preferential or regional trade agreements presents a potential instrument to govern trade-related environmental impacts between specific countries or regions (Kehoe et al. 2020). Recently, researchers have advocated shifting focus on the relation between trade and the environment away from merely mitigating the negative impacts of trade, and toward focusing on how to harness the positive environmental effects of trade through, for example, the use of so-called “trade-and-environment agreements” (Roux et al. 2021; https://ieep.eu/news/a-cup-of-trade-and-environment- agreement-tea/). In theory, environmental provisions in trade agreements can oblige parties to uphold environmental law and implement “Multilateral Environmental Agreements”; increase cooperation, transparency, and participation in environmental matters; and trigger the uptake of voluntary sustainability standards and public regulations targeted at sustainability issues of a specific sector or product. However, empirical evidence of the actual environmental effects of environmental provisions in trade agreements is scarce and inconclusive (Berger et al. 2020). Trade agreements can suffer from resolution mismatches. For example, Berger et al. (2020) reviewed 48 preferential trade agreements of five emerging economies and found that three- quarters of the agreements make reference to general environmental goals in their preamble or other chapters. However, these provisions are not of substantive nature, meaning that they do not imply any substantive rights or obligations in environmental matters to the parties. Additionally, some countries restate their commitment to ratify or implement Multilateral Environmental Agreements in their trade agreements, thus, only restating the pledges already made elsewhere. If countries only make commitments to general environmental goals and international conventions without defining concrete actions in their trade agreements, they are unlikely to address the specific social and ecological problems of telecoupling in particular social-ecological systems. Moreover, if the needs and priorities of local communities are overlooked or deprioritized, social mismatches may arise. Failure to recognize the economic, social, and environmental concerns of Although trade agreements do address specific flows at the scale of telecoupled relations, they pose a risk of leakage because trade Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety org/vol28/iss2/art24/ countries, in principle, have no jurisdiction over producing countries, where sustainability problems appear. affected communities can also induce a boundary mismatch. Due diligence obligations and laws The proliferation of due diligence policies shows that public sector actors increasingly govern social and environmental conduct beyond their own borders. Due diligence policies are a clear example of “rescaling” or “territorial extension”, whereby states or groups of states extend their regulatory influence to actions abroad (Scott 2020). Although due diligence laws are implemented within formal administrative boundaries on a jurisdictional scale, they govern extra-jurisdictional processes by obliging transnational companies to monitor their supply chains and to rectify unsustainable impacts. Due diligence policies tend to be applied at scales applicable to telecoupled systems because they address flows that extend beyond jurisdictional boundaries. Social and environmental provisions in trade agreements For example, a trade ban may prove ineffective if it does not recognize the economic concerns of local communities, who may derive little economic benefit from the ban, and hence have little incentive for conservation or sustainable resource use (Abensperg-Traun 2009). Consequently, the resource may be sold illegally or into alternative markets, creating leakage effects that limit the effectiveness of the trade ban. For instance, Busch et al. (2022) estimated that a European ban on importing high-deforestation palm oil from Indonesia would have only minor effects on deforestation because, among other reasons, non-participating countries would absorb the high-deforestation palm oil. More research is needed on how to avoid mismatches when designing trade agreements and trade bans. However, due diligence policies may suffer from resolution mismatches because they do not target any particular locality, but rather general social-environmental problems, irrespective of their local manifestation. This situation can lead to social mismatches. The EU Timber Regulation, for example, demands that timber is sourced legally according to the laws of the producer country. However, such policies that are reliant on local laws risk endorsing certification systems that neglect the rights of certain local communities (Bartley 2014) and work against sustainability by incentivizing a regulatory “race to the bottom” among exporting countries (dos Reis et al. 2021). Furthermore, if mandatory due diligence laws require companies to report on risk mitigation in their supply chains, companies may focus their reporting on issues that are not key priorities for local stakeholders. For example, under the French Duty of Vigilance Law, companies have focused on environmental issues such as deforestation in the soy and beef supply chains while neglecting other issues such as biodiversity loss, pesticide use, water scarcity, and water pollution. The companies prioritize labor rights, whereas the rights to health, land, water, and food may be more important for local stakeholders (Schilling-Vacaflor 2021). Institutional interplay Although we focus on how specific institutions can define and address what they conceive as mismatches, in practice, telecoupled systems are typically governed by several institutions, which interact horizontally at the same level of social organization or vertically across levels (Fig. 5). Institutions influence the decision- making, commitments, behavior, and effects of one another (Oberthür and Gehring 2006). Institutional interplay is based either on functional linkages that occur when developments in one issue area unavoidably affect another issue area, such as between institutions on agricultural production and land use, or it is based on political linkages that arise when actors recognize interdependencies and deliberately forge institutional interactions (Young 2005). For example, the EU’s Forest Law Enforcement Governance and Trade (FLEGT) initiative interacts with private certification schemes and public legal timber regulations in partner countries (Overdevest and Zeitlin 2014). FLEGT promotes better enforcement of forest law and the establishment of export licencing systems in partner countries to identify, monitor, and export legally harvested timber products destined for EU markets. Additionally, the FLEGT initiative, adopted in 2003, encouraged U.S. environmental activists to advocate for an extension of the U.S. Lacey Act from fish and wildlife to plants, leading to amendment of the Lacey Act in 2008. This example highlights how institutional interactions can lead to the convergence of separate national or regional governance regimes. The convergence between FLEGT and the U.S. Lacey Act ensured that illegally harvested timber is not simply diverted from one market to another (Overdevest and Zeitlin 2014). Research on telecoupling highlights the need to combine traditional place-based governance approaches with flow-based governance, which “considers a place in light of its relationships with other places, by tracking and managing where key flows start, progress, and end” (Liu et al. 2018:65). Flows are dynamic, and their origin and destination may change over time as a result of, for example, changing infrastructure, market demand, or biophysical conditions (dos Reis et al. 2023). Flow-based governance arrangements such as certification schemes, zero- deforestation commitments, and due diligence laws are designed to govern commodity flows, irrespective of changing trading relationships between supply chain actors. However, flow-based governance may generate new forms of social exclusion, inequality, and ecological simplification in places of production if transnational notions of sustainability do not match with local needs and realities (Newig et al. 2020). Landscape or jurisdictional approaches to supply chain governance If preventable human rights violations or environmental damages occur, the company can be held liable and can be required to remedy the harm (Schilling-Vacaflor 2021). Additionally, the European Commission proposed a Directive on sustainable corporate governance that covers human rights and environmental due diligence (Schilling- Vacaflor and Lenschow 2023). In sum, due diligence laws attempt to alleviate the boundary mismatch that occurs because importing Landscape and jurisdictional approaches aim to avoid the boundary mismatches that commonly affect public and private governance initiatives that focus exclusively at farm or supply- chain scales. This narrow focus can create “islands of good practice” while surrounding areas continue with business as usual (UNDP 2019:12). Many of the social-ecological problems that sustainability initiatives such as voluntary sustainability standards target manifest in the wider landscape, leading to mismatches between the scale of the intervention and the scale of the sustainability challenges being addressed (Sonderegger et al. 2022). For example, where companies seek to reduce commodity- driven deforestation by certifying some of their own or their Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety.org/vol28/iss2/art24/ suppliers’ farms or plantations, deforestation may shift to non- certified areas (Heilmayr et al. 2020). Jurisdictional and landscape approaches are assumed to reduce the risk of leakages (and thus boundary mismatches) because they target entire jurisdictions or landscapes rather than a selected smaller area. In terms of certification and standard-setting, landscape and jurisdictional approaches have been introduced to upscale governance to reduce the risk that commodity sourcing produces ungoverned impacts beyond the production area or unit (e.g., farms). Sustainable cocoa initiatives, for example, are evolving in their focus from the farm level to sector, landscape, and jurisdictional levels (Carodenuto 2019, Parra-Paitan et al. 2022, 2023). Empirical evidence on the effectiveness of landscape and jurisdictional approaches is scant, however, given their recent emergence (Bager 2021, von Essen and Lambin 2021). al. 2008, Bodin et al. 2017, Enqvist et al 2020). Bergsten et al. (2014:1) argue that “boundary mismatches are impossible to resolve if the focal ecological processes are not contained within the spatial jurisdiction of either a single high-level actor responsible for the whole area or by several lower level actors who collaborate” and thus jointly build a comprehensive governance system at a larger scale. Landscape or jurisdictional approaches to supply chain governance This idea suggests that studying telecoupled systems from the perspective of polycentric governance, defined as systems of overlapping jurisdictions with formally independent but interlinked centers of decision-making, could yield valuable insights into how to resolve mismatches in global land and agricultural commodity governance. Beyond examining the effectiveness of single governance institutions in isolation, a more systematic evaluation of the interplay and potential synergies between different governance interventions can advance the understanding of how to design governance solutions that match the scale of the problem at hand. Jurisdictional and landscape-based certification and sourcing also have limitations. Governance at the landscape level remains limited to a certain regionally confined scale and may not address all potentially relevant telecoupled dynamics such as migrant worker flows or illicit financial flows (Sonderegger et al. 2022). Additionally, the risk of leakage persists because neighboring jurisdictions may have weaker environmental protections (von Essen and Lambin 2021). Non-compliant production may shift to neighboring places with fewer restrictions (Meyfroidt et al. 2018), and commodities from non-compliant neighbors might be laundered into the more tightly regulated jurisdiction (Gibbs et al. 2016, Boshoven et al. 2021). A social-ecological network approach can be used to study collaborative natural resource governance across jurisdictional boundaries (Janssen et al. 2006, Bodin and Tengö 2012, Barnes et al. 2019). Studies could adopt such an approach to represent telecoupled systems as networks of social actors and ecological resources connected through commodity flows and institutional or social linkages. Although it is difficult to account for different kinds of social actors and the processing of commodities (e.g., from cocoa bean to chocolate bar) with this approach, it can help to capture how material, information, and communication flows connect different ecosystems, actors, and institutions (Janssen et al. 2006, Bodin and Tengö 2012). This approach is particularly suited to the analysis of landscape-scale responses to boundary mismatches because it highlights horizontal institutional interplay, as demonstrated, for example, in research on an agricultural landscape in Madagascar (Bodin and Tengö 2012) and wetlands in Sweden (Bergsten et al. 2014). CONCLUSION The governance of telecoupled systems is beset with problems of fit. Because most social and environmental problems in a globalizing world are neither purely local nor global in scale, addressing these problems requires governance responses that transcend political borders to match the spatial scale of the problem while also being sensitive to local context. Here, we applied the established concepts of institutional fit and governance mismatches to complex sustainability issues arising due to telecoupling. We identified two types of mismatches that are pertinent in the governance of telecoupled systems. First, boundary mismatches occur when governance institutions neglect social-ecological problems that transcend established jurisdictional boundaries, either because the institutional design fails to cover the full scale of the problem or because the intervention induces leakages. Second, resolution mismatches arise when governance institutions have a coarser resolution than is suitable to address the social-ecological problem they aim to address. Because of a lack of governance precision, governance instruments are too general to be effectively implemented and enforced. In the context of land and global agricultural commodity governance, approaches such as due diligence laws and policies, landscape and jurisdictional approaches to supply chain governance, and environmental provisions in trade agreements present important steps toward creating institutional fit in the governance of telecoupled systems. We do not claim that rescaling governance institutions to perfectly match telecoupled social-ecological systems will necessarily solve telecoupled sustainability issues, or even that it is attainable in all circumstances. Rather, we acknowledge that the risk of mismatches persists and identifying an “optimal spatial scale” may not be possible. Any attempt to resolve boundary or resolution mismatches comes with the risk of creating new mismatches, and because material flows, immaterial connections, and spillover relations are dynamic (dos Reis et al. 2020), governing telecoupled systems requires recognizing constantly evolving problem structures and continuously evaluating and adapting governance initiatives. However, even if it were possible to create institutional fit, there would be no guarantee of effective governance, due to implementation or enforcement problems. Nonetheless, we see substantial value in distinguishing different types of mismatches in telecoupled settings to be more productive in devising multiple, well-aligned, and adaptive governance arrangements that are better equipped to bring about the required change toward social and environmental sustainability. Institutional interplay This idea highlights that flow-based governance can cover the full spatial scale of telecoupled systems, but their flow specificity comes at the cost of place specificity. Evidence suggests that the effectiveness of flow-based governance benefits from synergistic place-based governance (zu Ermgassen et al. 2022). For example, governments can support the implementation of zero-deforestation commitments by providing additional disincentives for deforestation through, for example, credit restrictions for non- compliant individuals and companies, and through anti- Creating effective collaborative ties between institutions has been repeatedly proposed as a solution to rectify mismatches (Galaz et Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Fig. 5. Schematic illustration of institutional interactions in a telecoupled system. The circles denote the governance scales of different institutions. Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 yandsociety.org/vol28/iss2/art24/ Fig. 5. Schematic illustration of institutional interactions in a telecoupled system. The circles denote the governance scales of different institutions. Fig. 5. Schematic illustration of institutional interactions in a telecoupled system. The circles denote the governance scales of different institutions. the risk of infringing environmental or human rights is high, and shift to places with stricter governance to meet consumer demands for more transparency and due diligence (Gardner et al. 2019). This effect increases the risk of unintentionally marginalizing small-scale producers in these regions by excluding them from international value chains and the economic benefits of the global economy (Zhunusova et al. 2022). The most vulnerable people and countries may become subject to extraterritorial control and externally imposed notions of sustainability if actors of the Global North seek to govern environmental and social issues beyond their own borders. corruption measures that improve the reliability of geospatial forest information on which private governance schemes depend (Garrett et al. 2019). More research is needed to investigate the interplay between institutions that focus on the full spatial extent of the problem and institutions that are adapted to the local context. LITERATURE CITED Abensperg-Traun, M. 2009. CITES, sustainable use of wild species and incentive-driven conservation in developing countries, with an emphasis on southern Africa. Biological Conservation 142(5):948-963. https://doi.org/10.1016/j.biocon.2008.12.034 Andonova, L. B., and R. B. Mitchell. 2010. The rescaling of global environmental politics. Annual Review of Environment and Resources 35:255-282. https://doi.org/10.1146/annurev- [1] Bergsten et al. (2014) note that the two types of mismatches may overlap, for example, when jurisdictional boundaries compel actors to govern ecological processes at too fine a scale. [2] However, we acknowledge that the different types may overlap or be nested in reality, depending on which governance institution is taken as the analytical vantage point. For example, what appears as a spillover of one governance institution may be an induced leakage of another governance intervention. Bager, S. 2021. Delivering zero deforestation: how governance interventions in agro-food commodity supply chains can foster sustainable land use. Dissertation. Université catholique de Louvain, Ottignies-Louvain-la-Neuve, Belgium. https://dial. uclouvain.be/pr/boreal/object/boreal:254482 [3] For example, the Convention on Biological Diversity, United Nations Convention to Combat Desertification, United Nations Framework Convention on Climate Change, and New York Declaration on Forests are not flow specific. Bager, S. L., and E. F. Lambin. 2020. Sustainability strategies by companies in the global coffee sector. Business Strategy and the Environment 29(8):3555-3570. https://doi.org/10.1002/bse.2596 Bager, S. L., U. M. Persson, and T. N. P. dos Reis. 2021. Eighty- six EU policy options for reducing imported deforestation. One Earth 4(2):289-306. https://doi.org/10.1016/j.oneear.2021.01.011 CONCLUSION We acknowledge support by the German Research Foundation (DFG) and the Open Access Publication Fund of Leuphana University Lüneburg. Author Contributions: Barnes, M. L., Ö. Bodin, T. R. McClanahan, J. N. Kittinger, A. S. Hoey, O. G. Gaoue, and N. A. J. Graham. 2019. Social- ecological alignment and ecological conditions in coral reefs. Nature Communications 10:2039. http://dx.doi.org/10.1038/ s41467-019-09994-1 J. C. coordinated the development and conceptualization of the study and wrote the original draft of the manuscript, as well as the revisions. G. S., J. N., P. M., and E. Challies contributed to conceptualizing the study and drafting, revising, and editing, the manuscript; authors’ names are listed according to the degree of contribution. S. B., L. B. L., E. Corbara, C. F., A. F. P., P. L., C. P. P., S. Q., N. R., and J. Z. contributed to discussing the concepts at two workshops and revising and editing the manuscript; authors’ names are listed alphabetically to indicate equal contributions. All authors approved the final manuscript. Bartley, T. 2014. Transnational governance and the re-centered state: sustainability or legality? Regulation and Governance 8 (1):93-109. https://doi.org/10.1111/rego.12051 Bastos Lima, M. G. 2021. The politics of bioeconomy and sustainability. Springer, Cham, Switzerland. https://doi. org/10.1007/978-3-030-66838-9 Data Availability: Data/code sharing is not applicable to this article because no new data/code were created or analyzed in this study. CONCLUSION Looking at land-based commodity flows through the lens of boundary and resolution mismatches helps us to better anticipate potential governance weaknesses arising from a lack of governance precision or extent, and hence, enables better policy debates. Our analysis indicates that complementary interventions at various spatial scales, rather than single interventions, are needed to govern telecoupled systems effectively. Scaling or rescaling governance to match the scale of telecoupled systems is an inherently political process. The scale at which a given problem is perceived and framed influences the scale at which it is addressed (Newig and Moss 2017). Rescaling governance can entrench, rather than restructure, existing power relations and global inequalities. For instance, companies may stop sourcing from places with weak public governance, where The most pressing and challenging future research question is how to align multiple governance institutions to govern Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety.org/vol28/iss2/art24/ telecoupled systems. Advancing understanding of institutional mismatches in telecoupled systems requires interdisciplinary research, which itself needs to grapple with the challenge of bridging scales embedded in different research approaches, problem definitions, and perspectives (Friis et al. 2023). While we have focused on spatial mismatches in the governance of telecoupled systems, future investigations could analyze the occurrence and implications of temporal mismatches. Telecoupled systems are dynamic, and the spatiotemporal connections between regions and actors can change over time (dos Reis et al. 2020, 2023, Leijten et al. 2022), requiring adaptive governance responses. Additionally, investigating to what extent governance institutions fit with the complete life cycle of products merits further research because the spatial scale of governance expands when the temporal scale of governance is upscaled to the product life cycle. The task, albeit formidable, is to design governance systems in which effective institutional interplay offsets institutional mismatches of single institutions. E. Corbara acknowledges that this article contributes to the “María de Maeztu” Programme for Units of Excellence of the Spanish Ministry of Science and Innovation (CEX2019-000940-M). C. F. acknowledges support through the Carlsberg Foundation Reintegration Fellowship grant CF19-0526. We acknowledge support by the German Research Foundation (DFG) and the Open Access Publication Fund of Leuphana University Lüneburg. E. Corbara acknowledges that this article contributes to the “María de Maeztu” Programme for Units of Excellence of the Spanish Ministry of Science and Innovation (CEX2019-000940-M). C. F. acknowledges support through the Carlsberg Foundation Reintegration Fellowship grant CF19-0526. Acknowledgments: Disentangling intangible social- ecological systems. Global Environmental Change 22(2):430-439. https://doi.org/10.1016/j.gloenvcha.2012.01.005 Cox, M. 2014. Understanding large social-ecological systems: introducing the SESMAD project. International Journal of the Commons 8(2):265-276. https://doi.org/10.18352/ijc.406 Boshoven, J., L. C. Fleck, S. Miltner, N. Salafsky, J. Adams, A. Dahl-Jørgensen, G. Fonseca, D. Nepsted, K. Rabinovitch, and F. Seymour. 2021. Jurisdictional sourcing: leveraging commodity supply chains to reduce tropical deforestation at scale. A generic theory of change for a conservation strategy, v 1.0. Conservation Science and Practice 3(5):e383. https://doi.org/10.1111/csp2.383 Cumming, G. S., D. H. M. Cumming, and C. L. Redman. 2006. Scale mismatches in social-ecological systems: causes, consequences, and solutions. Ecology and Society 11(1):14. https://doi.org/10.5751/ES-01569-110114 Busch, J., O. Amarjargal, F. Taheripour, K. G. Austin, R. N. Siregar, K. Koenig, and T. W. Hertel. 2022. Effects of demand- side restrictions on high-deforestation palm oil in Europe on deforestation and emissions in Indonesia. Environmental Research Letters 17(1):014035. https://doi.org/10.1088/1748-9326/ ac435e Dawson, N. M., B. Coolsaet, E. J. Sterling, R. Loveridge, N. D. Gross-Camp, S. Wongbusarakum, K. K. Sangha, L. M. Scherl, H. Phuong Phan, N. Zafra-Calvo, W. G. Lavey, P. Byakagaba, C. J. Idrobo, A. Chenet, N. J. Bennett, S. Mansourian, and F. J. Rosado-May. 2021. The role of Indigenous peoples and local communities in effective and equitable conservation. Ecology and Society 26(3):19. https://doi.org/10.5751/ES-12625-260319 Busck-Lumholt, L. M., J. Coenen, J. Persson, A. Frohn Pedersen, O. Mertz, and E. Corbera. 2022a. Telecoupling as a framework to support a more nuanced understanding of causality in land system science. Journal of Land Use Science 17(1):386-406. https://doi.org/10.1080/1747423X.2022.2086640 Del Gatto, F., B. Ortiz-von Halle, B. Buendía, and C. H. Keong. 2009. Trade liberalisation and forest verification: learning from the US-Peru Trade Promotion Agreement. VERIFOR Briefing Paper. Overseas Development Institute, London, UK. https:// cdn.odi.org/media/documents/4482.pdf Busck-Lumholt, L. M., E. Corbera, and O. Mertz. 2022b. How are institutions included in Integrated Conservation and Development Projects? Developing and testing a diagnostic approach on the World Bank’s Forest and Community project in Salta, Argentina. World Development 157:105956. https://doi. org/10.1016/j.worlddev.2022.105956 Denier, L., S. Scherr, S. Shames, P. Chatterton, L. Hovani, and N. Stam. 2015. The little sustainable landscapes book: achieving sustainable development through integrated landscape management. Global Canopy Programme, Oxford, UK. https://globalcanopy. org/insights/publication/the-little-sustainable-landscapes-book/ Denier, L., S. Scherr, S. Shames, P. Chatterton, L. Hovani, and N. Stam. 2015. The little sustainable landscapes book: achieving sustainable development through integrated landscape management. Global Canopy Programme Oxford UK https://globalcanopy Carlson, A. K., J. G. Zaehringer, R. D. Acknowledgments: Bastos Lima, M. G., U. M. Persson, and P. Meyfroidt. 2019. Leakage and boosting effects in environmental governance: a framework for analysis. Environmental Research Letters 14 (10):105006. https://doi.org/10.1088/1748-9326/ab4551 We warmly thank all participants of the two synthesis workshops of the COUPLED project, led by Jonas Østergaard Nielsen and Kathrin Trommler in January and September 2021, for their useful comments and valuable feedback. Special thanks go to Floris Leijten, Pin Pravalprukskul, Finn Mempel, Tiago Reis, and Joel Persson for sharing their empirical reflections and instructive insights from their research. This work was funded by the Marie Skłodowska-Curie Actions (MSCA) grant agreement 765408 from the European Commission: COUPLED ‘Operationalising Telecouplings for Solving Sustainability Challenges for Land Use’. Berger, A., D. Blümer, C. Brandi, and M. Chi. 2020. Towards greening trade? Environmental provisions in emerging markets’ preferential trade agreements. Pages 61-81 in A. Negi, J. A. Pérez- Pineda, and J. Blankenbach, editors. Sustainability standards and global governance: experiences of emerging economies. Springer, Singapore. https://doi.org/10.1007/978-981-15-3473-7_4 Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety org/vol28/iss2/art24/ Bergsten, A., D. Galafassi, and Ö. Bodin. 2014. The problem of spatial fit in social-ecological systems: detecting mismatches between ecological connectivity and land management in an urban region. Ecology and Society 19(4):6. https://doi. org/10.5751/ES-06931-190406 Challies, E., J. Newig, and A. Lenschow. 2014. What role for social-ecological systems research in governing global teleconnections? Global Environmental Change 27:32-40. https:// doi.org/10.1016/j.gloenvcha.2014.04.015 Challies, E., J. Newig, and A. Lenschow. 2014. What role for social-ecological systems research in governing global teleconnections? Global Environmental Change 27:32-40. https:// doi.org/10.1016/j.gloenvcha.2014.04.015 Coleman, E. A., B. Schultz, V. Ramprasad, H. Fischer, P. Rana, A. M. Filippi, B. Güneralp, A. Ma, C. Rodriguez Solorzano, V. Guleria, R. Rana, and F. Fleischman. 2021. Limited effects of tree planting on forest canopy cover and rural livelihoods in northern India. Nature Sustainability 4(11):997-1004. https://doi. org/10.1038/s41893-021-00761-z Bodin, Ö. 2017. Collaborative environmental governance: achieving collective action in social-ecological systems. Science 357(6352):eaan1114. https://doi.org/10.1126/science.aan1114 Bodin, Ö., B. Crona, M. Thyresson, A.-L. Golz, and M. Tengö. 2014. Conservation success as a function of good alignment of social and ecological structures and processes. Conservation Biology 28(5):1371-1379. https://doi.org/10.1111/cobi.12306 Cotta, B., J. Coenen, E. Challies, J. Newig, A. Lenschow, and A. Schilling-Vacaflor. 2022. Environmental governance in globally telecoupled systems: mapping the terrain towards an integrated research agenda. Earth System Governance 13:100142. https:// doi.org/10.1016/j.esg.2022.100142 Bodin, Ö., and M. Tengö. 2012. Acknowledgments: Ghoddousi, D. Loibl, and P. Rufin. 2023. Enabling spaces for bridging scales: scanning solutions for interdisciplinary human-environment research. Sustainability Science, 18,125-1269 Dressel, S., G. Ericsson, and C. Sandström. 2018. Mapping social- ecological systems to understand the challenges underlying wildlife management. Environmental Science and Policy 84:105-112. https://doi.org/10.1016/j.envsci.2018.03.007 Friis, C., and J. Ø. Nielsen. 2017. Land-use change in a telecoupled world: the relevance and applicability of the telecoupling framework in the case of banana plantation expansion in Laos. Ecology and Society 22(4):30. https://doi.org/10.5751/ES-09480-220430 Eakin, H., R. DeFries, S. Kerr, E. F. Lambin, J. Liu, P. J. Marcotullio, P. Messerli, A. Reenberg, X. Rueda, S. R. Swaffield, B. Wicke, and K. Zimmerer. 2014. Significance of telecoupling for exploration of land-use change. Pages 141-161 in K. C. Seto and A. Reenberg, editors. Rethinking global land use in an urban era. MIT Press, Cambridge, Massachusetts, USA. https://doi. org/10.7551/mitpress/9780262026901.003.0008 Friis, C., and J. Ø. Nielsen, editors. 2019. Telecoupling: exploring land-use change in a globalised world. Palgrave Macmillan, Cham, Switzerland. https://doi.org/10.1007/978-3-030-11105-2 Eakin, H., X. Rueda, and A. Mahanti. 2017. Transforming governance in telecoupled food systems. Ecology and Society 22 (4):32. https://doi.org/10.5751/ES-09831-220432 Friis, C., J. Ø. Nielsen, I. Otero, H. Haberl, J. Niewöhner, and P. Hostert. 2016. From teleconnection to telecoupling: taking stock of an emerging framework in land system science. Journal of Land Use Science 11(2):131-153. https://doi.org/10.1080/1747423X.2015.1096423 Ekstrom, J. A., and B. I. Crona. 2017. Institutional misfit and environmental change: a systems approach to address ocean acidification. Science of the Total Environment 576:599-608. https://doi.org/10.1016/j.scitotenv.2016.10.114 Galaz, V., P. Olsson, T. Hahn, C. Folke, and U. Svedin. 2008. The problem of fit among biophysical systems, environmental and resource regimes, and broader governance systems: insights and emerging challenges. Pages 147-186 in O. R. Young, L. A. King, H. Schroeder, editors. Institutions and environmental change: principal findings, applications, and research frontiers. MIT Press, Cambridge, Massachussets, USA. https://doi.org/10.7551/ mitpress/9780262240574.003.0005 Ekstrom, J. A., and O. R. Young. 2009. Evaluating functional fit between a set of institutions and an ecosystem. Ecology and Society 14(2):16. https://doi.org/10.5751/ES-02930-140216 Enqvist, J. P., M. Tengö, and Ö. Bodin. 2020. Are bottom-up approaches good for promoting social-ecological fit in urban landscapes? Ambio 49(1):49-61. https://doi.org/10.1007/ Gardner, T. A., M. Benzie, J. Börner, E. Dawkins, S. Fick, R. Garrett, J. Godar, A. Grimard, S. Lake, R. K. Larsen, N. Mardas, C. L. McDermott, P. Meyfroidt, M. Osbeck, M. Persson, T. Sembres, C. Suavet, B. Strassburg, A. Trevisan, C. West, and P. Wolvekamp. Acknowledgments: Garrett, R. F. Bicudo Silva, P. R. Furumo, A. N. Raya Rey, A. Torres, M. G. Chung, Y. Li, and J. Liu. 2018. Toward rigorous telecoupling causal attribution: a systematic review and typology. Sustainability 10(12):4426. https://doi.org/10.3390/su10124426 Dorninger, C., A. Hornborg, D. J. Abson, H. von Wehrden, A. Schaffartzik, S. Giljum, J.-O. Engler, R. L. Feller, K. Hubacek, and H. Wieland. 2021. Global patterns of ecologically unequal exchange: implications for sustainability in the 21st century. Ecological Economics 179:106824. https://doi.org/10.1016/j. ecolecon.2020.106824 Carodenuto, S. 2019. Governance of zero deforestation cocoa in West Africa: new forms of public-private interaction. Environmental Policy and Governance 29(1):55-66. https://doi. org/10 1002/eet 1841 dos Reis, T. N. P., V. G. de Faria, G. Russo Lopes, G. Sparovek, C. West, R. Rajão, M. Napolitano Ferreira, M. M. S. Elvira, and R. S. T. do Valle. 2021. Trading deforestation—why the legality of forest-risk commodities is insufficient. Environmental Research Letters 16(12):124025. https://doi.org/10.1088/1748-9326/ ac358d Carrasco, L. R., J. Chan, F. L. McGrath, and L. T. P. Nghiem. 2017. Biodiversity conservation in a telecoupled world. Ecology and Society 22(3):24. https://doi.org/10.5751/ES-09448-220324 Cash, D. W., W. N. Adger, F. Berkes, P. Garden, L. Lebel, P. Olsson, L. Pritchard, and O. Young. 2006. Scale and cross-scale dynamics: governance and information in a multilevel world. Ecology and Society 11(2):8. https://doi.org/10.5751/ES-01759-110208 dos Reis, T. N. P., P. Meyfroidt, E. K. H. J. zu Ermgassen, C. West, T. Gardner, S. Bager, S. Croft, M. J. Lathuillière, and J. Godar. 2020. Understanding the stickiness of commodity supply chains is key to improving their sustainability. One Earth 3(1):100-115. https://doi.org/10.1016/j.oneear.2020.06.012 Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Foley, P. 2017. The territorialization of transnational sustainability governance: production, power and globalization in Iceland’s fisheries. Environmental Politics 26(5):915-937. https://doi.org/10.1080/09644016.2017.1343767 dos Reis, T. N. P., V. Ribeiro, R. D. Garrett, T. Kuemmerle, P. Rufin, V. Guidotti, P. C. Amaral, and P. Meyfroidt. 2023. Explaining the stickiness of supply chain relations in the Brazilian soybean trade. Global Environmental Change 78:102633. https:// doi.org/10.1016/j.gloenvcha.2022.102633 Folke, C., L. Pritchard Jr., F. Berkes, J. Colding, and U. Svedin. 2007. The problem of fit between ecosystems and institutions: ten years later. Ecology and Society 12(1):30. https://doi.org/10.5751/ ES-02064-120130 Dou, Y., R. F. B. da Silva, H. Yang, and J. Liu. 2018. Spillover effect offsets the conservation effort in the Amazon. Journal of Geographical Sciences 28(11):1715-1732. https://doi.org/10.1007/ s11442-018-1539-0 Friis, C., M. Hernández-Morcillo, M. Baumann, C. Coral, T. Frommen, A. Acknowledgments: Grund, B. Hedelin, N. Hernández-Mora, F. Hüesker, D. Huitema, K. Irvine, A. Klinke, L. Lange, D. Loupsans, M. Lubell, C. Maganda, P. Matczak, M. Parés, H. Saarikoski, L. Slavíková, S. Van der Arend, and Y. Von Korff. 2016. Transforming European water governance? Participation and river basin management under the EU Water Framework Directive in 13 member States. Water 8(4):156. https://doi. org/10.3390/w8040156 Leijten, F., S. Sim, H. King, and P. H. Verburg. 2021. Local deforestation spillovers induced by forest moratoria: evidence from Indonesia. Land Use Policy 109:105690. https://doi. org/10.1016/j.landusepol.2021.105690 F. Hüesker, D. Huitema, K. Irvine, A. Klinke, L. Lange, D. Loupsans, M. Lubell, C. Maganda, P. Matczak, M. Parés, H. Saarikoski, L. Slavíková, S. Van der Arend, and Y. Von Korff. 2016. Transforming European water governance? Participation and river basin management under the EU Water Framework Directive in 13 member States. Water 8(4):156. https://doi. org/10.3390/w8040156 Liu, J., Y. Dou, M. Batistella, E. Challies, T. Connor, C. Friis, J. D. A. Millington, E. Parish, C. L. Romulo, R. F. B. Silva, H. Triezenberg, H. Yang, Z. Zhao, K. S. Zimmerer, F. Huettmann, M. L. Treglia, Z. Basher, M. G. Chung, A. Herzberger, A. Lenschow, A. Mechiche-Alami, J. Newig, J. Roche, and J. Sun. 2018. Spillover systems in a telecoupled Anthropocene: typology, methods, and governance for global sustainability. Current Opinion in Environmental Sustainability 33:58-69. https://doi. org/10.1016/j.cosust.2018.04.009 Janssen, M. A., Ö. Bodin, J. M. Anderies, T. Elmqvist, H. Ernstson, R. R. J. McAllister, P. Olsson, and P. Ryan. 2006. Toward a network perspective of the study of resilience in social- ecological systems. Ecology and Society 11(1):15. https://doi. org/10.5751/ES-01462-110115 Kastner, T., K.-H. Erb, and H. Haberl. 2015. Global human appropriation of net primary production for biomass consumption in the European Union, 1986–2007. Journal of Industrial Ecology 19(5):825-836. https://doi.org/10.1111/ jiec.12238 Liu, J., V. Hull, M. Batistella, R. DeFries, T. Dietz, F. Fu, T. W. Hertel, R. C. Izaurralde, E. F. Lambin, S. Li, L. A. Martinelli, W. J. McConnell, E. F. Moran, R. Naylor, Z. Ouyang, K. R. Polenske, A. Reenberg, G. de Miranda Rocha, C. S. Simmons, P. H. Verburg, P. M. Vitousek, F. Zhang, and C. Zhu. 2013. Framing sustainability in a telecoupled world. Ecology and Society 18 (2):26. https://doi.org/10.5751/ES-05873-180226 Kehoe, L., T. N. P. dos Reis, P. Meyfroidt, S. Bager, R. Seppelt, T. Kuemmerle, E. Berenguer, M. Clark, K. F. Davis, E. K. H. J. zu Ermgassen, K. N. Farrell, C. Friis, H. Acknowledgments: How to track corporations across space and time. Ecological Economics 169:106492. https:// doi.org/10.1016/j.ecolecon.2019.106492 Kininmonth, S., A. Bergsten, and Ö. Bodin. 2015. Closing the collaborative gap: aligning social and ecological connectivity for better management of interconnected wetlands. Ambio 44:138-148. https://doi.org/10.1007/s13280-014-0605-9 Grabs, J., F. Cammelli, S. A. Levy, and R. D. Garrett. 2021. Designing effective and equitable zero-deforestation supply chain policies. Global Environmental Change 70:102357. https://doi. org/10.1016/j.gloenvcha.2021.102357 Kobrin, S. J. 2009. Sovereignty@Bay: globalization, multinational enterprise, and the international political system. Pages 183-204 in A. M. Rugman, editor. The Oxford handbook of international business. Second edition. Oxford University Press, Oxford, UK. https://doi.org/10.1093/oxfordhb/9780199234257.003.0007 Gupta, J. 2014. ‘Glocal’ politics of scale on environmental issues: climate change, water and forests. Pages 140-156 in F. Padt, P. Opdam, N. Polman, and C. Termeer, editors. Scale-sensitive governance of the environment. Wiley, Chichester, UK. https:// doi.org/10.1002/9781118567135.ch9 Laroche, P. C. S. J., C. J. E. Schulp, T. Kastner, and P. H. Verburg. 2021. Assessing the contribution of mobility in the European Union to rubber expansion. Ambio 51(3):770-783. https://doi. org/10.1007/s13280-021-01579-x Hagerman, S. M., L. M. Campbell, N. J. Gray, and R. Pelai. 2021. Knowledge production for target-based biodiversity governance. Biological Conservation 255:108980. https://doi.org/10.1016/j. biocon.2021.108980 Lebel, L., P. Garden, and M. Imamura. 2005. The politics of scale, position, and place in the governance of water resources in the Mekong region. Ecology and Society 10(2):18. https://doi. org/10.5751/ES-01543-100218 Heilmayr, R., K. M. Carlson, and J. J. Benedict. 2020. Deforestation spillovers from oil palm sustainability certification. Environmental Research Letters 15(7):075002. https://doi. org/10.1088/1748-9326/ab7f0c Lebel, L., E. Nikitina, C. Pahl-Wostl, and C. Knieper. 2013. Institutional fit and river basin governance: a new approach using multiple composite measures. Ecology and Society 18(1):1. https://doi.org/10.5751/ES-05097-180101 Higgins, V., and C. Richards. 2019. Framing sustainability: alternative standards schemes for sustainable palm oil and South- South trade. Journal of Rural Studies 65:126-134. https://doi. org/10.1016/j.jrurstud.2018.11.001 Leijten, F., T. N. P. dos Reis, S. Sim, P. H. Verburg, and P. Meyfroidt. 2022. The influence of company sourcing patterns on the adoption and effectiveness of zero-deforestation commitments in Brazil’s soy supply chain. Environmental Science and Policy 128:208-215. https://doi.org/10.1016/j.envsci.2021.10.032 Jager, N. W., E. Challies, E. Kochskämper, J. Newig, D. Benson, K. Blackstock, K. Collins, A. Ernst, M. Evers, J. Feichtinger, O. Fritsch, G. Gooch, W. Grund, B. Hedelin, N. Hernández-Mora, Jager, N. W., E. Challies, E. Kochskämper, J. Newig, D. Benson, K. Blackstock, K. Collins, A. Ernst, M. Evers, J. Feichtinger, O. Fritsch, G. Gooch, W. Acknowledgments: 2019. Transparency and sustainability in global commodity supply chains. World Development 121:163-177. https://doi.org/10.1016/j.worlddev.2018.05.025 Epstein, G., I. Pérez, M. Schoon, and C. L. Meek. 2014. Governing the invisible commons: ozone regulation and the Montreal Protocol. International Journal of the Commons 8 (2):337-360. https://doi.org/10.18352/ijc.407 Epstein, G., J. Pittman, S. M. Alexander, S. Berdej, T. Dyck, U. Kreitmair, K. J. Rathwell, S. Villamayor-Tomas, J. Vogt, and D. Armitage. 2015. Institutional fit and the sustainability of social- ecological systems. Current Opinion in Environmental Sustainability 14:34-40. https://doi.org/10.1016/j.cosust.2015.03.005 Garrett, R. D., S. Levy, K. M. Carlson, T. A. Gardner, J. Godar, J. Clapp, P. Dauvergne, R. Heilmayr, Y. le Polain de Waroux, B. Ayre, R. Barr, B. Døvre, H. K. Gibbs, S. Hall, S. Lake, J. C. Milder, L. L. Rausch, R. Rivero, X. Rueda, R. Sarsfield, B. Soares-Filho, and N. Villoria. 2019. Criteria for effective zero-deforestation commitments. Global Environmental Change 54:135-147. https://doi.org/10.1016/j.gloenvcha.2018.11.003 Ayre, R. Barr, B. Døvre, H. K. Gibbs, S. Hall, S. Lake, J. C. Milder, L. L. Rausch, R. Rivero, X. Rueda, R. Sarsfield, B. Soares-Filho, and N. Villoria. 2019. Criteria for effective zero-deforestation commitments. Global Environmental Change 54:135-147. https://doi.org/10.1016/j.gloenvcha.2018.11.003 European Commission. 2022. Green Deal: EU agrees law to fight global deforestation and forest degradation driven by EU production and consumption. Press release. European Commission, Brussels, Belgium. https://ec.europa.eu/commission/ presscorner/detail/en/ip_22_7444 Garrett, R. D., S. A. Levy, F. Gollnow, L. Hodel, and X. Rueda. 2021. Have food supply chain policies improved forest conservation and rural livelihoods? A systematic review. Environmental Research Letters 16(3):033002. https://doi. org/10.1088/1748-9326/abe0ed European Union. 2018. Directive (EU) 2018/2001 of the European Parliament and of the Council of 11 December 2018 on the promotion of the use of energy from renewable sources. Official Journal of the European Union L 328/82.http://data. europa.eu/eli/dir/2018/2001/oj Geist, H. J., and E. F. Lambin. 2002. Proximate causes and underlying driving forces of tropical deforestation. Bioscience 52 (2):143-150. https://doi.org/10.1641/0006-3568(2002)052[0143: PCAUDF]2.0.CO;2 Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety.org/vol28/iss2/art24/ the EU-Mercosur trade agreement fails the sustainability test. One Earth 3(3):268-272. https://doi.org/10.1016/j.oneear.2020.08.013 Gibbs, H. K., J. Munger, J. L’Roe, P. Barreto, R. Pereira, M. Christie, T. Amaral, and N. F. Walker. 2016. Did ranchers and slaughterhouses respond to zero-deforestation agreements in the Brazilian Amazon? Conservation Letters 9(1):32-42. https://doi. org/10.1111/conl.12175 Kingdon, J. W. 1984. Agendas, alternatives, and public policies. Pearson, London, UK. Goldstein, B., and J. P. Newell. 2020. Meemken, E.-M., C. B. Barrett, H. C. Michelson, M. Qaim, T. Reardon, and J. Sellare. 2021. Sustainability standards in global agrifood supply chains. Nature Food 2(10):758-765. https://doi. org/10.1038/s43016-021-00360-3 Acknowledgments: Haberl, T. Kastner, K. L. Murtough, U. M. Persson, A. Romero-Muñoz, C. O’Connell, V. V. Schäfer, M. Virah-Sawmy, Y. le Polain de Waroux, and J. Kiesecker. 2020. Inclusion, transparency, and enforcement: How Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety org/vol28/iss2/art24/ Melnykovych, M., M. Nijnik, I. Soloviy, A. Nijnik, S. Sarkki, and Y. Bihun. 2018. Social-ecological innovation in remote mountain areas: adaptive responses of forest-dependent communities to the challenges of a changing world. Science of the Total Environment 613-614:894-906. https://doi.org/10.1016/j.scitotenv.2017.07.065 Newell, J. P., and J. Simeone. 2014. Russia’s forests in a global economy: how consumption drives environmental change. Eurasian Geography and Economics 55(1):37-70. https://doi. org/10.1080/15387216.2014.926254 Newig, J., A. Adzersen, E. Challies, O. Fritsch, and N. Jager. 2013. Comparative analysis of public environmental decision-making processes – a variable-based analytical scheme. Discussion paper 37/13. Institute for Environmental and Sustainability Communication, Lüneburg, Germany. https://doi.org/10.2139/ ssrn.2245518 Mempel, F., and E. Corbera. 2021. Framing the frontier – tracing issues related to soybean expansion in transnational public spheres. Global Environmental Change 69:102308. https://doi. org/10.1016/j.gloenvcha.2021.102308 Meyfroidt, P., J. Börner, R. Garrett, T. Gardner, J. Godar, K. Kis- Katos, B. S. Soares-Filho, and S. Wunder. 2020. Focus on leakage and spillovers: informing land-use governance in a tele-coupled world. Environmental Research Letters 15(9):090202. https://doi. org/10.1088/1748-9326/ab7397 Newig, J., E. Challies, B. Cotta, A. Lenschow, and A. Schilling- Vacaflor. 2020. Governing global telecoupling toward environmental sustainability. Ecology and Society 25(4):21. https://doi.org/10.5751/ES-11844-250421 Newig, J., A. Lenschow, E. Challies, B. Cotta, and A. Schilling- Vacaflor. 2019. What is governance in global telecoupling? Ecology and Society 24(3):26. https://doi.org/10.5751/ES-11178-240326 Meyfroidt, P., A. de Bremond, C. M. Ryan, E. Archer, R. Aspinall, A. Chhabra, G. Camara, E. Corbera, R. DeFries, S. Díaz, J. Dong, E. C. Ellis, K.-H. Erb, J. A. Fisher, R. D. Garrett, N. E. Golubiewski, H. R. Grau, J. M. Grove, H. Haberl, A. Heinimann, P. Hostert, E. G. Jobbágy, S. Kerr, T. Kuemmerle, E. F. Lambin, S. Lavorel, S. Lele, O. Mertz, P. Messerli, G. Metternicht, D. K. Munroe, H. Nagendra, J. Ø. Nielsen, D. S. Ojima, D. C. Parker, U. Pascual, J. R. Porter, N. Ramankutty, A. Reenberg, R. Roy Chowdhury, K. C. Seto, V. Seufert, H. Shibata, A. Thomson, B. L. Turner II, J. Urabe, T. Veldkamp, P. H. Verburg, G. Zeleke, and E. K. H. J. zu Ermgassen. 2022. Ten facts about land systems for sustainability. Acknowledgments: Proceedings of the National Academy of Sciences 119(7):e2109217118. https://doi.org/10.1073/pnas.2109217118 Golubiewski, H. R. Grau, J. M. Grove, H. Haberl, A. Heinimann, P. Hostert, E. G. Jobbágy, S. Kerr, T. Kuemmerle, E. F. Lambin, Newig, J., and T. Moss. 2017. Scale in environmental governance: moving from concepts and cases to consolidation. Journal of Environmental Policy and Planning 19(5):473-479. https://doi. org/10.1080/1523908X.2017.1390926 gy S. Lavorel, S. Lele, O. Mertz, P. Messerli, G. Metternicht, D. K. Munroe, H. Nagendra, J. Ø. Nielsen, D. S. Ojima, D. C. Parker, U. Pascual, J. R. Porter, N. Ramankutty, A. Reenberg, R. Roy Chowdhury, K. C. Seto, V. Seufert, H. Shibata, A. Thomson, B. L. Turner II, J. Urabe, T. Veldkamp, P. H. Verburg, G. Zeleke, and E. K. H. J. zu Ermgassen. 2022. Ten facts about land systems for sustainability. Proceedings of the National Academy of Sciences 119(7):e2109217118. https://doi.org/10.1073/pnas.2109217118 S. Lavorel, S. Lele, O. Mertz, P. Messerli, G. Metternicht, D. K. Niewöhner, J., J. Ø. Nielsen, I. Gasparri, Y. Gou, M. Hauge, N. Joshi, A. Schaffartzik, F. Sejersen, K. C. Seto, and C. Shughrue. 2016. Conceptualizing distal drivers in land use competition. Pages 21-40 in J. Niewöhner, A. Bruns, P. Hostert, T. Krueger, J. Ø. Nielsen, H. Haberl, C. Lauk, J. Lutz, and D. Müller, editors. Land use competition: ecological, economic and social perspectives. Springer, Cham, Switzerland. https://doi. org/10.1007/978-3-319-33628-2_2 Meyfroidt, P., R. Roy Chowdhury, A. de Bremond, E. C. Ellis, K.-H. Erb, T. Filatova, R. D. Garrett, J. M. Grove, A. Heinimann, T. Kuemmerle, C. A. Kull, E. F. Lambin, Y. Landon, Y. le Polain de Waroux, P. Messerli, D. Müller, J. Ø. Nielsen, G. D. Peterson, V. Rodriguez García, M. Schlüter, B. L. Turner II, and P. H. Verburg. 2018. Middle-range theories of land system change. Global Environmental Change 53:52-67. https://doi.org/10.1016/ j.gloenvcha.2018.08.006 Meyfroidt, P., R. Roy Chowdhury, A. de Bremond, E. C. Ellis, K.-H. Erb, T. Filatova, R. D. Garrett, J. M. Grove, A. Heinimann, T. Kuemmerle, C. A. Kull, E. F. Lambin, Y. Landon, Y. le Polain T. Kuemmerle, C. A. Kull, E. F. Lambin, Y. Landon, Y. le Polain de Waroux, P. Messerli, D. Müller, J. Ø. Nielsen, G. D. Peterson, T. Kuemmerle, C. A. Kull, E. F. Lambin, Y. Landon, Y. le Polain de Waroux, P. Messerli, D. Müller, J. Ø. Nielsen, G. D. Peterson, de Waroux, P. Messerli, D. Müller, J. Ø. Nielsen, G. D. Peterson, Nyström, M., J.-B. Jouffray, A. V. Acknowledgments: Norström, B. Crona, P. Søgaard Jørgensen, S. R. Carpenter, Ö. Bodin, V. Galaz, and C. Folke. 2019. Anatomy and resilience of the global production ecosystem. Nature 575(7781):98-108. https://doi.org/10.1038/s41586-019-1712-3 V. Rodriguez García, M. Schlüter, B. L. Turner II, and P. H. Verburg. 2018. Middle-range theories of land system change. Global Environmental Change 53:52-67. https://doi.org/10.1016/ j.gloenvcha.2018.08.006 Oberlack, C., S. Boillat, S. Brönnimann, J.-D. Gerber, A. Heinimann, C. Ifejika Speranza, P. Messerli, S. Rist, and U. Wiesmann. 2018. Polycentric governance in telecoupled resource systems. Ecology and Society 23(1):16. https://doi.org/10.5751/ ES-09902-230116 Meyfroidt, P., T. K. Rudel, and E. F. Lambin. 2010. Forest transitions, trade, and the global displacement of land use. Proceedings of the National Academy of Sciences 107 (49):20917-20922. https://doi.org/10.1073/pnas.1014773107 Moss, T. 2012. Spatial fit, from panacea to practice: implementing the EU Water Framework Directive. Ecology and Society 17(3):2. https://doi.org/10.5751/ES-04821-170302 Oberthür, S., and T. Gehring. 2006. Institutional interaction in global environmental governance: the case of the Cartagena Protocol and the World Trade Organization. Global Environmental Politics 6(2):1-31. https://doi.org/10.1162/ glep.2006.6.2.1 Moss, T., and J. Newig. 2010. Multilevel water governance and problems of scale: setting the stage for a broader debate. Environmental Management 46(1):1-6. https://doi.org/10.1007/ s00267-010-9531-1 Ostrom, E. 1990. Governing the commons: the evolution of institutions for collective action. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511807763 Munroe, D. K., M. Batistella, C. Friis, N. I. Gasparri, E. F. Lambin, J. Liu, P. Meyfroidt, E. Moran, and J. Ø. Nielsen. 2019. Governing flows in telecoupled land systems. Current Opinion in Environmental Sustainability 38:53-59. https://doi.org/10.1016/j. cosust.2019.05.004 Ostrom, E., M. A. Janssen, and J. M. Anderies. 2007. Going beyond panaceas. Proceedings of the National Academy of Sciences 104(39):15176-15178. https://doi.org/10.1073/pnas.0701886104 cosust.2019.05.004 cosust.2019.05.004 Nagendra, H., and E. Ostrom. 2012. Polycentric governance of multifunctional forested landscapes. International Journal of the Commons 6(2):104-133. https://doi.org/10.18352/ijc.321 Overdevest, C., and J. Zeitlin. 2014. Assembling an experimentalist regime: transnational governance interactions in the forest sector. Regulation and Governance 8(1):22-48. https:// doi.org/10.1111/j.1748-5991.2012.01133.x Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Padt, F., P. Opdam, N. Polman, and C. Termeer, editors. 2014. Scale-sensitive governance of the environment. Wiley, Chichester, UK. https://doi.org/10.1002/9781118567135 Muñoz, and A. Yanosky. 2022. The geography of international conservation interest in South American deforestation frontiers. Conservation Letters 15(1):e12859. https://doi.org/10.1111/ conl.12859 Parra-Paitan, C., and P. H. Verburg. 2022. Accounting for land use changes beyond the farm-level in sustainability assessments: the impact of cocoa production. Science of the Total Environment 825:154032. https://doi.org/10.1016/j.scitotenv.2022.154032 Roux, N., T. Kastner, K.-H. Erb, and H. Haberl. 2021. Does agricultural trade reduce pressure on land ecosystems? Decomposing drivers of the embodied human appropriation of net primary production. Ecological Economics 181:106915. https://doi.org/10.1016/j.ecolecon.2020.106915 Parra-Paitan, C., E. K. H. J. zu Ermgassen, P. Meyfroidt, and P. H. Verburg. 2023. Large gaps in voluntary sustainability commitments covering the global cocoa trade. Global Environmental Change 81:102696. https://doi.org/10.1016/j. gloenvcha.2023.102696 Roux, N., L. Kaufmann, M. Bhan, J. Le Noe, S. Matej, P. Laroche, K. Thomas, A. Bondeau, H. Haberl, and K.-H. Erb. 2022. Embodied HANPP of feed and animal products: tracing pressure on ecosystems along trilateral livestock supply chains 1986–2013. SSRN Preprint. https://doi.org/10.2139/ssrn.3998990 Partzsch, L. 2020. Alternatives to multilateralism: new forms of social and environmental governance. MIT Press, Cambridge, Massachusetts, USA. https://doi.org/https://doi.org/10.7551/ mitpress/12610.001.0001 Sayer, J., T. Sunderland, J. Ghazoul, J.-L. Pfund, D. Sheil, E. Meijaard, M. Venter, A. K. Boedhihartono, M. Day, C. Garcia, C. van Oosten, and L. E. Buck. 2013. Ten principles for a landscape approach to reconciling agriculture, conservation, and other competing land uses. Proceedings of the National Academy of Sciences 110(21):8349-8356. https://doi.org/10.1073/pnas.1210595110 Pedersen, A. F., J. Ø. Nielsen, C. Friis, and J. B. Jønsson. 2021a. Mineral exhaustion and its livelihood implications for artisanal and small-scale miners. Environmental Science and Policy 119:34-43. https://doi.org/10.1016/j.envsci.2021.02.002 Schilling-Vacaflor, A. 2021. Integrating human rights and the environment in supply chain regulations. Sustainability 13 (17):9666. https://doi.org/10.3390/su13179666 Pedersen, A. F., J. Ø. Nielsen, F. Mempel, S. L. Bager, J. B. Jønsson, and E. Corbera. 2021b. The ambiguity of transparency in the artisanal and small-scale mining sector of Tanzania. Extractive Industries and Society 8(4):101004. https://doi.org/10.1016/j. exis.2021.101004 Schilling-Vacaflor, A., and A. Lenschow. 2023. cosust.2019.05.004 C Quintero A Pacheco A Popp J Rouet-Leduc F Schnabel U. C. Lehmann, B. Lenzner, A. Marques, E. Motivans Švara, L. C. Quintero, A. Pacheco, A. Popp, J. Rouet-Leduc, F. Schnabel, C. Quintero, A. Pacheco, A. Popp, J. Rouet-Leduc, F. Schnabel, J. Siebert, I. R. Staude, S. Trogisch, V. Švara, J.-C. Svenning, G. Pe’er, K. Raab, D. Rakosy, M. Vandewalle, A. S. Werner, C. Wirth, H. Xu, D. Yu, Y. Zinngrebe, and A. Bonn. 2022. Biodiversity post-2020: closing the gap between global targets and national- level implementation. Conservation Letters 15(2):e12848. https:// doi.org/10.1111/conl.12848 J. Siebert, I. R. Staude, S. Trogisch, V. Švara, J.-C. Svenning, G. Shkaruba, A., and V. Kireyeu. 2013. Recognising ecological and institutional landscapes in adaptive governance of natural resources. Forest Policy and Economics 36:87-97. https://doi. org/10.1016/j.forpol.2012.10.004 p doi.org/10.1111/conl.12848 Qin, S., T. Kuemmerle, P. Meyfroidt, M. Napolitano Ferreira, G. I. Gavier Pizarro, M. E. Periago, T. N. P. dos Reis, A. Romero- cosust.2019.05.004 Hardening foreign corporate accountability through mandatory due diligence in the European Union? New trends and persisting challenges. Regulation and Governance, in press. https://doi.org/10.1111/ rego.12402 Pendrill, F., U. M. Persson, J. Godar, and T. Kastner. 2019. Deforestation displaced: trade in forest-risk commodities and the prospects for a global forest transition. Environmental Research Letters 14(5):055003. https://doi.org/10.1088/1748-9326/ab0d41 Schilling-Vacaflor, A., A. Lenschow, E. Challies, B. Cotta, and J. Newig. 2021. Contextualizing certification and auditing: soy certification and access of local communities to land and water in Brazil. World Development 140:105281. https://doi. org/10.1016/j.worlddev.2020.105281 Perino, A., H. M. Pereira, M. Felipe-Lucia, H. J. Kim, H. S. Kühl, M. R. Marselle, J. N. Meya, C. Meyer, L. M. Navarro, R. van Klink, G. Albert, C. D. Barratt, H. Bruelheide, Y. Cao, A. Chamoin, M. Darbi, M. Dornelas, N. Eisenhauer, F. Essl, N. Farwig, J. Förster, J. Freyhof, J. Geschke, F. Gottschall, C. Guerra, P H T Hi kl U J b T K L K ll I Küh G Perino, A., H. M. Pereira, M. Felipe-Lucia, H. J. Kim, H. S. Kühl, M. R. Marselle, J. N. Meya, C. Meyer, L. M. Navarro, R. van Klink, G. Albert, C. D. Barratt, H. Bruelheide, Y. Cao, A. Chamoin, M. Darbi, M. Dornelas, N. Eisenhauer, F. Essl, N. Farwig, J. Förster, J. Freyhof, J. Geschke, F. Gottschall, C. Guerra, P. Haase, T. Hickler, U. Jacob, T. Kastner, L. Korell, I. Kühn, G. U. C. Lehmann, B. Lenzner, A. Marques, E. Motivans Švara, L. C. Quintero, A. Pacheco, A. Popp, J. Rouet-Leduc, F. Schnabel, J. Siebert, I. R. Staude, S. Trogisch, V. Švara, J.-C. Svenning, G. Pe’er, K. Raab, D. Rakosy, M. Vandewalle, A. S. Werner, C. Wirth, H. Xu, D. Yu, Y. Zinngrebe, and A. Bonn. 2022. Biodiversity post-2020: closing the gap between global targets and national- level implementation. Conservation Letters 15(2):e12848. https:// doi.org/10.1111/conl.12848 Chamoin, M. Darbi, M. Dornelas, N. Eisenhauer, F. Essl, N. Farwig, J. Förster, J. Freyhof, J. Geschke, F. Gottschall, C. Guerra, Scott, J. 2020. Reducing the European Union’s environmental footprint through ‘territorial extension’. Pages 65-85 in V. Mauerhofer, D. Rupo, and L. Tarquinio, editors. Sustainability and law: general and specific aspects. Springer, Cham, Switzerland. https://doi.org/10.1007/978-3-030-42630-9_5 P. Haase, T. Hickler, U. Jacob, T. Kastner, L. Korell, I. Kühn, G. Š U. C. Lehmann, B. Lenzner, A. Marques, E. Motivans Švara, L. U. C. Lehmann, B. Lenzner, A. Marques, E. Motivans Švara, L. doi.org/10.1111/conl.12848 Sikor, T., G. Auld, A. J. Bebbington, T. A. Benjaminsen, B. S. Gentry, C. Hunsberger, A.-M. Izac, M. E. Margulis, T. Plieninger, H. Schroeder, and C. Upton. 2013. Global land governance: from territory to flow? Current Opinion in Environmental Sustainability 5(5):522-527. https://doi.org/10.1016/j.cosust.2013.06.006 Persson, J., S. Ford, A. Keophoxay, O. Mertz, J. Ø. Nielsen, T. Vongvisouk, and M. Zörner. 2021. Large differences in livelihood responses and outcomes to increased conservation enforcement in a protected area. Human Ecology 49(5):597-616. https://doi. org/10.1007/s10745-021-00267-4 Persson, J., S. Ford, A. Keophoxay, O. Mertz, J. Ø. Nielsen, T. Vongvisouk, and M. Zörner. 2021. Large differences in livelihood responses and outcomes to increased conservation enforcement in a protected area. Human Ecology 49(5):597-616. https://doi. org/10.1007/s10745-021-00267-4 Persson, J., and O. Mertz. 2019. Discursive telecouplings. Pages 313-336 in C. Friis and J. Ø. Nielsen, editors. Telecoupling: exploring land-use change in a globalised world. Palgrave Macmillan, Cham, Switzerland. https://doi.org/10.1007/978-3-0 30-11105-2_17 Persson, J., and O. Mertz. 2019. Discursive telecouplings. Pages 313-336 in C. Friis and J. Ø. Nielsen, editors. Telecoupling: exploring land-use change in a globalised world. Palgrave Macmillan, Cham, Switzerland. https://doi.org/10.1007/978-3-0 30-11105-2_17 Sonderegger, G., A. Heinimann, V. Diogo, and C. Oberlack. 2022. Governing spillovers of agricultural land use through voluntary sustainability standards: a coverage analysis of sustainability requirements. Earth System Governance 14:100158. https://doi. org/10.1016/j.esg.2022.100158 30-11105-2_17 Qin, S., T. Kuemmerle, P. Meyfroidt, M. Napolitano Ferreira, G. I. Gavier Pizarro, M. E. Periago, T. N. P. dos Reis, A. Romero- Qin, S., T. Kuemmerle, P. Meyfroidt, M. Napolitano Ferreira, G. I. Gavier Pizarro, M. E. Periago, T. N. P. dos Reis, A. Romero- Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 https://www.ecologyandsociety.org/vol28/iss2/art24/ Ecology and Society 28(2): 24 andsociety.org/vol28/iss2/art24/ Sonderegger, G., C. Oberlack, J. C. Llopis, P. H. Verburg, and A. Heinimann. 2020. Telecoupling visualizations through a network lens: a systematic review. Ecology and Society 25(4):47. https:// doi.org/10.5751/ES-11830-250447 chains on smallholders, indigenous peoples, and local communities in producer countries outside the EU. Forest Policy and Economics 143:102817. https://doi.org/10.1016/j.forpol.2022.102817 zu Ermgassen, E. K. H. J., M. G. Bastos Lima, H. Bellfield, A. Dontenville, T. Gardner, J. Godar, R. Heilmayr, R. Indenbaum, T. N. P. dos Reis, V. Ribeiro, I.-O. Abu, Z. Szantoi, and P. Meyfroidt. 2022. Addressing indirect sourcing in zero deforestation commodity supply chains. Science Advances 8(17): eabn3132. https://doi.org/10.1126/sciadv.abn3132 zu Ermgassen, E. K. H. J., M. G. Bastos Lima, H. Bellfield, A. Dontenville, T. Gardner, J. Godar, R. Heilmayr, R. Indenbaum, T. zu Ermgassen, E. K. H. J., M. G. Bastos Lima, H. Bellfield, A. Dontenville, T. Gardner, J. Godar, R. Heilmayr, R. Indenbaum, T. N. P. dos Reis, V. Ribeiro, I.-O. Abu, Z. Szantoi, and P. Meyfroidt. 2022. Addressing indirect sourcing in zero deforestation commodity supply chains. Science Advances 8(17): eabn3132. https://doi.org/10.1126/sciadv.abn3132 doi.org/10.1111/conl.12848 N. P. dos Reis, V. Ribeiro, I.-O. Abu, Z. Szantoi, and P. Meyfroidt. 2022. Addressing indirect sourcing in zero deforestation commodity supply chains. Science Advances 8(17): eabn3132. https://doi.org/10.1126/sciadv.abn3132 Soterroni, A. C., F. M. Ramos, A. Mosnier, J. Fargione, P. R. Andrade, L. Baumgarten, J. Pirker, M. Obersteiner, F. Kraxner, G. Câmara, A. X. Y. Carvalho, and S. Polasky. 2019. Expanding the soy moratorium to Brazil’s Cerrado. Science Advances 5(7): eaav7336. https://doi.org/10.1126/sciadv.aav7336 United Nations Development Programme (UNDP). 2019. Value beyond value chains: guidance note for the private sector. Version 1.0. United Nations Development Programme, New York, New York, USA. https://jaresourcehub.org/wp-content/uploads/2020/09/ VBV-Guidance-Note.pdf von Essen, M., and E. F. Lambin. 2021. Jurisdictional approaches to sustainable resource use. Frontiers in Ecology and the Environment 19(3):159-167. https://doi.org/10.1002/fee.2299 Wiegant, D., M. Peralvo, P. van Oel, and A. Dewulf. 2020. Five scale challenges in Ecuadorian forest and landscape restoration governance. Land Use Policy 96:104686. https://doi.org/10.1016/ j.landusepol.2020.104686 World Trade Organization. 2021. Matrix on trade-related measures pursuant to selected Multilateral Environmental Agreements: note by the Secretariat. Revision. World Trade Organization, Geneva, Switzerland. https://docs.wto.org/dol2fe/ Pages/SS/directdoc.aspx?filename=q:/WT/CTE/W160R9.pdf&Open= True Young, O. R. 2002a. The institutional dimensions of environmental change: fit, interplay, and scale. MIT Press, Cambridge, Massachusetts, USA. https://doi.org/10.7551/ mitpress/3807.001.0001 Young, O. R. 2002b. Institutional interplay: the environmental consequences of cross-scale interactions. Pages 263-292 in E. Ostrom, T. Dietz, N. Dolšak, P. C. Stern, S. Stonich, and E. U. Weber, editors. The drama of the commons. National Academies Press, Washington, D.C., USA. https://doi.org/10.17226/10287 Young, O. R. 2005. Science plan: institutional dimensions of global environmental change. IHDP Report 16. International Human Dimensions Programme, Bonn, Germany. Zaehringer, J. G., F. Schneider, A. Heinimann, and P. Messerli. 2019. Co-producing knowledge for sustainable development in telecoupled land systems. Pages 357-381 in C. Friis and J. Ø. Nielsen, editors. Telecoupling: exploring land-use change in a globalised world. Palgrave Macmillan, Cham, Switzerland. https://doi.org/10.1007/978-3-030-11105-2_19 Zaehringer, J. G., G. Wambugu, B. Kiteme, and S. Eckert. 2018. How do large-scale agricultural investments affect land use and the environment on the western slopes of Mount Kenya? Empirical evidence based on small-scale farmers’ perceptions and remote sensing. Journal of Environmental Management 213:79-89. https://doi.org/10.1016/j.jenvman.2018.02.019 Zhunusova, E., V. Ahimbisibwe, L. T. H. Sen, A. Sadeghi, T. Toledo-Aceves, G. Kabwe, and S. Günter. 2022. Potential impacts of the proposed EU regulation on deforestation-free supply
https://openalex.org/W4283802930	https://pure.rug.nl/ws/files/229766784/Ecology_and_Evolution_2022_Kruitwagen_Evolution_of_parasitoid_host_preference_and_performance_in_response_to_an.pdf	English	null	Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap	Ecology and evolution	2,022	cc-by	14,752	University of Groningen University of Groningen R E S E A R C H A R T I C L E R E S E A R C H A R T I C L E Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap en \| Leo W. Beukeboom \| Bregje Wertheim \| G. Sander van Doorn Groningen Institute for Evolutionary Life Sciences, University of Groningen, Groningen, The Netherlands Abstract Abstract The invasion of a novel host species can create a mismatch in host choice and off- spring survival (performance) when native parasitoids attempt to exploit the invasive host without being able to circumvent its resistance mechanisms. Invasive hosts can therefore act as evolutionary trap reducing parasitoids' fitness and this may eventu- ally lead to their extinction. Yet, escape from the trap can occur when parasitoids evolve behavioral avoidance or a physiological strategy compatible with the trap host, resulting in either host-range expansion or a complete host-shift. We developed an individual based model to investigate which conditions promote parasitoids to evolve behavioral preference that matches their performance, including host-trap avoidance, and which conditions lead to adaptations to the unsuitable hosts. The model was in- spired by solitary endo-parasitoids attacking larval host stages. One important aspect of these conditions was reduced host survival during incompatible interaction, where a failed parasitization attempt by a parasitoid resulted not only in death of her off- spring but also in host killing. This non-reproductive host mortality had a strong influ- ence on the likelihood of establishment of novel host–parasitoid relationship, in some cases constraining adaptation to the trap host species. Moreover, our model revealed that host-search efficiency and genetic variation in host-preference play a key role in the likelihood that parasitoids will include the suboptimal host in their host range, or will evolve behavioral avoidance resulting in specialization and host-range conserva- tion, respectively. Hence, invasive species might change the evolutionary trajectory of native parasitoid species, which is important for predicting biocontrol ability of native parasitoids towards novel hosts. Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap Kruitwagen, Astrid; Beukeboom, Leo W; Wertheim, Bregje; van Doorn, G Sander Published in: Ecology and Evolution DOI: 10.1002/ece3.9030 Publication date: 2022 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Kruitwagen, A., Beukeboom, L. W., Wertheim, B., & van Doorn, G. S. (2022). Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap. Ecology and Evolution, 12(7), Article e9030. https://doi.org/10.1002/ece3.9030 Citation for published version (APA): Kruitwagen, A., Beukeboom, L. W., Wertheim, B., & van Doorn, G. S. (2022). Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap. Ecology and Evolution, 12(7), Article e9030. https://doi.org/10.1002/ece3.9030 Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date: 24-10-2024 Download date: 24-10-2024 Received: 7 November 2021 \| Revised: 24 May 2022 \| Accepted: 27 May 2022 Received: 7 November 2021 \| Revised: 24 May 2022 \| Accepted: 27 May 2022 DOI: 10.1002/ece3.9030 DOI: 10.1002/ece3.9030 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2022 Th A th E l d E l ti bli h d b J h Wil & S Ltd 2 of 18 \| 1 \| INTRODUCTION Such release could in return also be positive: for example, reduced parasitism by scelionine egg para- sitoids when attacking the unsuitable host H. halys could facilitate population growth of its suitable stink bug host Podisus maculiventris, which is an important biocontrol agent. As host availability and suit- ability have a direct impact on parasitoid reproductive success, they can also shape the evolution of the parasitoid behavior and parasit- ization strategies. Hence, attacking and killing unsuitable hosts can change host population abundance of both suitable and unsuitable hosts (Heimpel et al., 2003; Kaser et al., 2018) and might thus alter the strength and direction of selection in response to the host trap. Maladapted parasitoid populations can be “rescued” by un- dergoing evolutionary change. Studies have shown that some par- asitoid populations harbor genetic variation for host-choice and host use/virulence (Benoist et al., 2020; Desjardins et al., 2010; Henter, 1995; König et al., 2015). This would allow parasitoids to escape from an evolutionary trap by evolving avoidance behavior or physiological compatibility (i.e., improved host use/performance) (Keeler & Chew, 2008). Both mechanisms would promote parasitoid persistence, but only adaption to efficient utilization of the novel host would reduce the impact of the invader and would thus be fa- vored from a conservation and biological control perspective. Note that alterations in host-choice and host-use might also arise due to plastic changes; parasitoids might avoid suboptimal host (patches) through learning or overcome host resistance through superparasit- ism. Insight in the response to an evolutionary trap is also relevant for the evolution of host specialization or host-range expansion. Inclusion of the suboptimal host into the parasitoid's repertoire of hosts species can eventually result in specialization and host-shift to this novel host, whereas behavioral avoidance results in physiologi- cal host-range conservation. The impact of variation in host-suitability on parasitoid ecology and evolution has previously been modeled in 1-host/1-parasitoid (Fellowes & Travis, 2000; Sasaki & Godfray, 1999; Tuda & Bonsall, 1999) and 2-host/1-parasitoid systems (Heimpel et al., 2003; Kaser et al., 2018; Tuda & Bonsall, 1999). However, it is unknown how parasitoid induced non-reproductive host mortality interacts with genetic variation in host preference behavior and parasitiza- tion strategy, that is, their physiological compatibility with different host species. 2 of 18 \| 1 \| INTRODUCTION KRUITWAGEN et al. 2 of 18 1 \| INTRODUCTION for example, been documented in native scelionine egg parasitoid wasps (Hymenoptera: Scelionidae) attacking the invasive agricul- tural pest Halyomorpha halys (Abram et al., 2014, 2016) and figitid larval parasitoids (Hymenoptera: Figitidaea) attacking Drosophila su- zukii (Kruitwagen et al., 2021). Non-reproductive host killing is thus a third outcome of a host–parasitoid interaction, but has received little attention in host–parasitoid studies despite its common nature (Abram et al., 2019). Parasitoids are insects that lay eggs in or on other insects, and whose immature stages develop in or on a host that is eventually killed. They however sometimes accept hosts for oviposition that are unsuitable for their offspring to survive (Heimpel et al., 2003; Thompson, 1988). One explanation of such “bad motherhood” is the invasion of novel host species for which native parasitoid species do not have pre-adapted mechanisms to overcome host resistance, and/or to recognize these hosts as unsuitable (Thompson, 1988; Yoon & Read, 2016). For instance, native parasitoids are reported to attack the exotic species, Harmonia axyridis, Halyomorpha halys, Drosophila suzukii, and Cydalima perspectalis but their offspring perform relatively poorly with low survival (Konopka et al., 2018; Martini et al., 2019; Romero et al., 2020). Exploiting suboptimal hosts can therefore result in reduced fitness when parasitoids lose resources, time and/or eggs, exploring and attacking them. When parasitoids exploit suboptimal hosts even when alternative suit- able ones are present, the host is considered an “evolutionary trap” (Schlaepfer et al., 2002; Schlenke et al., 2007). Evolutionary traps can have large ecological impact on ecosystems: they can reduce the parasitoid population size and ultimately drive it to local extinction (Kokko & Sutherland, 2001; Yoon & Read, 2016). Empirical and theoretical studies show that non-reproductive host killing can influence the population size of the unsuitable host when suitable hosts are present to sustain the parasitoid population (Abram et al., 2016; Huang et al., 2017; Kaser et al., 2018; Münster- Swendsen, 2002). However, the inclusion of the unsuitable host in the parasitoid's repertoire can reduce the overall reproductive success of the parasitoid and can consequently feedback on host– parasitoid population dynamics and relative host species abundance via complex direct and indirect interactions (Kaser et al., 2018). For instance, time and/or eggs lost by attacking unsuitable hosts can re- duce parasitoid population size and release the suitable hosts from parasitism (“enemy release”). Correspondence Correspondence Astrid Kruitwagen, Groningen Institute for Evolutionary Life Sciences, University of Groningen, Nijenborgh 7, 9700 CC Groningen, The Netherlands. Email: ajkruitwagen@gmail.com Funding information Nederlandse Organisatie voor Wetenschappelijk Onderzoek, Grant/ Award Number: ALWGR.2015.6 K E Y W O R D S biological control, ecological trap, evolution, exotic species, host–parasitoid interactions, parasitism T A X O N O M Y C L A S S I F I C A T I O N Evolutionary ecology g © 2022 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. \| 1 of 18 www.ecolevol.org Ecology and Evolution. 2022;12:e9030. https://doi.org/10.1002/ece3.9030 2.1 \| Model description 휆i(t) = 1 ni ri Hi(t) e − Hi(t) Ki . We simulate the evolution of two evolving traits, host preference behavior and parasitization strategy (i.e., physiological compatibility with two host species), in a population of parasitoids that has access to an original host species (host species 1) and a novel invader (host species 2). Generations are discrete and non-overlapping for both host species and the parasitoid; we denote their population densi- ties at the start of generation t by H1(t), H2(t), and P(t), respectively. Here, ni is the number of patches for host species i in the environment, ri is the maximum per-capita offspring production rate, Hi(t) denotes the total number of adult host individuals before density regulation (which is equal to the number of surviving hosts at the end of the pre- vious time step), and Ki reflects the carrying capacity of host species i with respect to within-species competition (Table 1); this parameter determines the host population density at equilibrium in the absence of parasitization, H∗ i = Ki ∙헅헇(ri ). 2.2 \| Host population dynamics The developing host offspring that have been deposited in a patch can be parasitized during a vulnerable period in their develop- ment. The host survives if no parasitoid successfully parasitizes the host before the end of the sensitive period. For simplicity, we model host development as a first-order process, so that the length of the sensitive period can be sampled from an exponential distribution with mean 휏Hi d , the average length of the sensitive developmental period for host i. Here, and also later on, more sophisticated waiting time distributions can be used if desirable, for example, based on available data, but at the cost of introducing additional parameters. However, irrespective of how waiting times are modeled precisely, variation in the length of the sensitive period introduces variation in the vulnerability of hosts to parasitoid attack, which is an important The environment is structured into patches that each contain a small subpopulation of host individuals. We assume that the two host species occur in the same area but in two different microhab- itats, so that they occur in separate patches and do not compete for resources with each other. The density of each host popula- tion is regulated at the global scale by density-dependent survival, which acts before the start of each parasitoid generation. The sur- viving hosts reproduce, and distribute their offspring randomly over the host patches for their species, after which the host adults die. 2 \| MATERIALS AND METHODS 2.1 \| Model description species i (i = 1 or 2) is distributed following a Poisson distribution with mean: 2 of 18 \| 1 \| INTRODUCTION Using an individual-based model, we here address the questions: (1) how does genetic variation for host preference and parasitization strategy influence the evolution of generalization and specialization in response to an evolutionary trap and (2) how does non-reproductive host mortality (i.e., an incompatible inter- action that results in host death without parasitoid offspring) in- fluences the evolutionary response of the parasitoid? We explore the evolution of parasitization strategy and preference behavior in a 2-host/1-parasitoid system under various costs of exerting host preference and generalist parasitization efficiencies, to identify conditions that would favor trap avoidance and/or adaptation to the unsuitable host. The model was inspired by solitary koinobiont endo-parasitoids attacking larval host stages, such as the larval par- asitoid Leptopilina heterotoma. Models that focus on host–parasitoid dynamics typically as- sume that a compatible host–parasitoid interaction results in host mortality and yields parasitoid offspring, whereas an incompat- ible interaction (such as a host trap) results in host survival and no parasitoid offspring (Hassell, 2000; Hassell & Pacala, 1990; Heimpel et al., 2003). However, a mismatch in parasitization strat- egy and host suitability (incompatibility) can also result in “non- reproductive host mortality,” that is, parasitoids kill the unsuitable host but their offspring do not develop (Abram et al., 2016). This has, KRUITWAGEN et al. 3 of 18 2.3 \| Parasitoid behavior If a host is found by a parasitoid before it has completed the sen- sitive period of development (휏Hi d ), it will be parasitized, incrementing the parasitoid's reproductive investment by one unit. The parasit- oid's reproductive investment reflects energy invested in parasitiza- tion and egg production. Hosts can be parasitized multiple times; we assume that parasitoids do not discriminate against hosts that were previously attacked by another individual. Hence, our model allows for superparasitism: a host can be attacked multiple times by differ- ent conspecific females. The incidence of self-superparasitism will be low as we assumed parasitoids to leave a patch once it encoun- ters a host individual for a second time. Many parasitoids are indeed reported to be able to discriminate between already parasitized hosts (e.g., Ueno & Tanaka, 1994; Varaldi et al., 2005). Conspecific superparasitism is relevant as theoretical and empirical studies show that this can occur in parasitoid wasps, including L. heterotoma, when the parasitization mark for example only last shortly and/or is not detectable by conspecific females (Hofsvang, 1988; Van Alphen & Visser, 1990; Visser et al., 1992), or even as an adaptive strategy for optimal patch usage (Visser et al., 1992). Parasitoid reproductive success is taken to be limited by the abil- ity to locate and exploit host patches (i.e., parasitoids are limited by time, not primarily by the availability of resources needed for egg- production). This is relevant as most parasitoids seem to live rela- tively short and die before exhaustion of all their eggs and/or are able to replenish their egg supply during their life (Ellers et al., 2000; Wajnberg, 2006; Wajnberg et al., 2016). In the individual-based sim- ulations, we therefore keep track of the time budget of each para- sitoid individual as it progresses through consecutive stages of the parasitization cycle (Figure 1a). The first step in the parasitization behavior is that the parasitoid must search the environment to locate a host patch. Searching para- sitoids are assumed to move through the environment randomly and encounter host patches at a constant rate 1∕휏s, where 휏s represents the average searching time needed to locate a host patch. Prreject patch (q) = ⎧ ⎪ ⎨ ⎪⎩ max(0,−q) if patch is occupied by host 1 max(0,+q) if patch is occupied by host 2 Accordingly, individuals with q = 0 accept all patches (i.e., they select a host patch at random), whereas, on the extremes of the scale, individuals with q = + 1 or q = −1 exclusively accept patches occupied by host species 1 or 2, respectively, and thus avoid the other host (Figure 1b). Note that such selectivity comes at the costs of an increase in the expected search time needed to locate an acceptable patch (a parasitoid that rejects a patch has to resume searching). thus assumed that hosts are distributed within the patch and are not clustered together. This can apply to parasitoids attacking the larval stage, but not egg masses since parasitoids do not require additional search time between encountering individual eggs once they have found the egg mass. factor promoting the stabilization and persistence of host–parasitoid systems (Hassell, 2000; Hassell & Pacala, 1990). factor promoting the stabilization and persistence of host–parasitoid systems (Hassell, 2000; Hassell & Pacala, 1990). 2.3 \| Parasitoid behavior After lo- cating a patch, the parasitoid may decide to reject it, depending on which host species is occupying the patch and the host preference trait of the parasitoid individual, q, a quantitative character that can range in value on a continuous scale between −1 and +1 (Figure 1, Table 1). In particular, the probability that the parasitoid rejects the patch is given by: After leaving a host patch, parasitoids need to recuperate for a period of time to restore their energy reserves allocated to repro- duction: we assumed that this recovery period consists of a sum of exponentially distributed waiting times, each with expected length 휏r for each unit of reproductive investment that the individual spent in the previous host patch (Figure 1a, Table 1). After recovering from previous reproductive investment, parasitoids are allowed to resume their search for another host patch, until no exploitable hosts are available in the environment anymore. This thus corresponds to a Holling type 2 functional response, in which the attack rate per host individual decelerates with increasing host density due to an increase in the total host handling time. Prreject patch (q) = ⎧ ⎪ ⎨ ⎪⎩ max(0,−q) if patch is occupied by host 1 max(0,+q) if patch is occupied by host 2 2.2 \| Host population dynamics The intrinsic population dynamics of the host are described by a stochastic variant of the Ricker model (Ricker, 1954), where the total number of offspring in a local subpopulation of host TA B LE 1 Parameters and initial values used in simulations Parameter Interpretation Parameter value(s) Parasitoid 휇헊 Mutation rate of host-preference 0;0.001 휇S Mutation rate of parasitization strategy 0;0.01 휏햯1 햽; 휏햯2 햽 Development time of parasitoid offspring for specialists (S = 1, S = 2) 20 휏햯3 햽 Development time of parasitoid offspring for generalists (S = 3) 46.51; 45.45; 44.44; 43.48; 42.55 휏s Search time, time for parasitoids to find a patch with hosts 3;7;11 휏a Time after which the parasitoid abandons the patch when failing to locate another host 1 휏e Host–parasitoid encounter rate 0.02 휏r Time needed for parasitoid to recover after foraging and localizing the next host-patch 0.5 P(t) Initial parasitoid population size 500 Host r1 Growth rate host 1 4 r2 Growth rate host 2 4 K1 Carrying capacity coefficient host 1 1000 K2 Carrying capacity coefficient host 2 1000 휏Hi d Time for a host to mature 10 n1 Number patches of host 1 20 n2 Number patches of host 2 20 TA B LE 1 Parameters and initial values used in simulations Parameter Interpretation Parameter value(s) Parasitoid 휇헊 Mutation rate of host-preference 0;0.001 휇S Mutation rate of parasitization strategy 0;0.01 휏햯1 햽; 휏햯2 햽 Development time of parasitoid offspring for specialists (S = 1, S = 2) 20 휏햯3 햽 Development time of parasitoid offspring for generalists (S = 3) 46.51; 45.45; 44.44; 43.48; 42.55 휏s Search time, time for parasitoids to find a patch with hosts 3;7;11 휏a Time after which the parasitoid abandons the patch when failing to locate another host 1 휏e Host–parasitoid encounter rate 0.02 휏r Time needed for parasitoid to recover after foraging and localizing the next host-patch 0.5 P(t) Initial parasitoid population size 500 Host r1 Growth rate host 1 4 r2 Growth rate host 2 4 K1 Carrying capacity coefficient host 1 1000 K2 Carrying capacity coefficient host 2 1000 휏Hi d Time for a host to mature 10 n1 Number patches of host 1 20 n2 Number patches of host 2 20 TA B LE 1 Parameters and initial values used in simulations TA B LE 1 Parameters and initial values used in simulations KRUITWAGEN et al. 4 of 18 2.4 \| Parasitoid offspring development Hence, host preference q = 0 indicates that parasitoids accept all host patches (i.e., random host searching), whereas, on the extremes of the scale, individuals with q = + 1 or q = −1 exclusively accept patches occupied by host species 1 or 2, respectively, and thus avoid the other host FI G U R E 1 Stages of parasitoid host-searching behavior (a) and the effect of parasitoid host preference genotype (q) on the probability of interacting with host species i after it has accepted a patch assuming equal numbers of host patches of each host species (b). When the probability equals 0.5, parasitoids interact with both host species with equal probability and thus select hosts at random (horizontal gray line). Hence, host preference q = 0 indicates that parasitoids accept all host patches (i.e., random host searching), whereas, on the extremes of the scale, individuals with q = + 1 or q = −1 exclusively accept patches occupied by host species 1 or 2, respectively, and thus avoid the other host FI G U R E 2 Stages of host exploitation by parasitoid wasps (parasitization) and the outcome of host–parasitoid interactions in terms of host and parasitoid offspring survival. When the parasitoid is incompatible with the host, two different scenarios are considered when the parasitoid exploits the host: Under scenario 1 the parasitoid offspring dies but the host survives, whereas under scenario 2 both the parasitoid offspring and host die FI G U R E 2 Stages of host exploitation by parasitoid wasps (parasitization) and the outcome of host–parasitoid interactions in terms of host and parasitoid offspring survival. When the parasitoid is incompatible with the host, two different scenarios are considered when the parasitoid exploits the host: Under scenario 1 the parasitoid offspring dies but the host survives, whereas under scenario 2 both the parasitoid offspring and host die event from happening but would also constitute a form of incompat- ibility (Fatouros et al., 2020; Vinson, 1990). of their offspring in either host species, resulting in a lower offspring survival probability (Straub et al., 2011). Parasitoid offspring can only develop if they are compatible with the host (Table 2, Figure 2). 2.4 \| Parasitoid offspring development The outcome of the host–parasitoid interaction is dependent on the parasitization strategy (S) of the parasitoid, and the timing of host and parasitoid offspring development (Figure 2). In addition, we consider two different scenarios in modeling incompatible host–parasitoid interactions: one with non-reproductive host-killing, the other with- out (Table 2, Figure 2). The parasitization strategy S is modeled as a discrete character, with three possible trait values, reflecting the following tactics: parasitoids with S = 1 or S = 2 are specialist para- sitoids that exhibit specific adaptations to, respectively, host species 1 or 2, and that are incompatible with the other host species; parasi- toids with S = 3 are generalists that are compatible with both hosts. As explained below, we assume that the broader host range of these generalists trades-off against a lower rate of development efficiency Once an individual accepts a host patch, it proceeds to exploit the available hosts living there (Figure 1a). Parasitoids encounter hosts in a random sequence, and continue to search the patch until they encounter a host individual for the second time, or if they have been searching for longer than 휏a time units since their last encounter with a host individual (휏a is the time threshold for abandoning the patch when failing to locate another host). Time intervals between encounters with a given host individual are drawn from an exponential distribution with mean 휏e (i.e., 1∕휏e corresponds to the per-capita host–parasitoid encounter rate). As a consequence of the stochasticity in the order and timing of host encounters, parasitoids tend to interact with a variable subset of the host individuals in a patch, rarely with all of them. Note, we KRUITWAGEN et al. 5 of 18 \| 5 of 18 \| 5 of 18 KRUITWAGEN et al. FI G U R E 1 Stages of parasitoid host-searching behavior (a) and the effect of parasitoid host preference genotype (q) on the probability of interacting with host species i after it has accepted a patch assuming equal numbers of host patches of each host species (b). When the probability equals 0.5, parasitoids interact with both host species with equal probability and thus select hosts at random (horizontal gray line). FI G U R E 2 Stages of host exploitation by parasitoid wasps (parasitization) and the outcome of host–parasitoid interactions in terms of host and parasitoid offspring survival. When the parasitoid is incompatible with the host, two different scenarios are considered when the parasitoid exploits the host: Under scenario 1 the parasitoid offspring dies but the host survives, whereas under scenario 2 both the parasitoid offspring and host die 2.4 \| Parasitoid offspring development Irrespective of compatibility, the rate of parasitoid development is taken to depend only on the strategy S of the parasitoid parent: offspring of a generalist parasitoid are assumed to develop more slowly than the offspring of specialists, such that their expected time of development 휏P3 d is larger than for specialists (휏P1 d and 휏P2 d ; we assume these to be equal for simplicity) and thus have a lower suc- cess rate of completion of development and offspring survival (e.g., Desneux et al., 2009; Raymond et al., 2016). In all cases, their devel- opmental time is drawn from an exponential distribution. The eventual fate of a parasitized host is determined by the relative timing of its infestation relative to the hosts' sensitive period: if a para- sitoid infests the host before it reaches the end of the sensitive period, parasitization will result in killing of the host (Figure 2). As we focused on solitary parasitoids, this event is associated with the emergence of a single new parasitoid individual, unless the host is incompatible (note that incompatible hosts are killed only in model scenario 2, with non- reproductive host-killing). When a host has been attacked multiple times by a conspecific parasitoid (superparasitism), the first parasitoid offspring that completes development is decisive for the fate of the host; the offspring of the other parasitoids are always inviable. Finally, when the host reaches the end of the sensitive developmental period before any parasitoid infests it the host survives. After the outcome of all interactions has been decided, the para- sitoid offspring replace the parental generation (such that their total number sets the value of P [t + 1]), and surviving hosts are collected to determine H1(t + 1) and H2(t + 1), the density of hosts at the start of the next time step. The model was implemented as a stochastic individual- and event- based simulation in the programming language C++ (Appendix S1), based on a modification of the Gillespie algorithm for stochastic simulations (Gillespie, 2007). Data produced by the simulation (pop- ulation densities of hosts and parasitoids; frequencies of parasitiza- tion tactics and distribution of host preference) were analyzed in R (version 4.0.1) (R Core Team, 2020). 2.4 \| Parasitoid offspring development 6 of 18 TA B LE 2 Overview of potential outcome of host–parasitoid interactions under two different scenarios. Parasitization strategy (S) 1 or 2, are specialist parasitoids only compatible with host 1 and 2 respectively, and S = 3 are generalist parasitoids compatible with both host species (host 1 and 2) aOutcome under the assumption that parasitoid development completes before the end of the vulnerable developmental period of the host; otherwise, the host survives. bDevelopmental rate of generalist (S = 3) assumed to be slower than specialists (S = 1,2), with the magnitude of the difference being dependent on 휏P3 d , relative to 휏햯1 햽and 휏햯2 햽. Compatible interactions do not have an inherent advantage in developmental rate over incompatible interactions. Travis, 2000; Sasaki & Godfray, 1999), host preference and parasiti- zation strategy were each genetically encoded by a single haploid locus, and we assumed reproduction to be clonal. Simulations were initialized with a small initial population of parasitoid individuals, all with S = 1 (pre-adapted to the original host (host 1); incompatible with the invader (host 2) and q = 0 (no initial host discrimination). Host population densities were initialized at their equilibrium den- sity in the absence of parasitization (H∗ i = Ki∙ln(ri )). Genetic varia- tion in host preference and parasitization strategy was introduced by a low rate of mutation (Dieckmann & Law, 1996; Geritz et al., 1997; Metz et al., 1992): mutations in host preference occurred with probability 휇q per reproductive event, and changed the off- spring's q-value to qoffspring = min(1,max(0,qparent + Δq)), where the mutational effect size Δq was sampled from a standard normal dis- tribution; mutations in the parasitization strategy occurred with a low probability 휇S per reproductive event, and changed the tactic from Sparent to Soffspring with probabilities Pr [ Sparent →Soffspring ], given by Pr[1 →3] = 1, Pr[2 →3] = 1 and Pr[3 →1] = Pr[3 →2] = 1∕2 (all conditional on the occurrence of a mutation). Accordingly, specialists can mutate to the generalist tactic, which can mutate to either one of the specialist tactics with equal probability, but specialists cannot mutate to become specialized for the other host species in a single mutational step. TA B LE 2 Overview of potential outcome of host–parasitoid interactions under two different scenarios. Parasitization strategy (S) 1 or 2, are specialist parasitoids only compatible with host 1 and 2 respectively, and S = 3 are generalist parasitoids compatible with both host species (host 1 and 2) 2.4 \| Parasitoid offspring development We modeled the two main mechanisms to cope with the evolutionary trap (behavior avoidance or adaptation), by allowing the population to evolve under different mutation probabilities of either parasitization strategy alone or par- asitization strategy and host-species preference and tested how the 2.4 \| Parasitoid offspring development This comprises the ability of the para- sitoid to utilize the hosts' resources for its own growth, its ability to inactivate hemocytes or regulate the host physiology or behav- ior to facilitate the development of the parasitoid offspring or its ability to otherwise suppress or evade the immune system of the host. Incompatibility indicates that the parasitoid offspring does not succeed until the end of its development and dies. This would correspond to endo-parasitoids that come into contact with the in- ternal immune system of the host, such as larval parasitoids attack- ing Drosophila (Rizki & Rizki, 1990). Egg and pupal parasitoids often do not have to deal with cellular immune response (but see Reed et al., 2007; Yang et al., 2019), as eggs and pupal stages mostly rely on their external barrier, such as egg coating and the thickness of the chorion or puparium wall. This may prevent the entire parasitization The outcome of an incompatible host–parasitoid interaction dif- fers between scenario 1 and 2 (Table 2). In the model scenario with- out non-reproductive host killing (scenario 1), either the parasitoid offspring develops when it is compatible with the host, or the host survives the parasitoid attack when they are incompatible. Thus, in this scenario, either the host or the new parasitoid offspring survives the interaction. In the other model scenario with non-reproductive host killing (scenario 2), parasitoid offspring can initially develop in the incompatible hosts to the point that the host is killed, but par- asitoid offspring cannot successfully complete development in an incompatible host. Host death can come about due to, for exam- ple, damage inflicted by parasitoid attack and/or self-harm through mounting an immune defense. Hence, when the host is parasitized, it always dies, whether or not it is compatible. Thus, in this scenario, either the host or the parasitoid survives parasitization. 6 of 18 \| Parasitization strategy S Host species Outcomea Compatible interactions 1 1 Fast parasitoid developmentb; emergence of parasitoid offspring; host dies. 2 2 3 1, 2 Slow parasitoid developmentb; emergence of parasitoid offspring; host dies. Incompatible interactions 1 2 Scenario 1—without non-reproductive host killing: No interaction; host survives. Scenario 2—with non-reproductive host killing: Fast parasitoid developmentb, but no emergence of viable parasitoid offspring; host dies. 2 1 aOutcome under the assumption that parasitoid development completes before the end of the vulnerable developmental period of the host; otherwise, the host survives. 2.4 \| Parasitoid offspring development bDevelopmental rate of generalist (S = 3) assumed to be slower than specialists (S = 1,2), with the magnitude of the difference being dependent on 휏P3 d , relative to 휏햯1 햽and 휏햯2 햽. Compatible interactions do not have an inherent advantage in developmental rate over incompatible interactions. 6 of 18 \| Parasitization strategy S Host species Outcomea Compatible interactions 1 1 Fast parasitoid developmentb; emergence of parasitoid offspring; host dies. 2 2 3 1, 2 Slow parasitoid developmentb; emergence of parasitoid offspring; host dies. Incompatible interactions 1 2 Scenario 1—without non-reproductive host killing: No interaction; host survives. Scenario 2—with non-reproductive host killing: Fast parasitoid developmentb, but no emergence of viable parasitoid offspring; host dies. 2 1 aOutcome under the assumption that parasitoid development completes before the end of the vulnerable developmental period of the host; otherwise, the host survives. bDevelopmental rate of generalist (S = 3) assumed to be slower than specialists (S = 1,2), with the magnitude of the difference being dependent on 휏P3 d , relative to 휏햯1 햽and 휏햯2 햽. Compatible interactions do not have an inherent advantage in developmental rate over incompatible interactions. T o u P s h g b 6 of 18 \| KRUITWAGEN et al. Parasitization strategy S Host species Outcomea Compatible interactions 1 1 Fast parasitoid developmentb; emergence of parasitoid offspring; host dies. 2 2 3 1, 2 Slow parasitoid developmentb; emergence of parasitoid offspring; host dies. Incompatible interactions 1 2 Scenario 1—without non-reproductive host killing: No interaction; host survives. Scenario 2—with non-reproductive host killing: Fast parasitoid developmentb, but no emergence of viable parasitoid offspring; host dies. 2 1 aOutcome under the assumption that parasitoid development completes before the end of the vulnerable developmental period of the host; otherwise, the host survives. bDevelopmental rate of generalist (S = 3) assumed to be slower than specialists (S = 1,2), with the magnitude of the difference being dependent on 휏P3 d , relative to 휏햯1 햽and 휏햯2 햽. Compatible interactions do not have an inherent advantage in developmental rate over incompatible interactions. TA B LE 2 Overview of potential outcome of host–parasitoid interactions under two different scenarios. Parasitization strategy (S) 1 or 2, are specialist parasitoids only compatible with host 1 and 2 respectively, and S = 3 are generalist parasitoids compatible with both host species (host 1 and 2) KRUITWAGEN et al. 3 \| RESULTS Multiple traits can be involved in host-range and host adaptation, which might evolve independently in response to selection. We therefore first investigated how genetic variation in parasitization tactic (S), the parasitoid's host-use strategy that determines its physiological compatibility with either of the host species or both, would influence (physiological) host-range evolution. Next, we also considered genetic variation in host-preference behavior (q), the parasitoid's inherent choice to lay eggs in certain hosts. We did this under different costs by examining the influence of (1) the relative developmental success of generalist parasitoids (determined by different parameter ratios 휏P1 d ∕휏P3 d ) and (2) parasitoid search time (depending on parameter 휏헌). Search time may vary depending on environmental conditions: a low search time might reflect a resource rich-environment in which patches are clustered, whereas a rela- tively high search time might reflect a resource-poor environment in which host-habitat patches are sparse and/or more difficult to locate. The latter conditions imply a higher evolutionary costs of exerting host preference, as the decision to reject a patch would ne- cessitate a large time investment for finding a more suitable patch, compromising the amount of time remaining for reproduction. Interestingly, the parasitoid search efficiency also influences evolution of specialization. Generalization tended to evolve at low and/or medium search efficiencies while specialization tended to occur at high search efficiencies (Figure 3a). Even at the lowest gen- eralist efficiency, generalists can sometimes invade the population when search efficiency is low as shown in the example timeseries in Figure 3b. Although note that this rarely occurred across repli- cate simulates under these extreme parameter values (Figure 3a). An explanation is that higher search efficiencies increase the number of host-patches a parasitoid can visit during its life. This increases the chance of a specialist to find a patch with suitable hosts and give them a competitive advantage even when costs of generalists are relatively low. As such, the search efficiency of the parasitoid can exceed the costs of a narrow host use and facilitate evolution of specialization on the novel host. In contrast, at low/medium search efficiency, generalists have an advantage as they can successfully exploit hosts on every patch they visit, and thus will always find a patch with suitable hosts. 3.1.2 \| Scenario 2: Incompatible interaction results in host-killing 3.1.2 \| Scenario 2: Incompatible interaction results in host-killing 3 \| RESULTS Yet, independent of the parasitoid search efficiency and the magnitude of the generalist-specialist trade-off, genetic variation in host-use under scenario 1 consistently allowed parasitoids to adapt to the “trap” host, establishing a novel host– parasitoid relationship (Figure 3a), either by generalization or by spe- cialization. Note that parasitoids with a physiological specialist tactic can still make maladaptive host choices when accepting patches of their unsuitable host due to their random host selection behavior. Initially, all parasitoids were specialists of host species 1, mean- ing that exploitation of this host results in one parasitoid offspring and host mortality. Parasitoids were not able to discriminate be- tween patches of host 1 or 2 (no preference, random host patch se- lection) and lost time exploiting unsuitable host patches. As such, host 2 acted as an “evolutionary trap.” When specialists attempt to exploit an incompatible host, this either had no effect on host sur- vival (model scenario 1), or resulted in host death without parasitoid offspring, that is, non-reproductive host killing (scenario 2) (Table 2, Figure 2). \| costs of exerting host preference and the developmental costs of broadening the host-range influenced the evolution of the parasit- oid. These trade-offs were investigated by altering the host patch search time 휏s and the likelihood of successful infestation by gener- alist parasitoids, as determined by the parameter 휏P3 d (Table 1). We did this under two scenarios: one in which incompatible interaction had no effect on host survival and one in which incompatible inter- action resulted in host death (Table 2, Figure 2). Each combination of parameter settings was replicated 30 times and run for 1000 time steps (generations). (scenario 1, Table 2, Figure 2). This can occur for example when the host is unsuitable for development of the parasitoid offspring and the host survives the attack by encapsulation of the parasitoid egg (Kacsoh & Schlenke, 2012; Vinson, 1990). Figure 3 shows example simulations of relative trait values for parasitization strategy over 1000 generations in which host preference is not allowed to evolve and parasitoids exhibit random host patch selection under low and high efficiencies of generalists and host searching. A mutation of parasitization strategy (S) first results in the invasion of a generalist parasitization strategy which is able to overcome the unsuitability of the novel host and thus able to reproduce on host 1 as well as host 2 for reproduction. Note that we assumed that a specialist of host 1 cannot directly shift to the alternative specialist strategy and must first become a generalist by mutation. However, when general- ists are less efficient in parasitization (and without non-reproductive host killing), selection ultimately favors the evolution of a polymor- phism of two specialists, after the specialist for host 2 has emerged by mutation from the generalist strategy (Figure 3b). Hence, as ex- pected, low generalist efficiencies result in a consistently higher rela- tive frequency of specialists across replicate simulations (Figure 3a). 2.5 \| Genetic assumptions, initial conditions and simulation details Hymenoptera parasitoids are haplodiploid, but for simplicity and comparability with other host–parasitoid models (e.g., Fellowes & KRUITWAGEN et al. 7 of 18 3.1.1 \| Scenario 1: Incompatible interaction has no effect on host survival Panel (a) shows relative frequencies of specialist and generalist parasitization strategies (±SE) after 1000 generations (n = 30) at various generalist efficiencies and host-patch search efficiencies (low, 휏헌= 11; medium, 휏헌= 7; high, 휏헌= 3). Hence, which genotype(s) dominate depends on the replicate. Panel (b) shows four examples time series of parasitization strategies at low and high generalist efficiencies (휏햯3 햽= 45.5, 휏햯3 햽= 42.5 resp.) and low and high host-patch search efficiencies (휏헌= 11, 휏헌= 3 resp. ) have relative low host use efficiency (Figure 4), conditions which pro- moted the evolution of specialists of the unsuitable host under sce- nario 1. The ability to kill unsuitable hosts allows specialists of host 1 to compete with generalists by reducing the number of available hosts for generalists to exploit hindering the evolution of generaliza- tion. Consequently, the evolution of specialization on the (initial) un- suitable host 2 is hampered as well, because specialists of host 2 can only evolve through mutation of the generalist strategy. Hence, non- reproductive host killing can increase the competitive advantage of specialists over generalists and limit the evolution of specialization on host species 2 resulting in physiological host-range conservation. This also means that non-reproductive host killing can be maintained and expressed under these conditions when attacking unsuitable hosts. have relative low host use efficiency (Figure 4), conditions which pro- moted the evolution of specialists of the unsuitable host under sce- nario 1. The ability to kill unsuitable hosts allows specialists of host 1 to compete with generalists by reducing the number of available hosts for generalists to exploit hindering the evolution of generaliza- tion. Consequently, the evolution of specialization on the (initial) un- suitable host 2 is hampered as well, because specialists of host 2 can only evolve through mutation of the generalist strategy. Hence, non- reproductive host killing can increase the competitive advantage of specialists over generalists and limit the evolution of specialization on host species 2 resulting in physiological host-range conservation. This also means that non-reproductive host killing can be maintained and expressed under these conditions when attacking unsuitable hosts. due to physiological costs; e.g., self-harm through mounting an im- mune defense. 3.1.1 \| Scenario 1: Incompatible interaction has no effect on host survival Second, we consider the scenario where an incompatible host– parasitoid interaction results in host killing instead of host survival (Scenario 2, Figure 2, Table 2). This might occur when the host dies First, we consider the scenario where an incompatible interac- tion between a host and parasitoid does not result in host-killing 8 of 18 KRUITWAGEN et al. 8 of 18 \| KRUITWAGEN et al. FI G U R E 3 Evolution of parasitization strategy (S) with fixed host-preference (q = 0, i.e., random dispersal) under scenario 1 in which incompatible interactions have no effect on host survival. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but incompatible with host 2 and vice-versa for specialist 2. Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (‘generalist efficiency’). At generation 0, only specialists of host 1 occur (red) but also attack hosts that do not match their parasitization strategy due to random host searching. They can evolve a generalist strategy (gray) and subsequently mutate to become host use specialist of host 2 (blue). Panel (a) shows relative frequencies of specialist and generalist parasitization strategies (±SE) after 1000 generations (n = 30) at various generalist efficiencies and host-patch search efficiencies (low, 휏헌= 11; medium, 휏헌= 7; high, 휏헌= 3). Hence, which genotype(s) dominate depends on the replicate. Panel (b) shows four examples time series of parasitization strategies at low and high generalist efficiencies (휏햯3 햽= 45.5, 휏햯3 햽= 42.5 resp.) and low and high host-patch search efficiencies (휏헌= 11, 휏헌= 3 resp. ) FI G U R E 3 Evolution of parasitization strategy (S) with fixed host-preference (q = 0, i.e., random dispersal) under scenario 1 in which incompatible interactions have no effect on host survival. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but incompatible with host 2 and vice-versa for specialist 2. Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (‘generalist efficiency’). At generation 0, only specialists of host 1 occur (red) but also attack hosts that do not match their parasitization strategy due to random host searching. They can evolve a generalist strategy (gray) and subsequently mutate to become host use specialist of host 2 (blue). 3.1.1 \| Scenario 1: Incompatible interaction has no effect on host survival Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (“generalist efficiency”). At generation = 0, only specialists of host 1 occur (red) but also attack hosts FI G U R E 4 Evolution of parasitization strategy (S) with fixed host-preference (q = 0, i.e., random dispersal) under scenario 2 in which incompatible interactions results in host-killing. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but are incompatible with host 2 and vice-versa for specialist 2. Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (“generalist efficiency”). At generation = 0, only specialists of host 1 occur (red) but also attack hosts that do not match their parasitization strategy. They can evolve a generalist strategy (gray) and subsequently mutate to become host use specialist of host 2 (blue). Panel (a) shows relative frequencies of specialist and generalist parasitization strategies (±SE) after 1000 generations (n = 30) at various generalist efficiencies and host-patch search efficiencies (low, 휏헌= 11; medium, 휏헌= 7; high, 휏헌= 3). Hence, which genotype(s) dominate depends on the replicate. Panel (b) shows four examples time series of parasitization strategies at low and high generalist efficiencies (휏햯3 햽= 45.5, 휏햯3 햽= 42.5 resp.) and low and high host-patch search efficiencies (휏헌= 11, 휏헌= 3 resp.) 3.1.1 \| Scenario 1: Incompatible interaction has no effect on host survival A similar pattern arises for a subset of conditions: (1) relatively low developmental costs of broadening the physiologi- cal host-range resulted in the evolution of a generalist parasitization tactic and (2) high search efficiencies combined with low efficiency of generalists promoted parasitization specialists (Figure 4). However, in contrast to the simulations without host killing (scenario 1), either specialists of host 1 or generalists dominated after 1000 generations, but specialists of the novel host did not evolve (Figure 4). In other words, non-reproductive killing of unsuitable hosts can prevent spe- cialists of host 2 to evolve, constraining adaptation and the establish- ment of novel host–parasitoid relationship. This appeared in particular when parasitoids exhibited a high search efficiency and generalists \| 9 of 18 KRUITWAGEN et al. FI G U R E 4 Evolution of parasitization strategy (S) with fixed host-preference (q = 0, i.e., random dispersal) under scenario 2 in which incompatible interactions results in host-killing. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but are incompatible with host 2 and vice-versa for specialist 2. Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (“generalist efficiency”). At generation = 0, only specialists of host 1 occur (red) but also attack hosts that do not match their parasitization strategy. They can evolve a generalist strategy (gray) and subsequently mutate to become host use specialist of host 2 (blue). Panel (a) shows relative frequencies of specialist and generalist parasitization strategies (±SE) after 1000 generations (n = 30) at various generalist efficiencies and host-patch search efficiencies (low, 휏헌= 11; medium, 휏헌= 7; high, 휏헌= 3). Hence, which genotype(s) dominate depends on the replicate. Panel (b) shows four examples time series of parasitization strategies at low and high generalist efficiencies (휏햯3 햽= 45.5, 휏햯3 햽= 42.5 resp.) and low and high host-patch search efficiencies (휏헌= 11, 휏헌= 3 resp.) 9 of 18 KRUITWAGEN et al. 9 of 18 \| FI G U R E 4 Evolution of parasitization strategy (S) with fixed host-preference (q = 0, i.e., random dispersal) under scenario 2 in which incompatible interactions results in host-killing. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but are incompatible with host 2 and vice-versa for specialist 2. For example, whereas at t ~ 500 parasitoids with high search efficiency evolved as spe- cialists with a behavioral preference for their suitable host, a com- plete avoidance of their unsuitable hosts generally did not evolve at t = 1000 when search efficiency was low (Figure 5). Low search efficiency can therefore favor random host acceptance (prefer- ence = 0) over host avoidance, giving individuals with a generalist parasitization strategy an advantage. Note that host-preference of generalists highly fluctuates when generalists occur at low frequency and are not evolving (Figure 5a). The erratic pattern is therefore simply the result of the low number of parasitoids with generalist parasitization tactic that stochastically mutated among parental genetic background. In conclusion, genetic variation in host preference (1) promotes parasitoids with a specialist host use strategy and (2) can facilitate the evolution and co-existence of two distinct specialist strategies in response to the evolutionary trap: one evolving trap avoidance; the other evolving a preference and specialization for the novel host. a mismatch in host choice and performance (Figure 4). Only when search efficiency was high and costs of being a generalist low, evo- lution of generalist strategy occurred, although a complete physi- ological host-shift did not occur by evolution of specialists of host 2. When non-reproductive host killing occurred, interestingly, heritable variation in both host preference and performance also changed the evolutionary outcome. In this case, it resulted in evolution of special- ists of the ‘trap’ host (host 2) allowing them to co-occur with spe- cialist of host 1 (Figure 6). In other words, genetic variation in both traits increases the chance of establishment of novel host–parasitoid interaction and allows specialization to the unsuitable host by a com- plete physiological host-shift instead of host-range expansion. This indicates that in the situation of parasitoids killing unsuitable hosts, evolution of avoidance behavior matters. This is because parasi- toids with a specialist host use tactic of host 1 can evolve a com- plete avoidance of the trap; this reduces killing of unsuitable host 2 and removal of resources for parasitoids with a genotype enabling to reproduce on the trap host. Hence, evolution of host-preference can decrease the strength of competition through non-reproductive host killing when specialists evolve host-preference that matches their parasitization strategy, facilitating evolution and persistence of specialists of host 2. FI G U R E 5 Evolution of both parasitization strategy (S) and host-preference (q) under scenario 1 in which incompatible interactions have no effect on host survival. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but are incompatible with host 2 and vice-versa for specialist 2. Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (“generalist efficiency”). At generation = 0, only specialists of host 1 occur (red) and exhibit a random host searching (i.e., no preference, q = 0) and thus also attack hosts that do not match their parasitization strategy (host 2). They can evolve a generalist strategy (gray) and subsequently mutate to become host use specialist of host 2 (blue). Preference of 0 indicates random host searching and parasitoids can evolve preference for host-1 (p > 0) or preference for host 2 (p < 0). Panel (a) shows relative frequencies of specialist and generalist parasitization strategies (±SE) and preference genotypes after 1000 generations (n = 30) at various generalist efficiencies and host-patch search efficiencies (low, 휏헌= 11; medium, 휏헌= 7; high, 휏헌= 3). Hence, which genotype(s) dominate depends on the replicate. Panel (b) shows four examples time series at low and high generalist efficiencies (휏햯3 햽= 45.5, 휏햯3 햽= 42.5 resp.) and low and high host-patch search efficiencies (휏헌= 11, 휏헌= 3 resp.) the evolution of a generalist towards specialist tactic with de- creasing generalist efficiencies and increasing search efficiencies. However, the conditions for evolution of specialization are relaxed by combined evolution of host preference and parasitization tac- tic (Figure 5). The evolution of parasitization specialists was fa- cilitated because specialists gain an advantage over generalists by being able to select patches that matches their performance. As such, specialists are able to avoid their unsuitable host (preference ≠ 0) in presence of genetic variation in host-preference (Figure 5). Only at low host-search efficiencies and high generalist efficien- cies, parasitoids able to reproduce on both hosts were selected, thus resulting in adaptation by host-range expansion instead of a physiological host-shift. Low search efficiency increases the time for individuals to find suitable host patches, making it costly to be choosy, promoting random host acceptance. Specialists of host 1 that exhibit non-reproductive host killing of host 2, however, tend to dominate and constrain physiological host-range expansion and specialization on the (initial) suboptimal host when search efficiency is low and when generalists exhibit rel- ative high efficiencies (Figure 6). First consider that the individuals' time budget shapes the trade-off between random dispersal and the optimization of behavior preference with parasitization host use strategy. It follows that when search time constrains evolution of optimal host-preference, frequent killing of incompatible hosts im- pairs the evolution and persistence of generalists. Next, as a physi- ological host-shift is based on sequential evolution of a generalist to specialist strategy, a high cost of being a generalist further reduces the chance that specialists of host-2 can appear and persist. Hence, when parasitoids exhibit non-reproductive host killing, a complete physiological host-shift is most likely to occur when (1) specialists of host 1 exhibit preference for host 1 reducing intra-specific com- petition and (2) when search efficiency is high, reducing the costs of being choosy and allowing a behavioral preference for host 2 to evolve. 3.2.1 \| Scenario 1: Incompatible interaction has no effect on host survival Figure 5a shows the relative frequency of parasitization tactic and host-preference after 1000 generations under scenario 1 in which an incompatible interaction has no effect on host survival (Figure 2, Table 2). Figure 5b shows example simulations of rela- tive trait values for both parasitization strategies and host prefer- ence over 1000 generations, with various levels of generalist host exploitation efficiency and host searching efficiency. The evo- lutionary response to the unsuitable host trap was qualitatively similar to the situation with fixed random host selection behavior (Figure 3): consistent adaptation towards the ‘trap’ host among replicates by change in parasitization strategy with a switch from We next allowed both parasitization strategy (S) and the behavioral trait host preference to evolve (q). Host preference relies on indi- viduals rejecting host patches, which is a costly decision if the time needed to find a patch is high. Parasitoids that exhibit an avoidance for either host species can therefore increase their total repro- ductive output by selecting suitable host-patches that match their parasitization tactic (performance) and thus lose less time exploiting unsuitable hosts. The evolution of host species preference can how- ever be constrained by the search efficiency of individuals in a given environment as this sets their time budget to find their preferred hosts (Figure 1). 10 of 18 \| KRUITWAGEN et al. KRUITWAGEN et al. 10 of 18 KRUITWAGEN et al. 11 of 18 4 \| DISCUSSION Kruitwagen et al., 2021; Liu et al., 2015; Zhou et al., 2019), but lit- tle is known about its influence on ecological and evolutionary pro- cesses. We found that non-reproductive host killing can hamper adaptation to the trap in conditions which in fact promoted host- range evolution in parasitoids when they were not able to kill un- suitable hosts. Moreover, we show that the establishment of a novel host–parasitoid relationship does not necessarily have to occur at the behavioral level through changing host-preference behavior, as a physiological strategy can evolve that enables reproduction on the trap. However, evolved behavioral preference promotes adap- tation to the trap by a complete host-shift and allows parasitoids to evolve avoidance of their unsuitable host, minimizing suboptimal host choices. The rising number of exotic (invasive) species calls for understand- ing and predicting not only the—short-term—ecological impact on native species in the invaded range but also their impact over evo- lutionary time (Mooney & Cleland, 2001; Schlaepfer et al., 2005; Strauss et al., 2006). Invasive species can act as an evolutionary trap for parasitoids when they are unsuitable for reproduction but indis- tinguishable from suitable hosts (Schlaepfer et al., 2002; Schlaepfer et al., 2005). Such incompatible interactions reduce parasitoid fit- ness and results in a mismatch between host choice and offspring survival (performance), that is, “bad motherhood” (Thompson, 1988; Yoon & Read, 2016). Numerous studies have reported a negative relationship between host choice and performance when native parasitoids attack a non-native host species (Konopka et al., 2018; Martini et al., 2019; Romero et al., 2020). One such example is the solitary koinobiont endo-parasitoid L. heterotoma attacking the inva- sive D. suzukii fruit-fly. Although this fruit-fly species is unsuitable for parasitoid offspring development due to its high resistance (Kacsoh & Schlenke, 2012), this incompatible host–parasitoid interaction sometimes results in host mortality (Kruitwagen et al., 2021). Other examples in which parasitoids show low performance in exotic host species include native scelionine egg parasitoids attacking H. halys, euphorinae parasitoids attacking adult H. axyridis and exoris- tinae larval parasitoids attacking C. perspectalis (Firlej et al., 2012; Konopka et al., 2018; Martini et al., 2019). FI G U R E 6 Evolution of both parasitization strategy (S) and host-preference (q) under scenario 2 in which incompatible interactions result in host killing. Specialist 1 (red) represents parasitoids able to reproduce on host 1, but are incompatible with host 2 and vice-versa for specialist 2. Generalists (gray) are able to reproduce on both host species but vary in respect to their host use efficiency relative to specialists (“generalist efficiency”). At generation = 0, only specialists of host 1 occur (red) and exhibit a random host searching (i.e., no preference, q = 0) and thus also attack hosts that do not match their parasitization strategy (host 2). They can evolve a generalist strategy (gray) and subsequently mutate to become host use specialist of host 2 (blue). Preference of 0 indicates random host searching and parasitoids can evolve preference for host 1 (p > 0) or preference for host 2 (p < 0). Panel (a) shows relative frequencies of specialist and generalist parasitization strategies (±SE) and preference genotypes after 1000 generations (n = 30) at various generalist efficiencies and host-patch search efficiencies (low, 휏헌= 11; medium, 휏헌= 7; high, 휏헌= 3). Hence, which genotype(s) dominate depends on the replicate. Panel (b) shows four examples time series at low and high generalist efficiencies (휏햯3 햽= 45.5, 휏햯3 햽= 42.5 resp.) and low and high host-patch search efficiencies (휏헌= 11, 휏헌= 3 resp. ) 3.2.2 \| Scenario 2: Incompatible interaction results in host-killing As explained above, when parasitization strategy alone was allowed to evolve under scenario 2, the parasitoid's potential to adapt to the unsuitable host was hindered resulting in physiological host- range conservation (domination of specialists of host 1), remaining KRUITWAGEN et al. 12 of 18 12 of 18 \| KRUITWAGEN et al. KRUITWAGEN et al. 13 of 18 for time-limited parasitoids, females should maximize their host encounter rate to maximize their reproductive output (Comins & Hassell, 1979; Wajnberg, 2006). Our findings are in agreement with other studies (McNamara et al. (1993); e.g., Barrette et al. (2010)) that found that optimal foraging may include acceptance of less profitable hosts/prey to maximize reproductive output when travel time between patches is long. This however might not be the case for egg-limited parasitoids (such as synovigenic insects) as they are pre- dicted to maximize the quality of hosts they accept rather than their host encounter rate (Fletcher et al., 1994; Minkenberg et al., 1992). A thorough understanding of possible outcomes under egg-limitation would require a similar modeling approach as outlined in this paper. host was high as a result of a high efficiency to find host patches and random-host selection behavior. This makes sense as a high encoun- ter rate increases the parasitoid impact on the host population size and thus their competitive advantage. We assumed a fixed high intrinsic growth rate and equal number of host-patches of each host to support the parasitoid population. In nature however, variation in environmental factors such as climate conditions and resource quality and quantity might cause temporal variation in the hosts' reproduction potential and carrying capacity (e.g., Behrman et al., 2015; Lue et al., 2018). Moreover, host-species population dynamics might be altered when hosts share (some) resources resulting in inter-specific competition. As the host com- munity structure influences the magnitude and direction of selec- tion of the parasitoid, future modeling studies could focus on how more complex host population dynamics influence the role of non- reproductive host killing on the evolutionary trajectory of the par- asitoid. Some relevant factors that could be included are stochastic changes in reproductive potential of the hosts, the relative number of host patches and inter-specific competition between the host species. It would therefore be very interesting to implement field survey data to increase the predictive value of future models. 4.3 \| Genetic assumptions and evolutionary trajectory We assumed haploid genotypes with two single unlinked loci deter- mining host use and host preference with clonal inheritance. It has to be considered however that many parasitoids reproduce sexually through haplodiploidy (arrhenotoky) meaning that unfertilized eggs arise through meiosis and develop into males, and females arise after mating through fusion of two gametes and are diploid (Godfray & Cook, 1997; Heimpel & De Boer, 2008). Moreover, recombination between traits can occur in sexually reproducing parasitoid females at a frequency that seems to be similar to diploid higher eukary- otes (Beukeboom et al., 2010; Niehuis et al., 2010). Consequently, in nature, this may retard/prevent adaptation by specialization under joint evolution of host preference and parasitization strategy through continuous recombination between genotypes (Doebeli & Dieckmann, 2000; Felsenstein, 1981) or may speed up selec- tion by bringing together the two traits (Felsenstein, 1974; Marais & Charlesworth, 2003) depending on, for example, recombination frequency and population size. As our model does not allow for re- combination, a similar outcome of the model would be observed if traits interact via epistatic or pleiotropic effects in a sexual repro- ducing parasitoid species with multilocus genetics, such as when the expression of host preference would be developmentally coupled to the expression of parasitization strategy. Alternatively, match- ing host-preference with parasitization performance might arise through simultaneous but independent selection on both traits. In fact, the latter seems more likely as empirical studies indicate that physiological traits have no pleiotropic effect on behavior (host choice) in parasitoids and may therefore evolve parallel to each other (Dubuffet et al., 2006; Rolff & Kraaijeveld, 2001). 4.1 \| Non-reproductive host killing constrains evolution Even when parasitoids are not able to exploit hosts for reproduc- tion, their attack can still reduce host survival rate due to, for exam- ple, wounding or immune defense costs (Abram et al., 2016, 2019; Kruitwagen et al., 2021). Previous empirical and modeling studies showed that magnitude of non-reproductive host killing can influence host–parasitoid population dynamics via direct and indirect interac- tions (Heimpel et al., 2003; Kaser et al., 2018). We found that when incompatible host interactions result in host killing it also influences the ability of the parasitoid to adapt to a trap host. This adaptation entails establishing a novel compatible host–parasitoid relationship. While adaptation to the trap consistently occurred when parasitoids did not affect survival of their unsuitable host, adaptation can be con- strained when parasitoid attack is destructive for the host. This is be- cause the removal of resources through non-reproductive host killing increases the strength of competition with (generalist or specialist) genotypes compatible with the trap host and can therefore hamper their evolution. In fact, when parasitoids exhibit non-reproductive host killing of unsuitable hosts, their strategy appeared to be an “evo- lutionary dead end”: parasitoids remained attacking the trap without changing their host use tactic to include the host into their reper- toire. This occurred in particular when (1) specialist parasitoids were more efficient in parasitizing their host compared to genotypes with a generalist host use strategy and (2) the encounter rate with the trap Using the L. heterotoma–D. suzukii system as inspiration, in this study, we explored which conditions could promote parasitoid ad- aptation through improved performance on the trap host and re- duce suboptimal host choices through change in host preference. Whereas trap avoidance would protect parasitoid populations from becoming extinct, adaptation would also enhance their capacity to suppress invasive hosts and thus their value for biological control. Using an individual based model, we show that the outcome of in- compatible host–parasitoid interaction matters for the potential of the parasitoid to establish a new viable host–parasitoid relationship with the trap host. While compatible interactions are always fatal for the host (reproductive host killing), we considered that attacking un- suitable hosts either leaves the host unharmed or, on the other ex- treme, is fatal for the host. The latter often occurs in host–parasitoid systems (e.g., Abram et al., 2016, 2019; Heimpel et al., 2003; KRUITWAGEN et al. 14 of 18 OPEN RESEARCH BADGES This article has earned an Open Data badge for making pub- licly available the digitally-shareable data necessary to repro- duce the reported results. The data is available at: https://doi. org/10.5061/dryad.31zcrjdp6. The model code can be found in the Appendix S1. CONFLICT OF INTEREST The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpreta- tion of data; in the writing of the manuscript, or in the decision to publish the results. Astrid Kruitwagen https://orcid.org/0000-0001-5952-1589 Astrid Kruitwagen https://orcid.org/0000-0001-5952-1589 that they are also influenced by variation in environmental condi- tions. For example, we found in a previous study that heritability of attack rate and non-reproductive host killing in L. heterotoma with regard to the invasive D. suzukii host is h2 = 0.2 (Kruitwagen et al., 2021). Hence, these genetic factors might change the evo- lutionary trajectory by altering the rate of evolution and/or which traits are selected (Fellowes & Travis, 2000; Kawecki, 1998). For instance, on the one extreme, lack of evolution in host-preference under expression of non-reproductive host killing may seriously constrain evolution of specialization as we have shown in our model with universal host acceptance (random host searching). and editing (equal). Leo W. Beukeboom: Conceptualization (equal); methodology (equal); writing – review and editing (equal). Bregje Wertheim: Conceptualization (equal); methodology (equal); writing – review and editing (equal). G. Sander van Doorn: Conceptualization (equal); methodology (lead); writing – review and editing (equal). REFERENCES Abram, P., Brodeur, J., Burte, V., & Boivin, G. (2016). Parasitoid-induced host egg abortion: An underappreciated component of biological control services provided by egg parasitoids. Biological Control, 98, 52–60. Abram, P., Brodeur, J., Urbaneja, A., & Tena, A. (2019). Nonreproductive effects of insect parasitoids on their hosts. Annual Review of Entomology, 64, 259–276. Abram, P., Gariepy, T., Boivin, G., & Brodeur, J. (2014). An invasive stink bug as an evolutionary trap for an indigenous egg parasitoid. Biological Invasions, 16(7), 1387–1395. Barrette, M., Boivin, G., Brodeur, J., & Giraldeau, L.-A. (2010). Travel time affects optimal diets in depleting patches. Behavioral Ecology and Sociobiology, 64(4), 593–598. Behrman, E., Watson, S., O'Brien, K., Heschel, M., & Schmidt, P. (2015). Seasonal variation in life history traits in two drosophila species. Journal of Evolutionary Biology, 28(9), 1691–1704. Benoist, R., Capdevielle-Dulac, C., Chantre, C., Jeannette, R., Calatayud, P. A., Drezen, J. M., Dupas, S., Le Rouzic, A., Le Ru, B., Moreau, L., Van Dijk, E., Kaiser, L., & Mougel, F. (2020). Quantitative trait loci involved in the reproductive success of a parasitoid wasp. Molecular Ecology, 29(18), 3476–3493. 5 \| CONCLUSION This model together with previous studies shows that genetics, be- havioral, and physiological factors can constrain parasitoids to evolve matching of host-preference and host-performance explaining why parasitoids' host selection behavior is not always in agreement with their physiological host use capacity and thus results in a discrep- ancy in their behavioral and physiological host-range (e.g., Cronin et al., 2001; Thompson, 1988). This study adds that non-reproductive host killing is important for the evolutionary host–parasitoid dynam- ics and outcome in the situation of a novel unsuitable host species. Moreover, our study underlines that not only the parasitoid genetics should be studied, but also the parasitoid behavior, in particular the magnitude of non-reproductive host killing and host-finding ability, in their natural environment to predict whether and how parasitoids might adapt to the trap. This together with insight in time allocation “decisions” of a particular host–parasitoid system is in fact of great value for pest control: time allocated to traveling relative to resi- dency on different patches will determine their success to suppress pest populations (Mills & Wajnberg, 2008; Wajnberg et al., 2016). Hence, whether and how parasitoids evolve influences their host- range, and thus their ability to regulate novel hosts and their value for biological control. DATA AVAILABILITY STATEMENT Please see Dryad, https://doi.org/10.5061/dryad.31zcrjdp6. Please see Dryad, https://doi.org/10.5061/dryad.31zcrjdp6. ACKNOWLEDGMENTS We would like to thank Yuanshu Pu for her help in the initial model exploration. This work is part of the research program NWO-Green with project number ALWGR.2015.6, which is partly financed by the Netherlands Organisation for Scientific Research (NWO), and partly by Koppert Biological Systems. Here we assumed that non-heritable factors influence the length of the vulnerable period in which hosts can be parasitized. However, hosts could also evolve counter resistance to parasitoid attack, which can consequently result in a co-evolutionary arms race in virulence (degree of harm) and resistance (Fellowes & Travis, 2000; Kawecki, 1998; Sasaki & Godfray, 1999). Such change in level of virulence might come about through an increase in venom produc- tion resulting in a higher incidence of (non-)reproductive host killing (Cavigliasso et al., 2019; Colinet et al., 2010; Poirié et al., 2009). An interesting next step would therefore be to incorporate parasitoid virulence and host resistance in this model to further investigate the evolutionary response of the parasitoid under different outcomes of host–parasitoid compatibilities. 4.2 \| Bad motherhood and evolution of host- preference The “jack of all trades are masters of none” assumption postulates a generalist-specialist trade-off (Futuyma & Moreno, 1988). Indeed, we found that when generalists bear significant costs, it is more profitable to specialize to the trap by adopting a specialist host-use tactic. Although this can result in a new compatible host–parasitoid relationship, when parasitoids lose their ability to use their original host but do not (yet) evolve behavioral avoidance of these original hosts, this can still result in suboptimal host choices, that is, their physiological and behavioral host-range do not match. We found that ‘bad motherhood’ can be lifted when parasitoids also exhibit ge- netic variation in host preference. Joint evolution of host-preference and performance (host use tactic) enables parasitoids to maximize their reproduction by reducing their time spend in suboptimal host patches, resulting in host choices that match their performance. This is relevant as traits that influence host localization and pref- erence in various parasitoid species are found to be influenced by genetics (e.g., Desjardins et al., 2010; Dubuffet et al., 2006; Hopper et al., 2019; Rolff & Kraaijeveld, 2001). The model showed that evo- lution of host preference promotes evolution of host-use specialists by increasing their advantage over generalists and reducing subop- timal host decisions. Likewise, resource-consumer studies showed that evolution of habitat choice qualitatively changes adaptation by promoting specialist over generalist and allowing different special- ists to coexists (Ravigné et al., 2009; Rueffler et al., 2007). Another factor to keep in mind is that both host-preference and performance are often considered to be complex behavioral and physiological traits (Vinson, 1998; Vinson & Iwantsch, 1980), po- tentially controlled by multiple (linked) loci (Desjardins et al., 2010; Hawthorne & Via, 2001; Huang et al., 2021; Werren et al., 2010). Moreover, studies show that behavior and physiological traits gen- erally have low heritability (Dochtermann et al., 2019; Kruitwagen et al., 2021; Mousseau & Roff, 1987; Stirling et al., 2002), suggesting Despite genetic variation in host-preference, we found that ‘bad motherhood’ can still persist when parasitoid efficiency to locate host patches is low. According to optimal foraging models KRUITWAGEN et al. 15 of 18 Godfray, H., & Cook, J. M. (1997). Mating systems of parasitoid wasps. In B. J. Crespi & J. C. Choe (Eds.), The evolution of mating systems in insects and arachnids (pp. 211–225). Cambridge University Press. Beukeboom, L., Niehuis, O., Pannebakker, B., Koevoets, T., Gibson, J., Shuker, D., van de Zande, L., & Gadau, J. (2010). A comparison of re- combination frequencies in intraspecific versus interspecific map- ping populations of Nasonia. Heredity, 104(3), 302–309. Hassell, M. P. (2000). Host–parasitoid population dynamics. Journal of Animal Ecology, 69(4), 543–566. Cavigliasso, F., Mathé-Hubert, H., Kremmer, L., Rebuf, C., Gatti, J.-L., Malausa, T., Colinet, D., & Poirié, M. (2019). Rapid and differential evolution of the venom composition of a parasitoid wasp depend- ing on the host strain. Toxins, 11(11), 629. https://doi.org/10.3390/ toxins11110629 Hassell, M. P., & Pacala, S. W. (1990). Heterogeneity and the dynamics of host—Parasitoid interactions. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 330(1257), 203–220. Hawthorne, D. J., & Via, S. (2001). Genetic linkage of ecological special- ization and reproductive isolation in pea aphids. Nature, 412(6850), 904–907. Colinet, D., Schmitz, A., Cazes, D., Gatti, J.-L., & Poirié, M. (2010). The or- igin of intraspecific variation of virulence in an eukaryotic immune suppressive parasite. PLoS Pathogens, 6(11), e1001206. Heimpel, G. E., & De Boer, J. G. (2008). Sex determination in the Hymenoptera. Annual Review of Entomology, 53, 209–230. Comins, H., & Hassell, M. (1979). The dynamics of optimally forag- ing predators and parasitoids. The Journal of Animal Ecology, 48, 335–351. Heimpel, G. E., Neuhauser, C., & Hoogendoorn, M. (2003). Effects of parasitoid fecundity and host resistance on indirect interactions among hosts sharing a parasitoid. Ecology Letters, 6(6), 556–566. Cronin, J. T., Abrahamson, W. G., & Craig, T. P. (2001). Temporal variation in herbivore host-plant preference and performance: Constraints on host-plant adaptation. Oikos, 93(2), 312–320. Henter, H. J. (1995). The potential for coevolution in a host-parasitoid system. II. Genetic variation within a population of wasps in the ability to parasitize an aphid host. Evolution, 49(3), 439–445. Desjardins, C. A., Perfectti, F., Bartos, J. D., Enders, L. S., & Werren, J. H. (2010). The genetic basis of interspecies host preference differences in the model parasitoid Nasonia. Heredity, 104(3), 270–277. Hofsvang, T. (1988). Mechanisms of host discrimination and intraspecific competition in the aphid parasitoid Ephedrus cerasicola. Entomologia Experimentalis et Applicata, 48(3), 233–239. Desneux, N., Barta, R. J., Hoelmer, K. A., Hopper, K. R., & Heimpel, G. E. (2009). Multifaceted determinants of host specificity in an aphid parasitoid. Oecologia, 160(2), 387–398. Hopper, K. R., Oppenheim, S. J., Kuhn, K. L., Lanier, K., Hoelmer, K. A., Heimpel, G. E., Meikle, W. G., O'Neil, R. J., Voegtlin, D. G., Wu, K., Woolley, J. B., & Heraty, J. M. (2019). Counties not countries: Variation in host specificity among populations of an aphid parasit- oid. Evolutionary Applications, 12(4), 815–829. Dieckmann, U., & Law, R. (1996). The dynamical theory of coevolu- tion: A derivation from stochastic ecological processes. Journal of Mathematical Biology, 34(5), 579–612. Huang, J., Chen, J., Fang, G., Pang, L., Zhou, S., Zhou, Y., Pan, Z., Zhang, Q., Sheng, Y., Lu, Y., Liu, Z., Zhang, Y., Li, G., Shi, M., Chen, X., & Zhan, S. (2021). Two novel venom proteins underlie divergent parasitic strategies between a generalist and a specialist parasite. Nature Communications, 12(1), 1–16. Dochtermann, N. A., Schwab, T., Anderson Berdal, M., Dalos, J., & Royauté, R. (2019). The heritability of behavior: A meta-analysis. Journal of Heredity, 110(4), 403–410. Doebeli, M., & Dieckmann, U. (2000). Evolutionary branching and sym- patric speciation caused by different types of ecological interac- tions. The American Naturalist, 156(S4), S77–S101. Huang, J., Hua, H., Wang, L., Zhang, F., & Li, Y. (2017). Number of attacks by Trichogramma dendrolimi (Hymenoptera: Trichogrammatidae) affects the successful parasitism of Ostrinia furnacalis (lepidop- tera: Crambidae) eggs. Bulletin of Entomological Research, 107(6), 812–819. Dubuffet, A., Álvarez, C. I. R., Drezen, J. M., Van Alphen, J. J., & Poirié, M. (2006). Do parasitoid preferences for different host species match virulence? Physiological Entomology, 31(2), 170–177. Kacsoh, B. Z., & Schlenke, T. A. (2012). High hemocyte load is associated with increased resistance against parasitoids in Drosophila suzukii, a relative of D. melanogaster. PLoS One, 7(4), e34721. Ellers, J., Sevenster, J. G., & Driessen, G. (2000). Egg load evolution in parasitoids. The American Naturalist, 156(6), 650–665. Fatouros, N., Cusumano, A., Bin, F., Polaszek, A., & Van Lenteren, J. (2020). How to escape from insect egg parasitoids: A review of potential factors explaining parasitoid absence across the Insecta. Proceedings of the Royal Society. B, Biological sciences, 287(1931), 20200344. Kaser, J. M., Nielsen, A. L., & Abram, P. (2018). Biological control effects of non-reproductive host mortality caused by insect parasitoids. Ecological Applications, 28(4), 1081–1092. Kawecki, T. J. (1998). AUTHOR CONTRIBUTIONS Astrid Kruitwagen: Conceptualization (lead); formal analysis (lead); methodology (equal); writing – original draft (lead); writing – review 16 of 18 KRUITWAGEN et al. Liu, W.-X., Wang, W.-X., Zhang, Y.-B., Wang, W., Lu, S.-L., & Wan, F.-H. (2015). Adult diet affects the life history and host-killing behavior of a host-feeding parasitoid. Biological Control, 81, 58–64. Rolff, J., & Kraaijeveld, A. (2001). Host preference and survival in se- lected lines of a drosophila parasitoid, Asobara tabida. Journal of Evolutionary Biology, 14(5), 742–745. Lue, C.-H., Borowy, D., Buffington, M. L., & Leips, J. (2018). Geographic and seasonal variation in species diversity and community compo- sition of frugivorous drosophila (Diptera: Drosophilidae) and their Leptopilina (Hymenoptera: Figitidae) parasitoids. Environmental Entomology, 47(5), 1096–1106. Romero, V., Zaviezo, T., & Grez, A. A. (2020). The invasive coccinellid Harmonia axyridis (coleoptera: Coccinellidae) is a less suitable host for parasitism than resident species. Ciencia e Investigación Agraria: Revista Latinoamericana de Ciencias de la Agricultura, 47(3), 312–323. Rueffler, C., Van Dooren, T. J., & Metz, J. A. (2007). The interplay be- tween behavior and morphology in the evolutionary dynamics of resource specialization. The American Naturalist, 169(2), E34–E52. Marais, G., & Charlesworth, B. (2003). Genome evolution: Recombination speeds up adaptive evolution. Current Biology, 13(2), R68–R70. Sasaki, A., & Godfray, H. (1999). A model for the coevolution of resistance and virulence in coupled host–parasitoid interactions. Proceedings of the Royal Society of London. Series B: Biological Sciences, 266(1418), 455–463. Martini, A., Di Vitantonio, C., & Dindo, M. L. (2019). Acceptance and suit- ability of the box tree moth Cydalima perspectalis as host for the tachinid parasitoid Exorista larvarum. Bulletin of Insectology, 72(1), 150–160. Schlaepfer, M. A., Runge, M. C., & Sherman, P. W. (2002). Ecological and evolutionary traps. Trends in Ecology & Evolution, 17(10), 474–480. McNamara, J. M., Houston, A. I., & Weisser, W. W. (1993). Combining prey choice and patch use—What does rate-maximizing predict? Journal of Theoretical Biology, 164(2), 219–238. Schlaepfer, M. A., Sherman, P. W., Blossey, B., & Runge, M. C. (2005). Introduced species as evolutionary traps. Ecology Letters, 8(3), 241–246. Metz, J. A., Nisbet, R. M., & Geritz, S. A. (1992). How should we de- fine ‘fitness’ for general ecological scenarios? Trends in Ecology & Evolution, 7(6), 198–202. Schlenke, T. A., Morales, J., Govind, S., & Clark, A. G. (2007). Contrasting infection strategies in generalist and specialist wasp parasitoids of Drosophila melanogaster. PLoS Pathogens, 3(10), e158. Mills, N. J., & Wajnberg, E. (2008). Optimal foraging behavior and effi- cient biological control methods. In E. Wajnberg, C. Bernstein, & J. van Alphen (Eds.), Behavioural ecology of insect parasitoids–from theoretical approaches to field applications (pp. 3–30). Blackwell Publishing. Stirling, D., Réale, D., & Roff, D. (2002). Selection, structure and the heri- tability of behaviour. Journal of Evolutionary Biology, 15(2), 277–289. Straub, C. S., Ives, A. R., & Gratton, C. (2011). Evidence for a trade-off between host-range breadth and host-use efficiency in aphid para- sitoids. The American Naturalist, 177(3), 389–395. Minkenberg, O. P., Tatar, M., & Rosenheim, J. A. (1992). Egg load as a major source of variability in insect foraging and oviposition behav- ior. Oikos, 65, 134–142. Strauss, S. Y., Lau, J. A., & Carroll, S. P. (2006). Evolutionary responses of natives to introduced species: What do introductions tell us about natural communities? Ecology Letters, 9(3), 357–374. Mooney, H. A., & Cleland, E. E. (2001). The evolutionary impact of inva- sive species. Proceedings of the National Academy of Sciences of the United States of America, 98(10), 5446–5451. Thompson, J. N. (1988). Evolutionary ecology of the relationship be- tween oviposition preference and performance of offspring in phy- tophagous insects. Entomologia Experimentalis et Applicata, 47(1), 3–14. Mousseau, T. A., & Roff, D. A. (1987). Natural selection and the heritabil- ity of fitness components. Heredity, 59(2), 181–197. Münster-Swendsen, M. (2002). Population cycles of the spruce needle- miner in Denmark driven by interactions with insect parasitoids. In Population cycles: The case for trophic interactions (pp. 29–43). Oxford University Press. Tuda, M., & Bonsall, M. B. (1999). Evolutionary and population dynamics of host–parasitoid interactions. Population Ecology, 41(1), 81–91. Ueno, T., & Tanaka, T. (1994). Can a female parasitoid recognize a previ- ously rejected host? Animal Behaviour, 47, 988–990. Niehuis, O., Gibson, J. D., Rosenberg, M. S., Pannebakker, B. A., Koevoets, T., Judson, A. K., Desjardins, C. A., Kennedy, K., Duggan, D., Beukeboom, L. W., van de Zande, L., Shuker, D. M., Werren, J. H., & Gadau, J. (2010). Recombination and its impact on the genome of the haplodiploid parasitoid wasp Nasonia. PLoS One, 5(1), e8597. Van Alphen, J. J., & Visser, M. E. (1990). Superparasitism as an adaptive strategy for insect parasitoids. Annual Review of Entomology, 35(1), 59–79. Varaldi, J., Fouillet, P., Boulétreau, M., & Fleury, F. (2005). Superparasitism acceptance and patch-leaving mechanisms in parasitoids: A com- parison between two sympatric wasps. Animal Behaviour, 69(6), 1227–1234. Poirié, M., Carton, Y., & Dubuffet, A. (2009). Virulence strategies in par- asitoid Hymenoptera as an example of adaptive diversity. Red queen meets Santa Rosalia: Arms races and the evolution of host specialization in organisms with parasitic life- styles. The American Naturalist, 152(4), 635–651. Fellowes, M., & Travis, J. M. (2000). Linking the coevolutionary and population dynamics of host–parasitoid interactions. Population Ecology, 42(2), 195–203. Keeler, M. S., & Chew, F. S. (2008). Escaping an evolutionary trap: Preference and performance of a native insect on an exotic invasive host. Oecologia, 156(3), 559–568. Felsenstein, J. (1974). The evolutionary advantage of recombination. Genetics, 78(2), 737–756. Kokko, H., & Sutherland, W. J. (2001). Ecological traps in changing en- vironments: Ecological and evolutionary consequences of a be- haviourally mediated Allee effect. Evolutionary Ecology Research, 3(5), 603–610. Felsenstein, J. (1981). Skepticism towards Santa Rosalia, or why are there so few kinds of animals? Evolution, 35, 124–138. Firlej, A., Girard, P.-A., Brehélin, M., Coderre, D., & Boivin, G. (2012). Immune response of Harmonia axyridis (coleoptera: Coccinellidae) supports the enemy release hypothesis in North America. Annals of the Entomological Society of America, 105(2), 328–338. König, K., Krimmer, E., Brose, S., Gantert, C., Buschlüter, I., König, C., Klopfstein, S., Wendt, I., Baur, H., Krogmann, L., & Steidle, J. L. (2015). Does early learning drive ecological divergence during speciation processes in parasitoid wasps? Proceedings of the Royal Society B: Biological Sciences, 282(1799), 20141850. Fletcher, J. P., Hughes, J. P., & Harvey, I. F. (1994). Life expectancy and egg load affect oviposition decisions of a solitary parasit- oid. Proceedings of the Royal Society of London. Series B: Biological Sciences, 258(1352), 163–167. Konopka, J. K., Poinapen, D., Gariepy, T., & McNeil, J. N. (2018). Understanding the mismatch between behaviour and develop- ment in a novel host-parasitoid association. Scientific Reports, 8(1), 1–11. Futuyma, D. J., & Moreno, G. (1988). The evolution of ecological special- ization. Annual Review of Ecology and Systematics, 19(1), 207–233. Geritz, S. A., Metz, J. A., Kisdi, É., & Meszéna, G. (1997). Dynamics of adaptation and evolutionary branching. Physical Review Letters, 78(10), 2024. Kruitwagen, A., Wertheim, B., & Beukeboom, L. (2021). Artificial se- lection for non-reproductive host killing in a native parasitoid on the invasive pest, Drosophila suzukii. Evolutionary Applications, 14, 1993–2011. Gillespie, D. T. (2007). Stochastic simulation of chemical kinetics. Annual Review of Physical Chemistry, 58, 35–55. KRUITWAGEN et al. 17 of 18 evolutionary insights from the genomes of three parasitoid Nasonia species. Science, 327(5963), 343–348. Comptes Rendus Biologies, 332(2–3), 311–320. Vinson, S. B. (1990). How parasitoids deal with the immune system of their host: An overview. Archives of Insect Biochemistry and Physiology, 13(1–2), 3–27. R Core Team. (2020). R: A language and environment for statistical comput- ing. R Foundation for Statistical Computing. Vinson, S. B. (1998). The general host selection behavior of parasitoid Hymenoptera and a comparison of initial strategies utilized by lar- vaphagous and oophagous species. Biological Control, 11(2), 79–96. Ravigné, V., Dieckmann, U., & Olivieri, I. (2009). Live where you thrive: Joint evolution of habitat choice and local adaptation facilitates specialization and promotes diversity. The American Naturalist, 174(4), E141–E169. Vinson, S. B., & Iwantsch, G. (1980). Host suitability for insect parasit- oids. Annual Review of Entomology, 25(1), 397–419. Raymond, L., Plantegenest, M., Gagic, V., Navasse, Y., & Lavandero, B. (2016). Aphid parasitoid generalism: Development, assessment, and implications for biocontrol. Journal of Pest Science, 89(1), 7–20. Visser, M. E., Luyckx, B., Nell, H., & Boskamp, G. J. (1992). Adaptive su- perparasitism in solitary parasitoids: Marking of parasitized hosts in relation to the pay-off from superparasitism. Ecological Entomology, 17(1), 76–82. Reed, D. A., Luhring, K. A., Stafford, C. A., Hansen, A. K., Millar, J. G., Hanks, L. M., & Paine, T. D. (2007). Host defensive response against an egg parasitoid involves cellular encapsulation and melanization. Biological Control, 41(2), 214–222. Wajnberg, E. (2006). Time allocation strategies in insect parasitoids: From ultimate predictions to proximate behavioral mechanisms. Behavioral Ecology and Sociobiology, 60(5), 589–611. Ricker, W. E. (1954). Stock and recruitment. Journal of the Fisheries Board of Canada, 11(5), 559–623. Wajnberg, E., Roitberg, B. D., & Boivin, G. (2016). Using optimality mod- els to improve the efficacy of parasitoids in biological control pro- grammes. Entomologia Experimentalis et Applicata, 158(1), 2–16. Rizki, R., & Rizki, T. (1990). Parasitoid virus-like particles destroy Drosophila cellular immunity. Proceedings of the National Academy of Sciences of the United States of America, 87(21), 8388–8392. Werren, J. H., Richards, S., Desjardins, C. A., Niehuis, O., Gadau, J., Colbourne, J. K., & Group, N. G. W. (2010). Functional and 18 of 18 KRUITWAGEN et al. SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article. Yang, L., Wan, B., Wang, B.-B., Liu, M.-M., Fang, Q., Song, Q.-S., & Ye, G.-Y. (2019). The pupal ectoparasitoid Pachycrepoideus vindemmiae regulates cellular and humoral immunity of host Drosophila melan- ogaster. Frontiers in Physiology, 10, 1282. https://doi.org/10.3389/ fphys.2019.01282 How to cite this article: Kruitwagen, A., Beukeboom, L. W., Wertheim, B., & van Doorn, G. S. (2022). Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap. Ecology and Evolution, 12, e9030. https://doi.org/10.1002/ece3.9030 How to cite this article: Kruitwagen, A., Beukeboom, L. W., Wertheim, B., & van Doorn, G. S. (2022). Evolution of parasitoid host preference and performance in response to an invasive host acting as evolutionary trap. Ecology and Evolution, 12, e9030. https://doi.org/10.1002/ece3.9030 Yoon, S., & Read, Q. (2016). Consequences of exotic host use: Impacts on lepidoptera and a test of the ecological trap hypothesis. Oecologia, 181(4), 985–996. Zhou, J., Meng, L., & Li, B. (2019). Non-reproductive effects of two par- asitoid species on the oriental armyworm Mythimna separata on wheat and maize plants. BioControl, 64(2), 115–124.
https://openalex.org/W4247050112	https://revista.uemg.br/index.php/educacaoemfoco/article/download/3635/2048	Portuguese	null	Editorial	Educação em Foco	2,019	cc-by	417	Amanda Tolomelli Brescia Amanda Tolomelli Brescia Educação em Foco, ano 22, n. 36 - jan./abr. 2019 - p. 5 Editorial Trazemos à leitura o primeiro número de 2019. Começamos este ano com muitos desafios a serem enfrentados no campo das políticas públicas, em especial para a área de Educação. Desafios não faltam também quando tratamos de editoração de periódicos científicos, divulgação científica e o fazer científico dentro das universidades, neste cenário, a revista Educação em Foco apresenta algumas mudanças em seu modelo político/administrativo para este ano! Nosso periódico até o início deste ano era composto por uma editora-chefe, a professora Lana Mara Siman, uma editora gerente, a professora Amanda Tolomelli e uma equipe de editores que auxiliavam pontualmente nas questões pertinentes ao fazer editorial desta revista. A partir deste ano nosso formato foi ampliado, se pudermos chamar assim, e duas equipes (que trabalham constantemente juntas) foram formadas: uma equipe de editores-chefes, composta pelas professoras Amanda Tolomelli Brescia e Vera Lúcia Nogueira e pelo professor José Eustáquio de Brito; e por uma equipe de editores-adjuntos, que estão trabalhando semanalmente na leitura dos pareceres (feitos com tanto esmero pela nossa equipe de pareceristas) e tomada de decisões editoriais pontuais que estão chegando aos nossos autores. A composição desta equipe contempla diferentes áreas de conhecimento e formação inicial, sendo Cristina Alves Menezes Rocha, Cristiane França, Evandro Alexandre da Silva Costa, Fernando Zanetti e Luciano Andrade Ribeiro, todos docentes da Universidade do Estado de Minas Gerais. Esta mudança ocorre na tentativa de proporcionar maior agilidade e qualidade em todos os processos editoriais, passando desde o processo inicial de detecção de plágio dos artigos e contemplação das normas de submissão até à decisão final de aceitação ou rejeição do trabalho, sendo possível que em breve todo esse processo de tramitação do trabalho, da recepção até a publicação, não ultrapasse o período de 3 (três) meses. Outro aspecto que estamos desenvolvendo no periódico e que visa à qualidade das pesquisas publicadas e sua amplitude maior de horizonte é a internacionalização. Diversos trabalhos têm sido recebidos de países estrangeiros de língua portuguesa e um trabalho de divulgação deste periódico junto a programas de pós-graduação destes países tem sido realizado. Em breve também teremos novidades! Desejamos que os ensinamentos que aqui estão, nestes trabalhos, tragam elucidações para sua prática pedagógica, coragem e esperança de uma educação sempre de muita qualidade! Abraços! Abraços! Amanda Tolomelli Brescia Abril/2019. Amanda Tolomelli Brescia Abril/2019. Educação em Foco, ano 22, n. 36 - jan./abr. 2019 - p. 5 5 5
https://openalex.org/W2082587498	https://hal-amu.archives-ouvertes.fr/hal-01215957/file/ptpmcrender.pdf	English	null	Photographic but not line-drawn faces show early perceptual neural sensitivity to eye gaze direction	Frontiers in human neuroscience	2,015	cc-by	16,558	To cite this version: Alejandra Rossi, Francisco J. Parada, Marianne Latinus, Aina Puce. Photographic but not line-drawn faces show early perceptual neural sensitivity to eye gaze direction. Frontiers in Human Neuroscience, 2015, 9 (185), 10.3389/fnhum.2015.00185. hal-01215957 Photographic but not line-drawn faces show early perceptual neural sensitivity to eye gaze direction Alejandra Rossi, Francisco J. Parada, Marianne Latinus, Aina Puce HAL Id: hal-01215957 https://amu.hal.science/hal-01215957v1 Submitted on 15 Oct 2015 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. ORIGINAL RESEARCH published: 10 April 2015 doi: 10.3389/fnhum.2015.00185 Photographic but not line-drawn faces show early perceptual neural sensitivity to eye gaze direction Alejandra Rossi 1,2†, Francisco J. Parada 2,3,4†, Marianne Latinus 3,5 and Aina Puce 1,2,3 1 Cognitive Science Program, Indiana University, Bloomington, IN, USA, 2 Program in Neuroscience, Indiana University, Bloomington, IN, USA, 3 Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN, USA, 4 Department of Psychiatry, Harvard Medical School, Boston, MA, USA, 5 Institut de Neurosciences de la Timone, UMR7289, CNRS, Aix-Marseille Université, Marseille, France Our brains readily decode facial movements and changes in social attention, reﬂected in earlier and larger N170 event-related potentials (ERPs) to viewing gaze aversions vs. direct gaze in real faces (Puce et al., 2000). In contrast, gaze aversions in line- drawn faces do not produce these N170 differences (Rossi et al., 2014), suggesting that physical stimulus properties or experimental context may drive these effects. Here we investigated the role of stimulus-induced context on neurophysiological responses to dynamic gaze. Sixteen healthy adults viewed line-drawn and real faces, with dynamic eye aversion and direct gaze transitions, and control stimuli (scrambled arrays and checkerboards) while continuous electroencephalographic (EEG) activity was recorded. EEG data from 2 temporo-occipital clusters of 9 electrodes in each hemisphere where N170 activity is known to be maximal were selected for analysis. N170 peak amplitude and latency, and temporal dynamics from Event-Related Spectral Perturbations (ERSPs) were measured in 16 healthy subjects. Real faces generated larger N170s for averted vs. direct gaze motion, however, N170s to real and direct gaze were as large as those to respective controls. N170 amplitude did not differ across line-drawn gaze changes. Overall, bilateral mean gamma power changes for faces relative to control stimuli occurred between 150–350 ms, potentially reﬂecting signal detection of facial motion. Our data indicate that experimental context does not drive N170 differences to viewed gaze changes. Low-level stimulus properties, such as the high sclera/iris contrast change in real eyes likely drive the N170 changes to viewed aversive movements. Edited by: Davide Rivolta, University of East London, UK Reviewed by: Christine Parsons, University of Oxford, UK John Towler, Birkbeck, University of London, UK Correspondence: Alejandra Rossi, Cognitive Science Program, Indiana University, 1900 East Tenth St., Bloomington, IN 47405, USA alejandrarossic@gmail.com †These authors have contributed equally to this work. Edited by: Davide Rivolta, University of East London, UK Reviewed by: Christine Parsons, University of Oxford, UK John Towler, Birkbeck, University of London, UK *Correspondence: Alejandra Rossi, Cognitive Science Program, Indiana University, 1900 East Tenth St., Bloomington, IN 47405, USA alejandrarossic@gmail.com Keywords: N170 ERP, real faces, line-drawn faces, gaze aversion, apparent motion †These authors have contributed equally to this work. †These authors have contributed equally to this work. Keywords: N170 ERP, real faces, line-drawn faces, gaze aversion, apparent motion Introduction Received: 29 April 2014 Accepted: 19 March 2015 Published: 10 April 2015 Successful social behavior requires evaluating incoming sensory information and merging it with situationally relevant behavioral responses. Though a part of our social life may rely on purely reflexive behaviors, specialized neural activity is needed in evaluating social cues (Stanley and Adolphs, 2013). Over the past two decades social neuroscience, the study of social and cognitive influences on biological processes (Cacioppo, 1994; Cacioppo et al., 2000; Ochsner and Lieberman, 2001), has aimed to generate a brain-based understanding of social behaviors. An individual’s social cognitive understanding of the world is likely to not be context-invariant, however, the effects of task and experimental context on social cognition are seldom studied. In the case of social attention, in daily life a gaze change will occur in the context of Citation: Indeed, N170 is larger to ambiguous face-like stimuli that are perceived as faces relative to the same stimuli when they are not seen as faces (George et al., 1996; Sagiv and Bentin, 2001; Bentin and Golland, 2002; Latinus and Taylor, 2005, 2006). These effects have been proposed to be driven by stimulus context by a number of investigators (Bentin and Golland, 2002; Latinus and Taylor, 2006). N170 modulation to dynamic facial movements is not exclusive to eyes: larger N170s occur to mouth opening vs. closing movements—potentially reflecting a response to a pending utterance (Puce et al., 2000), and this effect occurs in both real and line-drawn faces (Puce et al., 2003; Rossi et al., 2014). Unlike in dynamic mouth motion, N170s to gaze aversions are strongly modulated by stimulus type: real faces show N170 differences to averted vs. direct gaze (Puce et al., 2000), whereas line-drawn faces do not (Rossi et al., 2014). These differences beg the question about effects of stimulus-driven context effects on the N170 elicited to dynamic facial movements. Hence, here we recorded N170 ERPs to dynamic gaze transitions to both real and line-drawn dynamic face images, and scrambled controls within the same experiment (using an experimental structure similar to that of Puce et al., 2003). We performed a standard ERP peak analysis, focusing on N170, and reasoned that if stimulus-context effects were driving N170 modulation, we would expect to observe larger N170s to gaze aversion vs. direct gaze for both real and line-drawn faces. In contrast, if the N170 effect was driven by low-level stimulus features only in the eye stimuli e.g., high iris/sclera contrast in the real images of faces, then the N170 effect would be seen only to dynamic images using real faces and not to line-drawn face images. This N170 modulation would not be predicted to occur for real control stimuli, in line with our previous studies. Finally, if the N170 effect was driven by a general low-level effect of local stimulus contrast change (occurring in both face and control stimuli), then we might expect to observe larger N170s to the real faces and their respective controls, relative to the line-drawn stimuli. As well as examining averaged ERP activity, we also investigated oscillatory electroencephalographic (EEG) behavior post-motion onset to all stimulus types at electrode sited generating maximal N170 activity in a frequency range of 5–50 Hz. Citation: In a laboratory setting, experimental context can be created within a trial, across trials or conditions, or across experimental sessions. Context effects could potentially be driven by the characteristics of the stimuli (bottom-up), or by task demands/instructions to subjects (top-down). One particularly striking experimental context effect has been reported to viewing faces. It has long been known that the N170 event-related potential (ERP) is strongly driven by the physical or structural characteristics of a face stimulus (Bentin et al., 1996). In an elegant experimental manipulation, Bentin and Golland (2002) recorded an N170 ERP evoked to (static) schematic line-drawings of faces, scrambled versions of the same faces, and line-drawings of common objects. In their design different subject groups were exposed to the stimuli with different block orders. The scrambled, or jumbled, versions of the line-drawn face stimuli had recognizable features, whose position relative to the outline of the face was altered. As expected, N170s were elicited to all stimulus categories, and were significantly larger to the intact schematic faces in both experiments. Critical to the current discussion, significantly larger N170s occurred to jumbled schematic faces but only when that stimulus block directly followed the schematic face block (Bentin and Golland, 2002), indicating how important stimulus-induced context effects can be in a laboratory setting. In a different study, N170 amplitude elicited to Moonee faces decreased by priming with photographic images of the same individuals represented in the Moonee faces (Jemel et al., 2003). The strongest priming effect occurred to images that were the actual photographic image of the Moonee face stimulus (a bottom-up effect), however, priming was also observed to different real images of the same individual relative to the Moonee faces (top-down effect) (Jemel et al., 2003). As a third example of the importance of experimental context effects, differences in the lateralization of N170 to faces can occur as a function of stimulus conditions used in the experiment. For example, the classic right lateralization of N170 is seen when faces are randomly presented among other object classes (e.g., Bötzel et al., 1995; Bentin et al., 1996; Eimer, 1998; Itier and Taylor, 2004) compared to a bilateral or even left-lateralization pattern when faces are presented in series with other faces (Deffke et al., 2007). These findings caution how important experimental context can be on N170s elicited to faces (Maurer et al., 2008). Citation: Rossi A, Parada FJ, Latinus M and Puce A (2015) Photographic but not line-drawn faces show early perceptual neural sensitivity to eye gaze direction. March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 1 Rossi et al. ERPs to changing gaze directed emotions and actions from not only one’s self, but from others around us. This social environment, with many multisensory cues and continually changing context, is difficult to reproduce in a controlled laboratory setting. However, even in a controlled laboratory setting, experimental context can potentially modulate neural responses to particular stimulus conditions or tasks, and may underlie some of the differences observed between studies in the literature. In a laboratory setting, experimental context can be created within a trial, across trials or conditions, or across experimental sessions. Context effects could potentially be driven by the characteristics of the stimuli (bottom-up), or by task demands/instructions to subjects (top-down). Isolated eyes evoke larger and delayed N170s relative to full faces (Bentin et al., 1996; Jemel et al., 1999; Puce and Perrett, 2003). Hence, the context of the face itself (e.g., outline and other face parts) may affect the neural response elicited to the eye stimulus—an effect that does not occur to presenting other face parts in isolation. Due to its sensitivity to dynamic gaze transitions (Puce et al., 2000; Conty et al., 2007), N170 has been posited to be a neural marker of communicative intent (Puce, 2013). Relevant for the present study, N170s to dynamic gaze aversions are larger and earlier than those to gaze transitions looking directly at the observer (Puce et al., 2000; Watanabe et al., 2002; but see Conty et al., 2007). This effect occurs to full images of faces, and isolated eyes (Puce et al., 2000), suggesting that N170 signals changes in social attention, and reflects the potential salience of gaze direction (Puce and Perrett, 2003; Conty et al., 2007). directed emotions and actions from not only one’s self, but from others around us. This social environment, with many multisensory cues and continually changing context, is difficult to reproduce in a controlled laboratory setting. However, even in a controlled laboratory setting, experimental context can potentially modulate neural responses to particular stimulus conditions or tasks, and may underlie some of the differences observed between studies in the literature. Stimulus Creation Real faces had been originally created from still 8-bit color photographs of posed direct and extreme averted (30 degree) gaze positions in both left and right directions. The stimulus face was superimposed on a background of concentric grayscale circles of different luminance. The images were originally created to be presented sequentially to depict dynamic gaze transitions, and mouth motion (see Puce et al., 1998). The corresponding non-facial motion controls consisted of a colored checkerboard pattern that was constructed from hues taken from the original head. Separate corresponding control stimuli were created to that sequential presentation resulted in checks alternating their position in the same regions of the visual field as the eyes in the real face (Puce et al., 1998), and to ensure that subjects did not visualize a ‘‘face’’ in the dynamic control stimulus. Ideally, ERSP and ERP analyses performed in parallel could more completely characterize neural activity to different task demands and conditions. However, the relationship between oscillatory EEG activity and ERP activity complex (see Rossi et al., 2014) and is not typically studied. Our previous comparisons of ERP and ERSP activity to viewed dynamic eye and mouth movements in dynamic line-drawn faces showed statistically significant differences for apparent motion in the beta and gamma bands between facial motion conditions, which differed timing and frequency content relative to control motion stimuli (Rossi et al., 2014). Given our previous study, here, we expected to observe oscillatory EEG changes in beta and gamma bands that would occur at different post-motion onset times for facial and control motion stimuli. White line-drawn faces on a black background had been originally created from a multimarker recording of facial expressions using specialized biological motion creation software from which lines were generated between some of the point lights [Elite Motion Analysis System (BTS, Milan, Italy)]. The black and white control stimuli had originally been created by extracting line segments from the line-drawn face and spatially re-arranging them in the visual space in an earlier version of Photoshop (Adobe Systems, Inc.), so that the face was no longer recognizable (Puce et al., 2003). The existing line-drawn faces were modified for Rossi et al. (2014) in Photoshop CS5 (Adobe Systems, Inc.) by adding a schematic iris to the face which when spatially displaced could signal a gaze change on the stimulus face. Stimulus Creation A direct gaze consisted of a diamond-shaped schematic pupil positioned in the center of each schematic eye. Averted gaze consisted of an arrow-shaped schematic pupil that was moved to the extremity of the schematic eye (Figure 1). Thus, by toggling the two schematic eye conditions, observers reported seeing a convincing ‘‘direct’’ vs. ‘‘averted’’ gaze transition in the line-drawn face. Similarly, line-drawn control stimuli were created, using a rearranged ‘‘scramble’’ of the lines making Citation: Oscillatory EEG activity expressed both as a function of EEG frequency and time relative to stimulus onset and/or execution of motor response can be identified using time-frequency decomposition of EEG signals, and displayed as Event-Related Spectral Perturbation (ERSP) plots (Makeig et al., 2004). Changes in a given EEG frequency band can occur from more than one process or underlying mechanism (e.g., see Sedley and Cunningham, 2013). Modulation of alpha band (8–12 Hz) power has been linked to changes in attentional state (Worden et al., 2000; Sauseng et al., 2005; Thut et al., 2006; Fries et al., 2008), and performance on visual perception tasks (Ergenoglu et al., 2004; Babiloni et al., 2006; Thut et al., 2006). Alpha may act as an inhibitory brain signal (Klimesch, 2012), which might enable timing of processing, and gated access to knowledge, and orientation in time, place, and context (Basar et al., 1997; Palva and Palva, 2007; Klimesch, 2012). Increases in beta band power (12–30 Hz) may reflect maintenance of current behaviorally relevant sensorimotor or cognitive states (Engel and Fries, 2010), whereas gamma band power (>30 Hz) increases may facilitate cortical processing, cognitive control and perceptual awareness (Ray and Cole, 1985; Tallon-Baudry and Bertrand, 1999; Grossmann et al., 2007; Engel and Fries, 2010; but see Sedley and Cunningham, 2013). Stimuli Participants viewed four different types of visual displays alternating between natural images of facial motion and respective motion controls, as well as motion of line-drawn faces and their respective motion controls. The real face, with eyes averting and looking directly at the observer, had a respective motion control that consisted of a colored checkerboard with checks moving towards the left and right in the same visuospatial position as the eyes in the real face (similar to that used in Puce et al., 1998). A line-drawn face, with eyes averting and looking at the observer, had a respective motion control in which line segments in the scrambled stimulus moved with a similar spatial excursion to the eyes, with the same number of pixels contributing to the motion (Figure 1). Citation: Previous studies evaluating facial motion effects have focused exclusively on averaged ERPs, which represent linearly summed EEG trials that are phase-locked to a relevant event (e.g., motion onset), and that are independent of ongoing EEG activity (Jervis et al., 1983). It has been proposed that the transient phase-resetting of ongoing oscillatory EEG activity underlies ERP generation (Brandt and Jansen, 1991). However, March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 2 Rossi et al. ERPs to changing gaze were right-handed, as assessed by the Edinburgh Handedness Inventory (mean: R64.6, SD: 19) (Oldfield, 1971). were right-handed, as assessed by the Edinburgh Handedness Inventory (mean: R64.6, SD: 19) (Oldfield, 1971). oscillatory EEG activity that is not phase-locked can also occur to a stimulus event, and will not be seen in an averaged ERP (Makeig et al., 2004). Oscillatory EEG activity expressed both as a function of EEG frequency and time relative to stimulus onset and/or execution of motor response can be identified using time-frequency decomposition of EEG signals, and displayed as Event-Related Spectral Perturbation (ERSP) plots (Makeig et al., 2004). Changes in a given EEG frequency band can occur from more than one process or underlying mechanism (e.g., see Sedley and Cunningham, 2013). Modulation of alpha band (8–12 Hz) power has been linked to changes in attentional state (Worden et al., 2000; Sauseng et al., 2005; Thut et al., 2006; Fries et al., 2008), and performance on visual perception tasks (Ergenoglu et al., 2004; Babiloni et al., 2006; Thut et al., 2006). Alpha may act as an inhibitory brain signal (Klimesch, 2012), which might enable timing of processing, and gated access to knowledge, and orientation in time, place, and context (Basar et al., 1997; Palva and Palva, 2007; Klimesch, 2012). Increases in beta band power (12–30 Hz) may reflect maintenance of current behaviorally relevant sensorimotor or cognitive states (Engel and Fries, 2010), whereas gamma band power (>30 Hz) increases may facilitate cortical processing, cognitive control and perceptual awareness (Ray and Cole, 1985; Tallon-Baudry and Bertrand, 1999; Grossmann et al., 2007; Engel and Fries, 2010; but see Sedley and Cunningham, 2013). oscillatory EEG activity that is not phase-locked can also occur to a stimulus event, and will not be seen in an averaged ERP (Makeig et al., 2004). Frontiers in Human Neuroscience \| www.frontiersin.org Participants h l Seventeen healthy participants provided written informed consent to participate in the study. All participants had normal or corrected-to-normal vision, and were free of a history of neuropsychiatric disorders. The study protocol was approved by the Institutional Review Board at Indiana University, Bloomington (IRB 1202007935). High-density (256 channel) EEG and behavioral data were collected from all participants, and data from 1 individual had to be excluded from further analysis due to a large amount of artifactual EEG contamination from facial/neck muscle activity, as well as line noise. Hence, data from 16 participants (7 males, 9 females) with an average age of 26 years (range 21–34 years) were submitted for analysis. The 16 participants March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 3 Rossi et al. ERPs to changing gaze FIGURE 1 \| Materials and methods. (A) Illustration of the visual stimulation sequence, with example stimuli. Each viewing condition had a 20 s duration and consisted of the alternation of the three images with different gaze direction: direct, away to the left and away to the right. The REAL CNTL condition consists of apparent motion of a checkerboard pattern to simulate eyes moving toward or away from the participant depicted in the REAL FACE condition. A similar correspondence exists (highlighted on red circles for display purposes only) for the LINE CNTL and LINE FACE conditions, where line-drawn images are alternated to display apparent motion in a control and face stimulus. (B) Timeline for an experimental trial. An example of a trial sequence in the REAL FACE stimulus block. Participants pressed a mouse button whenever they saw a target (a negative image of the stimulus). FIGURE 1 \| Materials and methods. (A) Illustration of the visual stimulation sequence, with example stimuli. Each viewing condition had a 20 s duration and consisted of the alternation of the three images with different gaze direction: direct, away to the left and away to the right. The REAL CNTL condition consists of apparent motion of a checkerboard pattern to simulate eyes moving toward or away from the participant depicted in the REAL FACE condition. A similar correspondence exists (highlighted on red circles for display purposes only) for the LINE CNTL and LINE FACE conditions, where line-drawn images are alternated to display apparent motion in a control and face stimulus. (B) Timeline for an experimental trial. Participants h l An example of a trial sequence in the REAL FACE stimulus block. Participants pressed a mouse button whenever they saw a target (a negative image of the stimulus). All stimulus types were always presented in each experimental run, with a run consisting of the repeated presentation of the following alternating 20 s stimulus blocks (Figure 1A; following the procedure used in Puce et al., 2003): up the eye movements on the face stimuli, ensuring that all stimuli presented would be equiluminant, and have similar motion excursions, as well as contrast and spatial frequency characteristics (Figure 1A). On debriefing post-experiment, subjects did not report seeing ‘‘eye’’ stimuli in the line-drawn control stimulus. 1. REAL FACE. Three versions of male face with eyes directly looking at the observer, eyes averted to the left, or eyes averted to the right (Figure 1B) were presented in alternation to produce apparent motion depicting change in gaze position from a direct gaze to an averted gaze position to the left or to the right [gaze aversions: eyes-away] and from an averted gaze back to a direct gaze position [eyes-direct] similar to that used in Puce et al. (1998). 1. REAL FACE. Three versions of male face with eyes directly looking at the observer, eyes averted to the left, or eyes averted to the right (Figure 1B) were presented in alternation to produce apparent motion depicting change in gaze position from a direct gaze to an averted gaze position to the left or to the right [gaze aversions: eyes-away] and from an averted gaze back to a direct gaze position [eyes-direct] similar to that used in Puce et al. (1998). For all stimulus types, the effect of smooth movement was generated and no side-switch transition was possible (e.g., eyes looking to the right followed by eyes looking to the left). Negative-contrast versions (inverse colors) of all the stimulus versions were constructed to be used as infrequently presented targets (Figure 1B). 2. REAL CONTROL. Checkerboard patterns were alternated to produce an apparent motion stimulus varying in the same part of the visual field as the eyes in the REAL FACE similar to that used in Puce et al. (1998). Frontiers in Human Neuroscience \| www.frontiersin.org Procedure Participants viewed the stimuli displayed on a 24-inch monitor (Dell Ultra Sharp U2412M, refresh rate of 60 Hz) resulting in an overall visual angle of 5 × 3 (vertical × horizontal) degrees. Participants completed four experimental runs in total; each run lasted approximately 6 min to allow participants to remain still for the EEG recording and maintain their level of alertness. After each run, participants had a self-paced break. 3. LINE FACE. Three versions of line-drawn face stimuli were alternated to change gaze position to look directly at participants [eyes-direct] or avert gaze either to the left or right [eyes-away] as for the REAL FACE condition, similar to that used in Rossi et al. (2014). March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 4 ERPs to changing gaze Rossi et al. 4. LINE CONTROL. The spatially ‘‘scrambled’’ versions of the LINE FACE were alternated to produce apparent motion in the same part of the visual field as the eyes in the LINE FACE condition, similar to that used in Rossi et al. (2014). 4. LINE CONTROL. The spatially ‘‘scrambled’’ versions of the LINE FACE were alternated to produce apparent motion in the same part of the visual field as the eyes in the LINE FACE condition, similar to that used in Rossi et al. (2014). R2010b (The Mathworks, Natick, MA). EEG data were first segmented into 1700 ms epochs: 572 ms pre-stimulus baseline and 1082 ms after apparent motion onset. EEG amplitude at each trial was normalized relative to the pre-stimulus baseline based on the event-markers, identifying each trial type. ERP data were displayed using a 200 ms pre-motion onset and 600 ms after the motion transition—see Figure 3. [A manufacturer- specified latency correction factor was applied to all behavioral data and epoched ERP data. In our case, given a sampling rate of 500 Hz, a correction of 18 ms was made, as per manufacturer guidelines]. Stimulus onset asynchrony was randomly varied between 1000 and 1500 ms on each experimental trial (i.e., between two consecutive apparent motion onsets). A total of 210 trials were acquired per stimulus type (LINE FACE, LINE CONTROL, REAL FACE, REAL CONTROL). The experiment was run using Presentation Version 14 (NeuroBehavioral Systems, 2010). Procedure Participant reaction times and accuracy were logged, and time stamps for different stimulus types (as well as button press responses for detected target stimuli) for each trial were automatically sent to the EEG system and stored in the EEG file. EEG epochs were first visually inspected to identify and exclude bad channels from each individual subject EEG dataset. The electrodes identified as bad differed between subjects; average number of ‘‘bad’’ electrodes was 22 ± 2.15 (standard error of mean) out of 256 channels. Epochs with very large artifacts (e.g., very large subject movements and channel drifts) were manually rejected prior to subjecting the EEG data subsequent artifact detection analyses. Participants were instructed to press a button indicating the presence of a target stimulus. Target stimuli were negative- contrast versions of all stimuli used in the experiment (Figure 1B). Targets were randomly assigned to each alternating block (20% of trials). Trials with targets were not included in ERP/EEG analyses. Similarly, so as to remove potential confounds created by changes in stimulus type (i.e., for the first stimulus of each block, as well as for stimuli immediately following targets), trials following a target and the first stimuli of each block were not included in ERP/EEG analyses. The purpose of the target detection task was to keep participants attentive. All participants completed a short practice run (36 trials) at the beginning of the session and were given feedback regarding detection of target stimuli. All participants completed the practice run with 100% accuracy. EEG trials from the practice run were not included in subsequent analyses. Independent Component Analysis (ICA) was used to identify and subtract components representing artifacts such as eye movements, eye blinks, carotid pulse, muscle activity and line-noise (Bell and Sejnowski, 1995; Delorme and Makeig, 2004). This allowed trials with eyeblinks to be adequately corrected, and allowed these trials to be included in the analysis. A total of 32 ICA components were generated for each participant’s EEG dataset. Eyeblinks, cardiac artifact and muscle activity were identified in isolated ICA components. Following removal of artifactual ICA components and reconstitution of the EEG signal, interpolation of bad channels was performed to regenerate a 256-channel EEG dataset. Bad channels were interpolated using a spherical interpolation: electrical activity was interpolated with respect to the surrounding nearest neighbor electrodes. EEG/ERP Data Analyses Two temporo-occipital 9 electrode clusters including equivalent 10–10 system sites P07/P9 and P08/P10 were chosen for further analyses, based on inspection of the grand averaged data from the current study and previously reported maxima in N170 amplitudes that used 4 electrode clusters for 64- and 128-channel EEG derivations, and P09 and P10 for smaller electrode arrays of 10–10 system sites (Puce et al., 2000, 2003; Carrick et al., 2007; Brefczynski-Lewis et al., 2011; Rossi et al., 2014; Figures 2, 3). Averaged data from the 9 electrodes in each hemispheric cluster were used in all subsequent ERP analyses. Similarly, single- trial EEG data recorded from the same 9 electrodes in each hemispheric cluster were used for ERSP analysis. EEG Data Acquisition and Preprocessing EEG Data Acquisition Data were re-referenced to a common average reference. ERP components such as the N170 and the vertex positive potential (VPP) amplitude have previously been shown to be very sensitive to reference location (Joyce and Rossion, 2005). The average reference has been suggested as being optimal as it captures finer hemispheric differences and shows the most symmetry between positive and negative ERP peaks for face-related stimuli (Joyce and Rossion, 2005). Only behaviorally correct EEG trials, i.e., no false alarms for targets, were included in subsequent analyses. q A Net Amps 300 high-impedance EEG amplifier and NetStation software (V4.4) were used to record EEG from a 256-electrode HydroCel Geodesic Sensor Net (Electrical Geodesics Inc.) while the participant sat in a comfortable chair and performed the task in a dimly lit, humidified room. Continuous 256- channel EEG data were recorded with respect to a vertex reference using a sampling rate of 500 Hz and bandpass filter of 0.1–200 Hz (the ground electrode was sited on the midline parietal scalp). Stimulus delivery and subject behavioral responses were time-stamped onto all EEG files. Impedances were maintained below 60 kΩ as per the manufacturer’s recommended guidelines. Impedances were tested at the beginning of the experimental session and then once more at the half-way point of the experiment, allowing any high-impedance electrode contacts to be adjusted if necessary. Frontiers in Human Neuroscience \| www.frontiersin.org EEG Data Preprocessing EEG data were first exported from EGI Net Station software as simple binary files. The same pre-processing procedure was applied to the ERP and ERSP analyses. All EEG pre-processing procedures were performed using functions from the EEGLAB toolbox (Delorme and Makeig, 2004) running under MATLAB March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 5 ERPs to changing gaze Rossi et al. FIGURE 3 \| Group data: ERPs from left (L) and right (R) occipitotemporal electrode clusters as a function of stimulus type. An epoch of activity is shown from 200 ms pre-stimulus onset to 600 ms post-stimulus. Legend: Line colors indicate corresponding stimulus type shown LINE CONTROLS in blue, LINE FACES in red, REAL CONTROLS in green and REAL FACES in gray. The vertical black bar superimposed on the ERP waveforms denotes motion onset. Vertical and horizontal calibration bars denote amplitude in microvolts and time in milliseconds, respectively. and latencies were extracted for each condition, each subject, and FIGURE 2 \| Group data: Topographic voltage maps of peak N170 activity. The N170 is distributed across the bilateral occipitotemporal scalp and appears in all conditions, including the control conditions. The topographic maps are displayed in a top-down view with nose at top and left hemisphere on the left. Color scale calibration bars depict amplitude in microvolts. Red circles on the maps depict the 9 electrodes in each hemispheric cluster that provided input for N170 statistical analyses. Small black dots depict additional sensor locations. FIGURE 3 \| Group data: ERPs from left (L) and right (R) occipitotemporal electrode clusters as a function of stimulus type. An epoch of activity is shown from 200 ms pre-stimulus onset to 600 ms post-stimulus. Legend: Line colors indicate corresponding stimulus type shown LINE CONTROLS in blue, LINE FACES in red, REAL CONTROLS in green and REAL FACES in gray. The vertical black bar superimposed on the ERP waveforms denotes motion onset. Vertical and horizontal calibration bars denote amplitude in microvolts and time in milliseconds, respectively. FIGURE 3 \| Group data: ERPs from left (L) and right (R) occipitotemporal electrode clusters as a function of stim FIGURE 3 \| Group data: ERPs from left (L) and right (R) \| p ( ) g ( ) occipitotemporal electrode clusters as a function of st An epoch of activity is shown from 200 ms pre-stimulus onset to 600 ms post-stimulus. Frontiers in Human Neuroscience \| www.frontiersin.org EEG Data Preprocessing Legend: Line colors indicate corresponding stimulus type shown LINE CONTROLS in blue, LINE FACES in red, REAL CONTROLS in green and REAL FACES in gray. The vertical black bar superimposed on the ERP waveforms denotes motion onset. Vertical and horizontal calibration bars denote amplitude in microvolts and time in milliseconds, respectively. and latencies were extracted for each condition, each subject, and each electrode cluster independently. Analysis of Event-Related Spectral Perturbations All analyses were performed using custom in-house routines written using the EEGLAB toolbox (Delorme and Makeig, 2004) running under MATLAB. Artifact-free, behaviorally correct EEG segments were convolved with a linearly increased Morlet- wavelet on a trial-by-trial basis for each condition and subject. Specifically, the length of the wavelet increased linearly from 1 to 12 cycles across the frequency range of 5–50 Hz (theta, alpha, beta, and low-gamma). The linear increment of wavelet cycles is a commonly used practice when calculating spectral components in neurophysiological data, so that temporal resolution can be comparable for lower and higher EEG frequencies (Le Van Quyen et al., 2001) (for a detailed account on spectral analyses of EEG see Herrmann et al., 2005). After the EEG signals in each trial were convolved with a Morlet wavelet, they were transformed into power, and the resulting values were then averaged across trials. We analyzed the spectral power of components in the theta (5–8 Hz), alpha (8–12 Hz), beta (12–30 Hz), and low-gamma (30–50 Hz) EEG frequency- bands as they evolved over the post-movement epoch. In order to extract even-related spectral power from raw power, a standard baseline procedure was applied in a trial by trail basis (Grandchamp and Delorme, 2011). The window used as baseline comprised data points between −200 and 0 ms pre-stimulus range. E 2 \| Group data: Topographic voltage maps of peak N170 FIGURE 2 \| Group data: Topographic voltage maps of peak N170 activity. The N170 is distributed across the bilateral occipitotemporal scalp and appears in all conditions, including the control conditions. The topographic maps are displayed in a top-down view with nose at top and left hemisphere on the left. Color scale calibration bars depict amplitude in microvolts. Red circles on the maps depict the 9 electrodes in each hemispheric cluster that provided input for N170 statistical analyses. Small black dots depict additional sensor locations. Analysis of Event-Related Potentials y A digital 40 Hz infinite impulse response (IIR) low-pass filter was applied to the artifact-free behaviorally correct EEG data. Average ERPs were generated for each of the eight conditions and for each subject (about 200 trials per condition per subject on average). The ERPs from all subjects were averaged to generate a grand-average set of ERP waveforms for each condition and EEG channel. Data from each 9 electrode temporo-occipital cluster were extracted and the average time-course for each electrode cluster was generated for subject and condition, and was subsequently used for calculating N170 amplitude and latency. In line with previous work (Puce et al., 2003), we focused on the N170 as a neural marker of the perception of facial motion. In our data, consistent with previous studies, the N170 showed a lateralized posterior scalp distribution (Figure 2). Using an automated peak detection procedure within a search time window of 150–250 ms after apparent motion onset, N170 peak amplitudes Induced activity is defined as EEG activity that is elicited to the stimulus, but may not be precisely time- or phase-locked to the stimulus transition (in this case apparent motion onset). However, each individual EEG epoch will also contain evoked March 2015 \| Volume 9 \| Article 185 6 ERPs to changing gaze Rossi et al. activity, hence a calculation of ‘‘total power’’ (i.e., sum of evoked and induced activity) in each frequency band was made for EEG epochs in our study (see Tallon-Baudry et al., 1996). activity, hence a calculation of ‘‘total power’’ (i.e., sum of evoked and induced activity) in each frequency band was made for EEG epochs in our study (see Tallon-Baudry et al., 1996). cluster mass being recorded. Finally, temporal clusters in the observed data were deemed significant if their mass exceeded the maximum cluster mass of 95% of all bootstrap replicates (corresponding to a significance level of 0.05). This method allowed correction for multiple comparisons. Thus, the cluster mass statistic identified temporal regions with significant differences while avoiding false-positives arising from multiple comparisons (Pernet et al., 2011). This approach is comparable to false-discovery rate (FDR; Benjamini and Hochberg, 1995). As we have previously noted differences between facial motion stimulus type for eye and mouth movements (Rossi et al., 2014), we performed a similar analysis and generated differential ERSP plots between pairs of conditions: LINE CONTROL Direct vs. Away, LINE FACE Direct vs. Analysis of Event-Related Potentials Away, REAL CONTROL Direct vs. Away and REAL FACE Direct vs. Away. Behavioral Data Differences in temporo-occipital N170 peak amplitude and latency were evaluated using a 4-way repeated-measures ANOVA with Hemisphere (Left, Right), Configuration (Face, Control), Stimulus Type (Real, Line) and Motion (Away, Toward) as within-subjects factors using SPSS for MAC 18.0 (SPSS Inc.). Significant main effects were identified at P values of less than 0.05 (after Greenhouse-Geisser correction). Contrasts were evaluated using the Bonferroni criterion to correct for multiple comparisons with P values of less than 0.05 identifying significant effects. Participants identified the target stimulus (image negative) with 99% accuracy by button press. Mean reaction time for target detection by stimulus type was 557 ± 94 ms (s.d.) for REAL faces, 577 ± 112 ms for REAL controls, 607 ± 113 ms for LINE faces and 560 ± 92 ms for LINE controls. A 2-way (Configuration × Stimulus Type) repeated-measures ANOVA showed a significant main effect of Stimulus Type (F(1,15) = 11.067, P < 0.001) and an interaction effect of Configuration by Stimulus (F(1,15) = 26.24, P < 0.001). Furthermore, to specifically assess the specificity of the effect to REAL FACE stimuli, we performed paired t-tests between each motion conditions for each stimulus condition. Four t-tests were performed (i.e., REAL FACE AWAY × REAL FACE DIRECT, REAL CONTROL AWAY × REAL CONTROL DIRECT, LINE FACE AWAY × LINE FACE AWAY, LINE CONTROL AWAY × LINE CONTROL DIRECT). The level of statistical significance (a priori two-tailed) was set at p < 0.05. For the significant main effect of Stimulus Type, REAL stimuli (real faces and real controls) generated faster responses relative to LINE stimuli (line-drawn faces and line-drawn scrambled controls, mean difference: 30 ± 9 ms). The significant interaction effect of Configuration by Stimulus type indicated that real faces generated a faster response compared to REAL controls (mean difference: 47 ± 11 ms), while the opposite was seen for line- drawn stimuli, line-drawn controls were identified fastest (mean difference: 47 ± 13). FACE AWAY × LINE FACE AWAY, LINE CONTROL AWAY × LINE CONTROL DIRECT). The level of statistical significance (a priori two-tailed) was set at p < 0.05. ERSP analysis T h The current behavioral task was used to help participants pay attention to the display. EEG epochs from these target trials were not included in subsequent analyses. To measure the complete temporal extent of effects over frequency, we used a bootstrap approach (N = 1000 bootstraps), identifying time-frequency data points of statistically significant differences based on data-driven 95% confidence intervals (as described and implemented in Pernet et al., 2011) from the data of the two 9 electrode clusters. Non-parametric permutation was used to estimate the distribution under the null hypothesis of no differences in oscillatory amplitude between the pair of conditions. Frontiers in Human Neuroscience \| www.frontiersin.org Peak Analysis of the N170 ERP y N170 amplitudes and latencies were extracted from each of the two temporo-occipital scalp electrode clusters for each participant and condition for subsequent statistical testing. N170 was maximal over the temporo-occipital scalp, as demonstrated by the topographic voltage maps (Figure 2) plotted at the time point at which the N170 was maximal in amplitude. N170 was elicited in all stimulus conditions (Figure 3). Due to our paired design, when a subject was selected randomly, results from all his or her conditions were included in that sample. For each condition, we averaged the data across (resampled) participants and computed differences between conditions. Thus, for each one of the observed mean differences between conditions for a given frequency at each time-point, a t-statistic was calculated. At this stage, time points were evaluated according to a threshold set if their t-statistic corresponded to a value below 0.05 according to the Student’s t-distribution. This procedure was repeated for all ERSP time-points at each frequency. Temporally contiguous threshold time points were grouped into temporal clusters. At each bootstrap iteration, the temporal cluster mass was computed as the sum of the t-statistics over consecutively significant time-points, with the maximum N170 latency and amplitude data for each condition and hemisphere are shown in Table 1. A 4-way repeated- measures ANOVA for N170 peak amplitude differences revealed a significant main effect for hemisphere (F(1,15) = 11.265, P < 0.001) and stimulus type (Real vs. Line; F(1,15) = 46.289, P < 0.001). The main effects for configuration (Face, Control) and motion (Away, Direct) were not significant. A significant interaction effect was observed between stimulus type and motion (F(1,15) = 5.143, P < 0.05). For the significant main effect of hemisphere, post hoc paired comparisons revealed that N170 amplitude was greater for the Right hemisphere relative to the Left Hemisphere (mean difference: 0.66 ± 0.26 µV). The main effect of March 2015 \| Volume 9 \| Article 185 7 ERPs to changing gaze Rossi et al. TABLE 1 \| Group N170 peak amplitude (µV) and latency (ms) data: Mean and Standard Errors (Std) as a function of hemisphere (Hem) and Condition. Peak Analysis of the N170 ERP stimulus type showed that N170 amplitude was greater for the line-drawn stimuli relative to the real stimuli (mean difference: 0.55 ± 0.08 µV, Figure 3). Post hoc comparisons for the interaction effect between stimulus type and motion revealed that among the REAL stimuli (i.e., real faces and real controls) the N170 for averted gaze was significantly larger than that to direct gaze (mean difference: 0.29 ± 0.13 µV) (Figure 3); an effect not seen for line-drawn stimuli. stimulus type showed that N170 amplitude was greater for the line-drawn stimuli relative to the real stimuli (mean difference: 0.55 ± 0.08 µV, Figure 3). Post hoc comparisons for the interaction effect between stimulus type and motion revealed that among the REAL stimuli (i.e., real faces and real controls) the N170 for averted gaze was significantly larger than that to direct gaze (mean difference: 0.29 ± 0.13 µV) (Figure 3); an effect not seen for line-drawn stimuli. One could argue that our 4-way ANOVA would reveal a 3-way interaction between stimulus type, configuration, and motion. This was not the case. When we compare our current data to those of Puce et al. (2000); we note that the authors also did not find interaction effects on N170 amplitude as assessed by means of 3-way ANOVAs performed at isolated hemispherically homologous electrode sites in a study that was performed using only 22 EEG electrodes. To try and investigate potential differences and similarities between the two studies, we further explored our current high-density EEG data by running paired t-tests on N170 amplitude for the Away and Direct motion transition for each stimulus type and configuration in the right occipitotemporal cluster (given that the main differences in the original study were reported in the right hemisphere). These analyses indicated that N170 amplitude was significantly larger for Away relative to Direct for REAL FACES (t(15) = −2.229, P = 0.04, mean difference = 0.46 µV), consistent with the difference reported in Puce et al., 2000. In contrast, N170 amplitudes for LINE faces were not significantly different [LINE FACE Away relative to LINE FACE Direct (t(0.15) = 0.411, P = 0.69, mean difference = 0.08 µV)]. Peak Analysis of the N170 ERP Hem Condition Peak Ampl (µV) Std Peak Lat (ms) Std Left REAL FACE away −1.45 0.24 209.68 10.97 REAL FACE direct −1.13 0.15 216.75 8.97 REAL CTRL away −0.12 0.26 207.25 10.12 REAL CTRL direct −1.14 0.20 200.87 8.91 LINE FACE away −1.66 0.25 236.06 8.89 LINE FACE direct −1.73 0.30 232.81 6.35 LINE CTRL away −1.63 0.16 234.37 9.65 LINE CTRL direct −1.87 0.24 233.18 7.28 Right REAL FACE away −2.24 0.25 218.12 7.72 REAL FACE direct −1.64 0.20 223.18 8.96 REAL CTRL away −1.86 0.28 203.06 7.52 REAL CTRL direct −1.69 0.26 200.81 7.25 LINE FACE away −2.54 0.34 246.81 7.37 LINE FACE direct −2.64 0.27 236.50 6.97 LINE CTRL away −2.18 0.32 236.62 8.44 LINE CTRL direct −2.42 0.36 231.87 5.87 Legend: Ampl = amplitude; Lat = latency. and latency (ms) data: Mean and Standard Errors (Std) as a function of hemisphere (Hem) and Condition. TABLE 1 \| Group N170 peak amplitude (µV) and latency (ms) data: Mean and Standard Errors (Std) as a function of hemisphere (Hem) and Condition. Legend: Ampl = amplitude; Lat = latency. The 4-way repeated-measures ANOVA for N170 latency revealed a significant main effect of Stimulus type (F(1,15) = 39.49, P < 0.001) and Configuration (F(1,15) = 7.773, P < 0.05). No other statistically significant main effects of hemisphere or motion, or interaction effects were observed for N170 latency. For the significant main effect of Stimulus type, post hoc comparisons indicated that the effect might have been driven by the shorter latencies to REAL faces compared to LINE faces (mean difference: 26 ± 4 ms, Figure 3). For the significant main effect of Configuration, post hoc comparisons suggested that N170s for CONTROL (both REAL and LINE) were shorter compared to FACE stimuli (mean difference: 9 ± 3 ms, Figure 3). To statistically evaluate these differences, four paired t-tests were performed (see Section Materials and Methods). None of these latency differences were found to be statistically significant [REAL FACE Away relative to REAL FACE direct (t(15) = 0.979, P = 0.343, mean difference = 6.06 ms; LINE FACE Away relative to LINE FACE direct (t(15) = 1.563, P = 0.139, mean difference = 6.78 ms; REAL CONTROL away relative to REAL CONTROL direct (t(15) = 1.140, P = 0.272, mean difference 4.31 ms; LINE CONTROL away relative to LINE CONTROL direct (t(15) = 0.457, P = 0.654, mean difference = 2.96 ms)]. Peak Analysis of the N170 ERP Similar comparisons across the control conditions were also not significant: REAL controls showed no differences between conditions [REAL CONTROL Away relative to Direct (t(15) = −0.85, P = 0.41, mean difference = −0.11 µV] and LINE controls also showed no significant differences in N170 amplitudes [LINE CONTROL Away relative to Direct (t(15) = 0.241, P = 0.12, mean difference = 1.6 µV]. Temporal Dynamics: ERSP Plots p y ERSP plots demonstrated clear activity in selected EEG bands in all conditions in the post-motion onset period (Figure 4). The activity profile was similar across all conditions, including activity spanning over multiple time points and frequency bands. A common feature across all conditions was a prolonged burst of activity in the theta (5–8 Hz) and alpha (8–12 Hz) frequency bands in both electrode clusters extending from ∼100–400 ms. Moreover, a consistent decrease in amplitude in the beta range (12–30 Hz) was also seen for most conditions extending from ∼150–400 ms. An additional feature in the ERSP plots was activity in the low-gamma band (30–50 Hz) peaking roughly March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 8 ERPs to changing gaze Rossi et al. FIGURE 5 \| Group data: Statistically signiﬁcant ERSP plot differences between stimulus conditions as a function of hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, respectively. LINE FACE, REAL FACE, LINE CONTROL and REAL CONTROL difference plots appear from top to bottom panels, respectively. For LINE FACE, LINE FACE Away was subtracted from LINE FACE Direct, and for REAL FACE, REAL FACE Away was subtracted from REAL FACE Direct. For LINE CONTROL, the ERSP plot from LINE CONTROL Away has been subtracted from LINE CONTROL Direct. For REAL CONTROL, REAL CONTROL Away was subtracted from REAL CONTROL Direct. Frequency (Hz) is displayed on the y-axis as a function of time (ms). The direction of the difference in spectral power is depicted by the color calibration bar at the right of the ﬁgure. Warm colors depict increased power for condition 1 (Direct) whilst cool colors indicate increased power for condition 2 (Away). Gray areas in the plot indicate regions where the differences between conditions were not signiﬁcant. The vertical broken line at time zero indicates the apparent motion stimulus onset. FIGURE 4 \| Group data: Event-Related Spectral Perturbation (ERSP) plots as a function of condition and hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, respectively. In each four-part display panel total ERSP activity is shown for each respective stimulus condition. The y-axis displays frequency (Hz) and the x-axis displays time (ms). Temporal Dynamics: ERSP Plots Power (decibels) of ERSP activity in decibels, being a default unit used in analysis packages such as EEGLAB (Delorme and Makeig, 2004), is depicted by the color calibration bar at the right of the ﬁgure. The vertical broken line at time zero indicates the apparent motion stimulus onset. FIGURE 5 \| Group data: Statistically signiﬁcant ERSP plot differences between stimulus conditions as a function of hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, respectively. LINE FACE, REAL FACE, LINE CONTROL and REAL CONTROL difference plots appear from top to bottom panels, respectively. For LINE FACE, LINE FACE Away was subtracted from LINE FACE Direct, and for REAL FACE, REAL FACE Away was subtracted from REAL FACE Direct. For LINE CONTROL, the ERSP plot from LINE CONTROL Away has been subtracted from LINE CONTROL Direct. For REAL CONTROL, REAL CONTROL Away was subtracted from REAL CONTROL Direct. Frequency (Hz) is displayed on the y-axis as a function of time (ms). The direction of the difference in spectral power is depicted by the color calibration bar at the right of the ﬁgure. Warm colors depict increased power for condition 1 (Direct) whilst cool colors indicate increased power for condition 2 (Away). Gray areas in the plot indicate regions where the differences between conditions were not signiﬁcant. The vertical broken line at time zero indicates the apparent motion stimulus onset. FIGURE 5 \| Group data: Statistically signiﬁcant ERSP plot differences between stimulus conditions as a function of hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, FIGURE 5 \| Group data: Statistically signiﬁcant ERSP plot differences between stimulus conditions as a function of hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, respectively. LINE FACE, REAL FACE, LINE CONTROL and REAL CONTROL difference plots appear from top to bottom panels, respectively. For LINE FACE, LINE FACE Away was subtracted from LINE FACE Direct, and for REAL FACE, REAL FACE Away was subtracted from REAL FACE Direct. For LINE CONTROL, the ERSP plot from LINE CONTROL Away has been subtracted from LINE CONTROL Direct. For REAL CONTROL, REAL CONTROL Away was subtracted from REAL CONTROL Direct. Frequency (Hz) is displayed on the y-axis as a function of time (ms). Line Face ERSP comparisons between gaze transitions for LINE FACE stimuli produced statistically significant differences in beta and gamma bands in the interval of 150–300 ms post-motion onset in both electrode clusters (Figure 5). Larger bilateral gamma amplitudes (∼30–∼40 Hz) were seen to direct gaze transitions (LINE FACE Direct) peaking at around 200 ms relative to gaze aversions (LINE FACE Away) (Figure 5 first row). For the opposite side of the contrast, larger bilateral gamma activity at ∼40 Hz occurred at a later point in time (at ∼300 ms for LINE FACE Away vs. Direct, Figure 5 first row). Significant differences in beta activity were only observed in the left temporo-occipital electrode cluster, with a relative larger decrement in beta amplitude for Away relative to Direct. Real Face REAL FACE stimuli elicited divergent significant differences across hemispheres (as shown in Figure 5 s row) relative to LINE FACES. In the left electrode cluster, statistically significant differences occurred at similar times and were confined to the same frequency range, and were similar for REAL and LINE face stimuli. Direct gaze changes in REAL FACES elicited larger gamma amplitudes peaking at ∼200 ms (∼30–45 Hz range) relative to gaze aversions, while Away gaze changes elicited stronger gamma power at ∼300 ms (∼40 Hz), and at ∼500 ms (∼35–45 Hz range). In the right electrode cluster, a later biphasic difference in gamma amplitude consisted of initial augmentation and then suppression of activity for direct relative to averted gaze. These effects occurred for frequencies ∼40 and ∼50 Hz and peaked between 400 ms (Figure 5, second row). Unlike in the left hemisphere, effects for REAL FACES occurred at later times relative to LINE FACES: gamma effects for REAL FACES occurred later in time relative to LINE FACES. around 200 ms after apparent motion onset for most conditions (Figure 4). Statistically significant differences between conditions were seen only in the beta (12–30 Hz) and low-gamma (30–50 Hz) frequency bands. Figure 5 displays masked time-frequency statistically significant differences between facial motion transitions and control motion transitions. We discuss these differences for each stimulus type below. Temporal Dynamics: ERSP Plots The direction of the difference in spectral power is depicted by the color calibration bar at the right of the ﬁgure. Warm colors depict increased power for condition 1 (Direct) whilst cool colors indicate increased power for condition 2 (Away). Gray areas in the plot indicate regions where the differences between conditions were not signiﬁcant. The vertical broken line at time zero indicates the apparent motion stimulus onset. FIGURE 4 \| Group data: Event-Related Spectral Perturbation (ERSP) plots as a function of condition and hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, respectively. In each four-part display panel total ERSP activity is shown for each respective stimulus condition. The y-axis displays frequency (Hz) and the x-axis displays time (ms). Power (decibels) of ERSP activity in decibels, being a default unit used in analysis packages such as EEGLAB (Delorme and Makeig, 2004), is depicted by the color calibration bar at the right of the ﬁgure. The vertical broken line at time zero indicates the apparent motion stimulus onset. FIGURE 4 \| Group data: Event-Related Spectral Perturbation (ERSP) plots as a function of condition and hemisphere. Left (L) and right (R) plots as a function of condition and hemisphere. Left (L) and right (R) occipitotemporal data are presented in left and right columns, respectively. In each four-part display panel total ERSP activity is shown for each respective stimulus condition. The y-axis displays frequency (Hz) and the x-axis displays time (ms). Power (decibels) of ERSP activity in decibels, being a default unit used in analysis packages such as EEGLAB (Delorme and Makeig, 2004), is depicted by the color calibration bar at the right of the ﬁgure. The vertical broken line at time zero indicates the apparent motion stimulus onset. Real Control However, relative to our original experimental question, based on the above findings we would argue that, stimulus- context effects from real faces were not present for viewed eye movements in impoverished faces, when both stimulus categories are presented within the same experiment. This suggests that the difference in N170 amplitude gaze changes in real faces might be driven by a different neural mechanism relative to N170 modulation by mouth movements. We previously reported to mouth opening vs. closing movements in both real and line-drawn faces produce N170 amplitude modulations (Puce et al., 2003). Consistent with what we had previously postulated, it appears that information from mouth movements can be accessed from both real and impoverished images, unlike information from the eyes that appears to require real faces. REAL CONTROL stimuli generated a much more diverse pattern of differences extending between ∼100 to ∼600 ms after movement onset in beta and gamma activity in the left hemisphere (Figure 5, fourth row) relative to LINE CONTROL stimuli. First, significantly stronger activity for the REAL CONTROL Away condition consisted of an early beta component peaking right before 200 ms (∼20–25 Hz), and a gamma component at ∼200 ms (∼45–50 Hz). In this same time range, gamma at ∼35–40 Hz between ∼200–300 ms, was significantly larger for Direct condition. Later in time, gamma burst extending from ∼40–50 Hz occurred between ∼500–600 ms and was stronger for the Away condition, while a lower frequency gamma component extending between ∼35–40 Hz was significantly stronger between ∼550–600 ms for the Direct condition (Figure 5, fourth row). Unlike the left electrode cluster, the right cluster showed more limited significant differences in oscillatory activity between conditions. Specifically, a very early gamma band response peaked at ∼100 ms, being stronger for the Direct condition, and a later higher frequency gamma component at ∼45 Hz peaking at ∼500 ms was stronger for the Away condition. Why would ERPs elicited to impoverished mouth movements behave so differently to those observed to gaze transitions? Bassili (1978, 1979) originally reported behavioral data to viewed emotional expressions on point-light face stimuli. Success in recognizing different emotions was driven by the subject focusing on either the upper or lower regions of the impoverished face (i.e., eyes/brows vs. mouth, respectively) (Bassili, 1979). The line-drawn stimuli in this study can also be regarded as biological motion stimuli (see Oram and Perrett, 1994, 1996). Real Control Most typically, these impoverished forms of stimuli are used to represent very effectively the articulated motion of the joints of the body—where information related to the type of activity being observed can be readily identified from seeing these minimalist displays (Blake and Shiffrar, 2007). There is a very large literature demonstrating the sensitivity of the human brain to biological motion stimuli (see reviews by Giese and Poggio, 2003; Puce and Perrett, 2003; Blake and Shiffrar, 2007). Despite this, very few research groups have studied brain responses to biological motion stimuli involving the face (see the original studies of Bassili, 1978, 1979). Similar to movements of the body, mouth movements are a type of articulated motion. Mouth opening and closing occurs due to the actions of the articulated mandible. Hence, we argue that our previously reported ERP data that demonstrate differences between mouth opening and closing are representing a brain response to articulated biological motion (e.g., see Beauchamp et al., 2002; Peuskens et al., 2005). Line Control Unlike for the face stimuli significant effects were effectively confined to the beta range, for LINE CONTROL stimuli identified significant bilateral differences that were confined to the beta band, consisting of brief periods between ∼300 to ∼400 ms after movement onset were observed bilaterally March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 9 ERPs to changing gaze Rossi et al. (Figure 5, third row). This difference was driven by stronger beta suppression in the Away condition in both hemispheres. (Figure 5, third row). This difference was driven by stronger beta suppression in the Away condition in both hemispheres. a number of interesting questions about the nature of stimuli and experimental designs, which we subsequently discuss in detail. Frontiers in Human Neuroscience \| www.frontiersin.org ERP Data: N170 Effects These, results raise March 2015 \| Volume 9 \| Article 185 10 ERPs to changing gaze Rossi et al. for both arrows and schematic eyes, but different ERP-related effects that most typically occur beyond the P100 and N170 that are specific to these visuospatial cueing tasks. Specifically, anterior and posterior negativities have been described to arrow and schematic gaze-cues respectively (Hietanen et al., 2008; Brignani et al., 2009). Interestingly, when real faces are used in a gaze-cueing paradigm, differences in early ERPs, such as P100 and N170 (P1 and N1) have been reported, producing larger amplitudes for valid trials (Schuller and Rossion, 2001). These experimental results, despite being generated in different experimental designs, are consistent with the current study in that schematic eyes do not elicit changes in earlier sensory ERPs such as N170. This finding bears further investigation, given that the schematic eyes in the visuospatial cueing studies did have contrast between ‘‘irises’’ and ‘‘sclera’’, unlike those in our current study. is thought to have evolved for the purposes of facilitating social interactions (Emery, 2000; Rosati and Hare, 2009). Human eyes are unique among primates with respect to this attribute, with most other species showing very little difference in contrast between irises and sclera (Rosati and Hare, 2009). The lack of demonstrated N170 differences to gaze aversions relative to direct gaze transitions in this and our previous study using line-drawn faces (Rossi et al., 2014) supports the idea that neural activity to eye gaze transitions in real faces might be triggered by low-level stimulus features. Specifically, changes in local visual contrast and increased eye white area (see also Whalen et al., 2004; Hardee et al., 2008) as irises/pupils move from a direct to an averted position likely drive the N170 differences previously reported by Puce et al. (2000) and also seen to the real faces in the current study. This is likely to be driven by the high-contrast human iris-sclera complex. We found only a 2-way interaction on N170 amplitude suggesting that a contribution of low-level visual features on the modulation of the N170 cannot be ruled out—as indicated by a discernable ERP to the motion control stimuli. The real control stimuli also presented a high local luminance/contrast difference in the same part of the visual space as did the eye stimulus in the real face. ERP Data: N170 Effects However, based on the paired t-test data the local high contrast effect cannot totally account for the observed modulation of N170 amplitude to gaze motion: no differences were seen between N170s to the ‘‘Away’’ and ‘‘Direct’’ transitions for real control stimuli. The local high-contrast of the iris/sclera complex probably contributes to the N170, but cannot explain differential effect seen on the N170 to the real face stimuli. It may well be that the actual configuration of the eye plays a role in the response. This, in some ways, could be regarded as a low-level feature also: a feature that is embedded in a more complex stimulus (the dynamic face). A set of studies with parametric manipulations of these variables would be required to get to the bottom of this effect. Additionally, it would be interesting to investigate the relative effect of iris/sclera contrast and the configuration effect on the N170 by using for instance faces where human eyes were replaced by non-human primate eyes (Dupierrix et al., 2014), or examining responses to non- human primate gaze changes, as typically the iris/sclera complex in the eyes of the non-human primate do not show these high local contrast differences (Emery, 2000; Rosati and Hare, 2009) A further point needs to be made on the issue of the schematic representation of faces. Our participants all reported that they found the gaze transitions in both types of faces to be compelling. However, some interesting differences in behavior and neurophysiology were observed for the impoverished stimulus categories. Subjects detected target stimuli that consisted of image negatives for all presented stimulus types. Participants were slower at detecting impoverished face targets relative to real face targets, and were the slowest for impoverished faces relative to impoverished controls. For real stimuli, face targets were identified faster relative to controls. We cannot directly relate our behavioral data to our ERP findings: the ERPs were recorded to trials where no behavioral response was recorded, so we can only speculate about the potential nature of our ERP findings to the impoverished stimuli. One consideration might be that the impoverished faces in the current study might not be treated as faces by the brain. A stimulus such as an impoverished face might be ambiguous, and would hence take a longer time to be evaluated and might require more detailed processing. ERP Data: N170 Effects Our main purpose for the experiment was to look for stimulus- induced context effects that might produce modulations of the N170 by gaze transition in line-drawn faces, when presented with real images of faces in the same experiment. For a stimulus context effect to be present, we would expect to observe parallel effects in the form of larger N170s to gaze aversions vs. direct gaze for both real and line-drawn faces. Given our experimental design, this would translate to a significant interaction effect of Motion [Away, Direct] × Config [Face, Control]. If, however, N170 modulation occurred only to gaze changes in real faces, then we would expect to see a significant interaction effect of Motion [Away, Direct] × Config [Face, Control] × Stimulus [Real, Line]. Finally, if N170 modulation was driven by general low-level effects of stimulus luminance and contrast, then we might expect to observe a significant main effect of Stimulus type [Real, Line]. Other facial movements involving the forehead and eyes do not require the movements of articulated joints, and gaze aversions also fall into this category. Changes in the eyes, either associated with gaze aversions, or with emotions such as fear, surprise, and happiness alter the amount of seen eye white area, which can modulate the brain’s response to these types of stimuli even when these observed changes are task-irrelevant (Whalen et al., 2004; Hardee et al., 2008). This is likely to be driven by the high-contrast human iris- sclera complex. A gaze change, such as a lateral gaze shift, produces a local visuospatial luminance/contrast change. This type of stimulus, which can readily be seen at a distance, Interestingly, our analysis generated effects that were more complex than predicted for N170 amplitude. We observed a significant main effect of Motion [Away, Direct] × Stim [Real, Line]—which was not what we had predicted. The nature of these differences was clarified with paired t-tests, which indicated that N170 was larger for averted vs. direct gaze only for real faces—consistent with our previous study (Puce et al., 2000). Also consistent with our previous work (Rossi et al., 2014) there was no effect of gaze aversion on N170 in our line-drawn face stimuli (or control stimuli). Having said that, there were other striking differences in the current dataset that resulted from our initial predictions not being upheld. Frontiers in Human Neuroscience \| www.frontiersin.org EEG Spectral Power Total EEG spectral power to the very brief apparent motion transition generated consistent prolonged bursts of activity in theta, beta and gamma EEG frequency bands which overall behaved similarly across conditions in a task requiring detecting negatives of the stimuli (see Figure 4). We expected to observe oscillatory EEG changes in beta and gamma bands that would occur at different post-motion onset times for facial and control motion stimuli when statistical comparisons were made between conditions, in line with our previous study where participants detected color changes in line drawn face and control motion stimuli (Rossi et al., 2014). Statistically significant differences between stimulus conditions were confined to the beta and gamma bands only. The main significant change in gamma activity occurred at ∼200 and ∼300 ms post-motion onset for LINE faces bilaterally, but only in the left electrode cluster for REAL faces. Direct gaze transitions elicited stronger gamma amplitudes at ∼200 ms (for LINE and REAL faces), whereas averted gaze elicited stronger gamma amplitudes at ∼300 ms. We speculate that these bursts of activity reflect processing of facial information, as these changes were not present in the respective control conditions (compare first and second rows of Figure 5). REAL controls showed a gamma amplitude increment (∼40–50 Hz) for the Direct condition, which occurred ∼100 ms later than the gamma burst seen for REAL faces (compare second and last rows, Figure 5). It may be that this gamma burst to the controls might be a general coherent motion effect in data that were sampled from our occipito-temporal electrode Caruana et al. (2014) performed an intracranial EEG study where ERPs and oscillatory activity were examined side-by-side. They presented epilepsy surgery patients with faces showing dynamic gaze changes while intracranial EEG was recorded from electrodes sited in the posterolateral temporal cortex. Intracranial N200 ERP activity (the analog of the scalp N170) and transient broadband high frequency gamma band activity (out to 500 Hz) occurring at around the same time as the N200 was significantly larger when patients viewed averted gaze relative to either direct gaze or lateral switching movements (Caruana et al., 2014). Lachaux et al. (2005) reported intracranial gamma-band activity in left fusiform gyrus, occipital gyrus and intraparietal sulcus for a static face detection task. ERP Data: N170 Effects Therefore, we chose to have congruous motion for all stimulus types. Interestingly, in doing so we may well have created a stimulus-context effect for the control stimuli—and have potentially allowed subjects to ‘‘see’’ eyes in the control stimuli (as we had previously experienced). This occurred only for the REAL CONTROL stimuli, and not for the LINE CONTROL stimuli—there were no differences related to movement direction. So, we might have actually created an unexpected effect of stimulus-context in this experiment, with the REAL FACES providing a context for the REAL CONTROLS (not unlike that seen by Bentin and Golland, 2002). Our original purpose for running the experiment was to explore context effects related to eye gaze changes in LINE FACES in the presence of REAL FACES. In this latter case, we can state that no effect of stimulus context was observed. Puce et al., 1998). In the current study we were concerned that differences in the type of presented motion (congruous vs. incongruous) across the stimulus conditions may have, in part, contributed to the differences in the neurophysiological response between faces and controls. Therefore, we chose to have congruous motion for all stimulus types. Interestingly, in doing so we may well have created a stimulus-context effect for the control stimuli—and have potentially allowed subjects to ‘‘see’’ eyes in the control stimuli (as we had previously experienced). This occurred only for the REAL CONTROL stimuli, and not for the LINE CONTROL stimuli—there were no differences related to movement direction. So, we might have actually created an unexpected effect of stimulus-context in this experiment, with the REAL FACES providing a context for the REAL CONTROLS (not unlike that seen by Bentin and Golland, 2002). Our original purpose for running the experiment was to explore context effects related to eye gaze changes in LINE FACES in the presence of REAL FACES. In this latter case, we can state that no effect of stimulus context was observed. In our previous study, we examined ERSP changes to impoverished line-drawn faces and controls only, where facial movements included eye and mouth movements in a color detection task that required a behavioral response for all presented stimuli (Rossi et al., 2014). ERP Data: N170 Effects In that study we also observed significant transient increases in the beta and gamma ranges to the facial motion stimuli, but these changes in oscillatory activity tended to occur at different time points relative to those seen in the current study. Gamma range changes to eye and mouth movements, if present, occurred much later in time relative to the current study e.g., after 400 ms post-motion onset relative to the changes at ∼150–300 ms in the current study and favored direct gaze and mouth closing movements. Beta range changes, if present, showed a short burst at ∼100 ms and a more prolonged burst between 350–550 ms favoring averted gaze (Rossi et al., 2014). In our previous study (Rossi et al., 2014) subjects viewed line-drawn face displays where the color of the lines could be either white or red, with subjects having to make a color decision on every seen motion transition. In our current study, as we had both line-drawn and real faces in the same experiment, we elected to use a target detection task where subjects had to identify negatives of all stimulus types, to try to ensure that equal attention was given to all stimulus types. Taken together, the ERSP findings of both studies would indicate that the significant differences in oscillatory behavior we observed at sites producing maximal ERP activity might be driven rather by task differences/decisions rather than the motion characteristics of the stimuli per se. Having said that, in both studies, changes in oscillatory activity were different across faces and respective controls, suggesting that changes in total oscillatory activity post-motion onset may reflect a complex mix of task and stimulus-related properties. At this point in time much more data are needed to make sense of these changes in oscillatory activity to facial motion stimuli and their relationship to ERP activity—it is not yet customary to perform both types of analysis in studies. The effects that are induced in N170 activity here underscore how important low-level stimulus manipulations related to luminance/contrast and also motion can be, and these have the potential to interact with task-related variables. Hence, control stimuli of multiple types may have to be used in an experiment so as to understand the nature of observed differences in neurophysiological data across different conditions. ERP Data: N170 Effects This might manifest as increased response time (for the detection of targets), as well as increased N170 latencies (which were seen as a main effect for line- drawn vs. real stimuli). Coupled with the longer latency is also an increased N170 amplitude (seen as a main effect for stimulus type for line faces and controls). The increased N170 latencies and amplitudes observed here might potentially reflect the more effortful processing that might be required of these stimuli. Viewing the gaze changes of another individual are thought to produce reflexive changes in one’s visuospatial attention (Hietanen et al., 2008; Itier and Batty, 2009). It could be argued that both the eyes and their respective scrambled controls might cue participants’ visual attention in the motion direction. This possibility has been discussed in the literature (Grossmann et al., 2007; Hadjikhani et al., 2008; Straube et al., 2010). Our line-drawn controls in their ‘‘averted’’ state looked like arrows (facing left and then right), and N170 did not differ between the ‘‘arrow’’ control and the direct gaze control condition (a diamond shape). Behavioral and ERP studies of visuospatial cueing paradigms using Posner-like tasks (Posner, 1980) in healthy subjects have demonstrated similar behavioral effects Some intriguing differences in N170 activity relating to data of our control stimuli need to be addressed. In our original study, the checkerboard controls had movements that were not congruous with one another i.e., checks changed in two locations corresponding to each eye in the real face, but moved in opposite lateral directions (Puce et al., 1998). The control stimuli were created deliberately in this fashion, as in the piloting of data for the earlier study, subjects reported a very convincing and persistent illusion of eyes that the checkerboard control stimuli created. This created the unwanted confound of visualization in the study. This effect was abolished by introducing a movement condition where checks reversed in opposite directions, and we used this control stimulus in our previous study (see March 2015 \| Volume 9 \| Article 185 11 Rossi et al. ERPs to changing gaze clusters in a task where negative images of stimuli had to be detected. Puce et al., 1998). In the current study we were concerned that differences in the type of presented motion (congruous vs. incongruous) across the stimulus conditions may have, in part, contributed to the differences in the neurophysiological response between faces and controls. Frontiers in Human Neuroscience \| www.frontiersin.org General Conclusions We would advocate that future studies evaluate ERP and ERSP activity in parallel, so that we can develop an understanding of the functional significance of each type of neurophysiological activity, and how one might affect the other. Gaze changes produced gamma activity irrespective of face type: direct gaze elicited more gamma at an earlier latency relative to averted gaze. Overall, our N170 ERP peak analysis argues for the idea that gaze changes/eye movements in impoverished line- drawn faces do not trigger the neural responses that have been associated to the perception of socially relevant facial motion (relevant for communicative behavior), replicating data in an earlier study (Rossi et al., 2014). In contrast, real faces in this study as well as others (e.g., Puce et al., 2000, 2003; Conty et al., 2007; Latinus et al., revised and resubmitted) show N170 sensitivity to gaze changes. Interestingly, differences between our real faces and control stimuli were not as strong as we had previously demonstrated (Puce et al., 2000), and this may be due to ability to potentially visualize our current controls as faces. Overall our data indicate that N170s elicited to social attention manipulations are not modulated by top- down processes (such as priming or context) for impoverished faces. In our study, it is not entirely clear if the gamma activity could be related to configurality, movement, or a combination of both. However, we believe that the gamma responses at ∼200 and ∼300 ms present both for LINE and REAL stimuli reflect a preferential response to facial movement. An alternative explanation is that as a gamma response was seen for LINE and REAL faces and controls, albeit at different post- motion onsets, these gamma responses might be correlates of motion perception in the horizontal axis (see Figure 5). However, in our previous study we included mouth movements (with a large vertical component) in an impoverished face, and recorded gamma activity to both stimulus types (Rossi et al., 2014). From the relatively few existing studies in the literature, it is clear that the relationship between oscillatory EEG activity, including gamma, and stimulus and task type is complex. Much further work to viewing motion stimuli with different attributes (e.g., linear vs. non-linear, inward vs. outward radial, looming vs. receding) where ERP and ERSP activity are directly compared will be needed to disentangle these issues. EEG Spectral Power Gamma band amplitude (40–200 Hz) significantly increased between 250–500 ms post- stimulus onset, and these condition differences were not present in the ERP (Lachaux et al., 2005). In contrast, intracranial N200 activity and high frequency gamma activity in ventral temporal cortex to viewing static faces has been documented (Engell and McCarthy, 2011). The presence of gamma activity Frontiers in Human Neuroscience \| www.frontiersin.org March 2015 \| Volume 9 \| Article 185 12 ERPs to changing gaze Rossi et al. in the ERSP predicted the presence and size of the N200 at a particular site. Relevant to this study, however, N200 activity was elicited to impoverished face stimuli, but notably gamma activity was absent (Engell and McCarthy, 2011), indicating that the relationship between intracranial ERP activity and gamma activity can be a very complex one. It is difficult to make a comparison between intracranial and scalp EEG studies, because with intracranial EEG high frequency gamma band activity can be sampled, whereas in scalp EEG studies the skull effectively acts as a low pass filter so that gamma frequencies typically will be recorded only under 100 Hz (Srinivasan et al., 1996). Interestingly, impoverished face stimuli, as used in Engell and McCarthy (2011) study and our study, do not appear to generate prominent gamma activity. In our study, stimuli tended to evoke sustained activity in frequency ranges below gamma), gamma activity was transient and was seen at around 200 ms post- motion onset in electrodes sited over lateral temporal scalp (Figure 4), similar to the broadband gamma reported in the intracranial EEG studies (e.g., Lachaux et al., 2005; Engell and McCarthy, 2011; Caruana et al., 2014). In another scalp EEG study, Zion-Golumbic and Bentin (2007), noted that activity in the 25–45 Hz range between 200–300 ms post-stimulus onset, was largest for (static) real faces compared to scrambled real faces in midline parieto-occipital locations (Zion-Golumbic and Bentin, 2007). Müller et al., 2003; Müller-Putz et al., 2007). However, the beta change for LINE FACES occurred earlier than that for LINE CONTROLS. We have previously seen similar decreases in the 20–30 Hz beta range for line-drawn faces producing identical lateral eye movements as those used in the current study (Rossi et al., 2014). However, in Rossi et al. (2014) decrements in beta spectral power occurred later (∼400–550 ms). EEG Spectral Power In these two studies, participants were asked to respond to very different stimulus attributes, and with different stimulus probability. In Rossi et al. (2014) participants responded on every trial to indicate the color of the line-drawn stimulus (which varied randomly from white to red, and vice-versa). In the current study, participants responded to infrequent targets that were negatives of all stimuli. Hence, at this stage it is not possible to clarify the nature of the observed beta power change to the apparent motion stimulus. We do believe that these differences were driven by task demands, and will have to explicitly test this in future studies. Frontiers in Human Neuroscience \| www.frontiersin.org References Does the face-specific N170 component reflect the activity of a specialized eye processor? Neuroreport 9, 2945–2948. doi: 10.1097/00001756- 199809140-00005 Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. Royal Stat. Soc. Series B Methodological 57, 289–300. Emery, N. J. (2000). The eyes have it: the neuroethology, function and evolution of social gaze. Neurosci. Biobehav. Rev. 24, 581–604. doi: 10.1016/s0149- 7634(00)00025-7 Bentin, S., Allison, T., Puce, A., Perez, E., and McCarthy, G. (1996). Electrophysiological studies of face perception in humans. J. Cogn. Neurosci. 8, 551–565. doi: 10.1162/jocn.1996.8.6.551 Engel, A. K., and Fries, P. (2010). Beta-band oscillations–signalling the status quo? Curr. Opin. Neurobiol. 20, 156–165. doi: 10.1016/j.conb.2010.02.015 Engell, A. D., and McCarthy, G. (2011). The relationship of gamma oscillations and face-specific ERPs recorded subdurally from occipitotemporal cortex. Cereb. Cortex 21, 1213–1221. doi: 10.1093/cercor/bhq206 Bentin, S., and Golland, Y. (2002). Meaningful processing of meaningless stimuli: the influence of perceptual experience on early visual processing of faces. Cognition 86, B1–B14. doi: 10.1016/s0010-0277(02)00124-5 Ergenoglu, T., Demiralp, T., Bayraktaroglu, Z., Ergen, M., Beydagi, H., and Uresin, Y. (2004). Alpha rhythm of the EEG modulates visual detection performance in humans. Brain Res. Cogn. Brain Res. 20, 376–383. doi: 10.1016/j.cogbrainres. 2004.03.009 Blake, R., and Shiffrar, M. (2007). Perception of human motion. Annu. Rev. Psychol. 58, 47–73. doi: 10.1146/annurev.psych.57.102904.190152 Bötzel, K., Schulze, S., and Stodieck, S. R. (1995). Scalp topography and analysis of intracranial sources of face-evoked potentials. Exp. Brain Res. 104, 135–143. doi: 10.1007/bf00229863 Fries, P., Womelsdorf, T., Oostenveld, R., and Desimone, R. (2008). The effects of visual stimulation and selective visual attention on rhythmic neuronal synchronization in macaque area V4. J. Neurosci. 28, 4823–4835. doi: 10. 1523/JNEUROSCI.4499-07.2008 Brandt, M. E., and Jansen, B. H. (1991). The relationship between prestimulus alpha-amplitude and visual evoked-potential amplitude. Int. J. Neurosci. 61, 261–268. doi: 10.3109/00207459108990744 George, N., Evans, J., Fiori, N., Davidoff, J., and Renault, B. (1996). Brain events related to normal and moderately scrambled faces. Cogn. Brain Res. 4, 65–76. doi: 10.1016/0926-6410(95)00045-3 Brefczynski-Lewis, J., Berrebi, M. E., McNeely, M. E., and Puce, A. (2011). In the blink of an eye: neural responses elicited to viewing the eye blinks of another individual. Front. Hum. Neurosci. 5:68. doi: 10.3389/fnhum.2011. 00068 Giese, M. A., and Poggio, T. (2003). Neural mechanisms for the recognition of biological movements. Nat. Rev. Neurosci. 4, 179–192. doi: 10.1038/ nrn1057 Brignani, D., Guzzon, D., Marzi, C. References Caruana, F., Cantalupo, G., Lo Russo, G., Mai, R., Sartori, I., and Avanzini, P. (2014). Human cortical activity evoked by gaze shift observation: an intracranial EEG study. Hum. Brain Mapp. 35, 1515–1528. doi: 10.1002/hbm. 22270 Babiloni, C., Vecchio, F., Bultrini, A., Luca Romani, G., and Rossini, P. M. (2006). Pre- and poststimulus alpha rhythms are related to conscious visual perception: a high-resolution EEG study. Cereb. Cortex 16, 1690–1700. doi: 10. 1093/cercor/bhj104 Conty, L., N’Diaye, K., Tijus, C., and George, N. (2007). When eye creates the contact! ERP evidence for early dissociation between direct and averted gaze motion processing. Neuropsychologia 45, 3024–3037. doi: 10.1016/j. neuropsychologia.2007.05.017 Basar, E., Schürmann, M., Basar-Eroglu, C., and Karakas, S. (1997). Alpha oscillations in brain functioning: an integrative theory. Int. J. Psychophysiol. 26, 5–29. doi: 10.1016/s0167-8760(97)00753-8 Deffke, I., Sander, T., Heidenreich, J., Sommer, W., Curio, G., Trahms, L., et al. (2007). MEG/EEG sources of the 170-ms response to faces are co-localized in the fusiform gyrus. Neuroimage 35, 1495–1501. doi: 10.1016/j.neuroimage. 2007.01.034 Bassili, J. N. (1978). Facial motion in the perception of faces and of emotional expression. J. Exp. Psychol. Hum. Percept. Perform. 4, 373–379. doi: 10. 1037//0096-1523.4.3.373 Delorme, A., and Makeig, S. (2004). EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J. Neurosci. Methods 134, 9–21. doi: 10.1016/j.jneumeth.2003.10.009 Bassili, J. N. (1979). Emotion recognition: the role of facial movement and the relative importance of upper and lower areas of the face. J. Pers. Soc. Psychol. 37, 2049–2058. doi: 10.1037//0022-3514.37.11.2049 Bassili, J. N. (1979). Emotion recognition: the role of facial movement and the relative importance of upper and lower areas of the face. J. Pers. Soc. Psychol. 37, 2049–2058. doi: 10.1037//0022-3514.37.11.2049 Beauchamp, M. S., Lee, K. E., Haxby, J. V., and Martin, A. (2002). Parallel visual motion processing streams for manipulable objects and human movements. Neuron 34, 149–159. doi: 10.1016/s0896-6273(02)00642-6 Beauchamp, M. S., Lee, K. E., Haxby, J. V., and Martin, A. (2002). Parallel visual motion processing streams for manipulable objects and human movements. Neuron 34, 149–159. doi: 10.1016/s0896-6273(02)00642-6 Dupierrix, E., de Boisferon, A. H., Méary, D., Lee, K., Quinn, P. C., Di Giorgio, E., et al. (2014). Preference for human eyes in human infants. J. Exp. Child Psychol. 23, 138–146. doi: 10.1016/j.jecp.2013.12.010 Bell, A. J., and Sejnowski, T. J. (1995). An information-maximization approach to blind separation and blind deconvolution. Neural Comput. 7, 1129–1159. doi: 10.1162/neco.1995.7.6.1129 Eimer, M. (1998). General Conclusions Taking together the findings of our previous and current studies, N170s to gaze changes appear to be generated by different processes relative to mouth movements. Eye movements/gaze changes in real faces generate local visuospatial luminance/contrast changes producing an N170 that altered by the luminance/contrast change that occurs between changing gaze conditions. When gaze changes are presented in an impoverished face, there is no differential luminance/contrast change and the N170 does not show modulation across gaze conditions. In contrast, mouth opening and closing movements are a type of articulated biological motion, whose moving form modulates N170 irrespective of whether the movement occurs in a real or an impoverished face (Puce et al., 2003; Rossi et al., 2014), therefore largely independent of luminance/contrast. Yet, the motion of both face parts happens to elicit ERP activity at the As noted earlier, LINE FACES and CONTROLS differ in configuration, but produce identical movements. A common spectral change between LINE FACES and CONTROLS was a decrement in beta power, maximal at ∼25 Hz for the conditions producing direct gaze (FACES) and a diamond shape (CONTROLS), albeit at slightly different time intervals (FACES from ∼215–300 ms, CONTROLS ∼300–400 ms). Since the common feature between these stimulus types is apparent motion of identical numbers of pixels, these beta components might be related to the encoding of the movement. Beta spectral power has been previously associated with the perception and production of movement per se (Pfurtscheller et al., 1996; March 2015 \| Volume 9 \| Article 185 13 ERPs to changing gaze Rossi et al. activity in healthy subjects and individuals with neuropsychiatric disorders. activity in healthy subjects and individuals with neuropsychiatric disorders. same latency with a similar scalp topography that likely reflects an aggregate of neural activity from various parts of motion- sensitive, as shown clearly by fMRI studies (e.g., Puce and Perrett, 2003). The functional dynamics from this very heterogeneous brain network will likely only be disentangled by aggregating the data from a number of different investigations such as functional connectivity using fMRI, intracranial EEG and scalp EEG/MEG studies which examine evoked and induced neurophysiological Acknowledgments AP, FJP, and AR were supported by NIH grant NS-049436. We also thank our participants, who made this experiment possible. AP, FJP, and AR were supported by NIH grant NS-049436. We also thank our participants, who made this experiment possible. References A correlate of an idling motor area? Electroencephalogr. Clin. Neurophysiol. 98, 281–293. doi: 10.1016/0013-4694(95)00258-8 Itier, R. J., and Taylor, M. J. (2004). N170 or N1? Spatiotemporal differences between object and face processing using ERPs. Cereb. Cortex 14, 132–142. doi: 10.1093/cercor/bhg111 Posner, M. I. (1980). Orienting of attention. Q. J. Exp. Psychol. 32, 3–25. doi: 10. 1080/00335558008248231 Jemel, B., George, N., Chaby, L., Fiori, N., and Renault, B. (1999). Differential processing of part-to-whole and part-to-part face priming: an ERP study. Neuroreport 10, 1069–1075. doi: 10.1097/00001756-199904060-00031 Puce, A. (2013). ‘‘Perception of nonverbal cues,’’ in The Oxford Handbook of Cognitive Neuroscience, Volume 2: The Cutting Edges, eds K. Ochsner and S. M. Kosslyn (Oxford: Oxford University Press), 148–164. Jemel, B., Pisani, M., Calabria, M., Crommelinck, M., and Bruyer, R. (2003). Is the N170 for faces cognitively penetrable? Evidence from repetition priming of Mooney faces of familiar and unfamiliar persons. Brain Res. Cogn. Brain Res. 17, 431–446. doi: 10.1016/s0926-6410(03)00145-9 Puce, A., Allison, T., Bentin, S., Gore, J. C., and McCarthy, G. (1998). Temporal cortex activation in humans viewing eye and mouth movements. J. Neurosci. 18, 2188–2199. Puce, A., and Perrett, D. (2003). Electrophysiology and brain imaging of biological motion. Philos. Trans. R. Soc. Lond. B Biol. Sci. 358, 435–445. doi: 10.1098/rstb. 2002.1221 Jervis, B. W., Nichols, M. J., Johnson, T. E., Allen, E., and Hudson, N. R. (1983). A fundamental investigation of the composition of auditory evoked potentials. IEEE Trans. Biomed. Eng. 30, 43–50. doi: 10.1109/tbme.1983. 325165 Puce, A., Smith, A., and Allison, T. (2000). ERPs evoked by viewing facial movements. Cogn. Neuropsychol. 17, 221–239. doi: 10.1080/0264329003 80580 Joyce, C., and Rossion, B. (2005). The face-sensitive N170 and VPP components manifest the same brain processes: the effect of reference electrode site. Clin. Neurophysiol. 116, 2613–2631. doi: 10.1016/j.clinph.2005.07.005 Puce, A., Syngeniotis, A., Thompson, J. C., Abbott, D. F., Wheaton, K. J., and Castiello, U. (2003). The human temporal lobe integrates facial form and motion: evidence from fMRI and ERP studies. Neuroimage 19, 861–869. doi: 10. 1016/s1053-8119(03)00189-7 Klimesch, W. (2012). α-band oscillations, attention and controlled access to stored information. Trends Cogn. Sci. 16, 606–617. doi: 10.1016/j.tics.2012.10.007 Ray, W. J., and Cole, H. W. (1985). Eeg alpha activity reflects attentional demands and beta-activity reflects emotional and cognitive-processes. Science 228, 750–752. doi: 10.1126/science.3992243 Lachaux, J. P., George, N., Tallon-Baudry, C., Martinerie, J., Hugueville, L., Minotti, L., et al. (2005). References A., and Miniussi, C. (2009). Attentional orienting induced by arrows and eye-gaze compared with an endogenous cue. Neuropsychologia 47, 370–381. doi: 10.1016/j.neuropsychologia.2008. 09.011 Grandchamp, R., and Delorme, A. (2011). Single-trial normalization for event- related spectral decomposition reduces sensitivity to noisy trials. Front. Psychol. 2:236. doi: 10.3389/fpsyg.2011.00236 Cacioppo, J. T. (1994). Social neuroscience: autonomic, neuroendocrine and immune responses to stress. Psychophysiology 31, 113–128. doi: 10.1111/j.1469- 8986.1994.tb01032.x Grossmann, T., Johnson, M. H., Farroni, T., and Csibra, G. (2007). Social perception in the infant brain: gamma oscillatory activity in response to eye gaze. Soc. Cogn. Affect. Neurosci. 2, 284–291. doi: 10.1093/scan/nsm025 Cacioppo, J. T., Berntson, G. G., Sheridan, J. F., and McClintock, M. K. (2000). Multilevel integrative analyses of human behavior: social neuroscience and the complementing nature of social and biological approaches. Psychol. Bull. 126, 829–843. doi: 10.1037/0033-2909.126.6.829 Hadjikhani, N., Hoge, R., Snyder, J., and de Gelder, B. (2008). Pointing with the eyes: the role of gaze in communicating danger. Brain Cogn. 68, 1–8. doi: 10. 1016/j.bandc.2008.01.008 Carrick, O. K., Thompson, J. C., Epling, J. A., and Puce, A. (2007). It’s all in the eyes: neural responses to socially significant gaze shifts. Neuroreport 18, 763–766. doi: 10.1097/wnr.0b013e3280ebb44b Hardee, J. E., Thompson, J. C., and Puce, A. (2008). The left amygdala knows fear: laterality in the amygdala response to fearful eyes. Soc. Cogn. Affect. Neurosci. 3, 47–54. doi: 10.1093/scan/nsn001 March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 14 ERPs to changing gaze Rossi et al. Pernet, C. R., Chauveau, N., Gaspar, C., and Rousselet, G. A. (2011). LIMO EEG: a toolbox for hierarchical LInear MOdeling of ElectroEncephaloGraphic data. Comput. Intell. Neurosci. 2011:831409. doi: 10.1155/2011/ 831409 Herrmann, C. S., Crigutsch, M., and Busch, N. A. (2005). ‘‘EEC oscillations and wavelet analysis,’’ in Event-Related Potentials: A Methods Handbook, ed T. C. Handy (Cambridge, MA: MIT Press), 229–258. Hietanen, J. K., Leppänen, J. M., Nummenmaa, L., and Astikainen, P. (2008). Visuospatial attention shifts by gaze and arrow cues: an ERP study. Brain Res. 1215, 123–136. doi: 10.1016/j.brainres.2008.03.091 Peuskens, H., Vanrie, J., Verfaillie, K., and Orban, G. A. (2005). Specificity of regions processing biological motion. Eur. J. Neurosci. 21, 2864–2875. doi: 10. 1111/j.1460-9568.2005.04106.x Itier, R. J., and Batty, M. (2009). Neural bases of eye and gaze processing: the core of social cognition. Neurosci. Biobehav. Rev. 33, 843–863. doi: 10.1016/j. neubiorev.2009.02.004 Pfurtscheller, G., Stancák, A. Jr., and Neuper, C. (1996). Post-movement beta synchronization. References The many faces of the gamma band response to complex visual stimuli. Neuroimage 25, 491–501. doi: 10.1016/j.neuroimage. 2004.11.052 Rosati, A. G., and Hare, B. (2009). Looking past the model species: diversity in gaze-following skills across primates. Curr. Opin. Neurobiol. 19, 45–51. doi: 10. 1016/j.conb.2009.03.002 Latinus, M., and Taylor, M. J. (2005). Holistic processing of faces: learning effects with Mooney faces. J. Cogn. Neurosci. 17, 1316–1327. doi: 10. 1162/0898929055002490 Rossi, A., Parada, F. J., Kolchinsky, A., and Puce, A. (2014). Neural correlates of apparent motion perception of impoverished facial stimuli: a comparison of ERP and ERSP activity. Neuroimage 98, 442–459. doi: 10.1016/j.neuroimage. 2014.04.029 Latinus, M., and Taylor, M. J. (2006). Face processing stages: impact of difficulty and the separation of effects. Brain Res. 1123, 179–187. doi: 10.1016/j.brainres. 2006.09.031 Sagiv, N., and Bentin, S. (2001). Structural encoding of human and schematic faces: holistic and part-based processes. J. Cogn. Neurosci. 13, 937–951. doi: 10. 1162/089892901753165854 Le Van Quyen, M., Foucher, J., Lachaux, J.-P., Rodriguez, E., Lutz, A., Martinerie, J., et al. (2001). Comparison of Hilbert transform and wavelet methods for the analysis of neuronal synchrony. J. Neurosci. Methods 111, 83–98. doi: 10. 1016/s0165-0270(01)00372-7 Sauseng, P., Klimesch, W., Stadler, W., Schabus, M., Doppelmayr, M., Hanslmayr, S., et al. (2005). A shift of visual spatial attention is selectively associated with human EEG alpha activity. Eur. J. Neurosci. 22, 2917–2926. doi: 10.1111/j.1460- 9568.2005.04482.x Makeig, S., Debener, S., Onton, J., and Delorme, A. (2004). Mining event-related brain dynamics. Trends Cogn. Sci. 8, 204–210. doi: 10.1016/j.tics.2004.03.008 Maurer, U., Rossion, B., and McCandliss, B. D. (2008). Category specificity in early perception: face and word n170 responses differ in both lateralization and habituation properties. Front. Hum. Neurosci. 2:18. doi: 10.3389/neuro.09.018. 2008 Schuller, A. M., and Rossion, B. (2001). Spatial attention triggered by eye gaze increases and speeds up early visual activity. Neuroreport 12, 2381–2386. doi: 10.1097/00001756-200108080-00019 Müller, G. R., Neuper, C., Rupp, R., Keinrath, C., Gerner, H. J., and Pfurtscheller, G. (2003). Event-related beta EEG changes during wrist movements induced by functional electrical stimulation of forearm muscles in man. Neurosci. Lett. 340, 143–147. doi: 10.1016/s0304-3940(03)00019-3 Sedley, W., and Cunningham, M. O. (2013). Do cortical gamma oscillations promote or suppress perception? An under-asked question with an over- assumed answer. Front. Hum. Neurosci. 7:595. doi: 10.3389/fnhum.2013.00595 Srinivasan, R., Nunez, P. L., Tucker, D. M., Silberstein, R. B., and Cadusch, P. J. (1996). References Spatial sampling and filtering of EEG with spline laplacians to estimate cortical potentials. Brain Topogr. 8, 355–366. doi: 10.1007/bf01186911 Müller-Putz, G. R., Zimmermann, D., Graimann, B., Nestinger, K., Korisek, G., and Pfurtscheller, G. (2007). Event-related beta EEG-changes during passive and attempted foot movements in paraplegic patients. Brain Res. 1137, 84–91. doi: 10.1016/j.brainres.2006.12.052 Stanley, D. A., and Adolphs, R. (2013). Toward a neural basis for social behavior. Neuron 80, 816–826. doi: 10.1016/j.neuron.2013.10.038 Ochsner, K. N., and Lieberman, M. D. (2001). The emergence of social cognitive neuroscience. Am. Psychol. 56, 717–734. doi: 10.1037/0003-066x.56.9.717 Straube, T., Dietrich, C., Mothes-Lasch, M., Mentzel, H. J., and Miltner, W. H. R. (2010). The volatility of the amygdala response to masked fearful eyes. Hum. Brain Mapp. 31, 1601–1608. doi: 10.1002/hbm.20960 Oldfield, R. C. (1971). The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9, 97–113. doi: 10.1016/0028-3932(71) 90067-4 Tallon-Baudry, C., and Bertrand, O. (1999). Oscillatory gamma activity in humans and its role in object representation. Trends Cogn. Sci. 3, 151–162. doi: 10. 1016/s1364-6613(99)01299-1 Oram, M. W., and Perrett, D. I. (1994). Responses of Anterior Superior Temporal Polysensory (STPa) neurons to ‘‘Biological Motion’’ stimuli. J. Cogn. Neurosci. 6, 99–116. doi: 10.1162/jocn.1994.6.2.99 Tallon-Baudry, C., Bertrand, O., Delpuech, C., and Pernier, J. (1996). Stimulus specificity of phase-locked and non-phase-locked 40 Hz visual responses in human. J. Neurosci. 16, 4240–4249. Oram, M. W., and Perrett, D. I. (1996). Integration of form and motion in the anterior superior temporal polysensory area (STPa) of the macaque monkey. J. Neurophysiol. 76, 109–129. Thut, G., Nietzel, A., Brandt, S. A., and Pascual-Leone, A. (2006). α-band electroencephalographic activity over occipital cortex indexes visuospatial attention bias and predicts visual target detection. J. Neurosci. 26, 9494–9502. doi: 10.1523/jneurosci.0875-06.2006 Palva, S., and Palva, J. M. (2007). New vistas for alpha-frequency band oscillations. Trends Neurosci. 30, 150–158. doi: 10.1016/j.tins.2007.02.001 Frontiers in Human Neuroscience \| www.frontiersin.org March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 15 Rossi et al. ERPs to changing gaze Watanabe, S., Miki, K., and Kakigi, R. (2002). Gaze direction affects face perception in humans. Neurosci. Lett. 325, 163–166. doi: 10.1016/s0304- 3940(02)00257-4 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Zion-Golumbic, E., and Bentin, S. (2007). Dissociated neural mechanisms for face detection and configural encoding: evidence from N170 and induced gamma-band oscillation effects. Cereb. Cortex 17, 1741–1749. doi: 10. 1093/cercor/bhl100 Frontiers in Human Neuroscience \| www.frontiersin.org References The Guest Associate Editor Davide Rivolta declares that, despite having collaborated with author Ania Puce, the review process was handled objectively and no conflict of interest exists. Whalen, P. J., Kagan, J., Cook, R. G., Davis, F. C., Kim, H., Polis, S., et al. (2004). Human amygdala responsivity to masked fearful eye whites. Science 306:2061. doi: 10.1126/science.1103617 Copyright © 2015 Rossi, Parada, Latinus and Puce. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Worden, M. S., Foxe, J. J., Wang, N., and Simpson, G. V. (2000). Anticipatory biasing of visuospatial attention indexed by retinotopically specific alpha-band electroencephalography increases over occipital cortex. J. Neurosci. 20:RC63. Zion-Golumbic, E., and Bentin, S. (2007). Dissociated neural mechanisms for face detection and configural encoding: evidence from N170 and induced gamma-band oscillation effects. Cereb. Cortex 17, 1741–1749. doi: 10. 1093/cercor/bhl100 March 2015 \| Volume 9 \| Article 185 Frontiers in Human Neuroscience \| www.frontiersin.org 16
https://openalex.org/W3164817065	https://www.e3s-conferences.org/articles/e3sconf/pdf/2021/34/e3sconf_uesf2021_09078.pdf	English	null	Instrumental records received in 11 storey steel frame building during a remote earthquake	E3S web of conferences	2,021	cc-by	3,807	* Corresponding author: lapin_1956@list.ru 1JSC “Kazakh Research and Design Institute of Construction and Architecture”, Almaty, 21 Solodovnikova, Kazakhstan 2Satbayev University, 050013,Almaty, 22a Satbayev st,,Kazakhstan 1JSC “Kazakh Research and Design Institute of Construction and Architecture”, Almaty, 21 Solodovnikova, Kazakhstan 2Satbayev University, 050013,Almaty, 22a Satbayev st,,Kazakhstan Abstract. On August 9, 2017, a strong earthquake with a magnitude Mw = 6.3 occurred in the Xinjiang Uygur Autonomous Region of China. For the first time in the last 10 years, representative instrumental data were obtained at the seismic station No. 11 ("Prospect Abay") in the city of Almaty, equipped with a digital instrumentation and measuring system. The building is eleven-story and is made in a steel frame. In the city of Almaty, there is a significant number of residential and public buildings made in a steel frame. Using the AT 1105 sensors and the RSM-8 recorder, high-quality instrumental records of accelerations in the basement and on the roof of the building were obtained, as well as the spectral curves. The frequency composition of the impact at the basement level has been determined. The prevailing period of fluctuation at the basement level is 0.3 sec, and at the roof level is close to 1.3 sec. It was found that during a remote earthquake, resonant vibrations of the building took place. The results of the work can be used in the design of buildings with a steel frame in the city of Almaty. © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) https://doi.org/10.1051/e3sconf/202125809078 1 Introduction The seismic hazard of the city of Almaty is usually associated with the seismically active zones of the Northern Tien Shan. The areas of the Trans-Ili and Kungey Alatau ridges are especially highly seismic. There are the main seismogenic zones located close to the city - Almaty, Trans-Ili and Chiliko-Kemin. Their seismic potential is very high – the possible magnitudes of the strongest earthquakes with sources in these zones are, respectively, from 7.5 to 8.5, and the intensity of earthquakes in the city can reach from 8 to 10 points [1-4]. The city was destroyed twice by strong earthquakes - Vernenskoye in 1887 with a magnitude of 7.2 and Keminskoye, in 1911 with a magnitude of 8.2. Moreover, the last earthquake had a class of seismic disasters and was felt over an area of 2 million square kilometers. The city has repeatedly experienced less severe earthquakes. The Zhalanash- Tyup earthquake of 1978 with a magnitude of 6.9, the Suusamyr earthquake of 1992 with a E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) https://doi.org/10.1051/e3sconf/202125809078 magnitude of 7.3, the Baisorun earthquake of 1990, M = 6.3 can be noted. On the territory of the city, they were felt with an intensity of 5-6 points [3]. magnitude of 7.3, the Baisorun earthquake of 1990, M = 6.3 can be noted. On the territory of the city, they were felt with an intensity of 5-6 points [3]. For a long time, there was a firm conviction that there were no earthquake foci on the territory of the city. It was believed that faults located on the territory of the city could enhance or weaken the effect of the passing seismic waves of earthquakes with foci in other seismogenic zones outside the city. In the last 15 years, the emergence of earthquake foci was noted in the city, the magnitude of which was noted within 2-3. In the city of Almaty, there are 10 engineering seismometric stations placed on buildings, which also recorded earthquakes with foci in the city. The stations are equipped with digital and analogue systems. It is planned to modernize the network of stations using modern measuring systems [4-9]. The purpose of the study is to obtain information on the spectral characteristics of ground vibrations, as well as the dynamic characteristics of the building, based on the analysis of instrumental records of the earthquake. 2 Method and subject of research The analysis of instrumental records of various earthquakes, obtained by the engineering and seismometric service of Kazakh Scientific Research and Design Institute of Civil Engineering and Architecture, using digital and analog accelerometers, continues. g g g g g According to the operational data of the Data Center of the RSE IGI, on August 9, 2017 at 05:27 am Astana time (August 8 at 11:27 pm GMT) in the border region of China and Kazakhstan, 100 km from the Dostyk settlement, 360 km from the city of Taldykorgan, 480 km from the city of Almaty, an earthquake occurred. Epicenter coordinates: 44.44 degrees north latitude, 82.92 degrees east longitude. Magnitude Mw = 6.3. The earthquake was felt in Almaty and was 4 points in Taldykorgan. y p y g The depth of the earthquake source, according to various sources, is 15-30 km. This is the strongest recorded earthquake in Almaty in 2017. It should be noted that for the city of Almaty, earthquakes with foci near and even in the city are more typical. The earthquake was recorded by almost all stations of the engineering and seismometric service of Kazakh Scientific Research and Design Institute of Civil Engineering and Architecture. Instrumental recordings of the earthquake turned out to be very interesting. Station No. 11 "Prospect Abay" is located on an 11-storey office building made in a metal frame (Fig. 1). It should be noted that this station is one of the very first ones in Almaty and has been operating since 1970. In 2010, the first modernization of the station took place - a digital instrument-measuring system was installed. 2 2 E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) https://doi.org/10.1051/e3sconf/202125809078 SECTION Fig. 1. View and vertical section of the building. Fig. 2. Section in the horizontal plane. A G F E B C D G E A B C D F I H SECTION A G F E B C D N A G F E B C D Fig. 1. View and vertical section of the building. Fig. 2. Section in the horizontal plane. G E A B C D F I H Fig. 1. View and vertical section of the building. G E B C D F I H Fig. 1. View and vertical section of the building. Fig. 1. View and vertical section of the building. H Fig. 2. 2 Method and subject of research Section in the horizontal plane. E3S Web of Conferences 258, 09078 (2021) UESF-2021 https://doi.org/10.1051/e3sconf/202125809078 The second upgrade was carried out in 2016: the ADXL accelerometers were replaced by the AT 1105. y The soils are boulders. The soils are boulders. h b ildi h di i f h i h The building has dimensions of 15x36 m, height 43.6 m. The building has dimensions of 15x36 m, height 43.6 m. From a constructive point of view, the building is frame-panel. External wall panels made of aluminum with glass and stemolite filling. The frame consists of longitudinal and transverse closed frames with a transverse and longitudinal column pitch of 6 m. The height of the basement is 3.4 m, the first floor is 4.2 m and all subsequent ones are 3.6 m. Concrete tape foundations. The basement walls are made of monolithic concrete. The columns of the frame are anchored to the walls of the basement with the help of metal shoes. Ceilings over the first and basement floors are made of monolithic reinforced concrete; over the rest of the floors, the ceilings are made of prefabricated reinforced concrete ribbed slabs of 3 - 6 m in size. The partitions are made of prefabricated expanded clay concrete panels, as well as from precast gypsum concrete and slag concrete slabs. At the top of the building there is a technical floor (12th floor), where part of the control and measuring equipment is located - the upper registration point (mark 40.2 m). The lowest registration point is located in the basement (Fig. 1). The devices are located at an elevation of - 4.4m (the basement itself is at an elevation of -3.4m). Figure 2 shows a section of the building in the horizontal plane. Station 11 is equipped with an RSM-8 recorder installed in standby mode. This is the first domestic 8-channel digital system of this type. In 2016, ADXL sensors at station No. 1 "Institute" were replaced by AT 1105 sensors of Russian production. This made it possible to record even rather weak seismic events. Thus, the research methods are experimental - the instrumental records of the engineering-seismometric service station installed in the standby mode are analyzed. 3 Results Table 1 shows the maximum values of accelerations for each of the components of instrumental recordings, effective durations, values of the spectral coefficient, values of the period of the maximum of the spectrum. Lines 4-6 correspond to the basement of the building, 1-3 - roofs. The distribution of acceleration over the height of the building shows that the building is deformed in accordance with the first mode of vibration. Table 1. Maximum values of accelerations and parameters of accelerograms. Component Acceleration, cm/s2 Effective duration, sec Spectral coefficient Spectrum maximum period, sec 1 96-11-0-u1-OX 14.31 42.75 5.33(3.43) 0.42(1.15) 2 96-11-0-u2-OY 14.52 52.86 4.67(4.60) 1.29(0.40) 3 96-11-0-u3-OZ 5.41 65.66 6.07 0.16 4 96-11-0-u6-OX 4.82 29.96 4.13 0.29 5 96-11-0-u7-OY 5.64 35.26 4.67 0.30 6 96-11-0-u8-OZ 2.33 42.99 4.13 0.28 Table 1. Maximum values of accelerations and parameters of accelerograms. E3S Web of Conferences 258, 09078 (2021) UESF-2021 https://doi.org/10.1051/e3sconf/202125809078 The spectral curves were determined by integrating the differential equation of the classical one-mass oscillator using the standard solvers of the MATLAB computer mathematics system. The functions were used to solve stiff problems by the implicit trapezoidal method or the implicit Runge-Kutta method. Accelerogram sampling step 0.0064 s. The decrement of fluctuations in determining the spectral curves is taken equal to 5% of the critical value. Figures 3-4 show accelerograms at the basement and roof level along three orthogonal axes. Accelerations are measured in m / s2. According to the amplitudes of accelerations in the basement level, the intensity of the ground shaking in the area of the seismic station is approximately three-point. Accelerations in the horizontal plane, both in the basement level and in the roof, are quite close. Figure 5-6 shows the spectral curves at the basement and roof levels. According to the spectral composition at the basement level along the OX axis, the prevailing period is within 0.29 sec. On the OY axis, the impact is high frequency with a period of 0.3 sec. Taking into account that the accelerometers are located at an elevation of -4.4 m in the pit in the basement of the building, it should be noted that this is apparently the predominant period of vibration of the foundation soil (boulder). At the roof level, the frequency composition is quite complex. The spectral curves in the horizontal plane are double humped. 3 Results Apparently, the period of natural vibrations of the steel frame along both axes is within 1.15-1.29 sec. The spectra are quite broadband and very rugged. This is apparently determined by the significant magnitude of a distant earthquake with a magnitude of Mw = 6.3. On the vertical axis, the prevailing periods of fluctuations are 0.16-0.28 sec, increasing towards the last floor. The values of the spectral coefficient at the basement level are approximately the sam I h ld b d i ifi ff i d i f 30 65 ( bl 1) H The values of the spectral coefficient at the basement level are approximately the same. It should be noted a significant effective duration from 30 sec to 65 sec (table 1). Here, h ff i d i i h d i f ill i i h li d f l h lf h The values of the spectral coefficient at the basement level are approximately the same. It should be noted a significant effective duration from 30 sec to 65 sec (table 1). Here, the effective duration is the duration of oscillations with an amplitude of at least half the i It should be noted a significant effective duration from 30 sec to 65 sec (table 1). Here, the effective duration is the duration of oscillations with an amplitude of at least half the maximum. 5 5 E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) https://doi.org/10.1051/e3sconf/202125809078 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OX axis acceleration, m/s2 time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 acceleration, m/s2 on, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OY axis accelerogram on the OZ axis time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OX axis acceleration, m/s2 time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 Accelerograms of the earthquake of 08/09/2017 (station No. 3 Results 4 The discussion of the results Dozens of steel-framed civil and administrative buildings have been built in Almaty. These are buildings with a height of 10 to 18 floors. Therefore, it is very to study the behavior of such buildings in a real earthquake. Unfortunately building has an engineering and seismometric service station installed on building According to the results of the analysis of the accelerograms of the remote beta coefficient XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 curve on the OX axis curve on the OY axis curve on the OZ axis curve on the OX axis curve on the OY axis curve on the OZ axis period, s period, s beta coefficient Fig. 5. Spectral curves seismic events 08/09/2017 (basement). beta coefficient XINJIANG EARTHQUAKE ON AUGUST 9, 2017 curve on the OX axis curve on the OY axis curve on the OZ axis period, s Fig. 5. Spectral curves seismic events 08/09/2017 (basement). beta coefficient XINJIANG EARTHQUAKE ON AUGUST 9, 2017 curve on the OX axis curve on the OY axis curve on the OZ axis period, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 Fig. 5. Spectral curves seismic events 08/09/2017 (basement). Fig. 6. Spectral curves seismic events 08/09/2017 (roof). XINJIANG EARTHQUAKE ON AUGUST 9, 2017 curve on the OX axis curve on the OY axis curve on the OZ axis period, s beta coefficient XINJIANG EARTHQUAKE ON AUGUST 9, 2017 Fig. 6. Spectral curves seismic events 08/09/2017 (roof). 3 Results 11 "Prospect Abai" XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OX axis acceleration, m/s2 time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 acceleration, m/s2 acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OY axis accelerogram on the OZ axis time, s time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 acceleration, m/s2 accelerogram on the OY axis time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 acceleration, m/s2 acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OZ axis time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 Fig. 3. Accelerograms of the earthquake of 08/09/2017 (station No. 11 "Prospect Abai", roof). 6 E3S Web of Conferences 258, 09078 (2021) UESF-2021 https://doi.org/10.1051/e3sconf/202125809078 acceleration, m/s2 acceleration, m/s2 m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OX axis accelerogram on the OY axis accelerogram on the OZ axis time, s time, s acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OX axis time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 Accelerograms of the earthquake of 08/09/2017 (station No. 1 "Prospect Abay", basem acceleration, m/s2 acceleration, m/s2 acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OX axis accelerogram on the OY axis accelerogram on the OZ axis time, s time, s time, s 7 Fig. 4. Accelerograms of the earthquake of 08/09/2017 (station No. 1 "Prospect Abay", basement). acceleration, m/s2 acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OY axis accelerogram on the OZ axis time, s time, s time, s time, s acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OY axis time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 acceleration, m/s2 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 accelerogram on the OZ axis time, s XINJIANG EARTHQUAKE ON AUGUST 9, 2017 XINJIANG EARTHQUAKE ON AUGUST 9, 2017 Fig. 4. Accelerograms of the earthquake of 08/09/2017 (station No. 1 "Prospect Abay", basement). 7 7 E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) https://doi.org/10.1051/e3sconf/202125809078 Fig. 5. Spectral curves seismic events 08/09/2017 (basement). Fig. 6. Spectral curves seismic events 08/09/2017 (roof). 5 Conclusion 1. Seismic event (earthquake in Xinjiang Uygur Autonomous Region, China) on August 9, 2017 on the territory of Almaty in the area of seismic station No. 11 "Prospect Abai" is an earthquake with an intensity of at least 3 points (3-4 points) . Accelerations in the horizontal plane at the roof level are approximately equal to the same values during the earthquake of August 16, 2014 [16]. 2. Acceleration in the horizontal plane at the basement level is practically the same. The period of the spectrum maximum characterizes the frequency characteristics of the foundation soil and has a value of 0.3 sec. 3. Accelerations in the horizontal plane are 3-4 times less than during the 5-6 point Zhalanashtyup earthquake on March 25, 1978, recorded at the same station. 3. Accelerations in the horizontal plane are 3-4 times less than during the 5-6 point Zhalanashtyup earthquake on March 25, 1978, recorded at the same station. 4. At the level of the roof on the horizontal plane, spectral curves with two maxima. 5. The digital instrumentation and measuring system RSM-8 with AT 1105 accelerometers is an effective means of registering earthquakes. Seismic station No. 11 "Prospect Abai" is in working order. 6. Accelerometers AT 1105 should be used in the modernization of digital instrumentation and measuring systems of the network of ISS stations of KazRDICA JSC. 6. Accelerometers AT 1105 should be used in the modernization of digital instrumentation and measuring systems of the network of ISS stations of KazRDICA JSC. 4 The discussion of the results Dozens of steel-framed civil and administrative buildings have been built in the city of Almaty. These are buildings with a height of 10 to 18 floors. Therefore, it is very important to study the behavior of such buildings in a real earthquake. Unfortunately, only one building has an engineering and seismometric service station installed on buildings. According to the results of the analysis of the accelerograms of the remote earthquake of August 9, 2017, the oscillation period of an 11-storey building with a steel frame of 1.29 - 1.30 sec can be considered established. It can be predicted that buildings with a height of up to 18 floors will have an oscillation period of presumably up to 1.5 seconds. It should also be noted that this 11-storey building is located between two tectonic faults, the distance to which is from 700 m to 900 m. Therefore, the building is a very 8 E3S Web of Conferences 258, 09078 (2021) UESF-2021 E3S Web of Conferences 258, 09078 (2021) https://doi.org/10.1051/e3sconf/202125809078 important object for observation, as it will allow to assess the influence of tectonic faults on the deformativity of the building. important object for observation, as it will allow to assess the influence of tectonic faults on the deformativity of the building. An important result is the establishment of the fact of resonant vibrations of the building during remote earthquakes. It should be noted that high-frequency vibrations of the foundation soil are predicted for earthquakes near the city of Almaty. Thus, even a weak but tangible earthquake delivers information about both the dynamic characteristics of the building and the spectral characteristics of ground motion. It is necessary to continue the installation of engineering and seismometric service stations on high-rise buildings, on buildings with new constructive solutions, including objects with various seismic isolation systems [10-15] and buildings with pile foundations [17]. Acknowledgements The research was carried out using the funds of the grant AP 05130702 of the Ministry of Education and Science of the Republic of Kazakhstan. 4. A.B. Sadykova, N.V. Silacheva, N.P. Stepanenko, Series of geology and technical sciences 1(445), 127-134 (2021) https://doi.org/10.32014/2021.2518-170X.18 References 1. N.N. Mikhaylova, I.N. Sokolova, A.E. Velikanov, A.N. Sokolov, Bulletin of the National Nuclear Center 3, 87-93 (2015) 2. N.N. Mikhailova, A.N. Uzbekov, News of the national academy of sciences of the republic of Kazakhstan. Series of geology and technical sciences 3(429), 146-155 (2018) ISSN 2518-170X, http://www.geolog-technical.kz/images/pdf/g20183/146- 155.pdf 3. R.M. Galperin, I.L. Nersesov, E.I. Galperin, Seismic regime of Almaty for 1972-1982 years (Science, Moscow, 1985) 3. R.M. Galperin, I.L. Nersesov, E.I. Galperin, Seismic regime of Almaty for 1972-1982 years (Science, Moscow, 1985) 4. A.B. Sadykova, N.V. Silacheva, N.P. Stepanenko, Series of geology and technical sciences 1(445), 127-134 (2021) https://doi.org/10.32014/2021.2518-170X.18 4. A.B. Sadykova, N.V. Silacheva, N.P. Stepanenko, Series of geology and technical sciences 1(445), 127-134 (2021) https://doi.org/10.32014/2021.2518-170X.18 9 9 E3S Web of Conferences 258, 09078 (2021) UESF-2021 https://doi.org/10.1051/e3sconf/202125809078 5. M. Deng, M. Wang, S. Jin, Q. Zhang, K. Chen, S.A. Cheng, Geoscientific Instrumentation Methods and Data Systems 8, 13-19 (2019) https://doi.org/10.5194/gi- 8-13-2019 6. X. Zhang, Q. Zhang, M. Wang, et al., Geoscientific Instrumentation Methods and Data Systems 6, 495-503 (2017) https://doi.org/10.5194/gi-6-495-2017 7. L. Yang, L. Stehly, A. Paul, Geophysical Journal International 214(2), 1136-1150 (2018) https://doi.org/10.1093/gji/ggy188 8. S. Qiao, H. Duan, Q. Zhang et al., Geoscientific Instrumentation Methods and Data Systems 7, 253-263 (2018) https://doi.org/10.5194/gi-7-253-2018 9. L. Groukamp, J.J. Rossow, C.J.S. Fourie, L. Cormbrinck, African Journal of Geology 114(3/4), 573-576 (2011) https://doi.org/10.2113/ gssajg.114.3-4.573 10. A.F. Bulat, V.I. Dyrda, M.I. Lysytsya, S.M. Grebenyuk, Strength of Materials 50(3), 387–395 (2018) http://doi.org/10.1007/s11223-018-9982-9 11. S. Ahmad, F. Ghani, A. Md Raghib, Construction & Building Materials 23, 146-152 (2009) https://doi.org/10.1016/j.conbuildmat.2008.01.012 12. V. I. Dyrda, A. Kobets, A.F. Bulat, et al., E3S Web of conference 168, 00042 (2020) https://doi.org/10.1051/e3sconf/202016800042 13. V. Lapin, 16th World Conference on Seismic Isolation, Energy Dissipation on Active Vibration Conference of Structures. 2020, p.325-332. http://doi.org/10.13753./2686- 7974-2019-16-482-527 14. A.F. Bulat, V.I. Dyrda, S.N. Grebenyuk, M.I. Klimenko, Strength of Materials 51(2), 183-192 (2019) https://doi.org/10.1007/s11223-019-00064-x 15. S. Yerzhanov, 16th World Conference on Seismic Isolation, Energy Dissipation on Active Vibration Conference of Structures, 955-963 (2020) http://doi.org/10.13753./2686-7974-2019-16-919-927 16. S.Y. Yerzhanov, V.P. Daugavet, V.A. Lapin, The collection "Researches of earthquake resistance of building and construction" 23(33), 145-163 (2015) 17. A.Zh. Zhussupbekov, J. Frankovská, J. Stacho, et al., 15th Asian Regional Conference on Soil Mechanics and Geotechnical Engineering, ARC 2015: New Innovations and Sustainability, 2704-2709 (2015)_https://doi.org/10.3208/jgssp.TC305-05 10
https://openalex.org/W4376279865	https://zenodo.org/record/7976751/files/20.pdf	English	null	GRAFTS AND GRAFTING TECHNIQUES IN IMPLANTS - A REVIEW	International journal of advanced research	2,023	cc-by	11,921	ISSN: 2320-5407 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Journal Homepage: -www.journalijar.com Article DOI: 10.21474/IJAR01/16729 DOI URL: http://dx.doi.org/10.21474/IJAR01/16729 Introduction:- t oduct o : Repairing and restoring bone defects has a long history. The surgeons of the pre-Incan period used gold and silver plates and shells as grafting materials to repair trephine holes in the skull 3000 years BC. Trephination – the removal of a circular section of bone from the skull – is the oldest known surgical intervention. Dr Philip Von Walter is credited in 1821 as being the first surgeon to use bone autografts for reconstruction of the remaining defect after a trephination in the skull1. The term „autograft‟ implies transplantation of bone tissue from one site to another, within the same individual. Bone transplantation has been a common surgical procedure since the early 1920s1. Causes of ridge defects: 1. Surgery 2. Trauma 3. Infection or 4. Congenital malformations. Causes of ridge defects: 1. Surgery 2. Trauma 3. Infection or 4. Congenital malformations. 4. Congenital malformations. GRAFTS AND GRAFTING TECHNIQUES IN IMPLANTS - A Review Dr. M. Harikrishna, Dr. K. Bhargavi, Dr. Y. Ravi Shankar, Dr. K. Srinivas, Dr. G. Vasavi Prasanna and Dr. Ch. Roja ……………………………………………………………… Bone graft is the gold standard materials used for repairing bony defects. There are various bone graft materials/biomaterials and methods available for bone augmentation. The present article explains in detail about bone grafts and the recent advancements in bone graft materials as well as bone augmentation in implant dentistry.The literature search was performed for relevant articles on bone grafts for dental implants in PubMed/Medline, Scopus, Google Scholar, and Science Direct published in English literature published between January 2005 and March 2023. Relevant studies on bone grafts for dental implants were also included and critically analyzed in this review. Various biomaterials can be used to augment bone for implant placement. Each graft procedure has its own advantages and disadvantages in each clinical application and should choose the graft material carefully to attain a high success rate and less morbidity. Copy Right, IJAR, 2023,. All rights reserved. Copy Right, IJAR, 2023,. All rights reserved. Corresponding Author:- Dr. K.Bhargavi Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 5. The insufficient quantity of bone is due to tooth loss which results in rapid resorption of alveolar bone due to lack of intraosseous stimulation by periodontal ligament (PDL) fibers, for example, pneumatization of maxillary sinus following tooth loss. 5. The insufficient quantity of bone is due to tooth loss which results in rapid resorption of alveolar bone due to lack of intraosseous stimulation by periodontal ligament (PDL) fibers, for example, pneumatization of maxillary sinus following tooth loss. Definition Bone Graft is defined as the augmentation or replacement of the portion of the maxillary or mandibular bone with an osteogenic material - GPT 9. Bone grafting is a surgical procedure that replaces missing bone with material from patient‟s own body, an artificial, synthetic, or natural substitute. Bone grafting is possible because bone tissue has the ability to regenerate completely if provided the space into which it has to grow. As natural bone grows, it generally replaces the graft material completely, resulting in a fully integrated region of new bone2.Therefore, for managing thin ridges to ridges with insufficient height and dealing with soft tissue malformation to produce brilliant pink aesthetics,grafting play a major role in the prognosis and outcome of dental implant therapy. Classifications Classification 2 Grafts are classified based on: 1. Structure 2. Activity of the graft 3. Type or source of the graft material Based on the Structures They can be classified as: 1. Particulate graft 2. Block graft 3. Bone putty Based on the activity 1. Osseoconductive 2. Oseeoinductive 3. Osteogenic Based on the source  Natural source: - Autogenous graft  Extraoral:  Iliac crest  Tibula  Fibula  Ribs  Intraoral  Symphysis  Ramus  Maxillary tuberosity  Exostosis - Isograft - Allograft  DFDBA  FDBA  Frozen graft  Xenograft  Bovine-derived  Porcine-derived  Coralline calcium carbonate Classification 2 Grafts are classified based on: 3. Type or source of the graft material Goals of osseous replacement : Goals of osseous replacement : 1. Maintenance of contour, 2. Elimination of dead space, 3. Reduce postoperative infection 4. Enhance bony and soft tissue healing. Corresponding Author:- Dr. K.Bhargavi 793 ISSN: 2320-5407 Based on the structure: Particulate graft g Small particles of autogenous grafts or synthetic graft material. g Small particles of autogenous grafts or synthetic graft material. They can either be harvested as small particles or as a block and then be ground to smaller particles. The particles are mixed with blood or blood components to improve the uptake and nutrient supply or graft3. They can either be harvested as small particles or as a block and then be ground to smaller particles. The particles are mixed with blood or blood components to improve the uptake and nutrient supply or graft3. The particles are mixed with blood or blood components to improve the uptake and nutrient supply or graft3. les are mixed with blood or blood components to improve the uptake and nutrient supply or graft3. Based on the Structures They can be classified as: 794 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407  Synthetic sources:  Alloplasts:  Hydroxyapatite  Calcium phosphate cements  Beta-tricalcium phosphate  Biphasic alloplastic materials  Bioactive glasses  Synthetic polymer  Ceramic and bioactive molecules  Synthetic sources:  Alloplasts:  Hydroxyapatite  Calcium phosphate cements  Beta-tricalcium phosphate  Biphasic alloplastic materials  Bioactive glasses  Synthetic polymer  Ceramic and bioactive molecules  Ceramic and bioactive molecules Classification of bone grafts based on material groups: Classification of bone grafts based on material groups: 1. Allograft‑based bone graft involves allograft bone, used alone or in combination with other materials (e.g., Grafton, Orthoblast). 1. Allograft‑based bone graft involves allograft bone, used alone or in combination with other materials (e.g., Grafton, Orthoblast). 2. Factor‑based bone graft are natural and recombinant growth factors, used alone or in combination with other materials such as transforming growth factor‑beta (TGF‑beta), platelet‑derived growth factor (PDGF), fibroblast growth factors (FGF), and bone morphogenic protein (BMP). 2. Factor‑based bone graft are natural and recombinant growth factors, used alone or in combination with other materials such as transforming growth factor‑beta (TGF‑beta), platelet‑derived growth factor (PDGF), fibroblast growth factors (FGF), and bone morphogenic protein (BMP). 3. Cell‑based bone grafts use cells to generate new tissue alone or are added onto a support matrix, for example, mesenchymal stem cells. 4. Ceramic‑based bone graft substitutes include calcium phosphate, calcium sulphate, and bioglass used alone or in combination; for example, OsteoGraf, ProOsteon, OsteoSet. 5. Polymer‑based bone graft uses degradable and nondegradable polymers alone or in combination with other materials, for example, open porosity polylactic acid polymer. 5. Polymer‑based bone graft uses degradable and nondegradable polymers alone or in combination with other materials, for example, open porosity polylactic acid polymer. Discussion:- Discussion:- Discussion:- Figure 1:- Types of Ordinary or activated bone grafts. Figure 1:- Types of Ordinary or activated bone grafts. Classification 2: Classification of bone grafts based on material groups: Based on the structure: Particulate graft Based on the structure: P i l f Advantages The material being pliable can be molded into any desired shape. I b k d i i h l hi The material being pliable can be molded into any desired shape. It can be packed into spaces without leaching. The material being pliable can be molded into a It can be packed into spaces without leaching. The material being pliable can be molded into any desired shape It can be packed into spaces without leaching. g p y It can be packed into spaces without leaching. Based on the activity of the graft The biologic mechanisms that provide a rationale for bone grafting are osteoconduction, osteoinduction, and osteogenesis. Osteoconduction Occurs when bone graft material serves as a scaffold for new bone growth, which is perpetuated by the native bone. Osteoblasts from the margin of defect that is being grafted, utilize the bone graft material as a framework upon which to spread and generate new bone. In the very least, a bone graft material should be osteoconductive. Disadvantages Price of the material is expensive Being a new material, long-term clinical trials are not available. Sinus lift surgeries Bone putty Clay like graft material comprising of multiple combinations of their synthetic or natural bone grafts in a resorbable matrix. Bone putty Clay like graft material comprising of multiple combinations of their synthetic or natural bone grafts in a resorbable matrix. Basically composed of demineralized bone, sodium hyaluronate along with beta tricalcium phosphate and hydroxyapatite. ISSN: 2320-5407 Contraindications Large horizontal defects Large defects Block grafts Block grafts are chunks of graft material that can be used to increase the height and width of the available bone The block can be stabilized with stainless steel or titanium screws. Autogenous block can be harvested from: Intraoral sites: Symphysis, ramus of the mandible. Extraoral sites: Iliac crest, fibula, tibia and ribs. Indications Vertical height augmentation Increasing width of bone area Contraindication Sinus lift surgeries Bone putty Clay like graft material comprising of multiple combinations of their synthetic or natural bone graft resorbable matrix. Basically composed of demineralized bone, sodium hyaluronate along with beta tricalcium phosphat hydroxyapatite. Advantages The material being pliable can be molded into any desired shape. It can be packed into spaces without leaching. Disadvantages Price of the material is expensive Being a new material, long-term clinical trials are not available. Based on the activity of the graft The biologic mechanisms that provide a rationale for bone grafting are osteoconduction, osteoinduction, osteogenesis. 3-walled and 2-walled defects. Indirect and direct sinus lift. Contraindications Large horizontal defects Large defects Block grafts Block grafts are chunks of graft material that can be used to increase The block can be stabilized with stainless steel or titanium screws. Autogenous block can be harvested from: Intraoral sites: Symphysis, ramus of the mandible. Extraoral sites: Iliac crest, fibula, tibia and ribs. Indications Vertical height augmentation Increasing width of bone area Contraindication Sinus lift surgeries Bone putty Clay like graft material comprising of multiple combinations o resorbable matrix. Basically composed of demineralized bone, sodium hyaluronate hydroxyapatite. Advantages The material being pliable can be molded into any desired shape. It can be packed into spaces without leaching. Disadvantages Price of the material is expensive B i i l l li i l i l il bl 3-walled and 2-walled defects. Indirect and direct sinus lift. Contraindications Block grafts Block grafts are chunks of graft material that can be used to increase the height and width of the available bone. Block grafts Block grafts are chunks of graft material that can be used to increase the height and width of the available bone. The block can be stabilized with stainless steel or titanium screws. Autogenous block can be harvested from: Indications Minor fenestrations and dehiscence. Minor fenestrations and dehiscence. 795 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 ISSN: 2320-5407 3-walled and 2-walled defects. Indirect and direct sinus lift. Contraindications Large horizontal defects Large defects Block grafts Block grafts are chunks of graft material that can be used to increase the height a The block can be stabilized with stainless steel or titanium screws. Autogenous block can be harvested from: Intraoral sites: Symphysis, ramus of the mandible. Extraoral sites: Iliac crest, fibula, tibia and ribs. Indications Vertical height augmentation Increasing width of bone area Contraindication Sinus lift surgeries Bone putty Clay like graft material comprising of multiple combinations of their syn resorbable matrix. Basically composed of demineralized bone, sodium hyaluronate along wi hydroxyapatite. Advantages The material being pliable can be molded into any desired shape. It can be packed into spaces without leaching. Disadvantages Price of the material is expensive Being a new material, long-term clinical trials are not available. Based on the activity of the graft The biologic mechanisms that provide a rationale for bone grafting are ost osteogenesis. Osteoconduction Occurs when bone graft material serves as a scaffold for new bone growth, whic Osteoblasts from the margin of defect that is being grafted, utilize the bone graft m to spread and generate new bone. In the very least, a bone graft material should be O t i d ti 3-walled and 2-walled defects. Indirect and direct sinus lift. Contraindications Large horizontal defects Large defects Block grafts Block grafts are chunks of graft material that c The block can be stabilized with stainless stee Autogenous block can be harvested from: Intraoral sites: Symphysis, ramus of the mandible. Extraoral sites: Iliac crest, fibula, tibia and ribs. Indications Vertical height augmentation Increasing width of bone area Contraindication Sinus lift surgeries Bone putty Clay like graft material comprising of mu resorbable matrix. Basically composed of demineralized bone hydroxyapatite. Advantages The material being pliable can be molded into It can be packed into spaces without leaching. Disadvantages Price of the material is expensive Being a new material, long-term clinical trials Based on the activity of the graft The biologic mechanisms that provide a ra osteogenesis. Osteoconduction Occurs when bone graft material serves as a s Osteoblasts from the margin of defect that is b to spread and generate new bone. In the very le 3-walled and 2-walled defects. Indirect and direct sinus lift. Osteopromotion p Involves enhancement of osteoinduction without possession of osteoinductive properties. For example, enamel matrix derivative enhances the osteoinductive effect of demineralized freeze‑dried bone allograft (DFDBA), but will not stimulate bone growth alone. Osteoinduction Involves stimulation of osteoprogenitor cells to differentiate into osteoblasts and then begins formation of new bone. The most widely studied type of osteoinductive cell mediators are BMPs. A bone graft material that is osteoconductive and osteoinductive will not only serve as a scaffold for currently existing osteoblasts but will also trigger formation of new osteoblasts, promoting faster integration of the graft4. 796 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 Isograft g Isografts are grafts harvested from an individual of the same species with the same genotype but different phenotype; for example, an identical twin. Isografts have a good success rate. Based on the source Natural source Autograft g Autologous or autogenous bone grafting involves utilizing bone obtained from same individual receiving the graft. Bone can be harvested from nonessential bones, such as from iliac crest, mandibular symphysis (chin area), and anterior mandibular ramus (coronoid process). When a block graft will be performed, autogenous bone is the most preferred because there is less risk of graft rejection as the graft is originated from the patient‟s body. It would be osteoinductive and osteogenic, as well as osteoconductive. Disadvantage of autologous grafts is that additional surgical site is required, another potential location for postoperative pain and complications5. All bones require blood supply in the transplanted site. Depending on where the transplant site is and size of the graft, an additional blood supply may be required. For these types of grafts, extraction of the part of the periosteum and accompanying blood vessels along with the donor bone is required. This kind of graft is known as a free flap graft. Osteogenesis It occurs when vital osteoblasts originating from bone graft material contributes to the growth of new bone along with bone formation. Xenograft 1. Xenografts are sourced from animals. 1. Xenografts are sourced from animals. 2. The sources of the graft can be: 3. Bovine- derived 4. Porcine – derived 5. Coralline calcium carbonate (non- resorbable) 6. Calcifying marine algae: Corallina officinalis (Flourohydroxyapatite) 6. Calcifying marine algae: Corallina officinalis (Flourohydroxyapatite) Allografts Allograft is derived from humans. The difference is that allograft is harvested from an individual other than the one receiving the graft. Allograft bone is taken from cadavers that have donated their bone so that it can be used for living people who are in need of it; it is typically sourced from a bone bank. There are three types of bone allograft available: (Figure 2) There are three types of bone allograft available: (Figure 2) 1. Fresh or fresh‑frozen bone 2. FDBA 3. DFDBA 1. Fresh or fresh‑frozen bone The use of allografts for bone repair often requires sterilization and deactivation of proteins normally found in healthy bone. Contained in the extracellular matrix of bone tissue are the full cocktail of bone growth factors, proteins, and other bioactive materials necessary for osteoinduction (Table 1) and successful bone healing; the desired factors and proteins are removed from the mineralized tissue by using a demineralizing agent such as hydrochloric acid. The mineral content of the bone is degraded, and the osteoinductive agents remain in a demineralized bone matrix (DBM)6. 797 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 ISSN: 2320-5407 Figure 2:- Manufacturing of FDBA and DFDBA. Figure 2:- Manufacturing of FDBA and DFDBA. Table 1:- Based on the activity of the graft. Type Osseoconductive Osseoinductive Osseogenic Alloplast ✔ x X Xenograft ✔ ✔/ X X Allograft ✔ ✔ X Autograft ✔ ✔ ✔ Table 1:- Based on the activity of the graft. Table 1:- Based on the activity of the graft. Synthetic variants y Flexible hydrogel‑hydroxyapatite (HA) composite which has a mineral to organic matrix ratio, approximating that of human bone.Artificial bone can be created from ceramics such as calcium phosphates (e.g., HA and tricalcium phosphate), bioglass and calcium sulphate are biologically active depending on solubility in physiological environment.5 These materials combine with growth factors, ions such as strontium or mixed with bone marrow aspirate to increase biological activity. The presence of elements such as strontium can result in higher bone mineral density (BMD) and enhanced osteoblast proliferation. Bioactive glass(bioglass) It is a biologically active silicate‑based glass, having high modulus and brittle nature; it has been used in combination with polymethylmethacrylate to form bioactive bone cement and with metal implants as a coating to form a calcium‑deficient carbonated calcium phosphate layer which facilitates the chemical bonding of implants to the surrounding bone. Different types of calcium phosphates are tricalcium phosphate, synthetic hydroxyapatite, and coralline hydroxyapatite; available in pastes, putties, solid matrices, and granules8.Such calcium phosphates products include Bio‑Oss and OsteoGraft. Both products use hydroxyapatite, either as a particulate (Bio‑Oss) or as blocks and particulates (OsteoGraft). Pro‑Osteon is a unique product based on sea coral, which is converted from calcium carbonate to calcium hydroxyapatite. The advantage of this material is that the structure of coral, which is similar to that of trabecular bone. Polymer‑based bone graft substitutes This can be divided into natural polymers and synthetic polymers. Subclassified into degradable and nondegradable types. Polymer‑based bone graft substitutes include the following: Healos is a natural polymer‑based product, a polymer‑ceramic composite consisting of collagen fibers coated with hydroxyapatite and indicated for spinal fusions. Healos is a natural polymer‑based product, a polymer‑ceramic composite consisting of collagen fibers coated with hydroxyapatite and indicated for spinal fusions. Cortoss is an injectable resin‑based product with applications for load‑bearing sites9. Cortoss is an injectable resin‑based product with applications for load‑bearing sites9. Alloplastic grafts Alloplastic grafts may be made from hydroxyapatite, a naturally occurring mineral (main mineral component of bone), made from bioactive glass. Hydroxyapatite is a synthetic bone graft, which is the most used now due to its osteoconduction, hardness, and acceptability by bone. Some synthetic bone grafts are made of calcium carbonate, which start to decrease in usage because it is completely resorbable in short time and makes breaking of the bone easier. Finally used is the tricalcium phosphate in combination with hydroxyapatite and thus giving effect of both, osteoconduction and resorbability7. 798 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Growth factors Growth factors enhanced grafts are produced using recombinant DNA technology. They consist of either human growth factors or morphogens (BMPs in conjunction with a carrier medium, such as collagen).The factors and proteins that exist in bone are responsible for regulating cellular activity. Growth factors bind to receptors on cell surfaces and stimulate intracellular environment to act. Generally, this activity translates to a protein kinase that induces a series of events resulting in transcription of messenger ribonucleic acid (mRNA) and ultimately into the formation of a protein to be used intracellularly or extracellularly. Some of Growth factors are TGF‑beta, insulin like growth factors I and II, PDGF, FGF, and BMPs. Cell‑based bone graft substitutes: Stem cells are cultured in the presence of various additives such as dexamethasone, ascorbic acid, and glycerophosphate to direct the undifferentiated cell towards osteoblast lineage.The addition of TGF‑beta and BMP‑2, BMP‑4, and BMP‑7 to the culture media can also influence the stem cells towards osteogenic lineage. Mesenchymal stem cells have also been seeded onto bioactive ceramics conditioned to induce differentiation to osteoblasts8. Ceramic‑based bone graft substitutes Majority of bone grafts available involve ceramics, either alone or in combination with another material (e.g., calcium sulfate, bioactive glass, and calcium phosphate). The use of ceramics, like calcium phosphates is calcium hydroxyapatite which is osteoconductive and osteointegrative; and in some cases, osteoinductive. They require high temperatures for scaffold formation and have brittle properties. 1. Calcium sulfate is also known as Plaster of Paris. It is biocompatible, bioactive, and resorbable after 30‑60 days. Significant loss of its mechanical properties occurs upon its degradation; therefore, it is a questionable choice for load‑bearing applications: 1. Calcium sulfate is also known as Plaster of Paris. It is biocompatible, bioactive, and resorbable after 30‑60 days. Significant loss of its mechanical properties occurs upon its degradation; therefore, it is a questionable choice for load‑bearing applications: g pp 2. OsteoSet is a tablet used for defect packing. It is degraded in approximately 60 days. 3. Allomatrix is Osteoset combined with DBM, forms a putty or injectable paste. OsteoSet is a calcium sulfate tablet used for bone defect sites, whereas allomatrix is a combination of calcium sulfate and DBM that forms an injectable paste or fable putty. 3. Allomatrix is Osteoset combined with DBM, forms a putty or injectable paste. OsteoSet is a calcium sulfate tablet used for bone defect sites, whereas allomatrix is a combination of calcium sulfate and DBM that forms an injectable paste or fable putty. Expanded PTFE (E-PTFE) p 1. It is manufactured by heating PTFE and then applying a force which causes an expansion of the microstructure. This creates pores in its structure. 1. It is manufactured by heating PTFE and then applying a force which causes an expansion of the microstructure. This creates pores in its structure. 1. It is manufactured by heating PTFE and then applyi microstructure. This creates pores in its structure. 2. E-PTFE membranes have two microstructures: a. Coronal portion and b. Occlusive portion Polytetrafluoroethylene(PTFE) • This was the first non-resorbable membrane to event to be used clinically (1984). Polytetrafluoroethylene (PTFE) is an inert, biocompatible membrane that prevents tissue ingrowth. There are different types of PTFE membranes that have been developed for use in grafting. They are: • This was the first non-resorbable membrane to event to be used clinically (1984). Polytetrafluoroethylene (PTFE) is an inert, biocompatible membrane that prevents tissue ingrowth. There are different types of PTFE membranes that have been developed for use in grafting. They are: 1. Expanded PTFE (E-PTFE) 2. High- density PTFE (D-PTFE) 3. Titanium-reinforced PTFE Biphasic alloplastic materials 1. These materials are referred to as biphasic because they are made of two different microstructural phases. For example, biphasic calcium phosphate consists of Hydroxyapatite (HA) and Beta- Tricalcium Phosphate (TCP) in different ratios. 2. Biphasic calcium phosphate also has a negative surface charge (Zeta potential) that has been found to have favourable bone formation characteristics. The negative charge accelerates bone growth. 799 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Degradable synthetic polymers, like natural polymers are resorbed by the body. The benefit of having the implant resorbed by the body is that, the body is able to heal itself completely without remaining foreign bodies. (Figure 3). Figure 3:- Bone grafts and its substitute materials in Dentistry. Degradable synthetic polymers, like natural polymers are resorbed by the body. The benefit of having the implant resorbed by the body is that, the body is able to heal itself completely without remaining foreign bodies. (Figure 3). Figure 3:- Bone grafts and its substitute materials in Dentistry. Barrier Membranes A barrier membrane is used primarily in implant, oral and periodontal surgery to prevent the epithelium from growing into an area where bone growth is desired. This method of preventing epithelial migration into a specific area is known as guided tissue regeneration (GTR). Membranes are classified into non-resorbable and resorbable products10. Non-resorbable Membranes Non-resorbable membranes, as the name suggests, do not resorb on their own and require a second surgical procedure for removal. High-density PTFE (D-PTFE) D- PTFE was developed in 1993, and is made of high density PTFE, with a submicron (0.2)  It also facilitates easy removal of the membrane without damaging the underlying structures. Occlusive portion: p Here the internodal distance is less than 8 pm to allow nutrient inflow. It also prevents the infiltration of other tissue cell types. These pores can harbour bacteria and limit the amount of bone regrowth that can occur. It is currently not being used for dental surgeries. However, it is being used extensively for digestive, cerebral and cardiovascular surgeries11. g 1. These pores can harbour bacteria and limit the amount of bone regrowth that can occur. p currently not being used for dental surgeries. How p 2. It is currently not being used for dental surgeries. However, 2. It is currently not being used for dental surgeries. However, 2. It is currently not being used for dental surgeries. However, 3. It is being used extensively for digestive, cerebral and cardiovascular surgeries. Titanium- reinforced PTFE - Titanium reinforcement of PTFE allows shaping of these membranes and can be used for socket preservation procedures Titanium mesh Titanium meshes were introduced in 1969 by Boyne and were mainly used to treat continuity defects in the maxilla and mandible. The rigidity of titanium aids in space maintenance and prevents collapse of the contour.Titanium meshes also: Prevent graft displacement, permit bending, contouring, andadaptation to anybony defect.Some of the drawbacks of using titanium are: Increased stiffness of titanium sometimes leads to mucosal irritation (Figure 4) Figure 4:- (A) Use of a titanium mesh as a structural scaffold and physical barrier in GBR for prevention of soft tissue cell migration and promotion of bone regeneration. (B) Use of a barrier membrane in GBR. Figure 4:- (A) Use of a titanium mesh as a structural scaffold and physical barrier in GBR for prevention of soft tissue cell migration and promotion of bone regeneration. (B) Use of a barrier membrane in GBR. Coronal portion: p Coronal portion is fixed which has an internodal distance of 25 pm that facilitates early clot formation and collagen fiber attachment that, in turn, stabilizes the membrane until it becomes fixed. 800 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 y They are non-resorbable membranes which are made of knitted nylon fabric, that is mechanically bonded onto a semipermeable silicone membrane and coated with collagen peptides. density PTFE, with a submicron (0.2 pm) pore size.This combination of high density and small pore size entirely eliminates bacterial infiltration and also reduces the amount of fiber attachment to the membrane12. Osseous coagulum Bone blend Collection of bone using rongeurs and trephines followed by trituration in an amalgam capsule to achieve a slushy osseous mass. Nylonmembranes They are non-resorbable membranes which are made of knitted nylon fabric, that is mechanically bonded onto a semipermeable silicone membrane and coated with collagen peptides. density PTFE, with a submicron (0.2 pm) pore size.This combination of high density and small pore size entirely eliminates bacterial infiltration and also reduces the amount of fiber attachment to the membrane12. Synthetic Poly-DL-lactide (Guidor, Resolute, Atrisorb) Collagen of porcine origin (Bioguide) Collagen of porcine origin (Bioguide) 801 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 802 Collagen from bovine achilles tendon (Biomend, Periogen) Polyglactin 910 (Vicryl mesh) Collagen from bovine corium (Avitene) Polylactic acid (Epiguide) Collagen from bovine peritoneum (Cargium membranes) Polydiaxonon (Mempol) Collagen of fish origins (Periocol) Carbonate (Inion) Oxidized cellulose membrane (Surgicel) - Alloderm (Allogenic skin matrix) Duramater Amniotic membrane Table 2:- Differences between Natural and Synthetic grafts. Harvesting Autogenous Grafts Intraoral 1. Cortical bone chips 2. Osseous coagulum 3. Bone blend 4. Symphysis 5. Ramus 6. Tuberosity Extraoral 1. Calvarium 2. Rib 3. Fibula 4. Iliac Intraoral 1. harvesting autologous bone graft include the mandibular symphysis, mandibular ramus and the maxillary tuberosity. 2. Graft can be acquired from: a. Cortical bone chips b. Osseous Coagulum c. Bone blend Cortical Bone Chips 1. Shavings of bone removed by osteotomy and osteoplasty 2. Larger shavings of bone are vulnerable to sequestration Osseous coagulum Bone particles harvested with a bur and mixed with blood Difficult to collect graft in a quantifiable manner and difficult to aspirate completely. Bone blend Collection of bone using rongeurs and trephines followed by trituration in an amalgam capsule to achieve a slushy osseous mass. Symphysis Primary source of quality cortical bone and is mostly used to augment vertical defects Advantages: 1. Conventional surgical access 2. Proximity of donor and recipient sites 3. Reduced operative and anesthesia time, making it ideal for outpatient implant surgery. 4. Absence of scarring, no extraoral donor site morbidity. ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Collagen from bovine achilles tendon (Biomend, Periogen) Polyglactin 910 (Vicryl mesh) Collagen from bovine corium (Avitene) Polylactic acid (Epiguide) Collagen from bovine peritoneum (Cargium membranes) Polydiaxonon (Mempol) Collagen of fish origins (Periocol) Carbonate (Inion) Oxidized cellulose membrane (Surgicel) - Alloderm (Allogenic skin matrix) Duramater Amniotic membrane Table 2:- Differences between Natural and Synthetic grafts. ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Advantages: It can be used either as a block or particulate graft. Ramus The harvested bone may be placed in an extraction site, an implant site defect, a buccal alveolar defect. It may be used as vertical onlay graft of the ridge or as an inlay graft13. ISSN: 2320-5407 ISSN: 2320-5407 Mental nerve injury oft tissue injury of cheeks, lips and tongue lock graft fracture and infection. Figure 5:- Block bone graft harvested from the mandibular symphysis. Figure 5:- Block bone graft harvested from the mandibular symphysis. Ramus The harvested bone may be placed in an extraction site, an implant site defect, a buccal alveolar defect. It may be used as vertical onlay graft of the ridge or as an inlay graft13. Symphysis Symphysis Primary source of quality cortical bone and is mostly used to augment vertical defects Advantages: 1. Conventional surgical access 2. Proximity of donor and recipient sites 3. Reduced operative and anesthesia time, making it ideal for outpatient implant surgery. 4. Absence of scarring, no extraoral donor site morbidity. 802 Int. J. Adv. Res. 11(04), 793-817 Advantages: 1. Large amount of bone 2. Corticocancellous bone Disadvantages: Disadvantages: 1. Chances of damaging the neurovascular bundle of the inferior alveolar nerve which can cause temporary or permanent paresthesia. Disadvantages: 1. Chances of damaging the neurovascular bundle of the inferior alveolar nerve which can cause temporary or permanent paresthesia. 2 Injury to the lingual nerve p p 2. Injury to the lingual nerve 3. Mandibular fracture. ular fracture. Figure 6:- Block bone graft harvested from mandibular lateral ramus area. Figure 6:- Block bone graft harvested from mandibular lateral ramus area. 803 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 Advantage: Since the boss is porous, it aids in vascular ingrowth. Iliac crest Iliac crest The iliac crest graft provides medullary and cortico-medullary bone in large quantities, enough for large maxillary reconstructions in width, height and bilateral elevation of the maxillary sinus floor. The iliac crest graft provides medullary and cortico-medullary bone in large quantities, enough for large maxillary reconstructions in width, height and bilateral elevation of the maxillary sinus floor. Disadvantages: g 1. Exposure of the maxillary sinus cavity 2. Possible buccosinususal fistula 3. Injury to the remaining teeth. Figure 7:- Block bone graft harvested from the maxillary tuberosity. Figure 7:- Block bone graft harvested from the maxillary tuberosity. Tuberosity y It is a rich source of medullary bone. Depending on the anatomy, it may serve as a source of autograft for small to medium sized defect.They are mostly used in sinus lift procedures and to cover up fenestrations.Since the bone is mostly cancellous, it is advisable to use it only as a particulate graft. Ribs Ribs Used for the mandibular reconstruction ofisolated mandibular defects and in rare cases, reconstruction of atrophic maxilla. Ribs Used for the mandibular reconstruction ofisolated mandibular defects and in rare cases, reconstruction of atrophic maxilla. Cranial vault Cranial vault Cranial vault The bone graft taken from the cranial vault provides large amount of cortical bone and small amount of medullary bone. The bone graft taken from the cranial vault provides large amount of cortical bone and small amount of medullary bone. Advantages: Advantages: 1. By having intramembranous origin, it displays lower rates of resorption due to embryological resemblance to the jaw. 2. Ease of access to large amount of cortical bone. Advantages: 1. By having intramembranous origin, it displays lower rates of resorption due to embryological resemblance to the jaw. 2. Ease of access to large amount of cortical bone. g 1. By having intramembranous origin, it displays lower rates of resorption due to embryological resemblance to the jaw. 2 E f l f i l b j 2. Ease of access to large amount of cortical bone. j 2. Ease of access to large amount of cortical bone. Disadvantages: 1. Pneumothorax 1. Pneumothorax 2. Thin bone is liable to fracture Disadvantages: 1. Postoperative morbidity 2. Temporary difficulty in walking 3. Due to its endochondral origin, this type of bone graft resorbs more rapidly than theintramembranous grafts. 3. Due to its endochondral origin, this type of bone graft resorbs more rapidly than theintramembranous grafts. 804 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Figure 8:- Iliac bone graft harvested to reconstruct the mandible. Figure 8:- Iliac bone graft harvested to reconstruct the mandible. Advantages: 1. Favourable shape 2. Shorter postoperative hospitalization time 3. Lesser donor site sequelae Advantages: vantages: The donor site has enough bone length available (25cm) to reconstruct any length of the mandible. 15 g g y g 2. Low donor site morbidity and abundant periosteal blood supply that permits multiple osteotomies15. Low donor site morbidity and abundant periosteal blood supply that permits multiple osteotomies15 Disadvantages: advantages: Damage to the parietal branch of the superficial temporal artery. g 1. Damage to the parietal branch of the superficial temporal artery. 2. Penetration of cranial cavity during removal of the graft, leading to irreparable damage. Fibula Fibula In patients who undergo large ablative surgeries of the jaws, reconstruction of the maxilla or mandible should be done using vascularized bone containing free flaps14. Bone sources h f The most frequently used site for bone harvest is the iliac crest. In general, this site can supply enough volume of both cortical and cancellous bone for different reconstructive purposes in the maxillofacial region. Other sites also used, although less commonly, are the tibia, fibula and the ribs. When only a small amount of bone is needed, it can be harvested from the chin or at the anterior aspect of the mandibular ramus. The last two sites can, however, only contribute compact bone18. A bone graft may be of two types: free vascularized (i.e. a graft with vessels to be connected with vessels at the recipient bed), or free non-vascularized. The vascularized bone graft may have higher chances of survival but, on the other hand, this type of grafting is more time- consuming and more expensive.(Table 3) Indications (i) replace missing bone; (ii) enhance bone formation, in order to restore form and function. Many types of materials have been used and tested to replace missing bone during the last century, for example, banked bone (allograft; bone from individuals within the same species), xenografts (bone derived from other species), ceramics such as hydroxyapatite, metals, corals and plastics17. Disadvantages: Disadvantages: 1. Risk of developing a deep vein thrombosis of the lower leg due to immobilization. 805 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 2. High cost.16 Figure 9:- Showing the anatomical structures of the tooth. 2. High cost.16 Figure 9:- Showing the anatomical structures of the tooth. Indications (i) replace missing bone; (ii) enhance bone formation, in order to restore form and function. Many types of materials have been used and tested to replace missing bone during the last century, for example, banked bone (allograft; bone from individuals within the same species), xenografts (bone derived from other species), ceramics h h d tit t l l d l ti 17 2. High cost.16 Figure 9:- Showing the anatomical structures of the tooth. Anatomy of the skeleton The fundamental bony skeleton of the jaws consist of a mandible and two maxillary bones. Because of the functional aspect of these structures and their atrophic changes during aging, anatomical features have specific importance to distinguish defects and determine the proper treatment plan. The quantity and quality of bone in the alveolar process and adjacent structures are the key elements of this issue. The anatomical knowledge of these structures is also a determinant factor when using them as donor sites for reconstruction. The alveolar bone of mandible and maxilla is a functional bony process which harbors teeth in a dentate human. After tooth loss, this bony structure loses its dimensions both vertically and horizontally. After atrophic sequences, the maxillary alveolar arch diameter decreases, despite the fact that the mandibular alveolar arch enlarges in diameter and a pseudo-class III relation may appear in severe atrophic alveolar ridges (Figure 10). The alveolar bone of mandible and maxilla is a functional bony process which harbors teeth in a dentate human. After tooth loss, this bony structure loses its dimensions both vertically and horizontally. After atrophic sequences, the maxillary alveolar arch diameter decreases, despite the fact that the mandibular alveolar arch enlarges in diameter and a pseudo-class III relation may appear in severe atrophic alveolar ridges (Figure 10). D1 demonstrates the thickest cortical bone and the most dense trabecular part and is usually located in anterior mandible; the maxillary alveolar arch diameter decreases, despite the fact that the mandibular alveolar arch enlarges in diameter and a pseudo-class III relation may appear in severe atrophic alveolar ridges (Figure 10). D1 demonstrates the thickest cortical bone and the most dense trabecular part and is usually located in anterior mandible; D1 demonstrates the thickest cortical bone and the most dense trabecular part and is usually located in anterior mandible; D4 demonstrates a large volume of low density trabecular bone and thin cortices and is located mainly in posterior maxilla. D2 and D3 with intermediate characteristics are located in posterior mandible and anterior maxilla r The maxillary tuberosity is located in the posterior maxillary bone on each side and contains low density D4 bone and attached to the pterygoid plates at the pterygomaxillary junction. It is located next to important anatomical structures- the pterygomaxillary fissure and pterygo‐ palatine fossa. 806 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Figure 10:- A, The atrophic changes of mandible. Anatomy of the skeleton B, The atrophic changes of maxilla. The quality of edentulous alveolar bone is classified to D1, D2, D3 and D4 based on cortical bone thickness and density of trabecular bone respectively. Figure 10:- A, The atrophic changes of mandible. B, The atrophic changes of maxilla. The quality of edentulous alveolar bone is classified to D1, D2, D3 and D4 based on cortical bone thickness and density of trabecular bone respectively. The maxillary sinus is a pyramidal cavity in each maxilla with a broad base medially and an apex laterally. Its size varies depending on the patient's age and presence of teeth. During the lifetime the sinus enlarges continuously and at the age about 12, the floor of the sinus is almost at the level of the nasal floor. Maxillary posterior teeth loss and sinus pneumatization are responsible for decreasing bone volume in this area. The maxillary sinus is a pyramidal cavity in each maxilla with a broad base medially and an apex laterally. Its size varies depending on the patient's age and presence of teeth. During the lifetime the sinus enlarges continuously and at the age about 12, the floor of the sinus is almost at the level of the nasal floor. Maxillary posterior teeth loss and sinus pneumatization are responsible for decreasing bone volume in this area. The mandible is the largest bone of the face and generally consists of thicker cortical bone compared to the maxilla. The anterior border of ramus as runs toward the mandibular body creates external oblique ridges bilaterally. The mandibular canal begins from the mandibular foramen at the middle medial surface of ramus horizontally and vertically and ends at the mental foramen on the buccal surface of the mandibular body near the apices of the premolar teeth on both sides. The least distance from the mandibular canal to the buccal cortex is in the distal part of the mandibular first molars. The canal course through the mandible usually makes a loop near the mental foramen with about a 3 mm diameter. The neurovascular bundle travels through this canal to supply sensation and blood to the mandibular teeth and some part of the chin. The buccal fat pads or Bichat's fat are located lateral to the buccinator muscles bilaterally and consist of four parts; body, temporal, buccal, and pterygoid extensions. Buccal fat pads are supplied by the temporal and transverse facial arteries. Anatomy of the skeleton The buccal fat pads are very useful structures in reconstruction of oral defects. Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Sinus augmentation procedures Dental implant placement in patients who are edentulous in the posterior maxilla can be difficult for many reasons, including inadequate posterior alveolar height and increased pneumatization of the maxillary sinus, and therefore close approximation of the maxillary sinus to crestal bone. The size of the maxillary sinus correlates with the degree of pneumatization.18If appropriate graft materials are being used (i.e., at least 50% autogenous bone with good cellular density) and principles to maximize bone are followed, bone grafting and implant placement can be performed at the same time (i.e., in a one-step method) in ridges with as little as 1 to 5 mm residual crestal bone density when compared with premaxillary or mandibular bone. Adjacent cortices consist of compact bone19. Figure 12:- Showing normal anatomy of the structures in a cadaver. Figure 13:- Showing the procedure of grafting. rials that are being used currently for antral floor augmentation include autogenous bone, bone alloplasts such as tricalcium phosphate (TCP), bioactive glass, and resorbable and non-resorbable s the superomedial sinus area.20 d Figure 12:- Showing normal anatomy of the structures in a cadaver. Figure 12:- Showing normal anatomy of the structures in a cadaver. Figure 13:- Showing the procedure of grafting. Figure 13:- Showing the procedure of grafting. Grafting materials that are being used currently for antral floor augmentation include autogenous bone, bone allografts, and alloplasts such as tricalcium phosphate (TCP), bioactive glass, and resorbable and non-resorbable artery), supplies the superomedial sinus area.20 Grafting materials that are being used currently for antral floor augmentation include autogenous bone, bone allografts, and alloplasts such as tricalcium phosphate (TCP), bioactive glass, and resorbable and non-resorbable artery), supplies the superomedial sinus area.20 Autogenous Bone Graft techniques g q Our bone grafting considerations include sinus augmentation procedures, anatomy, materials, techniques, and complications. Figure 11:- Showing Bone repair by autologous cells. Figure 11:- Showing Bone repair by autologous cells. 807 The Rationale for Bone Grafts Implant placement requires sufficient volume of bone and bone quality in a biological way. This is achieved by the particular implant design, which requires certain dimensional properties that is long lasting success.24 Indications for Bone Grafts in Implant treatment are: Indications for Bone Grafts in Implant treatment are: 1. In alveolar sockets after extraction 2. Refilling a local bony defect due to trauma or infection 3. Refilling a peri-implant defect due to peri- implantitis 4. Vertical augmentation procedures of the mandible and maxi 5. Horizontal augmentation in the mandible and maxilla The commonly used grafting approaches in implant are Particulate Bone grafting and Block grafting. The commonly used grafting approaches in implant are Particulate Bone grafting and Block graft Bone Grafts For Implant Dentistry Bone Grafts For Implant Dentistry In the present scenario, an Implant procedure is a treatment that replaces missing teeth, improves function, and enhances esthetics. However, cases of bone deficient sites where Implant placement is difficult, bone grafts play their role. In these situations, it is possible to place Implants with bone grafts only.23 Nasal Floor And Block Grafts For Premaxilla Augmentation g Anterior tooth loss usually compromises ideal bone volume. On the day of primary surgical techniques Antibiotics that are effective against both aerobic and anaerobic bacteria should be prescribed preoperatively and postoperatively. The patient‟s oral-facial area should be prepared and draped. Surgery can be performed with the patient sedated with intravenous medication unless the graft material is procured from the iliac crest, in which case general anaesthesia is used. A local anaesthetic, with a Vaso-constrictor for haemostasis, is infiltrated into the maxillary surgical site and the maxillary or mandibular donor sites (if autogenous bone will be harvested from an intraoral site).22 ) Figure 14,15:- Showing procedure of graft placement. Figure 14,15:- Showing procedure of graft placement. Bone Grafts For Implant Dentistry Grafting procedure g p Autogenous bone is harvested from the predetermined site and is mixed with reconstituted freeze-dried bone in a 1 : 2 ratio. This mixture then is packed into 1-cc tuberculin syringes and is set aside. The mixture is used to densely fill the sinus. After completely filling the maxillary sinus with the desired level of bone mixture, as above, the clinician repositions the mucoperiosteal flap, and the incisions are closed with interrupted non-resorbable sutures43. After the bone has matured, it is evaluated to ensure that there is sufficient quantity for implant placement. Implants then can 808 ISSN: 2320-5407 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 be placed in the mature graft material according to the surgical protocol for the particular implant system and can be allowed to integrate. Postoperative considerations are similar to those for most oral surgery and sinus manipulation procedures.21 Minimally Invasive Tunnel Technique It is considered to be a safe, simple and patient compliance method to augment bone. The subperiosteal tunneling approach is a minimally invasive procedure that allows the surgeon to allocate the graft in a space that is obtained between the soft tissues and the underlying bone through an access represented by a single incision on the mesial limit of the bone defect. This approach is believed to warrant minimal discomfort to the patient in the immediate postoperative phase in addition to ensuring a steady coverage of the graft during the healing time, with minimal risk of exposure, infection, and failure. Alveolar ridge splitting technique This technique is used to augment the atrophic maxilla and mandible before the implant placement. It is commonly referred as Ridge splitting or Bone spreading technique.This techniques is performed by gaining access to a ridge that is <3 mm wide by splitting the buccal and palatal bone flaps with a scalpel first by separating two cortices through its cancellous bone. This technique is employed in cases where there is an insufficient width to utilize round osteotomes. This procedure provides a quicker method wherein an atrophic ridge can be but it does induce immune rejection. Bone manipulation techniques These techniques are capable of manipulating the bone to alter their density to make it exceedingly durable and strong. These techniques mobilize vital bone with plastic bending, shaping, or condensation of tissue as a bone flap or bone-periosteal flap. This results in contour or dimensional changes, while preserving bone integrity and viability. Current techniques include inlay and onlay grafting, guided bone regeneration (GBR), bone expansion, bone splitting osteotomy, and different fixation devices such as bone screws, pins, titanium mesh, different augmentation materials, and different barrier membranes. Bone expansion technique This technique involves the manipulation of the bone to form a receptor site for an implant without the removal of any bone from the patient. The objective is to maintain the existing soft bone by pushing the buccal bony plates of the residual ridge laterally with minimal trauma. This technique takes the advantage of the softer bone quality found in Types III and IV maxillary bone by relocating the alveolar bone rather than losing the precious bone by drilling. The most common anatomic area in which ridge expansion is performed is in the narrow anterior maxilla, followed by posterior maxilla and then the anterior and posterior mandible, respectively. Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 This can be achieved by using commercially available non-resorbable titanium based and titanium reinforced membranes or the adjunct use of tenting screws and simultaneous placement of implants to prevent barrier membranes collapse into the space of the bony defects by the overlying soft tissue during healing. In some reports where autografts were utilized for GBR purposes, resorption tended to be higher with when no membrane was used. A clinical study reported a significantly less resorption of the block grafts was found when e-PTFE membranes were used to protect the graft. Ridge expansion/ ridge splitting techniques Ridge splitting is an alternative to the various techniques described for horizontal ridge augmentation, such as distraction osteogenesis. It was proven that both pervious mentioned procedures have a similar healing pattern and end results. In an area with a narrow ridge measuring 3 mm in bucco-lingual width or more, splitting of the alveolar bone is started by using either chisels, osteotomes, or piezo surgical devices to increase the horizontal ridge width. Buccal and lingual cortical plates or targeted sites should not be fused and some intervening cancellous bone between those cortical plates should be present to prevent a complete bone fracture and separation. Combination approaches With reference to the aforementioned GBR techniques, combining one or more of the previously described approaches can be utilized in cases where severe bone defects are present in order to optimize GBR outcomes. In many situations, a membrane may not be required, and the graft material alone can be effective. However, it is found that the use of barrier membranes to cover the grafting material can further improve the quality of regeneration by holding grafting material in proper location which particulate grafts are used, acting as space maintenance and minimizing alveolar bone resorption.25 Barrier membranes can be non-resorbable, such as expanded polytetrafluoroethylene (ePTFE) and titanium, or resorbable, such as polypeptides (collagen) and synthetic polymers (polylactide and polyglycolide). These membranes may be used in combination with block grafts and/or particulate graft materials. Membranes required for grafting of severe bony defect has to have a space maintenance property, which make them rigid enough to be shaped to the desired contour, height and width of future bone needed.26 809 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 ISSN: 2320-5407 Alveolar distraction osteogenesis It is eliminating the need for a second donor site and a second stage surgery. Ideal sites to perform this procedure includeclinical scenarios where there is a knife-edge ridge that widens further apically, and that consists of adequate cortical thickness but with some degree of interpositional lamellar bone. This is particularly true in the anterior region of the maxilla. In this technique, a defect is created when two bone segments are slowly separated under tension. One week after osteotomy and distractor placement (latency period), distraction of segments is advanced at a rate of 0.5-1mm/day until the desired separation is reached. A consolidation period of 5 days/mm of the space created should be maintained before device removal and implant placement. It allows for a vertical bone gain of 3- 810 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 ISSN: 2320-5407 20mm without the use of graft material and additional mucosal grafting is not required as the soft tissue follows bone distraction. Biophysical effects p y Mechanical loading and electromagnetic signals are important regulators of bone formation. The regenerative capacity of bone depends largely upon its capacity to recognize the functional environment that is required for the emergence and maintenance of structurally intact osseous tissue. Methods such as biophysical stimulation have thus been introduced with success in clinical practice. In addition to distraction osteogenesis, which is dealt with in the next section, exposure to an electromagnetic field and the application of ultrasonic waves are considered to be special forms of mechanical stimulation50. Autogenous teeth utoge ous teet Autogenous bone is considered an ideal material for the reconstruction of hard tissue defects, due to its ability to promote osteogenesis, osteoinduction, osteoconduction, and enhanced healing. Autogenous bone is considered an ideal material for the reconstruction of hard tissue defects, promote osteogenesis, osteoinduction, osteoconduction, and enhanced healing. Transplantation of cells Different donor sites require the grafting of different types and amounts of bone-cell-containing materials. Bone- forming tissues can be applied as periosteal flaps, as bulk grafts of cortico-spongiosal or vascularized bone, and as chips of cancellous or cortical bone. Periosteal flaps and chips of cancellous or cortical bone are used to treat small osseous defects if the local conditions are conducive to bone healing. Endochondral bone can be harvested from the ilium, the tibia or the ribs, and membranous bone from the facial skeleton. Animal studies indicate that membranous bone is less prone to resorption than endochondral bone Guided bone regeneration g It is also known as “membrane protected bone regeneration.” The concept of GBR implies the use of cell-occlusive membranes for space provision over a vertical or horizontal defect, promoting the in-growth of osteogenic cells while preventing migration of undesired cells from the overlying soft tissue. It also effectively stabilizes the blood coagulum and thereby allows for faster healing to occur. This technique can be used before or at the same time as implant placement. Growth factors Numerous growth factors have been widely tested in animal models. Amongst these, bone BMPs are considered to be unique as they induce osteogenic precursor cells into osteogenic cells and have shown incredible bone growth in many animals and also human clinical studies. Other growth factors besides BMPs that have been implicated during bone regeneration are also being investigated, including platelet-derived growth factor, transforming growth factor- β, insulin-like-growth factor-1, vascular endothelial growth factor, and fibroblast growth factor, among others. Local bone grafts g For the repairof alveolar defects local grafts harvested from either the mandibular symphysis or ramus are suitable. The advantages of local grafts harvested from the circum-mandibular wires that were unlikely to maintain rigid fixation grafting of the edentulous mandible in the past. These grafts were secured by donor and recipient sites and an ideal method for out-patient pre-implant surgery. Figure 21:- Showing The location and extent of the corico- cancellous graft taken from the body of the mandible as described by the various authors. Figure 21:- Showing The location and extent of the corico- cancellous graft taken from the body of the mandible as described by the various authors. Figure 21:- Showing The location and extent of the corico- cancellous graft described by the various authors. 811 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 Distant bone grafts g In situations where large amounts of autogenous bone are required for alveolar reconstruction, usually of the edentulous jaws prior to implant placement, distant donor sites are required, usually from the iliac crest. Cells Large numbers of cells capable of producing bone extracellular matrix are needed for the production of clinically- sized engineered tissues. Mesenchymal stem cells, which differentiate and form bone during normal development, have long been the primary cell source for engineering bone grafts. It has long been recognized that adult bone marrow stem cells (BMSC) form multiple mesenchymal tissues in vivo including bone, and have utility for engineering skeletal tissues. Functional Requirements Surgical interventions utilizing autografts and allografts have been shown to improve repair of bone defects in various degrees. However, none of currently used grafts has all the ideal characteristics: high osteo-inductive and angiogenic potentials, biological safety, low patient morbidity, no size restrictions, ready access to surgeons, long shelf life, and reasonable cost. The promise of tissue engineering is to combine the advances in the fields of biomaterials and cell biology towards bone grafts matching most or all of these characteristics. Maxillary sinus grafting (open technique) The sinus lift technique is a simple procedure and allows implant rehabilitation in the posterior atrophic maxilla with predictable results. Implant placement is carried out simultaneously to the grafting technique in the case of having enough residual alveolar bone to provide primary stability to the implants. In the case of extreme atrophic bone, not allowing adequate fixation for the implants, primary bone grafting is indicated and secondary placement of the implants after consolidation of the grafting material will be carried out . Bone Tissue Engineering g g To date, the field of bone tissue engineering has been focused on creating tissue grafts that have capacity to enhance osteogenesis in the site of the bone defect. Constructs have been assembled in vitro by seeding cells with osteogenic potential into biodegradable scaffolds, and either directly transplanted in vivo to assess their bone forming potential, or cultured in vitro to enable development of new tissue and the formation of “mature” bone-like grafts. Scaffolds The scaffold is crucial for the successful engineering of bone tissues as it provides a suitable environment for osteogenic cells to migrate, proliferate, differentiate, and promote new bone formation, and it also provides mechanical competence during the bone regeneration. There are a few requirements to be considered in the design and construction of 3D bone scaffolds. First, the scaffold must be biocompatible and degrade with time into non- toxic products. It should also be highly porous and permeable for cell seeding (in vitro) and infiltration (in vivo), nutrient transport, tissue ingrowth, and vascularization. The scaffold should be mechanically stable, having properties similar to those of the native bone. Finally, an ideal bone scaffold should also be osteoconductive (to recruit bone cells from the recipient), osteoinductive (to differentiate stem cells into bone-forming cells), and osseointegrative (to provide permanent and functional attachment to native bone). 812 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Figure 25:- Bone tissue engineering paradigm. Figure 26:- Tissue engineering procedure. Protein Based Bone Grafts Numerous studies of critical-size calvarial, alveolar, and mandibular defectmodels of BMPinduced osteogenesi have demonstrated a successful result in a wide variety of animal models. These have focused on the delivery o BMPs, typically rhBMP-2 and rhBMP-7, with a wide variety of delivery systems including demineralized bone matrix and other alloplastic materials along with the standard commercially available collagen sponge implants . Figure 25:- Bone tissue engineering paradigm. Figure 25:- Bone tissue engineering paradigm. Figure 26:- Tissue engineering procedure. Figure 26:- Tissue engineering procedure. Protein Based Bone Grafts Protein Based Bone Grafts Numerous studies of critical-size calvarial, alveolar, and mandibular defectmodels of BMPinduced osteogenesis have demonstrated a successful result in a wide variety of animal models. These have focused on the delivery of BMPs, typically rhBMP-2 and rhBMP-7, with a wide variety of delivery systems including demineralized bone matrix and other alloplastic materials along with the standard commercially available collagen sponge implants . Protein Based Bone Grafts Numerous studies of critical-size calvarial, alveolar, and mandibular defectmodels of BMPinduced osteogenesis have demonstrated a successful result in a wide variety of animal models. These have focused on the delivery of BMPs, typically rhBMP-2 and rhBMP-7, with a wide variety of delivery systems including demineralized bone matrix and other alloplastic materials along with the standard commercially available collagen sponge implants . 813 Int. J. Adv. Res. Scaffolds 11(04), 793-817 ISSN: 2320-5407 Figure 27:- Showing Mechanism of Bone Morphogenetic Protein Induction of Osteogenesis. Figure 27:- Showing Mechanism of Bone Morphogenetic Protein Induction of Osteogenesis. Gene- based Bone Grafts Gene- based Bone Grafts Bone tissue has significant potential for reparative regeneration. However, with “osteogenic insufficiency,” this process can take a long time and often does results in incomplete bone healing, despite the use of current surgical technologies. Osteogenic insufficiency is a pathological condition associated with low activity of systemic or local osteoinductive factors and/or a lack of cambial cells in the bone lesion area; therefore, the natural course of reparative osteogenesis may not provide complete histotypic and organotypic recovery. Therefore, treatment of patients with large bone defects, delayed consolidation, and non-unions remains extremely challenging and is followed by a prolonged loss of working ability, decreased quality of life, and even disability. Successful treatment requires restoration of the lostcambial reserve and/or osteoinductive factors, and classically involves the use of bone autografts, as the “golden standard.” However, well-known limitations and disadvantages of this approachpredetermine the development of acceptable alternatives, among which activated bone substitutes are quite promising. 814 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 ISSN: 2320-5407 Figure 28:- Scheme of intracellular Smad- mediated transduction pathway for BMP signals. Figure 28:- Scheme of intracellular Smad- mediated transduction pathway for BMP signals. Conclusion:- The introduction of endosteal implants, together with better understanding of primary bone healing, has contributed substantially to the successful application of bone grafts in pre-implant surgery.The techniques used to harvest and apply the grafts are not new and found, for a large part, their origin in conventional pre-prosthetic surgery.Sophisticated and the results are far better than in the past, simply because of the favourable loading conditions of the grafted bone by endosteal implants. The resorption as seen in the old days was largely the result of disuse atrophy and compression loading by a conventional, mucosally borne denture. Yet, fundamental rules, learned from the early experiences in pre-prosthetic surgery, are still valid today and the difference between success and failure is very small.A proper understanding of the function and healing of the graft is therefore mandatory when choosing the type of graft for effective pre-implant surgery of the jaws. References:- 1. Robert Alan Horowitz, Minas Leventis. Bone grafting: History, rationale, and selection of materials and Techniques. Compendium. Volume 35.2014. 2. Arash Khojasteh, Mohammad Esmaeelinejad and Farzad Aghdashi. Regenerative Techniques in Oral and Maxillofacial Bone Grafting.Oral and Maxillofacial Surgery. 3. Reham AL Jasserand Sebastiano Andreana .An overview of bone augmentation techniques. Clin Case Rep Rev, 2016. 3. Reham AL Jasserand Sebastiano Andreana .An overview of bone augmentation techniques. Clin Case Rep Rev, 2016. References:- 3. Reham AL Jasserand Sebastiano Andreana .An overview of bone augmentation techniques. Clin Case Rep Rev, 2016. 815 Int. J. Adv. Res. 11(04), 793-817 2014 35: 644-671. 17. Rusin Zhao, Bidez MW (1999) A scientific rationale for dental implant design. In: Misch CE (ed). Contemporary Implant Dentistry, St Louis, Mosby Publishers, USA, 2016.pp: 329-343. 18. Gilbert MW, Sangam CE .Force transfer in implant dentistry: Basic concepts and principles. Oral Implantol 2018: 264-274. 19. Navneet Kaur, Rakesh, Shakthi, Bidez MW (1999) Mechanical properties of trabecular bone in the human mandible: Implications for dental implant treatment planning and surgical placement. J Oral Maxillofac Surg 2018, 57: 700-706. 20. Haben Fesseha, Carlsson GE, Jemt TA (1996) A prospective 15-year follow-up study of mandibular fixed prostheses supported by osseointegrated implants. Clinical results and marginal bone loss. Clin Oral Implants Res. 2019, 7: 329-336. 21. Björn Johansson, Leonard BS, Block MS, lacono VJ (1998) Report of the Sinus Consensus Conference of 1996. Int J Oral Maxillofac Implants, 2019,13: 11-30. 22. PAUL FLETCHER, Desjardins RP, Keller E, Tolman DA (1996) Endosseous implants in an irradiated tissue bed. J Prosthet Dent, 2020 76: 45-49. 23. Alan S. Herfor, Andreasson L, Bjelkengren G .Oral implant rehabilitation in irradiated patients without adjunctive hyperbaric oxygen. Int J Oral Maxillofac Implants. 2021 13: 647-654Niimi A, Ueda M, Keller EE, Worthington P (1998) Experience with osseointegrated implants placed in irradiated tissues in Japan and in the United States. Int J Oral Maxillofac Implants 13: 407-411. rson DE, Overholser CD (1996) Dental care for patients receiving chemotherapy. J Am Dent Assoc 3. 24. Reham, Peterson DE, Overholser CD (1996) Dental care for patients receiving chemotherap 112: 198-203. 25. RuggieroSL,MehrotraB,RosenbergTJ,EngroffSL(2004)Osteonecrosis of the jaws associated with the use of bisphosphonates: A review of 63 cases. J Oral Maxillofac Surg 62: 527-534. 26. Marx RE (2003) Pamidronate (Aredia) and zolendronate (Zometa) induced avascular necrosis of the jaws: A growing epidemic. J Oral Maxillofac Surg 61: 1115-1117. 27. Stammberger H (1989) History of rhinology: Anatomy of the paranasal sinuses. Rhinology 27: 197-210. 28. ValentiniP, AbensurDJ(2003)Maxillarysinusgraftingwithanorganic bovine bone: A clinical report of long-term 27. Stammberger H (1989) History of rhinology: Anatomy of the paranasal sinuses. Rhinology 27: 197-210. 28. ValentiniP, AbensurDJ(2003)Maxillarysinusgraftingwithanorganic bovine bone: A clinical report of long-term results. Int J Oral Maxillofac Implants 18: 556-560. 29. Misch CE (1993) Bone density. In: Misch CE (ed). Contemporary Implant Dentistry. St Loui pp: 241-255. 30. Valentini P, Abensur DJ (2003) Maxillary sinus grafting with anorganic bovine bone: A clinical report. International Journal of dentistry. y 31. Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 Int. J. Adv. Res. 11(04), 793-817 4. Bone Grafting Techniques for Maxillary Implants text book-Karl-Erik Kahnberg. 5. RoleofBoneGraftsinImplant Surgery:AReview. International Journal of Dentistry.2017.International Journal of Scientific study. 6 y 6. Reham AL Jasserand Sebastiano Andreana .An overview of bone augmentation techniques. Reham AL Jasserand Sebastiano Andreana .An overview of bone augmentation techniques. 7. Bryant SR (1998) The effects of age, jaw site, and bone condition on oral implant outcomes. Int J Prosthodont 11: 470-490. 7. Bryant SR (1998) The effects of age, jaw site, and bone condition on oral implant outcomes. 11: 470-490. 8. An overview of bone augmentation techniques. Reham AL Jasserand Sebastiano Andreana. 9. Truhlar RS, Orenstein IH, Morris HF, Ochi S (1997) Distribution of bone quality in patients receiving endosseous dental implants. J Oral Maxillofac Surg 55: 38-45. p g 10. Adrash, Lekholm U, Sruthi, kranthi. A 15 year study of osseointegrated implants in the edentulous jaw. Int J Oral Surg. 2008. 10: 387-416. j g 11. Chandini , Benner KU, kaumudhi.Evaluation of different methods of indirect sinus floor elevation for elevation heights of 10 mm: An experimental ex vivo study. Clin Implant Dent Relat Res. 2009,13: 124-133. g enner KU, kaumudhi.Evaluation of different methods of indirect sinus floor elevation for elevation 0 A i l i d Cli I l D R l R 2009 13 124 133 j g 11. Chandini , Benner KU, kaumudhi.Evaluation of different methods of indirect sinus floor ele heights of 10 mm: An experimental ex vivo study. Clin Implant Dent Relat Res. 2009,13: 12 f 10 mm: An experimental ex vivo study. Clin Implant Dent Relat Res. 2009,13: 124-133. 12. Robert, Veltri M, Cagidiaco MC, Balleri P, Ferrari M.Sinus membrane elevation in combination with placement of blasted implants: A 3-year case report of sinus augmentation without grafting material. Int J Oral Maxillofac Surg. 2010,37: 966-969. 13. Bach Le, Bernal G, Riringcharassaeng K, Kan JY (2003) Clinical complications with implants and implant prostheses. J Prosthet Dent 2011,2: 121-132. 14. Matteo Chiapasco. Density of bone: Effect on treatment plans, surgical approach, healing loading. Int J Oral Implantol. 2011,6: 23-31. apasco. Density of bone: Effect on treatment plans, surgical approach, healing and progressive J Oral Implantol. 2011,6: 23-31. 15. Prasanna kumar, Chitra, Hari, Ash MM .A six-channel intra-oral transmitter for measuring occlusal forces. J Prosthet Dent 2014,16: 56. 16. Reham, Pretham, Sanjana Measurements of stress in mastication. J Dent Res. Int. J. Adv. Res. 11(04), 793-817 Int. J. Adv. Res. 11(04), 793-817 ISSN: 2320-5407 32. Misch CE (1999) Divisions of available bone. In: Misch CE (ed). Contemporary Implant Dentistry, St Louis: Mosby, USA, pp: 89-107. \ 33. Regev E, Smith RA, Perrot DH, Pogrel MA (1995) Maxillary sinus complications relate implants. Int J Oral Maxillofac Implants 10: 451-461. p p 34. Misch CE (1993) Treatment planning for edentulous maxillary posterior region. In: Misch CE (ed). Contemporary Implant Dentistry. St Louis: Mosby, pp: 241-255. p y p y y pp 35. Misch CE (1988) Bone character: Second vital implant criterion. Dent Today 7: 39. 36. Timmenga NM, Raghoebar GM, Boering G, van Weissenbruch R (1997) Maxillary sinus function after sinus lifts for the insertion of dental implants. J Oral Maxillofac Surg 55: 936-939. 37. Zimbler MS, Lebowitz RA, Glickman R (1998) Antral augmentation osseointegration and sinusitis: The otolaryngologist perspective. Am J Rhinol 12: 311-316. 38. Watelet JB, Van Cauwenberge P (1999) Applied anatomy and physiology of the nose and p Allergy 54: 14-25. gy 39. Peleg M, Chaushu G, Mazor Z (1999) Radiological findings of the post-sinusitis lift maxillary sinus: A computerized tomography follow- up. Periodontol 70: 1564-1573. p g p y p 40. Van den Berg JP, Ten Bruggenkate CM, Disch FJ, Tuinzing DB (2000) Anatomical aspects of sinus floor elevations. Clin Oral Implants Res 11: 256-265. 41. SambataroG,MantovaniM,ScottiA(2003)Rialzodelsenomascellare e implicazioni sulle sue funzioni. In: Chiapasco M, Romeo E (eds). La riabilitazione implanto-protesica nei casi complessi. Torino, Italy, pp: 292- 298. 42. Goldman L (1983) Cardiac risks and complications of noncardiac surgery. Ann Intern Med 98: 504-513. 43. Goldman L, Caldera DL, Nussbaum SR (1977) Multifactorial index of cardiac risk in noncardiac surgical procedures. N Engl J Med 297: 845-850. 44. Cygan R, Watzkin H (1987) Stopping and restarting medications in the perioperative period. J Gen Intern Med 2: 270. 45. Marx RE, Johnson RP (1987) Studies in the radiobiology of osteoradionecrosis and their c Oral Surg Oral Med Oral Pathol 64: 379-390. 46. EpsteinJB, WongFLW, Stevenson-MooreP(1987)Osteoradionecrosis: Clinical experience and classification. J Oral Maxillofac Surg 45: 104-110. 47. Nguyen TD, Panis X, Froissart D, Legros M, Coninx P, et al. (1988) Analysis of late complications after rapid hyperfractionated radiotherapy in advanced head and neck cancers. Int J Radiat Oncol Biol Physics 14: 23-25. 48. WidmarkG, SagneS, HeikelP(1989)Osteoradionecrosis of the jaws. Int J Oral Maxillofac Surg 18: 302-306. 49. Int. J. Adv. Res. 11(04), 793-817 Shillingburg HE, Hobo S, Whitsett LD (1997) Fundamentals of Fixed Prosthodontics. Chicago: Quintessence publishers. 31. Shillingburg HE, Hobo S, Whitsett LD (1997) Fundamentals of Fixed Prosthodontics. Chicago: Quintessence publishers. 816 Int. J. Adv. Res. 11(04), 793-817 Murray CG, Herson J, Daly TE, Zimmerman S (1980) Radiation necrosis of the mandible: A 10 year study. I. Factors influencing the onset of necrosis. Int J Radiat Oncol Biol Physics 6: 543-548. 50. Granstromm G, Bergstrom K, Tjellstrom A, Branemark Pl (1994) A detailed analysis of titani in irradiated tissue. Int J Oral Maxillofac Implants 9: 653-662. 51. Jacobsson MG, Jonsson AK, Albrektsson TO, Turesson IE (1985) Shortand long-term effects of irradiation on bone regeneration. Plast Reconstr Surg 76: 841-848. g g 52. Ueda M, Kaneda T, Takahashi H (1993) Effect of hyperbaric oxygen therapy on osseointegration of titanium implants in irradiated bone: A preliminary report. Int J Oral Maxillofac Implants 8: 41-44. 53. Watzinger F, Ewers R, Henninger A, Sudasch G, Babka A, et al. (1996) Endosteal implants in the irradiated lower jaw. J Craniomaxillofac Surg 24: 237-244. 54. Watzinger F, Ewers R, Henninger A, Sudasch G, Babka A, et al. (1996) Endosteal implants in the irradiated lower jaw. J Craniomaxillofac Surg 24: 237-244. 55. Jisander S, Grenthe B, Alberius P (1997) Dental implant survival in the irradiated jaw: A preliminary report. Int J Oral Maxillofac Implants 12: 643-648. 817
https://openalex.org/W4375856225	https://biblio.ugent.be/publication/01H7YWB6E6208YC3VJH2Y6H8GX/file/01H7YWCMVP5296PDZRH44QM89R.pdf	English	null	Improving de novo protein binder design with deep learning	Nature communications	2,023	cc-by	7,346	Improving de novo protein binder design with deep learning Nathaniel R. Bennett 1,2,3,8, Brian Coventry1,2,4,8, Inna Goreshnik1,2, Buwei Huang1,2,5, Aza Allen 1,2, Dionne Vafeados 1,2, Ying Po Peng1,2, Justas Dauparas 1,2, Minkyung Baek 1,2, Lance Stewart 1,2, Frank DiMaio1,2, Steven De Munck 6,7, Savvas N. Savvides 6,7 & David Baker 1,2,4 Nathaniel R. Bennett 1,2,3,8, Brian Coventry1,2,4,8, Inna Goreshnik1,2, Buwei Huang1,2,5, Aza Allen 1,2, Dionne Vafeados 1,2, Ying Po Peng1,2, Justas Dauparas 1,2, Minkyung Baek 1,2, Lance Stewart 1,2, Frank DiMaio Steven De Munck 6,7, Savvas N. Savvides 6,7 & David Baker 1,2,4 Recently it has become possible to de novo design high afﬁnity protein binding proteins from target structural information alone. There is, however, con- siderable room for improvement as the overall design success rate is low. Here, we explore the augmentation of energy-based protein binder design using deep learning. We ﬁnd that using AlphaFold2 or RoseTTAFold to assess the probability that a designed sequence adopts the designed monomer structure, and the probability that this structure binds the target as designed, increases design success rates nearly 10-fold. We ﬁnd further that sequence design using ProteinMPNN rather than Rosetta considerably increases computational efﬁciency. Methods for designing proteins which bind with high afﬁnity and speciﬁcity to protein targets of interest are of considerable importance in biomedicine for generating candidate therapeutics1, diagnostics2, and imaging reagents3, 4. Currently, the most widely used methods involve immunization of an animal with the target to elicit antibodies5, or screening high complexity random libraries of antibody6 or other scaffolds7 for binding activities. Although powerful, these methods require considerable experimental effort and do not provide sub- stantial control over the properties of the resulting binding molecules. Methods for computationally designing binders could potentially provide much faster routes to afﬁnity reagents having desired bio- physical properties that target speciﬁc surface patches, and there has been considerable progress in computational design of protein bind- ing proteins based on extension of binding motifs observed in protein structures8–12. Recently, a general Rosetta-based approach to designing binding proteins using only the structure of the target was developed and used to design binding proteins to 13 different target sites13. Given a speciﬁed region on a target of interest, the method designs sequences predicted to fold up into protein structures that have shape and chemical complementarity to the region. 1Department of Biochemistry, University of Washington, Seattle, WA, USA. 2Institute for Protein Design, University of Washington, Seattle, WA, USA. 3Molecular Engineering Graduate Program, University of Washington, Seattle, WA, USA. 4Howard Hughes Medical Institute, University of Washington, Seattle, WA, USA. 5Department of Bioengineering, University of Washington, Seattle, WA, USA. 6VIB-UGent Center for Inﬂammation Research, Ghent, Belgium. 7Unit for Structural Biology, Department of Biochemistry and Microbiology, Ghent University, Ghent, Belgium. 8These authors contributed equally: Nathaniel R. Bennett, Brian Coventry. e-mail: dabaker@uw.edu Article https://doi.org/10.1038/s41467-023-38328-5 Improving de novo protein binder design with deep learning directed to the likely correct structure, and there is a more accurate stopping criterion17). For accurate prediction of the structures of naturally occurring proteins, both AF2 and RF generally require mul- tiple sequence alignments (which contain rich co-evolutionary infor- mation on residues likely to be in contact, etc), but for de novo designed sequences, which are generally more stable and more regular than naturally occurring proteins, accurate predictions can be obtained from single sequences18, 19. There has also been progress in accuracy prediction for protein structure models; for example Dee- pAccuracyNet (DAN), which uses a representation consisting of 3D convolutions of local atomic environments20, achieved state-of-the-art performance in accuracy prediction in CASP14. y pp y g Encouraged by this accuracy, we set out to ﬁlter the entire set of Cao et al. designs based on the similarity of the AF2 or RF2 predicted monomer structure to the designed structure (dis- agreement is an indication of a possible Type I failure). For each designed sequence for each target, using AF2 or RF2 with a single sequence as input, we predicted the structure of the binder monomer. We found that the closer the prediction of the binder structure was to the Rosetta-designed structure in Cɑ RMSD, the more likely a binder was to be successful (Supplementary Fig. 4). We also found that the prediction conﬁdence metric pLDDT was predictive of success (Fig. 1d); the two metrics are quite corre- lated (Supplementary Fig. 5; the pLDDT of AF2 and RF2 were equally discriminative). These results suggest that Type I failures contribute to the low success rate of binder design, and that such failures can, in part, be identiﬁed by discrepancies between design models and AF2 or RF2 structure predictions. p y p We reasoned that these newly-developed DL methods could increase the success rate of Rosetta-based protein binder design. As noted above, while providing a general computational route to designing binders to arbitrary protein targets, the overall success rate is quite low. The approach has two primary failure modes (Fig. 1a): ﬁrst, the designed sequence may not fold to the intended monomer structure, and second, the designed monomer structure may not actually bind the target (Fig. 1b). Retrospective analysis of type II failures To estimate the likelihood of the designed binder structure forming an interface with the intended target, Cao et al. primarily used the dif- ference in energy of the bound complex and the unbound monomers allowing sidechain repacking as computed by Rosetta (Rosetta ddG), and despite the extensive use of this metric during the original cal- culations, Rosetta ddG remains an effective ﬁlter (Fig. 1e). We investi- gated the efﬁcacy of DAN in supplementing Rosetta in assessing the accuracy of the designed complex structure. We found DAN’s complex accuracy metric to be approximately as predictive of binder success as Rosetta ddG (Fig. 1e). We next investigated whether AF2 and RF2 complex prediction could be used to discriminate designs that form the intended complex structure from those that do not. We again began by evaluating the ability of AF2 and RF2 to reproduce the ﬁve experimentally determined minibinder structures from Cao et al. Given an MSA for the target protein and the single sequence of the designed binder, AF2 predicted the complex structure with binder Cɑ accuracy between 1.0Å−2.0 Å for three of ﬁve, and RF2 for four of ﬁve. The two structures that were not correctly predicted by AF2 were LCB1 and LCB3 which both target the SARS-CoV-2 Spike protein; AF2 was not able to correctly model a long loop in the Spike protein which caused the binders to be predicted as unbound. RF2 also predicted LCB1 as unbound (Supplementary Fig. 1). To enable AF2 to be used for binding prediction in cases where the target is incorrectly modeled, we investigated providing the target structure to the model as a template. We found this allowed AF2 to predict the correct COVID spike structures but caused all of the interfaces except FGFR2 to be predicted incorrectly (Supplementary Fig. 1). We next investigated initializing the AF2 pair representation with an encoding of the Rosetta binder structure; we call this protocol “AF2 initial guess” (see AF2 Initial Guess in Methods). Using AF2 with target template and an initial guess, AF2 is able to recapitulate the experimentally determined structures for all 5 minibinder interfaces with binder Cɑ accuracy between 1.0Å−2.0 Å RMSD (Supplementary Fig. 1). Notably, for all structures except LCB1 and LCB3, the AF2- predicted structures are closer to the experimentally determined structure than the original design models, even after extensive relaxation using Rosetta. Improving de novo protein binder design with deep learning The physically based Rosetta approach frames both the folding and binding problems in energetic terms; for the approach to succeed, the designed sequence must have as its lowest energy state in isolation the designed monomer structure, and the complex between this designed monomer structure and the target must have sufﬁciently low energy to drive formation of the design-target protein complex. The primary challenges in accurate design of both the monomer structure and the protein-protein inter- face are inaccuracies of the energy function which for computational tractability is generally represented as a sum of pairwise decom- posable terms (in Rosetta: Lennard Jones, hydrogen bonding, elec- trostatic, solvation, and bonded geometry), and the very large size of the space which must be sampled; if the energy function is inaccurate, or conformational sampling is incomplete, the designed sequence may not fold to the intended monomer structure and/or the monomer may not bind to the target as intended. Improving de novo protein binder design with deep learning While providing a gen- eral computational route to designing binders to arbitrary protein targets, the method requires screening of large numbers of compu- tationally designed binders to identify hits as only a small fraction typically have sufﬁciently high afﬁnity for experimental detection. In parallel with advances in physical model based protein binder design, deep learning methods have achieved unprecedented accu- racy in protein structure prediction. In contrast to Rosetta and other physically based molecular mechanics methods, which employ energy functions with one or two thousand parameters obtained from struc- tural and thermodynamic data on proteins and small molecules14, the deep learning structure prediction methods AlphaFold215 (AF2) and RoseTTAFold16 (RF) have hundreds of millions of parameters obtained by training on very large datasets of protein sequences and structures, and make no assumptions about pairwise decomposability or func- tional form. In place of the energy-guided stochastic conformational sampling approaches utilized by physically based approaches – molecular dynamics in many protein dynamics studies or Monte Carlo plus minimization in the case of Rosetta– the deep learning methods learn iterative transformations of representation of the sequence and possible structure that very rapidly converge on often quite accurate models (the successive transformations are analogous to the structure updates in traditional simulation, but are more concerted, more Nature Communications\| (2023) 14:2625 Nature Communications\| (2023) 14:2625 1 Article https://doi.org/10.1038/s41467-023-38328-5 directed to the likely correct structure, and there is a more accurate stopping criterion17). For accurate prediction of the structures of naturally occurring proteins, both AF2 and RF generally require mul- tiple sequence alignments (which contain rich co-evolutionary infor- mation on residues likely to be in contact, etc), but for de novo designed sequences, which are generally more stable and more regular than naturally occurring proteins, accurate predictions can be obtained from single sequences18, 19. There has also been progress in accuracy prediction for protein structure models; for example Dee- pAccuracyNet (DAN), which uses a representation consisting of 3D convolutions of local atomic environments20, achieved state-of-the-art performance in accuracy prediction in CASP14. the monomer structure with binder Cɑ accuracy between 0.2 Å−0.8 Å for all binders except for LCB1 which was predicted with 1.5 Å accuracy (Supplementary Fig. 1). An updated version of RoseTTAFold (RF222, 23) was also found to predict all monomer structures with binder Cɑ accuracy between 0.2 Å−0.8 Å, except for TrkA which was predicted with 1.8 Å accuracy (Supplementary Fig. 1). Retrospective analysis of type II failures In this work, we develop a deep learning-augmented de novo protein binder design protocol. We show retrospectively and pro- spectively that this improved protocol has nearly 10-fold higher suc- cess rate than the original energy-based method. Nature Communications\| (2023) 14:2625 Retrospective analysis of type I failures We began by investigating the ability of deep learning methods to discriminate binders from non-binders (a task we call ﬁltering) in the set of ~1 million experimentally characterized designs for 10 different targets described in Cao et al. 15,000–100,000 designs were experi- mentally tested for each target, and the number of actual binders ranged from 1 to 584. We ﬁrst focused on identifying Type I failures (Fig. 1) in which the designed sequence does not fold to the intended monomer structure. As a baseline, we used the Rosetta energy of the monomer, normalized by chain length (since energy is an extensive quantity). Not surprisingly as this metric was already used as a stringent ﬁlter in generating the input scaffold set for the Rosetta interface design calculations21, it provided little discriminatory power (Fig. 1d). In contrast, the deep learning-based accuracy prediction method DAN was able to partially discriminate binders from non-binders (Fig. 1d). While DAN is very fast, taking ~0.5 GPU seconds per monomer structure, AF2 structure predictions are relatively slow (~5 GPU sec- onds). As an initial test of the utility of AF2 for monomer structure modeling, we evaluated the ability of AF2 to predict the structures of the binder monomers for the ﬁve minibinder structures from Cao et al. for which structures have been solved experimentally (for designs in complex with TrkA, FGFR2, IL-7Rɑ, and the SARS-CoV-2 Spike protein). Given only the single sequence for the designed binder, AF2 predicted We used the AF2 initial guess approach and RF2 without a starting model to generate complex models for each designed sequence for each target, and compared the predicted structure of the complex to the designed complex structure. The Cɑ RMSD of Nature Communications\| (2023) 14:2625 2 2 Article https://doi.org/10.1038/s41467-023-38328-5 1 \| Monomer and protein complex structure prediction metrics distinguish iously designed binders from non-binders. a For binder design to be suc- ful, the designed sequence must fold to the designed binder monomer struc- (left), and this structure must form the designed interface with the target ein (right). b, c Design failure modes. b Type-1 Failures. The designed sequence not fold to the designed monomer structure. c Type-2 Failures. The designed sequence folds to the designed monomer structure but does not form the desig interface. Prospective analysis The retrospective analysis in Fig. 1 suggests incorporation of AF2 or RF2 into the design pipeline as a ﬁnal evaluation ﬁlter could con- siderably increase the design success rate. To directly test this hypothesis, we carried out binder design campaigns on four targets of considerable biological importance: ALK24, LTK24, IL10 receptor-ɑ (IL- 10Rɑ)25, and IL2 receptor-ɑ (IL-2Rɑ)26–29. As is clear from the retro- spective analysis of the Cao et al. data (Fig. 1d, e), binder success rate and the predictivity of metrics varies between targets: generating designs for new targets (where there is no a priori knowledge of which ﬁlters would be predictive) is the most unbiased approach for com- paring different design protocols. For IL-2Rɑ, two separate sites were targeted with independent campaigns. Using the Rosetta-based design protocol of Cao et al., we generated computational libraries of ~2 million designs for each target and ﬁltered these down to ~20,000 designs to be experimentally tested for each target: ~15,000 designs using the physically based ﬁlters of Cao et al. and ~5000 designs with AF2 pAE_interaction <10 (these designs were also ﬁltered by additional metrics as described in the Supplement). Synthetic genes were obtained for the ~80,000 designs, transformed into yeast, and the resulting library sorted for display of the proteins on yeast cells, fol- lowed by sorts at 1 μM target with avidity, and sorts at decreasing concentrations of target. The frequencyof each design at each sort was determined by deep sequencing, and SC50 values (the concentration where half of the expressing yeast-cells are collected) estimated as described in Cao et al. Designs with SC50 values better than 4 μM were considered successes; the number of successes for the four targets ranged from 1 to 17. For each target, several designs found to bind by YSD were expressed in E. coli and binding was conﬁrmed by single- concentration Biolayer Inferometry (BLI). All designs which showed binding by YSD also showed binding by BLI (Supplementary Fig. 9; for IL-10Rɑ where only one binder was identiﬁed, only this single design was screened by BLI). For all four targets, there was a considerably higher success rate (number of successes / number of designs tested) in the AF2-ﬁltered design set than in the Rosetta set (Fig. 2). Retrospective analysis of type I failures d, e The retrospective experimental success rate (YSD SC50 < 4 μM) the top 1% of designs selected according to different monomer (d) or protein complex (e) based metrics over 10 targets from Cao et al. Source data are provi as a Source Data ﬁle. sequence folds to the designed monomer structure but does not form the designed interface. d, e The retrospective experimental success rate (YSD SC50 < 4 μM) for the top 1% of designs selected according to different monomer (d) or protein complex (e) based metrics over 10 targets from Cao et al. Source data are provided as a Source Data ﬁle. Fig. 1 \| Monomer and protein complex structure prediction metrics distinguish previously designed binders from non-binders. a For binder design to be suc- cessful, the designed sequence must fold to the designed binder monomer struc- ture (left), and this structure must form the designed interface with the target protein (right). b, c Design failure modes. b Type-1 Failures. The designed sequence does not fold to the designed monomer structure. c Type-2 Failures. The designed and protein complex structure prediction metrics distinguish Nature Communications\| (2023) 14:2625 3 3 https://doi.org/10.1038/s41467-023-38328-5 Article the predicted complex to the Rosetta-designed complex model was predictive of design success in both cases (Fig. 1e). We obtained the best discrimination of binders from non-binders using the pAE prediction conﬁdence metrics produced by the two methods (Fig. 1e). For the IL7Ra, TrkA, FGFR2, InsulinR, and PDGFR datasets from Cao et al, the average pAE of interchain residue pairs (pAE_interaction) was extremely effective in identifying the experimentally conﬁrmed binders (Fig. 1e); conﬁdent predictions had very high success frequencies (see the Receiver Operator Characteristic (ROC) curves in Supplementary Fig. 6) with sharp increases in success rates for designs with pAE_interaction <10. AF2 had slightly better performance than RF2 (Fig. 1e), and we used this in the new design campaigns described in the following section. The excellent performance of both AF2 and RF2 on the binder discrimination task strongly suggest that Type II errors are primarily responsible for the low success rates of Cao et al. ﬁlter is computationally expensive (~30 GPU-seconds per design) and only ~2.3% of designs pass, so large numbers of prediction calculations must be run. Retrospective analysis of type I failures To enable the testing of large (~5,000) pools of designs, it is desirable to decrease the computational demand of the design pipeline, in particular to maximize the number of designs passing the AF2 ﬁlter a method can generate per unit compute time (the time to generate all designs and run AF2; we use a conversion factor of 100 CPU-s to 1 GPU-s because of the relative scarcity of GPU resources). ð1Þ Using this metric, we ﬁnd that Rosetta-design has an efﬁciency of about 7.6×10−7 successful designs per CPU-s equivalent. We investigated whether the recently developed deep learning graphical model based sequence design method ProteinMPNN30 could be used to increase the efﬁciency of the design pipeline. ProteinMPNN is very fast, generating a sequence for a minibinder backbone in ~2 CPU-s compared to ~350 CPU-s for Rosetta-design. We ﬁrst compared the experimental success rate of ProteinMPNN designs to Rosetta designs by generating sequences for backbones generated by AF2 for Rosetta designs to the four new targets that had low complex Cɑ RMSD to the AF2 prediction (~104 designs in total). Genes encoding designs with AF2 pAE_interaction <10 (~103 per method) were synthesized, and the binding evaluated by FACS followed by deep sequencing as described above. For each target, several designs from ProteinMPNN were expressed in E. coli and their binding was veriﬁed with BLI, we again found that all designs which bound by YSD showed binding by BLI (Supplementary Fig. 9). We found that the design success rate of ProteinMPNN and Rosetta-design were similar (Supplementary Fig. 7), thus the considerable increase in speed comes with no decrease in performance. Prospective analysis Physically based ﬁltering yielded successful binders for two targets: LTK and Site 1 of IL-2Rɑ; for these the AF2-ﬁltered libraries had 8- and 30-fold higher success rates, respectively. AF2-ﬁltered libraries also yielded success- ful binders to both ALK and IL-10Rɑ; physically based ﬁltering yielded no successful binders to either of these targets (Neither ﬁltering method was able to generate successful binders to Site 2 on IL-2Rɑ). Thus, AF2 ﬁltering performs as expected in prospective tests, increasing success rates (for targets where physically based ﬁltering is successful) and expanding the set of targets for which successful minibinders can be generated. Encouraged by the speed and performance of ProteinMPNN design, we next evaluated its efﬁciency in generating sequences pas- sing the AF2 cutoffs. ProteinMPNN design alone had an efﬁciency of 1.6 × 10−6 successful designs per CPU-s equivalent. The average of the fold efﬁciency improvement over all targets is ~5-fold greater for ProteinMPNN compared to Rosetta-design (Fig. 2c). Since unlike Rosetta, ProteinMPNN keeps the protein backbone ﬁxed, it is sensitive to the input backbone structure quality. Inspired by the very efﬁcient alternation between sequence optimization and structure reﬁnement in Rosetta ﬂexible backbone design31, we evaluated similar cycling between ProteinMPNN and Rosetta structure reﬁnement (FastRelax), hoping to converge on a high-quality backbone that would then allow ProteinMPNN to generate a high-quality sequence. This hybrid Pro- teinMPNN/Rosetta sequence design protocol (henceforth referred to as ProteinMPNN-FR) generated AF2 pAE_interaction <10 structures at a rate of ~6.6% with a throughput of 1 design per 120 CPU-s for an efﬁ- ciency of 2.2×10−6. The average per-target efﬁciency improvement of ProteinMPNN-FR over Rosetta-design is ~8-fold (Fig. 2c). Discussion These experiments show that by complementing physically based methods with deep learning-based approaches trained on large num- bers of protein structures, signiﬁcant improvements to the one-sided protein-interface design challenge can be achieved. Our retrospective and prospective studies suggest an increase in design success rate of ten fold. In contrast to Rosetta energy calculations and DAN structure accuracy measures, which operate on single protein structures (or with Rosetta relax calculations, structures very close to the query), struc- ture prediction calculations implicitly assess the ﬁt of the sequence with the desired target structure compared to all others. As observed previously32, such consideration of the overall folding landscape Increasing binder design pipeline compute efﬁciency with Pro- teinMPNN. While an effective predictor of binder success, the AF2 Nature Communications\| (2023) 14:2625 4 Article https://doi.org/10.1038/s41467-023-38328-5 poration of structure prediction metrics increases design success targets. a Results of Prospective Campaigns. For each target the SC50 shown for all designs which showed binding by YSD (like Kd’s, lower tter). The number of designs included in each library for each target is the bars in the top panel. The AF2-predicted structure of the top rget design is shown as a cartoon. No binders were identiﬁed to Site 2 of IL2 receptor-ɑ so this campaign is not included here or in panel C. b The experimental success rate for libraries ﬁltered by DL-based ﬁltering versus P cally based ﬁltering for the four prospective targets. c The computational efﬁc (the number of designs with pAE_interaction <10 per CPU-s) for the ProteinM sequence design plus Rosetta relax protocol outperforms that of the origina Rosetta sequence design protocol. Source data are provided as a Source Dat Fig. 2 \| Incorporation of structure prediction metrics increases design success rate on new targets. a Results of Prospective Campaigns. For each target the SC50 from YSD is shown for all designs which showed binding by YSD (like Kd’s, lower values are better). The number of designs included in each library for each target is indicated by the bars in the top panel. The AF2-predicted structure of the top scoring on-target design is shown as a cartoon. No binders were identiﬁed to Site 2 Fig. 2 \| Incorporation of structure prediction metrics increases design success of IL2 receptor-ɑ so this campaign is not included here or in panel C. Discussion b The experimental success rate for libraries ﬁltered by DL-based ﬁltering versus Physi- cally based ﬁltering for the four prospective targets. c The computational efﬁciency (the number of designs with pAE_interaction <10 per CPU-s) for the ProteinMPNN sequence design plus Rosetta relax protocol outperforms that of the original Rosetta sequence design protocol. Source data are provided as a Source Data ﬁle. Fig. 2 \| Incorporation of structure prediction metrics increases design success rate on new targets. a Results of Prospective Campaigns. For each target the SC50 from YSD is shown for all designs which showed binding by YSD (like Kd’s, lower values are better). The number of designs included in each library for each target is indicated by the bars in the top panel. The AF2-predicted structure of the top scoring on-target design is shown as a cartoon. No binders were identiﬁed to Site 2 Nature Communications\| (2023) 14:2625 5 5 Article https://doi.org/10.1038/s41467-023-38328-5 enables considerably more accurate assessment of the likelihood a design will fold and bind as intended compared to evaluation of only the depth of the designed energy well. Although the protocol reported hereisan order-of-magnitude improvement over the previous state-of- the-art, it is clear that much about interface energetics remains poorly understood; success rates among the targets remain low (<1%) and no binders were identiﬁed to Site 2 of IL2 receptor-ɑ. There is also con- siderable room for improvement in designing high afﬁnity; as with the original pipeline the initially generated binders are in the high nM afﬁnity range. Given the rate of progress in the ﬁeld, we anticipate further increases in design success rates and afﬁnities in the near future, which will make computational protein design methods even more powerful compared to empirical selection methods for gen- erating afﬁnity reagents and therapeutic candidates. While continued progress is nearly certain, an open question is whether this will come from integration of deep learning and physically based methods, or from deep learning alone–there are exciting times ahead! around 8 K per target. The third subset was all predictions with af2_complex_rmsd <5 Å. These designs were designated to be rede- signed and were typically about 12 K in scale. These AF2-predicted interfaces were then designed either with Rosetta or ProteinMPNN. Here, ProteinMPNN was used to generate a protein sequence from the input coordinates and no further optimi- zation was performed. Design and ﬁltering procedure for prospective study EBY100 yeast were grown in C-Trp-Ura media supplemented with 2% (w/v) glucose. Yeast cells were centrifuged and resus- pended in SGCAA media supplemented with 0.2% (w/v) glucose. Cells were resuspended to a concentration of 1×107 cells per ml and induced at 30°C for 16-24 hours. Cells were washed with PBSF (PBS with 1% (w/v) BSA) and then labeled with biotinylated target. To allow for the identiﬁcation of low afﬁnity binders, an initial sort with target avidity was performed for all libraries. In the avidity sort, the cells are incubated with biotinylated target, anti-c-Myc ﬂuorescein isothiocyanate (FITC, Miltenyi Biotech) and steptavidin-phycoerythrin (SAPE, ThermoFisher). To allow all SAPE molecules to display four biotinylated target molecules, the bio- tinylated target is provided at a 4x excess over the concentration of SAPE. When sorting without avidity, the cells are incubated ﬁrst with biotinylated target alone, then washed in PBSF and subse- quently incubated with SAPE and FITC. Each library was sorted against a titration of target concentrations. Sorts were performed using a Sony SH800S cell sorter with software version 2.1.5. The prospective study was performed at a time of rapid protocol discovery with a tight deadline for placing the gene-order. As such, not every experiment that could have been performed was performed. However, the comparison of Rosetta ﬁltering to AF2 ﬁltering was the main goal and the data required for this comparison was plentiful. The standard procedure from Cao et. al. was followed for the 4 targets starting with the following pdbs: IL2RA “1Z92”, “2B5I”, “3NFP”, “2ERJ”), IL10RA (“1LQS”), ALK (Privately communicated structure. Now “7NWZ”), LTK (Privately communicated structure. Now “7NX0”). The “recommended_scaffolds.list” from Cao et. al. were used and on the order of 10 M RifDock35 outputs were generated for each target with about 500 K FastDesigned. ~6 K motifs were extracted, grafted up to 10 M docks, and 500 K FastDesigned again. The resulting 1 M designs for each target were predicted by AF2. From this set of 1 M designs, 3 overlapping subsets were selected. The ﬁrst subset was the Rosetta-control group where the AF2 predic- tions were ignored and the top ~18 K per target were selected by the pareto-front method from Cao et al. looking at target_delta_sap, ddG, contact_patch, and contact_molec_sq5_apap_target. The second subset was the AF2-ﬁltered group where all designs passing pae_interaction <10 and af2_complex_rmsd <5 Å were included. AF2 initial guess The protein structure provided to the model as an initial guess is ﬁrst converted to AlphaFold atom positions. These positions are then provided, along with the standard model inputs into the AlphaFold Model Runner. In the AlphaFold class of the AlphaFold code, on the ﬁrst recycle, the prev_pos variable is initialized to the input AlphaFold atom positions as opposed to the standard initialization of all zeros. A script to run AF2 with an initial guess and the modiﬁed source code is provided here: https://github.com/nrbennet/dl_binder_design33. The AlphaFold model used in the script and in this work is conﬁgured to run with a reduced number of extra MSA sequences which speeds the inference of the network dramatically, as described in previous work34. Oligonucleotide libraries were ampliﬁed using Kapa HiFi poly- merase (Kapa Biosystems) with a qPCR machine (Bio-Rad, CFX96). The PCR product was run on a DNA agarose gel, the band with the correct size was cut out of the gel and cleaned (Qiagen QIAquick Clean up kit). The extracted DNA products were then re-ampliﬁed and puriﬁed fol- lowing the above protocol. The resulting DNA inserts and linearized pETcon3 vector were transformed into EBY100 yeast following an established protocol37. DNA library preparation DNA libraries were prepared in the manner described in Cao et al., we review this protocol here: The sequences of protein designs were padded to 65 amino acids through addition of a (S)n linker at the C-terminus. The protein sequences were reverse translated and codon optimized for Sacchar- omyces cerevisiae using DNAworks2.036. After reverse translation, DNA adapter sequences are added to the N (GGTGGATCAGGAGGTTCG) and C (GGAAGCGGTGGAAGTGG) terminus. Designs were purchased as oligonucleotide libraries from Agilent Technologies. Discussion The Rosetta-redesigned and ProteinMPNN- redesigned pools were predicted again by AF2 and were ﬁltered either with the Rosetta ﬁlters mentioned above or the AF2 ﬁlters mentioned above resulting in pools of sizes 9 K (Rosetta-Rosetta), 2 K (Rosetta- AF2), and 2 K (ProteinMPNN-AF2). The Rosetta ﬁlters weren’t used to ﬁlter ProteinMPNN designs because Rosetta models of ProteinMPNN outputs didn’t exist. ProteinMPNN FastRelax To prepare libraries for deep sequencing, yeast plasmids were isolated from 5 × 107 to 1 × 108 yeast cells by Zymoprep (Zymo Research). Two qPCR ampliﬁcations were then performed following the protocol in the above paragraph. Illumina adapters and 6-bp pool- speciﬁc barcodes were added in the second ampliﬁcation. The ﬁnal DNA product was puriﬁed by gel extraction. The libraries were sequenced using Illumina NextSeq sequencing. This protocol takes as input a protein complex structure. Pro- teinMPNN is then provided the complex structure with the binder sequence masked and asked to assign the binder a sequence. The new sequence is then threaded back onto the binder structure in the complex and the complex structure is relaxed using Rosetta Fas- tRelax. The relaxed complex structure can then be used as the input to ProteinMPNN to continue the cycle. A python script to perform this design technique is provided here: https://github.com/nrbennet/ dl_binder_design33. Yeast surface display Yeast surface display experiments were performed in the manner described in Cao et al., we review this protocol here: Nature Communications\| (2023) 14:2625 Article Article Article https://doi.org/10.1038/s41467-023-38328-5 Genes encoding the designed protein sequences were purchased from Integrated DNA Technologies (IDT). All genes included an N-terminal 8-His tag followed by a TEV cleavage site. The genes were cloned into modiﬁed pET-29b(+) E. coli plasmid expression vectors. Plasmids were transformed into chemically competent E. coli BL21(DE3) cells (NEB). Cells were either grown overnight in Studier autoinduction media supplemented with antibiotics or induced using the IPTG expression system and then grown overnight. Cells were then lysed by sonication and the protein samples were puriﬁed by immo- bilized metal afﬁnity chromatography (Qiagen) followed by size- exclusion fast protein liquid chromatography (Superdex 75 10/300 GL, GE Healthcare). Reporting summary Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. References 1. Nelson, A. L., Dhimolea, E. & Reichert, J. M. Development trends for human monoclonal antibody therapeutics. Nat. Rev. Drug Discov. 9, 767–774 (2010). 2. Brennan, D. J., O’Connor, D. P., Rexhepaj, E., Ponten, F. & Gallagher, W. M. Antibody-based proteomics: fast-tracking molecular diagnostics in oncology. Nat. Rev. Cancer 10, 605–617 (2010). 2. Brennan, D. J., O’Connor, D. P., Rexhepaj, E., Ponten, F. & Gallagher, W. M. Antibody-based proteomics: fast-tracking molecular diagnostics in oncology. Nat. Rev. Cancer 10, 605–617 (2010). Target protein preparation Expression and puriﬁcation of biotinylated ALK and LTK ectodo- mains. DNA encoding for the cytokine binding domains of ALK (ALKTG- EGFL, residues 648-1030) and LTK (LTKTG-EGFL, residues 63-420) were cloned in the pHLsec vector in frame with a N-terminal chicken RTPμ- like signal peptide sequence and a C-terminal Avi-tag followed by a caspase-3-cleavable Fc-Hisx6 tag38. Proteins were produced in HEK293S suspension cells maintained in growth medium consisting of 50% Freestyle (Thermoﬁsher) and 50% Ex-Cell (Sigma-Aldrich). Transient transfection was performed using linear 25 kDA polyethyleneimine (Polysciences) as transfection reagent. To allow speciﬁc in vivo biotinylation of the Avi-tag, both constructs were co-transfected with the pDisplay-BirA-ER plasmid in a 4:1 pHLsec:pDisplay stoichiometric ratio39. The growth medium was supplemented with D-biotin to a ﬁnal concentration of 100 μM to ensure complete biotinylation of the recognition sequence. After 4 days of expression, conditioned medium was clariﬁed by cen- trifugation and ﬁltered through a 0.22 μm ﬁlter prior to chromatographic steps. ngs_data_analysis/ ﬁles.ipd.uw.edu/pub/improving_dl_binders_2023/supplemental_ ﬁles/experimental_data_and_analysis.tar.gz Proteins were captured via their Fc tag on a protein A column (HiTrap Protein A HP, Cytiva) and eluted in HBS (20 mM HEPES, pH 7.4, 150 mM NaCl) after an on-column digestion with caspase-3 for 1 h at 37 °C and an additional 2-h incubation at room temperature. As a ﬁnal polishing step, recombinant proteins were concentrated and injected onto a Superdex 200 increase 10/300 GL (Cytiva) size-exclusion chromatography column pre-equilibrated with HBS. Puriﬁed biotiny- lated proteins were ﬂash frozen in liquid nitrogen and stored at −80 °C until further use. g _ _ / ﬁles.ipd.uw.edu/pub/improving_dl_binders_2023/supplemental_ ﬁles/design_models_silent.tar.gz The docks we used for the efﬁciency benchmark (6.1 GB) Contains these ﬁles: efﬁciency_benchmark_docks/ The docks we used for the efﬁciency benchmark (6.1 GB) Contains these ﬁles: efﬁciency_benchmark_docks/ ﬁles.ipd.uw.edu/pub/improving_dl_binders_2023/supplemental_ ﬁles/efﬁciency_benchmark_docks.tar.gz. Source data are provided with this paper. Biotinylated IL-10Rɑ was purchased from R&D Systems (AVI9044). Biotinylated IL-2Rɑ was purchased from Acro Biosystems (ILA-H82E6). Biolayer interferometry binding experiments Biolayer interferometry binding experiments Biolayer interferometry (BLI) measurements were performed on an Octet Red96 (ForteBio) or Octet R8 (Sartorius) instrument with Octet BLI Discovery 12.2.1.18 software, with streptavidin coated tips (Sar- torius Item no. 18-5019). The binding buffer consisted of 1X HBS-EP + buffer (Cytiva BR100669) supplemented with 1.0% w/v bovine serum albumin. 30-50 nM (depending on target availability) of target protein was loaded onto the tips. After target loading, a baseline measurement was performed in binding buffer alone for 120 s. The tips were then dipped in a solution of 500 nM (1000 nM for the IL-10Rɑ design) protein analyte in binding buffer for 600 s (association phase). The tips were then dipped back into binding buffer alone for 1000 s (dis- sociation phase). The Rosetta macromolecular modeling suite (https://www. rosettacommons.org) is freely available to academic and non- commercial users. Commercial licenses for the suite are available via the University of Washington Technology Transfer Ofﬁce. Scripts for running ProteinMPNN-FastRelax and AF2 with templating and initial guess are available at https://github.com/nrbennet/dl_binder_design33. Data availability The raw data from the prospective study, the raw scores of the ret- rospective analysis, the input structures and benchmarking scores for the efﬁciency study, and the raw data from the biolayer interferometry measurements are available at the following repository hosted by the Institute for Protein Design: Design and ﬁltering procedure for prospective study This set was typically Protein expression Proteins were expressed and puriﬁed in the manner described in Cao et al., we review this protocol here: Nature Communications\| (2023) 14:2625 6 https://doi.org/10.1038/s41467-023-38328-5 28. Wang, X., Rickert, M. & Garcia, K. C. Structure of the quaternary complex of interleukin-2 with Its α, ß, and γ c receptors. Science 310, 1159–1163 (2005). 4. Warram, J. M. et al. Antibody-based imaging strategies for cancer. Cancer Metast. Rev. 33, 809–822 (2014). 5. Gray, A. et al. Animal-free alternatives and the antibody iceberg. Nat. Biotechnol. 38, 1234–1239 (2020). 5. Gray, A. et al. Animal-free alternatives and the antibody iceberg. Nat. Biotechnol. 38, 1234–1239 (2020). 29. Rickert, M., Wang, X., Boulanger, M. J., Goriatcheva, N. & Garcia, K. C. The structure of interleukin-2 complexed with its alpha receptor. Science 308, 1477–1480 (2005). 6. Chao, G. et al. Isolating and engineering human antibodies using yeast surface display. Nat. Protoc. 1, 755–768 (2006). 30. Dauparas, J. et al. Robust deep learning–based protein sequence design using ProteinMPNN. Science 378, 49–56 (2022). 7. Hackel, B. J., Kapila, A. & Dane Wittrup, K. Picomolar afﬁnity ﬁbronectin domains engineered utilizing loop length diversity, recursive mutagenesis, and loop shufﬂing. J. Mol. Biol. 381, 1238–1252 (2008). 31. Leaver-Fay, A. et al. Rosetta3. in Methods in Enzymology vol. 487 545–574 (Elsevier, 2011). 8. Chevalier, A. et al. Massively parallel de novo protein design for targeted therapeutics. Nature 550, 74–79 (2017). 32. Norn, C. et al. Protein sequence design by conformational land- scape optimization. Proc. Natl Acad. Sci. USA 118, e2017228118 (2021). 9. Silva, D.-A. et al. De novo design of potent and selective mimics of IL-2 and IL-15. Nature 565, 186–191 (2019). 33. Bennett, N. R. et al. Improving de novo protein binder design with deep learning. nrbennet/dl_binder_design: v1.0.0 (Release). Zenodo https://doi.org/10.5281/zenodo.7730843 (2023). 10. Strauch, E.-M. et al. Computational design of trimeric inﬂuenza- neutralizing proteins targeting the hemagglutinin receptor binding site. Nat. Biotechnol. 35, 667–671 (2017). 34. Mirdita, M. et al. ColabFold: making protein folding accessible toall. Nat. Methods 19, 679–682 (2022). 11. Fleishman, S. J. et al. Computational design of proteins targeting the conserved stem region of inﬂuenza hemagglutinin. Science 332, 816–821 (2011). 35. Dou, J. et al. De novo design of a ﬂuorescence-activating β-barrel. Nature 561, 485–491 (2018). 12. Baran, D. et al. Principles for computational design of binding antibodies. Proc. Natl Acad. Sci. USA 114, 10900–10905 (2017). 36. Hoover, D. M. & Lubkowski, J. DNAWorks: an automated method for designing oligonucleotides for PCR-based gene synthesis. Nucleic Acids Res. 30, e43 (2002). 13. Cao, L. et al. https://doi.org/10.1038/s41467-023-38328-5 Design of protein-binding proteins from the target structure alone. Nature https://doi.org/10.1038/s41586-022- 04654-9 (2022). 37. Benatuil, L., Perez, J. M., Belk, J. & Hsieh, C.-M. An improved yeast transformation method for the generation of very large human antibody libraries. Protein Eng. Des. Sel. 23, 155–159 (2010). 14. Alford, R. F. et al. The rosetta all-atom energy function for macro- molecular modeling and design. J. Chem. Theory Comput. 13, 3031–3048 (2017). 38. Aricescu, A. R., Lu, W. & Jones, E. Y. A time- and cost-efﬁcient sys- tem for high-level protein production in mammalian cells. Acta Crystallogr. D Biol. Crystallogr. 62, 1243–1250 (2006). 15. Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021). 39. Howarth, M. et al. Monovalent, reduced-size quantum dots for imaging receptors on living cells. Nat. Methods 5, 397–399 (2008). 16. Baek, M. et al. Accurate prediction of protein structures and inter- actions using a three-track neural network. Science 373, 871–876 (2021). Acknowledgements 17. Baek, M. & Baker, D. Deep learning and protein structure modeling. Nat. Methods 19, 13–14 (2022). This work was supported with funds provided by The Donald and Jo Anne Petersen Endowment for Accelerating Advancements in Alzhei- mer’s Disease Research (N.R.B.), a gift from Microsoft (J.D., M.B., and D.B.), the Audacious Project at the Institute for Protein Design (A.A., D.V., and D.B.), a grant from DARPA supporting the Harnessing Enzymatic Activity for Lifesaving Remedies (HEALR) Program (HR001120S0052 contract HR0011-21-2-0012, I.G., F.D., Y.P.P., B.H., L.S., and D.B.), the Flanders Institute for Biotechnology (S.N.S), a Strategic Basic Research grant from Research Foundation Flanders (S.N.S.), and the Howard Hughes Medical Institute (B.C. and D.B.). We thank AWS and Microsoft for generous gifts of cloud computing credits and Texas Advanced Computing Center (TACC) The University of Texas at Austin for providing the CPU resources for all Rosetta design calculations. 18. Anishchenko, I. et al. De novo protein design by deep network hallucination. Nature 600, 547–552 (2021). 19. Wang, J. et al. Scaffolding protein functional sites using deep learning. Science 377, 387–394 (2022). 20. Hiranuma, N. et al. Improved protein structure reﬁnement guided by deep learning based accuracy estimation. Nat. Commun. 12, 1340 (2021). 21. Rocklin, G. J. et al. Global analysis of protein folding using massively parallel design, synthesis and testing. Science 357, 168–175 (2017). 22. Baek, M., McHugh, R., Anishchenko, I., Baker, D. & DiMaio, F. Accurate prediction of nucleic acid and protein-nucleic acid com- plexes using RoseTTAFoldNA. Preprint at https://doi.org/10.1101/ 2022.09.09.507333 (2022). Author contributions 23. Watson, J. L. et al. Broadly applicable and accurate protein design by integrating structure prediction networks and diffusion gen- erative models. Preprint at https://doi.org/10.1101/2022.12.09. 519842 (2022). N.R.B., B.C., and D.B. designed the research. N.R.B. and B.C. contributed equally. N.R.B. and B.C. developed the method. N.R.B. and B.C. designed the binders. I.G., B.H., A.A., Y.P.P., and D.V. performed the yeast screening. J.D. developed ProteinMPNN. M.B. and F.D. developed RoseTTAFold2. S.D.M. and S.N.S. solved the structures of ALK and LTK. All authors analyzed data. L.S. and D.B. supervised research. N.R.B., B.C., and D.B. wrote the manuscript with input from the other authors. All authors revised the manuscript. 24. De Munck, S. et al. Structural basis of cytokine-mediated activation of ALK family receptors. Nature 600, 143–147 (2021). 25. Jones, B. C. et al. Crystal structure of human cytomegalovirus IL-10 bound to soluble human IL-10R1. Proc. Natl Acad. Sci. USA 99, 9404–9409 (2002). 26. Stauber, D. J., Debler, E. W., Horton, P. A., Smith, K. A. & Wilson, I. A. Crystal structure of the IL-2 signaling complex: paradigm for a heterotrimeric cytokine receptor. Proc. Natl Acad. Sci. USA 103, 2788–2793 (2006). Statistics and reproducibility No statistical method was used to predetermine sample size. No data were excluded from the analyses. The experiments were not rando- mized. The Investigators were not blinded to allocation during experiments and outcome assessment. 3. Stern, L. A., Case, B. A. & Hackel, B. J. Alternative non-antibody protein scaffolds for molecular imaging of cancer. Curr. Opin. Chem. Eng. 2, 425–432 (2013). Nature Communications\| (2023) 14:2625 7 7 Article https://doi.org/10.1038/s41467-023-38328-5 Competing interests N.R.B., B.C., I.G., L.S., and D.B. are co-inventors on a United States Patent and Trademark Ofﬁce provisional patent application (63/490,479) that covers the binders designed in this study. The remaining authors declare no competing interests. 27. Yang, H. et al. Structural basis of immunosuppression by the ther- apeutic antibody daclizumab. Cell Res. 20, 1361–1371 (2010). Nature Communications\| (2023) 14:2625 8 Article Additional information Additional information Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Correspondence and requests for materials should be addressed to David Baker. Peer review information Nature Communications thanks Arne Elofsson and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review ﬁle is available. Nature Communications\| (2023) 14:2625 Reprints and permissions information is available at http://www.nature.com/reprints © The Author(s) 2023 Publisher’s note Springer Nature remains neutral with regard to jur- isdictional claims in published maps and institutional afﬁliations. Nature Communications\| (2023) 14:2625 9 9
https://openalex.org/W4367292481	https://acta-ape.org/wp-content/uploads/articles_xml/1982-0194-ape-36-eAPE01462/1982-0194-ape-36-eAPE01462.pdf	Portuguese	null	Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa	Acta Paulista de Enfermagem	2,023	cc-by	9,306	Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Technological strategies aimed at fall prevention in hospital environments: an integrative review Estrategias tecnológicas orientadas hacia la prevención de caídas en ambiente hospitalario: revisión integradora Renata Camargo Alves1 https://orcid.org/0000-0003-0955-7086 Rosana Maria Barreto Colichi1 https://orcid.org/0000-0002-8765-3965 Silvana Andrea Molina Lima1 https://orcid.org/0000-0001-9945-2928 Artigo de Revisão Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Technological strategies aimed at fall prevention in hospital environments: an integrative review Estrategias tecnológicas orientadas hacia la prevención de caídas en ambiente hospitalario: revisión integradora Renata Camargo Alves1 https://orcid.org/0000-0003-0955-7086 Rosana Maria Barreto Colichi1 https://orcid.org/0000-0002-8765-3965 Silvana Andrea Molina Lima1 https://orcid.org/0000-0001-9945-2928 Artigo de Revisão Artigo de Revisão Renata Camargo Alves1 https://orcid.org/0000-0003-0955-7086 Rosana Maria Barreto Colichi1 https://orcid.org/0000-0002-8765-3965 Silvana Andrea Molina Lima1 https://orcid.org/0000-0001-9945-2928 Como citar: Alves RC, Colichi RM, Lima SA. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa. Acta Paul Enferm. 2023;36:eAPE01462. DOI http://dx.doi.org/10.37689/acta-ape/2023AR01462 Descritores Descritores Segurança do paciente; Acidentes por quedas; Tecnologia; Hospitais; Prevenção de acidentes Keywords Patient safety; Accidental falls; Technology; Hospitals; Accident prevention Descriptores Seguridad del paciente; Accidentes por caídas; Tecnología; Hospitales; Prevención de accidentes Submetido 25 de Julho de 2022 Aceito 24 de Outubro de 2022 Autor correspondente Renata Camargo Alves E-mail: recamargoalves@gmail.com Editor Associado (Avaliação pelos pares): Ana Lucia de Moraes Horta (https://orcid.org/0000-0001-5643-3321) Escola Paulista de Enfermagem, Universidade Federal de São Paulo, São Paulo, SP, Brasil Descritores Segurança do paciente; Acidentes por quedas; Tecnologia; Hospitais; Prevenção de acidentes Keywords Patient safety; Accidental falls; Technology; Hospitals; Accident prevention Descriptores Seguridad del paciente; Accidentes por caídas; Tecnología; Hospitales; Prevención de accidentes Submetido 25 de Julho de 2022 Aceito 24 de Outubro de 2022 Autor correspondente Renata Camargo Alves E-mail: recamargoalves@gmail.com Editor Associado (Avaliação pelos pares): Ana Lucia de Moraes Horta (https://orcid.org/0000-0001-5643-3321) Escola Paulista de Enfermagem, Universidade Federal de São Paulo, São Paulo, SP, Brasil Descritores Segurança do paciente; Acidentes por quedas; Tecnologia; Hospitais; Prevenção de acidentes Keywords Patient safety; Accidental falls; Technology; Hospitals; Accident prevention Descriptores Seguridad del paciente; Accidentes por caídas; Tecnología; Hospitales; Prevención de accidentes Submetido 25 de Julho de 2022 Aceito 24 de Outubro de 2022 Autor correspondente Renata Camargo Alves E-mail: recamargoalves@gmail.com Editor Associado (Avaliação pelos pares): Ana Lucia de Moraes Horta (https://orcid.org/0000-0001-5643-3321) Escola Paulista de Enfermagem, Universidade Federal de São Paulo, São Paulo, SP, Brasil Resumo Como citar: Alves RC, Colichi RM, Lima SA. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa. Acta Paul Enferm. 2023;36:eAPE01462. Como citar: Alves RC, Colichi RM, Lima SA. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa. Acta Paul Enferm. 2023;36:eAPE01462. Objetivo: Identificar o conhecimento produzido sobre estratégias tecnológicas voltadas para prevenção de quedas de adultos no ambiente hospitalar. Métodos: Revisão integrativa da literatura realizada nas bases de dados Medical Literature Analysis and Retrieval System Online/National Library of Medicine (MEDLINE/ PubMed), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS), Scopus e Web of Science publicados no período de 2017 a 2022. Resultados: Foram incluídos 19 artigos, a maioria publicado em revistas de Enfermagem (42%). Os temas que emergiram da análise foram: tipos de estratégias tecnológicas, custos, privacidade do paciente e relevância dos profissionais de enfermagem no desenvolvimento, avaliação e utilização dos dispositivos. Conclusão: A pesquisa revelou uma gama de recursos tecnológicos voltados para prevenção de quedas disponíveis no mercado que podem ser incorporados nas práticas assistenciais. Cabe aos hospitais definirem qual estratégia aplica-se de forma mais adequada à sua realidade. Abstract Objective: To identify the knowledge produced about technological strategies aimed at fall prevention of adults in hospital environments. Methods: This is an integrative literature review conducted in the Medical Literature Analysis and Retrieval System Online/National Library of Medicine (MEDLINE/PubMed), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Latin American and Caribbean Literature in Health Sciences (LILACS), Scopus and Web of Science databases published from 2017 to 2022. Results: A total of 19 articles were included, most of them published in nursing journals (42%). The themes that emerged from the analysis were: types of technological strategies, costs, patient privacy and nursing professionals’ relevance in device development, assessment and use. Conclusion: The research revealed a range of technological resources aimed at fall prevention available in the market that can be incorporated into care practices. Hospitals must define which strategy applies most appropriately to their reality. Resumen Objetivo: Identificar el conocimiento producido sobre estrategias tecnológicas para la prevención de caídas de adultos en el ambiente hospitalario. 1Departamento de Enfermagem, Faculdade de Medicina de Botucatu, Universidade Estadual Paulista Júlio de Mesquita Filho, Botucatu, SP, Brasil. Conflitos de interesse: nada a declarar 1 1 Acta Paul Enferm. 2023; 36:eAPE01462. Acta Paul Enferm. 2023; 36:eAPE01462. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Métodos: Revisión integradora de la literatura realizada en las bases de datos Medical Literature Analysis and Retrieval System Online/National Library of Medicine (MEDLINE/ PubMed), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Literatura Latinoamericana y del Caribe en Ciencias de la Salud (LILACS), Scopus y Web of Science publicados durante el período de 2017 a 2022. Resultados: Se incluyeron 19 artículos, la mayor parte publicados en revistas de Enfermería (42 %). Los temas que surgieron del análisis fueron: tipos de estrategias tecnológicas, costos, privacidad del paciente y relevancia de los profesionales de enfermería en el desarrollo, evaluación y utilización de los dispositivos. Conclusión: La investigación reveló una variedad de recursos tecnológicos para la prevención de caídas disponibles en el mercado que pueden ser incorporados en las prácticas asistenciales. Cabe a los hospitales definir qué estrategia se aplica de forma más adecuada a su realidad. Introdução Apesar de sua importância, a incorporação de inovações na área da saúde ainda é um desafio, não só pela operacionalização tecnológica, mas pela in- tegração de questões humanísticas relacionadas ao cuidado, bem como aspectos comportamentais e culturais. Por isso, as aplicações dos resultados ad- vindos das inovações nos segmentos tecnológicos ainda apresentam obstáculos na enfermagem, sendo sua implementação prática realizada de forma mui- to lenta. Neste sentido, a divulgação de soluções, propostas inovadoras e patentes podem reduzir a lacuna entre teoria e aplicabilidade prática.(7) Fornecer assistência livre de danos e consolidar- -se como uma organização de alta confiabilidade é um marco almejado pelas instituições hospitalares e consequentemente, um grande desafio já que a prestação de cuidados de saúde é considerada uma atividade de alta complexidade. A busca por novos tratamentos, equipamentos mais eficientes e inova- ções em tecnologia representam uma valiosa contri- buição para as práticas hospitalares.(1,2) Com o intuito de construir uma cultura de se- gurança e mitigar a ocorrência de eventos adversos como a queda, os sistemas de saúde buscam adotar os princípios de uma organização de alta confiabi- lidade que incluem o desenvolvimento das lideran- ças, implementação de intervenções de melhoria da qualidade, o apoio a cultura justa e as recomenda- ções de organizações como a Joint Comission e do Institute for Healthcare Improvement (IHI).(3) Considerando a evolução científica atual, obser- vamos o surgimento de diversos recursos voltados aos acidentes por quedas com objetivo de detectar a ocorrência desses eventos. Entretanto, visando melhores padrões de qualidade, avançamos para a busca por dispositivos para previsão e prevenção dos mesmos, com maior potencial de melhorar a segurança e a qualidade de vida do paciente.(8) Neste sentido, o presente estudo objetivou identificar o conhecimento produzido sobre estratégias tecnoló- gicas voltadas para prevenção de quedas de adultos no ambiente hospitalar. Proposto pela IHI, as hierarquias das ações têm como finalidade auxiliar as equipes a identificar quais ações proporcionam o efeito mais forte para o sucesso e melhoria sustentada do sistema. Assim, diante à ocorrência de um evento adverso e adequa- da análise de causa raiz estabelecem-se ações mais fortes que mitiguem essas fragilidades e proporcio- nem melhoria contínua, reduzindo as chances de o evento se repetir.(3,4) Acta Paul Enferm. 2023; 36:eAPE01462. Métodos Trata-se de um estudo de revisão integrativa que tem como finalidade uma análise ampla da lite- ratura, sintetizando os resultados obtidos em pes- quisas acerca de tema ou questão por meio de uma questão norteadora.(9) Foram percorridas as etapas a seguir: 1) Identificação do tema apresentado em forma de questão; 2) Definição dos critérios de in- clusão e exclusão dos estudos; 3) Seleção da amos- tra; 4) Inclusão dos estudos selecionado; 5) Análise dos resultados com a identificação de similaridades As ações fortes requerem menos dependência humana e englobam categorias de atividades como questões arquitetônicas e físicas, implantação de novos dispositivos com teste de usabilidade, pa- dronização de equipamentos ou processos, envolvi- mento tangível da liderança, além da aplicação de estratégias tecnológicas, as quais vem apresentando impacto significativo na diminuição da prevalência dos acidentes por queda em ambiente hospitalar.(4-6) 2 Alves RC, Colichi RM, Lima SA Resultado da busca (n = 1445) LILACS (n = 14) SCOPUS (n = 321) CINAHL (n = 271) Medline (n = 726) Web of Science (n = 113) Selecionados para a leitura dos títulos e resumos (n = 1300) Artigos incluídos para revisão da análise (n = 19) Leitura na íntegra dos artigos selecionados potencialmente elegíveis (n = 410) Duplicados removidos (n = 145) Excluídos ao serem aplicados os critérios de exclusão (n = 890) Excluídos (n = 391) 4 (tipo de estudo), 1 (criança), 2 (detecção de queda), 2 (ambiente não hospitalar) e 382 (resumo) IDENTIFICAÇÃO SELEÇÃO ELEGIBILIDADE INCLUÍDO Figura 1. Fluxograma do processo de seleção de artigos para composição da amostra e conflitos; 6) Apresentação e discussão dos da- dos.(10) As recomendações do fluxograma Preferred Reporting Items for Systematic reviews and Meta- Analyses (PRISMA) foram utilizadas como referên- cia para execução da pesquisa e relatar os resultados. A busca nas bases de dados eletrônicas foi realizada no período de março a abril de 2022.(11) p ç Foram incluídos artigos publicados nos últi- mos cinco anos, no idioma inglês, português e es- panhol que abordassem a utilização de estratégias tecnológicas para prevenção de quedas de adultos no ambiente hospitalar. Excluíram-se dissertações, teses, documentos duplicados, editoriais, revisões, bem como aqueles não correspondentes à questão norteadora do estudo, tais como tecnologias exclu- sivas para público infantil ou não aplicáveis para o ambiente hospitalar. Métodos Para a definição do problema, a questão norteadora da revisão foi: Como se carac- teriza a produção de conhecimento sobre estratégias tecnológicas voltadas para prevenção de queda de adultos no ambiente hospitalar? A busca foi reali- zada pelo acesso online nas seguintes bases de dados: Medical Literature Analysis and Retrieval System Online/National Library of Medicine (MEDLINE/ PubMed), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Literatura Latino- Americana e do Caribe em Ciências da Saúde (LILACS), Scopus e Web of Science. Figura 1. Fluxograma do processo de seleção de artigos para composição da amostra tematizados em um quadro síntese com as seguintes informações: número do artigo, título, autores, ano e local em que foi publicado, base de dados, tipo de estudo, especificação tecnológica, principais acha- dos, limitações e escopo da revista. O estudo está vinculado ao projeto Registro de patentes relaciona- dos à enfermagem, CAAE: 47943621.5.0000.5411 e aprovado pelo Comitê de Ética em Pesquisa da Faculdade de Medicina de Botucatu da UNESP pelo parecer 4.841.423. No intuito de assegurar ampla busca de estudos primários, a estratégia compreendeu os descritores controlados da terminologia em saúde dos Descritores em Ciências da Saúde (DeCS): “Accidental Falls”, “Acidentes por Quedas”, “Accidentes por Caídas”, “Technology”, “Tecnologia”, “Tecnología”, “Hospitals”, “Hospitales” e “Hospitais”, que foram combinados utilizando-se os operadores booleanos OR e AND. A figura 1 apresenta o processo de sele- ção dos artigos até a obtenção final da amostra. Resultados Dos 1445 artigos localizados, 19 foram estabele- cidos como corpo de análise, sendo a maioria pu- blicados em 2017(12-18) e 2018(19-22) (58%), nos Estados Unidos(14,15,17-19,23-26) (47,4%), Austrália,(12,13) Alemanha,(27,28) Brasil(21,22) e Japão(16,20) (10,5%) e Reino Unido(29) e Singapura(30) (5,3%). Identificamos 14 estudos de usabilidade e apresentação de produ- tos,(13-16,18,19,21-28,31) dois estudos de casos controle,(17,20) um estudo retrospectivo comparativo,(30) um estudo controlado randomizado(29) e um prospectivo com- parativo.(12)A maioria dos artigos foi publicada em re- vistas especializadas em Enfermagem.(17,19,20,21,25,26,28,30) Inicialmente realizou-se a leitura de título e re- sumo de todos os artigos rastreados; em seguida, efe- tuou-se leitura detalhada dos artigos selecionados e realizada a escolha daqueles para compor a amostra final. A busca e seleção foram realizadas por reviso- res de forma independente por dois pesquisadores, recorrendo-se a um terceiro pesquisador para os ca- sos de divergência. Os dados foram coletados e sis- 3 3 Acta Paul Enferm. 2023; 36:eAPE01462. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa cinco para dispositivos vestíveis,(13,14,23,27,28) cinco para programas/ferramentas de tecnologia da informa- ção(12,16,20,22,24) e dois artigos eram voltados para proces- sos educativos.(17,26) O trabalho possibilitou identificar estratégias tecnológicas voltadas para prevenção de que- das no ambiente hospitalar, conforme demonstrado no quadro 1. (8/42,1%), seguido por Gerontologia e Geriatria(12,14,23,24) (4/22,1%) e Informática Médica(15,16,26) (3/15,8%). As demais publicações foram em revistas com escopo de Engenharia(13,27) (2/10,5%), Segurança do Paciente(18) (1/5,3%) e Sociologia Médica(29) (1/5,3%). Dentre as especificações das estratégias tecnológicas, sete artigos referiam-se a dispositivos para ambiente,(15,18,19,21,25,29,30) Quadro 1. Síntese dos estudos incluídos na revisão Sequência Autores/Ano/ País Objetivo Especificação tecnológica Principais achados 1 Teh RC, et al. 2017. Austrália(12) Avaliar a tecnologia da informação em saúde (HIT) comparada à ferramenta Fall Risk for Older Persons (FROP) na triagem de risco de quedas. Programa/ ferramentas de tecnologia da informação O sistema HIT apresentou a mesma aceitabilidade e pontuação em comparação com o FROP. 2 Wickramasinghe A, et al. 2017. Austrália(13) Analisar as respostas de etiqueta de identificação por radiofrequência (RFID) sem bateria mediante movimentos de saída do leito. Dispositivo vestível O uso da etiqueta de identificação sem bateria indicou eficácia da estrutura de monitoramento de saída do leito. 3 Razjouyan J, et al. 2017. EUA(14) Avaliar a viabilidade do uso de um sensor no peito durante a hospitalização para determinar o risco de queda. Resultados Dispositivo vestível O sensor demonstrou viabilidade da tecnologia vestível para monitorar a atividade física, sono, posturas e variação de frequência cardíaca como potenciais marcadores de risco de queda. 4 Balaguera HU, et al. 2017.EUA(15) Realizar uma avaliação de tecnologia, incluindo viabilidade, usabilidade e experiência do usuário, de um sistema de Intranet das Coisas (IoT) baseado em sensores médicos para facilitar a resposta de enfermagem às saídas de leito em um hospital de cuidados intensivos. Dispositivo para ambiente O sistema de IoT integrado ao fluxo de trabalho de enfermagem reduziu o risco de queda no leito do paciente. 5 Yokota S, et al. 2017. Japão(16) Avaliar o risco de queda de paciente internado por meio de máquina de vetores de suporte. Programa/ ferramentas de tecnologia da informação A adoção do método que se baseia em dados diários registrados no prontuário eletrônico foi capaz de avaliar o risco de queda sem gerar sobrecarga a equipe de enfermagem. 6 Opsahl AG, et al. 2017. EUA(17) Avaliar o uso de intervenções agrupadas de prevenção de quedas destacadas por um vídeo educativo de envolvimento do paciente/família. Processo educativo O programa de prevenção de quedas que inclui o vídeo educativo para pacientes e seus familiares resultou em tendências positivas para uma diminuição da taxa de queda no cenário de cuidados, além de proporcionar a família a oportunidade de participar da discussão e do processo educativo. 7 Potter P, et al. 2017. EUA(18) Relatar a primeira aplicação do sistema completo de sensor de detecção de queda para uso na prevenção de quedas em um ambiente de cuidados agudos. Dispositivo para ambiente O sistema de sensor de detecção de queda proporcionou à equipe uma melhor consciência de prevenção de quedas e ofereceu um nível de vigilância que os sistemas padrão de alerta de queda não oferecem. 8 Cournan M, et al. 2018. EUA(19) Comparar as taxas de quedas antes e após a implantação de um sistema de videomonitoramento, além de quantificar os custos economizados com a redução de quedas. Dispositivos para ambiente A implantação do videomonitoramento resultou na redução de quedas e de custos por diminuir a necessidade de acompanhantes (sitter), além de melhorar a satisfação do paciente, família e equipe. 9 Yokota S, et al. 2018. Japão(20) Investigar o efeito do uso de uma ferramenta de triagem de risco de queda em um sistema de prontuário eletrônico. Acta Paul Enferm. 2023; 36:eAPE01462. Resultados Processo educativo As 3 modalidades do Fall TIPS facilitaram efetivamente o envolvimento do paciente no processo de prevenção de quedas e provaram ser uma estratégia eficaz para a disseminação, permitindo que as instituições escolham a modalidade que mais se adequa à sua realidade. 16 Jähne-Raden N, et al. 2019. Alemanha(27) Apresentar o Inexpensive Node for Bed-Exit Detection (INBED), um sistema de sinalização para detecção de saída do leito e prevenção de quedas. Dispositivo vestível sem fio Dispositivo vestível que detecta movimentação do paciente e seus dados são encaminhados a uma central ligada à equipe de enfermagem através de uma infraestrutura sem fio. 17 Jähne-Raden N, eta al. 2019. Alemanha(28) Apresentar o Inexpensive Node for Bed Exit Detection (INBED), um sistema de sinalização abrangente para prevenção de quedas. Dispositivo vestível O Inexpensive Node for Bed Exit Detection (INBED), um sistema de sinalização abrangente para prevenção de quedas pode detectar vários tipos de movimento que são encaminhadas à equipe de enfermagem por rede. O sistema pode ajudar a aliviar a equipe de assistencial e proporcionar maior liberdade de movimento e privacidade aos pacientes. 18 Timmons S, et al. 2019. Reino Unido(29) Analisar as fragilidades de sensores de pressão de cama e de cadeira para prevenir quedas à beira do leito em idosos internados no hospital. Dispositivo para ambiente Os sensores de pressão de cama e cadeira não são rentáveis, pois não reduziram as quedas. 19 Seow JP, et al. 2022. Singapura(30) Examinar a eficácia de um sistema integrado de alarme de saída de leito de três modos na redução de quedas de pacientes internados em um ambiente hospitalar de cuidados agudos em Cingapura. Dispositivo para ambiente A utilização de sistema de alarmes de saída do leito está associado a redução da incidência de quedas. Figura 2. Tipos de estratégias tecnológicas Estratégias tecnológicas para prevenção de quedas Processo educativo Dispositivos vestíveis Dispositivos para ambientes Programa/ferramentas de tecnologia da informação Vídeo educativo Relógios Vestuários Alarmes Sensores Monitores Vídeo monitoramento Software Aplicativo Figura 2. Tipos de estratégias tecnológicas Estratégias tecnológicas para prevenção de quedas Processo educativo Dispositivos vestíveis Dispositivos para ambientes Programa/ferramentas de tecnologia da informação Vídeo educativo Relógios Vestuários Alarmes Sensores Monitores Vídeo monitoramento Software Aplicativo Estratégias tecnológicas para prevenção de quedas Figura 2. Resultados Processo educativo As 3 modalidades do Fall TIPS facilitaram efetivamente o envolvimento do paciente no processo de prevenção de quedas e provaram ser uma estratégia eficaz para a disseminação, permitindo que as instituições escolham a modalidade que mais se adequa à sua realidade. 16 Jähne-Raden N, et al. 2019. Alemanha(27) Apresentar o Inexpensive Node for Bed-Exit Detection (INBED), um sistema de sinalização para detecção de saída do leito e prevenção de quedas. Dispositivo vestível sem fio Dispositivo vestível que detecta movimentação do paciente e seus dados são encaminhados a uma central ligada à equipe de enfermagem através de uma infraestrutura sem fio. 17 Jähne-Raden N, eta al. 2019. Alemanha(28) Apresentar o Inexpensive Node for Bed Exit Detection (INBED), um sistema de sinalização abrangente para prevenção de quedas. Dispositivo vestível O Inexpensive Node for Bed Exit Detection (INBED), um sistema de sinalização abrangente para prevenção de quedas pode detectar vários tipos de movimento que são encaminhadas à equipe de enfermagem por rede. O sistema pode ajudar a aliviar a equipe de assistencial e proporcionar maior liberdade de movimento e privacidade aos pacientes. 18 Timmons S, et al. 2019. Reino Unido(29) Analisar as fragilidades de sensores de pressão de cama e de cadeira para prevenir quedas à beira do leito em idosos internados no hospital. Dispositivo para ambiente Os sensores de pressão de cama e cadeira não são rentáveis, pois não reduziram as quedas. 19 Seow JP, et al. 2022. Singapura(30) Examinar a eficácia de um sistema integrado de alarme de saída de leito de três modos na redução de quedas de pacientes internados em um ambiente hospitalar de cuidados agudos em Cingapura. Dispositivo para ambiente A utilização de sistema de alarmes de saída do leito está associado a redução da incidência de quedas. Figura 2. Tipos de estratégias tecnológicas Estratégias tecnológicas para prevenção de quedas Processo educativo Dispositivos vestíveis Dispositivos para ambientes Programa/ferramentas de tecnologia da informação Vídeo educativo Relógios Vestuários Alarmes Sensores Monitores Vídeo monitoramento Software Aplicativo Sequência Autores/Ano/ País Objetivo Especificação tecnológica Principais achados 15 Duckworth M, et al. 2019. EUA(26) Examinar se a modalidade Fall TIPS afeta o envolvimento do paciente no processo de prevenção de quedas em 3 etapas e sua eficácia. Resultados Tipos de estratégias tecnológicas Após análise, sob as perspectivas propostas pelas autoras, foram encontradas pesquisas sobre estraté- gias tecnológicas voltadas para prevenção de quedas do adulto sob vários aspectos, emergindo os seguin- tes temas: a) tipos de estratégias tecnológicas; b) custos; c) privacidade do paciente; d) relevância dos profissionais de enfermagem no desenvolvimento, avaliação e utilização dos dispositivos. Os tipos de estratégias tecnológicas são apresentados em quatro vertentes conforme demonstrado na figura 2. No tema custos, além da análise para implantação, são apresentados estudos que revelam dispositivos que podem reduzir gastos com cuidado, ou ainda aque- les dispendidos no tratamento dos danos decorrente de queda no ambiente hospitalar. A privacidade é abordada em questões relacionadas a aceitabilidade por parte do paciente, família e equipe, além de res- trições e aspectos legais. O último tema retrata a im- portância do papel do enfermeiro na avaliação e uti- lização dos dispositivos, como critério de aceitabili- dade dos produtos, além de apontar a necessidade de envolver estes profissionais no desenvolvimento de tecnologias assistivas, voltadas para prevenção de quedas no ambiente hospitalar. Resultados Programa/ ferramentas de tecnologia da informação A probabilidade de ocorrência de quedas diminuiu após a implementação da ferramenta, mas o seu uso não reduziu a ocorrência real de quedas, reforçando a necessidade de se manter as medidas preventivas adotadas. 10 Sakano LMN, et al. 2018. Brasil(21) Descrever a elaboração do dispositivo de prevenção de quedas para idosos durante o uso do sanitário. Dispositivo para ambiente O dispositivo “Abraço Seguro” foi avaliado positivamente pela equipe de saúde, pacientes e cuidadores. Mediante o benefício obtido, o dispositivo foi patenteado para que outras instituições possam usufruir do recurso. 11 Nuñez Filha MCA, et al. 2018. Brasil(22) Desenvolver uma ferramenta eletrônica web mobile que auxilie os profissionais de saúde na avaliação do risco de quedas do idoso a partir de instrumentos adaptados para o português do Brasil, validados para a população idosa e com amplo uso na literatura. Programa/ ferramentas de tecnologia da informação O aplicativo desenvolvido apresentou-se rápido, de fácil operação e possuindo os principais instrumentos para identificar o risco de quedas em idosos. 12 Baker PA, et al. 2021. EUA(23) Avaliar se as meias PUP® (Patient Is Up) Smart Socks previnem quedas de pacientes com alto risco de queda em ambiente hospitalar. Dispositivo vestível sem fio A tecnologia resultou em uma redução significativa nas taxas de queda entre pacientes com alto risco de queda. 13 Sun R, et al. 2019. EUA(24) Avaliar a validade preliminar e usabilidade de um sistema de avaliação de risco de queda baseado em câmera Kinect em um ambiente hospitalar. Programa/ ferramentas de tecnologia da informação O sistema oferece a avaliação de risco de queda objetiva, de baixo custo e autoguiada, além de ter sido aceito pelo público idoso. 14 Hebb A, et al. 2021. EUA(25) Determinar a satisfação da equipe e a aceitação da tecnologia de monitoramento contínuo de vídeo (CVM) em comparação com assistentes. Dispositivos para ambiente Compreender a satisfação da equipe e a aceitação da tecnologia é imperativo para os líderes de enfermagem e administração ao implementar novas tecnologias. Quadro 1. Síntese dos estudos incluídos na revisão Acta Paul Enferm. 2023; 36:eAPE01462. 4 Acta Paul Enferm. 2023; 36:eAPE01462. Alves RC, Colichi RM, Lima SA Sequência Autores/Ano/ País Objetivo Especificação tecnológica Principais achados 15 Duckworth M, et al. 2019. EUA(26) Examinar se a modalidade Fall TIPS afeta o envolvimento do paciente no processo de prevenção de quedas em 3 etapas e sua eficácia. Discussão Nosso estudo retrata a diversidade de estratégias que vem sendo desenvolvidas para prevenção de quedas, demonstrando principalmente a usabilidade destes recursos. Os achados do estudo revelaram que países desenvolvidos tendem a investir mais em tecnologia e inovação refletindo a preocupação com a segurança do paciente, bem como atendendo ao envelhecimento po- pulacional presente nestes países. No Brasil, apesar da 5 Acta Paul Enferm. 2023; 36:eAPE01462. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa parâmetros biológicos, incluindo ECG, frequên- cia respiratória, temperatura corporal e aceleração tridimensional.(13,23,27,28) Entre os softwares e aplica- tivos desenvolvidos como recurso para prevenção de quedas, observamos sistemas para triagem de riscos utilizando-se de câmera Kinect e máquina de vetores de suporte, além daqueles fundamentados em Health Information Technology (HIT), Fall Risk for Older Persons (FROP). Todos eles buscam parametrizar os riscos de quedas com a captação ou utilização de infor- mações armazenadas anteriormente em banco de da- dos como prontuário eletrônico. No entanto, existem softwares mais acessíveis que permitem o uso infor- matizado de escalas validadas e já utilizadas na prática profissional, tais como Escala de Equilíbrio de Berg, Escala de Tinetti, Índice Dinâmico da Marcha e o teste Timed up & go (TUG), oferecendo um escore final e possibilidade de emissão de relatórios.(12,16,20,22,24) O vídeo educativo, planejado e desenvolvido para uso da equipe assistencial, visa a adoção de medidas preventi- vas de quedas, permitindo o envolvimento dos fami- liares na discussão e sensibilização sobre o tema. Ainda sobre os processos educativos, o Fall TIPS Toolkit, desenvolvido nos EUA, baseia-se na avaliação de risco realizada pelo profissional de enfermagem que em segui- da elabora o plano de prevenção de quedas personaliza- do e estas informações são disponibilizadas em monitor beira leito para o paciente e familiares proporcionando engajamento, evidenciando a adesão de mais de 80% ao protocolo.(17,26) A ampliação da gama de dispositivos impacta diretamente na saúde da população à medida que são disponibilizados nas redes de hospitais. Discussão Ainda sobre os processos educativos, o Fall TIPS Toolkit, desenvolvido nos EUA, baseia-se na avaliação de risco realizada pelo profissional de enfermagem que em segui- da elabora o plano de prevenção de quedas personaliza- do e estas informações são disponibilizadas em monitor beira leito para o paciente e familiares proporcionando engajamento, evidenciando a adesão de mais de 80% ao protocolo.(17,26) A ampliação da gama de dispositivos impacta diretamente na saúde da população à medida que são disponibilizados nas redes de hospitais. Apesar da evolução significativa da incorporação de tecnologias em saúde no âmbito público brasileiro, aprimoramen- tos constantes são necessários para que este processo ocorra de forma sustentada e transparente, garantindo eficácia e segurança aos usuários.(38) escassez de investimentos na ciência inovativa, obser- va-se a tendência crescente de óbitos por quedas com o aumento da idade que podem ser evitados mediante a adoção de medidas preventivas.(13-17,19,22-26,28,32,33) Tipos de estratégias tecnológicas Com avanço tecnológico, observa-se aumento do emprego de estratégicas tecnológicas voltadas para o monitoramento e prevenção de quedas no ambiente hospitalar. Fato positivo, visto que a hospitalização e o tempo de internação aumentam consideravelmente o risco de queda, devido a mudança e desconhecimento do ambiente.(34-36) Dispositivos como hardwares e soft- wares, sensores e tecnologia da informação possibili- taram a integração do ambiente virtual com ativida- des cotidianas presenciais, além de contribuírem para a ampliação do acesso às informações e ao cuidado. (37) No estudo foi possível identificar dispositivos para ambientes, tais como monitoramento por vídeo que é um mecanismo não invasivo de prevenção de que- das de pacientes internados, em reabilitação ou ainda com deficiência cognitiva, podendo ainda ser utili- zadas tecnologias de sistemas de captura de imagem 3D com sensor detector de movimento. Sistemas in- tegrados de alarmes capazes de sinalizar a tentativa e a saída do paciente do leito, almofada sensível à pressão que é colocada sob o colchão da cama e a almofada da cadeira que transmite um alerta para um monitor central na enfermaria se a pressão na placa for reduzi- da e sensor sem fio colocados no leito para direcionar lesões por pressão e monitoramento de sinais vitais. Com avanço tecnológico, observa-se aumento do emprego de estratégicas tecnológicas voltadas para o monitoramento e prevenção de quedas no ambiente hospitalar. Fato positivo, visto que a hospitalização e o tempo de internação aumentam consideravelmente o risco de queda, devido a mudança e desconhecimento do ambiente.(34-36) Dispositivos como hardwares e soft- wares, sensores e tecnologia da informação possibili- taram a integração do ambiente virtual com ativida- des cotidianas presenciais, além de contribuírem para a ampliação do acesso às informações e ao cuidado. (37) No estudo foi possível identificar dispositivos para ambientes, tais como monitoramento por vídeo que é um mecanismo não invasivo de prevenção de que- das de pacientes internados, em reabilitação ou ainda com deficiência cognitiva, podendo ainda ser utili- zadas tecnologias de sistemas de captura de imagem 3D com sensor detector de movimento. Tipos de estratégias tecnológicas Sistemas in- tegrados de alarmes capazes de sinalizar a tentativa e a saída do paciente do leito, almofada sensível à pressão que é colocada sob o colchão da cama e a almofada da cadeira que transmite um alerta para um monitor central na enfermaria se a pressão na placa for reduzi- da e sensor sem fio colocados no leito para direcionar lesões por pressão e monitoramento de sinais vitais. Equipamentos mecânicos para contenção também representam importante papel na prevenção de que- das como o dispositivo desenvolvido para uso seguro do vaso sanitário.(15,18,19,21,25,29,30) Dentre os dispositivos vestíveis, encontramos o Inexpensive Node for Bed- Exit Detection (INBED) recurso a ser colocado na metade superior da perna do paciente para detectar vários tipos de movimentos, as meias inteligentes PUP que sinalizam quando o paciente sai da cama e pisa no chão, detectando a pressão pelos sensores da meia e acionando um alerta, a etiqueta de identificação por radiofrequência (RFID) sem bateria anexada ao tó- rax do paciente usando eletrodos padrão de eletro- cardiograma (ECG) para analisar os movimentos p vídeo educativo, planejado e desenvolvido para uso da equipe assistencial, visa a adoção de medidas preventi- vas de quedas, permitindo o envolvimento dos fami- liares na discussão e sensibilização sobre o tema. Ainda sobre os processos educativos, o Fall TIPS Toolkit, desenvolvido nos EUA, baseia-se na avaliação de risco realizada pelo profissional de enfermagem que em segui- da elabora o plano de prevenção de quedas personaliza- do e estas informações são disponibilizadas em monitor beira leito para o paciente e familiares proporcionando engajamento, evidenciando a adesão de mais de 80% ao protocolo.(17,26) A ampliação da gama de dispositivos impacta diretamente na saúde da população à medida que são disponibilizados nas redes de hospitais. Apesar da evolução significativa da incorporação de tecnologias em saúde no âmbito público brasileiro, aprimoramen- tos constantes são necessários para que este processo ocorra de forma sustentada e transparente, garantindo eficácia e segurança aos usuários.(38) Discussão Apesar da evolução significativa da incorporação de tecnologias em saúde no âmbito público brasileiro, aprimoramen- tos constantes são necessários para que este processo ocorra de forma sustentada e transparente, garantindo eficácia e segurança aos usuários.(38) parâmetros biológicos, incluindo ECG, frequên- cia respiratória, temperatura corporal e aceleração tridimensional.(13,23,27,28) Entre os softwares e aplica- tivos desenvolvidos como recurso para prevenção de quedas, observamos sistemas para triagem de riscos utilizando-se de câmera Kinect e máquina de vetores de suporte, além daqueles fundamentados em Health Information Technology (HIT), Fall Risk for Older Persons (FROP). Todos eles buscam parametrizar os riscos de quedas com a captação ou utilização de infor- mações armazenadas anteriormente em banco de da- dos como prontuário eletrônico. No entanto, existem softwares mais acessíveis que permitem o uso infor- matizado de escalas validadas e já utilizadas na prática profissional, tais como Escala de Equilíbrio de Berg, Escala de Tinetti, Índice Dinâmico da Marcha e o teste Timed up & go (TUG), oferecendo um escore final e possibilidade de emissão de relatórios.(12,16,20,22,24) O parâmetros biológicos, incluindo ECG, frequên- cia respiratória, temperatura corporal e aceleração tridimensional.(13,23,27,28) Entre os softwares e aplica- tivos desenvolvidos como recurso para prevenção de quedas, observamos sistemas para triagem de riscos utilizando-se de câmera Kinect e máquina de vetores de suporte, além daqueles fundamentados em Health Information Technology (HIT), Fall Risk for Older Persons (FROP). Todos eles buscam parametrizar os riscos de quedas com a captação ou utilização de infor- mações armazenadas anteriormente em banco de da- dos como prontuário eletrônico. No entanto, existem softwares mais acessíveis que permitem o uso infor- matizado de escalas validadas e já utilizadas na prática profissional, tais como Escala de Equilíbrio de Berg, Escala de Tinetti, Índice Dinâmico da Marcha e o teste Timed up & go (TUG), oferecendo um escore final e possibilidade de emissão de relatórios.(12,16,20,22,24) O vídeo educativo, planejado e desenvolvido para uso da equipe assistencial, visa a adoção de medidas preventi- vas de quedas, permitindo o envolvimento dos fami- liares na discussão e sensibilização sobre o tema. Custos No Brasil, os hospitais apresentam gastos consideráveis no tratamento de fratura do quadril em idosos. Estudo demonstrou que os valores variaram desde R$ 1.700,00 para atendimentos por paciente no SUS até R$ 24.000,00 para atendimentos por pacien- tes em operadoras de saúde. Uma das justificativas para este custo elevado é o tempo entre o trauma e a cirur- gia devido à indisponibilidade de salas cirúrgicas e ma- teriais necessários para o procedimento, além do longo tempo de preparo do paciente pré-operatório.(36,39-41) A validade, viabilidade e aceitabilidade de dispositivos destinados a prevenção de quedas pode ajudar a redu- zir os custos relacionados a este evento adverso no am- biente hospitalar, além de diminuir a carga de trabalho da equipe de enfermagem enquanto aumenta a quali- dade do atendimento e segurança do paciente. Alguns dispositivos direcionados ao monitoramento de vários pacientes de forma simultânea podem reduzir custos, além de que os valores investidos na aquisição e im- plantação de dispositivos podem ser recuperados em curto espaço de tempo em consequência das reduções de quedas e de verbas que seriam direcionadas à con- tratação de cuidadores.(14,19,25) Recursos que se utilizam de tecnologias de imagens disponíveis comercialmen- te apresentam-se como soluções mais econômicas, como câmeras de baixo custo com detecção de vídeo/ profundidade com rastreamento de movimento 3D, como Microsoft Kinect™ e Orbbec Astra. Além disso, utilização de sensores sem bateria que não requerem manutenção, também podem representar baixos in- vestimentos para sua implantação. As tecnologias ves- tíveis mais recentes agregam múltiplas funcionalidades apresentando baixo custo, otimizando espaço físico e equipes. Já produtos que utilizam ferramentas de tec- nologia da informação com foco em avaliação de risco Custos Já produtos que utilizam ferramentas de tec- nologia da informação com foco em avaliação de risco também podem apresentar baixo custo proporcionan- do uma intervenção eficaz e de baixo valor. Ideias sim- ples, sem utilização de tecnologias de última geração para a prevenção de quedas se revelam como soluções de baixo custo, podendo ser implantados em regiões com limitações financeiras.(12-14,21,22,24,27,28) Por outro lado, observa-se que alguns dos dispositivos analisados não se mostraram efetivos na diminuição dos custos, uma vez que, sua implementação requer investimen- tos dispendiosos ou ainda não são capazes de reduzir a incidência de quedas no ambiente hospitalar. Já outros estudos reforçam a necessidade de pesquisas adicionais para verificar o custo-benefício da utilização de seu sis- tema.(15,19,23,29,30,38) A utilização de inovações em saúde tem impacto na diminuição da mortalidade, princi- palmente, quando se trata de pacientes idosos e doen- ças incapacitantes, reforçando o fato de que a internet das coisas pode agir diretamente na redução dos cus- tos assistenciais. No entanto, um processo rigoroso de avaliação para incorporação de tecnologias certamente resultará em melhores benefícios clínicos ao paciente e orçamentários às instituições assistencial.(42,43) 30.969 (dólar canadense) comparado ao paciente que não apresentou este desfecho. Já a média do tempo de permanência hospitalar foi de 45 dias para os pacientes que apresentaram queda e de 11 dias, para os que não apresentaram. Estudo australiano revelou o aumento de tempo de internação em oito dias, além de gerar um custo adicional para o hospital de A$ 6.669 (dólar australiano). No Brasil, os hospitais apresentam gastos consideráveis no tratamento de fratura do quadril em idosos. Estudo demonstrou que os valores variaram desde R$ 1.700,00 para atendimentos por paciente no SUS até R$ 24.000,00 para atendimentos por pacien- tes em operadoras de saúde. Uma das justificativas para este custo elevado é o tempo entre o trauma e a cirur- gia devido à indisponibilidade de salas cirúrgicas e ma- teriais necessários para o procedimento, além do longo tempo de preparo do paciente pré-operatório.(36,39-41) 30.969 (dólar canadense) comparado ao paciente que não apresentou este desfecho. Já a média do tempo de permanência hospitalar foi de 45 dias para os pacientes que apresentaram queda e de 11 dias, para os que não apresentaram. Estudo australiano revelou o aumento de tempo de internação em oito dias, além de gerar um custo adicional para o hospital de A$ 6.669 (dólar australiano). Custos As quedas são responsáveis pelo aumento do custo do tratamento e do tempo de permanência hospita- lar, além de gerar alterações de aspectos emocionais na equipe de saúde e repercutir negativamente na cre- dibilidade da instituição perante a sociedade. Estudo canadense mostrou que o custo médio excedente despendido pelo hospital com queda grave foi de C$ 6 Acta Paul Enferm. 2023; 36:eAPE01462. Alves RC, Colichi RM, Lima SA 30.969 (dólar canadense) comparado ao paciente que não apresentou este desfecho. Já a média do tempo de permanência hospitalar foi de 45 dias para os pacientes que apresentaram queda e de 11 dias, para os que não apresentaram. Estudo australiano revelou o aumento de tempo de internação em oito dias, além de gerar um custo adicional para o hospital de A$ 6.669 (dólar australiano). No Brasil, os hospitais apresentam gastos consideráveis no tratamento de fratura do quadril em idosos. Estudo demonstrou que os valores variaram desde R$ 1.700,00 para atendimentos por paciente no SUS até R$ 24.000,00 para atendimentos por pacien- tes em operadoras de saúde. Uma das justificativas para este custo elevado é o tempo entre o trauma e a cirur- gia devido à indisponibilidade de salas cirúrgicas e ma- teriais necessários para o procedimento, além do longo tempo de preparo do paciente pré-operatório.(36,39-41) A validade, viabilidade e aceitabilidade de dispositivos destinados a prevenção de quedas pode ajudar a redu- zir os custos relacionados a este evento adverso no am- biente hospitalar, além de diminuir a carga de trabalho da equipe de enfermagem enquanto aumenta a quali- dade do atendimento e segurança do paciente. Alguns dispositivos direcionados ao monitoramento de vários pacientes de forma simultânea podem reduzir custos, além de que os valores investidos na aquisição e im- plantação de dispositivos podem ser recuperados em curto espaço de tempo em consequência das reduções de quedas e de verbas que seriam direcionadas à con- tratação de cuidadores.(14,19,25) Recursos que se utilizam de tecnologias de imagens disponíveis comercialmen- te apresentam-se como soluções mais econômicas, como câmeras de baixo custo com detecção de vídeo/ profundidade com rastreamento de movimento 3D, como Microsoft Kinect™ e Orbbec Astra. Além disso, utilização de sensores sem bateria que não requerem manutenção, também podem representar baixos in- vestimentos para sua implantação. As tecnologias ves- tíveis mais recentes agregam múltiplas funcionalidades apresentando baixo custo, otimizando espaço físico e equipes. Privacidade do paciente Embora as estratégias que se utilizam de tecnologias de imagem, possibilitando grandes avanços para prevenção de quedas, são apontadas restrições legais e dificuldades de aceitação por parte do paciente de- vido ao sentimento de exposição de sua intimidade. Por outro lado, há relatos que podem ser menos in- trusivos do que cuidadores, e podem levar a menos agitação do paciente.(18,19,25) Alguns dos dispositivos, como sensores, ainda que considerados menos inva- sivos, podem ter resistência para a sua implantação, já que pacientes mais velhos consideram a discri- ção como um critério chave de aceitação, podendo ser visto como um obstáculo para a transposição da tecnologia para a prática.(13,21,27,28) A adoção de tec- nologias mecanizadas pode oferecer soluções para ambientes que exigem maior privacidade como os banheiros, dispensando a captação de imagens e movimentos ou mesmo o acompanhamento de cuidadores.(18,21) Apesar dos benefícios advindos do emprego da tecnologia no campo da saúde é neces- sário, para sua implantação, estudos prévios regu- latórios e ainda levar em consideração as questões éticas relacionadas aos usuários da saúde, os quais tem direito à privacidade, ao conforto e a confiden- cialidade de toda e qualquer informação pessoal, ainda que pouco conhecidos pelos pacientes.(38,44,45) 7 7 7 Acta Paul Enferm. 2023; 36:eAPE01462. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Relevância dos profissionais de enfermagem no desenvolvimento, na avaliação e utilização dos dispositivos cuidados adequadas para todos os pacientes.(18) Reuniões das equipes assistenciais para análise de eventos adversos podem se tornar promissoras para a detecção de de- mandas e a concepção ou adequação de produtos vol- tados para prevenção de quedas. Torna-se importante a participação de profissionais de diferentes áreas como engenharia clínica, arquitetura, informática, hotelaria, segurança patrimonial, entre outras, já que podem cola- borar com ideias e conceitos complementares, buscando alternativas viáveis para o desenvolvimento de novas so- luções.(21,52) A parceria da equipe assistencial permanece como ponto fundamental para o desenvolvimento de ferramentas eficazes e aplicáveis na prática.(12,35) Apesar da relevância, a enfermagem ainda tem participação dis- creta na inovação dos produtos. Os enfermeiros deve- riam ser encorajados no desenvolvimento de materiais, equipamentos e serviços, com o devido registro no sis- tema de patentes. Privacidade do paciente Neste sentido, o ensino de empreen- dedorismo adequado a esse novo contexto profissional deve ser incorporado às grades curriculares do curso de graduação em enfermagem, incentivando ainda os estu- dantes a buscarem auxílio nos centros de transferência de tecnologia para registro de patentes e inovações tec- nológicas presentes em diversas universidades.(21,47,48,53) A predominância de estudos em revistas especializadas em Enfermagem destaca a relevância do profissional na prevenção deste tipo de incidente, através de medidas como a avaliação de risco, planejamento e prestação de cuidado individualizado, no processo educativo e na realização de rondas intencionais nos quartos.(17,20- 22,25,28,30,46) Em vários casos, a equipe de enfermagem considera útil o emprego de alertas e de sistemas que sejam adequados e integrados ao fluxo de trabalho, re- velando a aceitabilidade destes dispositivos tecnológicos. No entanto, observa-se a necessidade do envolvimento da equipe na implantação de novas tecnologias, bem como a investigação da satisfação dos profissionais vi- sando a melhor aceitação dos produtos.(15,19,25,47-49) A en- fermagem brasileira vem avançando cientificamente no campo da tecnologia, apropriando-se de recursos inova- dores e fortalecendo a prática assistencial, a educação e a gestão em saúde, apesar da escassez quando comparada ao contexto internacional.(50) Neste sentido, o emprego da tecnologia pode se tornar capaz de aliviar a carga de trabalho da equipe de enfermagem, otimizando o tem- po de cuidado e permitindo que o profissional exerça outras atividades enquanto o paciente encontra-se “pro- tegido” pelo dispositivo, como por exemplo, o Abraço Seguro.(21,27,41) Entretanto, a equipe pode se demonstrar insatisfeita com problemas decorrentes, como o número inadequado de profissionais para atender aos alarmes, o esquecimento do bipe no posto de enfermagem e a hie- rarquização de prioridades quando são acionados, já que muitas vezes poderiam estar atendendo outro paciente o que impossibilitaria se dirigir ao local em que o alarme foi acionado. Acta Paul Enferm. 2023; 36:eAPE01462. Conclusão 4. Institute for Healthcare Improvement (IHI). RCA2: Improving Root Cause Analyses and Actions to Prevent Harm. United States: IHI; 2015 [cited 2022 May 5]. Available from: http://www.ihi.org/resources/Pages/Tools/RCA2- Improving-Root-Cause-Analyses-and-Actions-to-Prevent-Harm.aspx A pesquisa identificou diversas estratégias tecnoló- gicas que podem ser aplicadas para prevenção de quedas de adultos em ambiente hospitalar, além de trazer reflexões sobre medidas aplicadas para pre- venção de quedas, revelando uma gama de recur- sos desenvolvidos e disponíveis no mercado, que podem ser incorporadas nas práticas assistenciais compondo políticas institucionais. Entre as estra- tégias tecnológicas ressalta-se aquelas relaciona- das a processo educativo, programas/ferramentas de tecnologia da informação, dispositivos vestí- veis e dispositivos para ambiente. Estudos reve- lam que alguns dispositivos podem reduzir gastos com cuidado, ou aqueles decorrentes de queda no ambiente hospitalar. No entanto, deve-se atentar para questões de privacidade do paciente, família e equipe, além de restrições e aspectos legais. Cabe aos hospitais definirem qual estratégia aplica-se de forma mais adequada à sua realidade, levando- -se em consideração o orçamento disponível para investimento, público atendido, quantitativo de equipe de enfermagem, setor onde será implan- tada, disponibilidade para capacitação da equipe antes do início do uso, perfil epidemiológico dos pacientes, entre outros. Torna-se relevante ampliar as discussões para a formulação de programas e políticas institucionais de prevenção de quedas, compostas por estratégias tecnológicas e interven- ções diferenciadas voltadas para a segurança do pa- ciente, ultrapassando a prática de ações de forma isolada nos ambientes hospitalares. Além disso, os membros da equipe de enfermagem devem ser en- corajados a participarem no desenvolvimento de materiais, equipamentos e serviços, revelando a importância de seu papel por meio da divulgação científica e pelo registro de patentes. 5. Ximenes MA, Brandão MG, Macêdo TS, Costa MM, Galindo Neto NM, Caetano JA, et al. Effectiveness of educational technology for preventing falls in a hospital environment. Acta Paul Enferm. 2022;35:eAPE01372. 6. Ferreira JM, Hammerschmitdt KS, Siewert JS, Alvarez AM, Locks MO, Heidmann IT. Gerontotechnology for the prevention of falls of the elderly with Parkinson. Rev Bras Enferm. 2019;72(Suppl 2):243–50. 7. Silva TI, Braz PR, Cavalcante RB, Alves M. Diffusion of innovations theory and its applicability in research studies on nursing and health. Texto Contexto Enferm. 2022;31:e20210322. 8. Bates DW, Levine D, Syrowatka A, Kuznetsova M, Craig KJ, Rui A, et al. The potential of artificial intelligence to improve patient safety: a scoping review. NPJ Digit Med. 2021;4(1):54. Review. 9. Whittemore R. Referências exigindo cada vez menos dependência humana em processos voltados para prevenção de eventos adver- sos, possibilitando desta forma que a equipe assis- tencial esteja voltada para o planejamento e entrega de cuidados individualizados pautados na qualidade e segurança do paciente. 1. Sousa P, Mendes W. Segurança do paciente: criando organizações de saúde seguras. 2a ed. Rio de Janeiro: FIOCRUZ; 2019. 68 p. 2. Silva MS,Lago PN, Machado VM, Campos EC, Anaisse SL, Cruz ML, et al. As facilidades e contribuições da tecnologia point of care no ambiente hospitalar. Res Society Development,. 2022;11(2):e45511226086. 3. Veazie S, Peterson K, Bourne D, Anderson J, Damschroder L, Gunnar W. Implementing High-Reliability Organization Principles Into Practice: a rapid evidence review. J Patient Saf. 2022;18(1):e320–8. Review. Privacidade do paciente Além disso, vale destacar que quando a instituição opta pelo uso de um dispositivo sonoro, de- ve-se levar em conta os efeitos de fadiga do alarme que pode resultar no atraso do atendimento ou na redução da taxa de resposta devido à exposição excessiva aos aler- tas.(23,29,30,51) O uso de sistemas de monitoramento pode trazer benefícios interessantes ao permitir a revisão dos vídeos de queda, possibilitando à equipe de enferma- gem estabelecer planos de intervenção individuais aos pacientes que necessitam de intervenções mais eficazes levando-se em consideração os motivos dos incidentes, b dl d l d d d g p Por ser uma revisão integrativa, pode-se ter des- considerado algum estudo que não contemplasse os critérios de inclusão, assim como estratégias im- plementadas, porém descritas em outras formas de publicação, apesar da adoção de critérios rigorosos para o levantamento da literatura. Outra limitação refere-se a escassez de artigos que respondessem à questão norteadora deste estudo, já que a maioria das pesquisas tendem a investigar dispositivos que visam sinalizar o momento do incidente, após sua ocorrência. O estudo sugere que outras pesquisas na área sejam realizadas, buscando soluções capazes de auxiliar na prevenção das quedas. Disponibilizamos aos profissionais da área da enfermagem informa- ções acerca das estratégias tecnológicas desenvol- vidas e suas possíveis aplicabilidades. Dados sobre custos e privacidade funcionam como informações adicionais para auxiliar no processo de tomada de decisão sobre os recursos mais indicados. Espera-se também trazer reflexões e inspirações para a cria- ção e aprimoramento de tecnologias em saúde de modo que atendam as predisposições do mercado, 8 Acta Paul Enferm. 2023; 36:eAPE01462. Alves RC, Colichi RM, Lima SA Conclusão Combining evidence in nursing research: methods and implications. Nurs Res. 2005;54(1):56–62. 10. Mendes KD, Silveira RC, Galvão CM. Revisão integrativa: método de pesquisa para a incorporação de evidências na saúde e na enfermagem. Texto Contexto Enferm. 2008;17(4):758–64. Review. 11. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. 12. Teh RC, Wilson A, Ranasinghe D, Visvanathan R. Use and clinical efficacy of standard and health information technology fall risk assessment tools. Australas J Ageing. 2017;36(4):327–31. 13. Wickramasinghe A, Ranasinghe DC, Fumeaux C, Hill KD, Visvanathan R. sequence learning with passive RFID Sensors for Real-Time Bed-Egress Recognition in Older People. IEEE J Biomed Health Inform. 2017;21(4):917–29. 14. Razjouyan J, Grewal GS, Rishel C, Parthasarathy S, Mohler J, Najafi B. Activity Monitoring and Heart Rate Variability as Indicators of Fall Risk: Proof-of-Concept for Application of Wearable Sensors in the Acute Care Setting. J Gerontol Nurs. 2017;43(7):53–62. 15. Balaguera HU, Wise D, Ng CY, Tso HW, Chiang WL, Hutchinson AM, et al. Using a medical intranet of things system to prevent bed falls in an acute care Hospital: a pilot study. J Med Internet Res. 2017;19(5):e150. 16. Yokota S, Endo M, Ohe K. Establishing a Classification System for High Fall-Risk Among Inpatients Using Support Vector Machines. Comput Inform Nurs. 2017;35(8):408–16. 17. Opsahl AG, Ebright P, Cangany M, Lowder M, Scott D, Shaner T. Outcomes of adding patient and family engagement education to fall prevention bundled interventions. J Nurs Care Qual. 2017;32(3):252–8. 18. Potter P, Allen K, Costantinou E, Klinkenberg WD, Malen J, Norris T, et al. Evaluation of sensor technology to detect fall risk and prevent falls in acute care. Jt Comm J Qual Patient Saf. 2017;43(8):414–21. 9 Acta Paul Enferm. 2023; 36:eAPE01462. Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa Estratégias tecnológicas voltadas para prevenção de quedas em ambiente hospitalar: revisão integrativa de quedas. Brasília (DF): Ministério da Saúde; Anvisa; Fiocruz; 2013 [citado2022 Set 20]. Disponível em: www.saude.mt.gov.br/upload/ controle.../protocolos_cp_n6_2013_prevencao.pdf 19. Cournan M, Fusco-Gessick B, Wright L. Improving patient safety through video monitoring. Rehabil Nurs. 2018;43(2):111–5. 20. Yokota S, Tomotaki A, Mohri O, Endo M, Ohe K. Evaluating the effectiveness of a fall risk screening tool implemented in an electronic medical record system. J Nurs Care Qual. 2018;33(4):E1–6. 36. Pasa TS, Magnago TS, Urbanetto JS, Baratto MA, Morais BX, Carollo JB. Conclusão Risk assessment and incidence of falls in adult hospitalized patients. Rev Lat Am Enfermagem. 2017;25:e2862. 21. Niwa LM, Radovich NM, Ciosak SI. Safe Embrace: technological innovation for elderly safety in the use of toilets. Rev Bras Enferm. 2018;71(Suppl 6):2833-6. Erratum in: Rev Bras Enferm. 2021;74(3):e2021n3e07. 37. Marengo LL, Kozyreff AM, Moraes FD, Maricato LI, Barberato-Filho S. Tecnologias móveis em saúde: reflexões sobre desenvolvimento, aplicações, legislação e ética. Rev Panam Salud Publica. 2022;46:e37. 22. Nuñez Filha MC, Pinto EB, Leite HJ. Desenvolvimento de um aplicativo para identificação do risco de quedas em idosos. Rev Pesq Fisio. 2018;8(3):354-60. 38. Lima SG, Brito C, Andrade CJ. O processo de incorporação de tecnologias em saúde no Brasil em uma perspectiva internacional. Cien Saude Colet. 2019;24(5):1709–22. 39. Zecevic AA, Chesworth BM, Zaric GS, Huang Q, Salmon A, McAuslan D, et al. Estimating the cost of serious injurious falls in a Canadian acute care hospital. Can J Aging. 2012;31(2):139-47. 23. Baker PA, Roderick MW, Baker CJ. PUP® (Patient Is Up) Smart Sock Technology Prevents Falls Among Hospital Patients With High Fall Risk in a Clinical Trial and Observational Study. J Gerontol Nurs. 2021;47(10):37-43. 40. Morello RT, Barker AL, Watts JJ, Haines T, Zavarsek SS, Hill KD, et al. The extra resource burden of in-hospital falls: a cost of falls study. Med J Aust. 2015;203(9):367. 24. Sun R, Aldunate RG, Paramathayalan VR, Ratnam R, Jain S, Morrow DG, et al. Preliminary evaluation of a self-guided fall risk assessment tool for older adults. Arch Gerontol Geriatr. 2019;82:94–9. 41. Loures FB, Chaoubah A, Oliveira VM, Almeida AM, Campos EM, Paiva EP. Análise econômica do tratamento cirúrgico de fratura do quadril em idosos. Rev Saude Publica. 2015;49:12. 25. Hebb A, Kistler M, George E, Zamboni B. Satisfaction and technology acceptance of staff regarding use of continuous video monitoring in comparison with sitters. J Nurs Adm. 2021;51(2):60–2. 42. Sadoughi F, Behmanesh A, Sayfouri N. Internet of things in medicine: a systematic mapping study. J Biomedical Informatics. 2020;103:103383. 26. Duckworth M, Adelman J, Belategui K, Feliciano Z, Jackson E, Khasnabish S, et al. Assessing the effectiveness of engaging patients and their families in the three-step fall prevention process across modalities of an evidence-based fall prevention toolkit: an implementation science study. J Med Internet Res. 2019;21(1):e10008. 43. Avelar AF, Santos LM. Technological innovation in health: back to origins [editorial]. Rev Bras Enferm. 2021;74(Suppl 5):e74Suppl501. 44. Brasil. Ministério da Saúde. Conselho Nacional de Saúde. Conclusão Carta dos direitos dos usuários da saúde. Brasília (DF): Ministério da Saúde; 2012 [citado 2022 Set 20]. Disponível em: http://bit.ly/2bZcyXd. 27. Jähne-Raden N, Kulau U, Marschollek M, Wolf KH. INBED: a Highly Specialized System for Bed-Exit-Detection and Fall Prevention on a Geriatric Ward. Sensors (Basel). 2019;19(5):1017. 45. Silva DN, Araújo JL, Silva RT, Nascimento EG. Privacidade e confidencialidade de usuários em um hospital geral. Rev Bioet. 2017;25(3):585–95. 28. Jähne-Raden N, Gütschleg H, Wolf MC, Kulau U, Wolf KH. Wireless sensor network for fall prevention on geriatric wards: a report. stud health Technol Inform. 2019;264:620–4. 46. Sena AC, Alvarez AM, Nunes SF, Costa NP. Nursing care related to fall prevention among hospitalized elderly people: an integrative review. Rev Bras Enferm. 2021;74(74 Suppl 2):e20200904. Review. 29. Timmons S, Vezyridis P, Sahota O. Trialling technologies to reduce hospital in-patient falls: an agential realist analysis. Sociol Health Illn. 2019;41(6):1104–19. 47. Souza CJ, Silvino ZR, Souza DF. Analysis of patent registries in Brazilian nursing and its relationship with the professional master’s degree. Rev Gaucha Enferm. 2020;41:e20190358. 30. Seow JP, Chua TL, Aloweni F, Lim SH, Ang SY. Effectiveness of an integrated three-mode bed exit alarm system in reducing inpatient falls within an acute care setting. Jpn J Nurs Sci. 2022;19(1):e12446. 48. Colichi RM, Lima SG, Bonini AB, Lima SA. Entrepreneurship and Nursing: integrative review. Rev Bras Enferm. 2019;72(Suppl 1):321– 30. Review. 31. Costa LF, Ramalho FA. A usabilidade nos estudos de uso da informação: em cena usuários e sistemas interativos de informação. Perspect Cien Inf. 2010;15(1):92–117. 49. Giuliano KK. Challenging precedent: critical care nursing and medical product innovation. Am J Crit Care. 2020;29(4):253–61. 32. Organização Mundial da Propriedade Intelectual (OMPI). Índice Global de Inovação 2021: resumo executivo. Genebra: OMPI; 2021 [citado2022 May 5]. Disponível em: https://www.wipo.int/edocs/ pubdocs/pt/wipo_pub_gii_2021_exec.pdf 50. Lima CS, Barbosa SF. Aplicativos móveis em saúde: caracterização da produção científica da enfermagem brasileira. Rev Eletr Enferm. 2019;21:53278. 33. Monteiro YC, Vieira MA, Vitorino PV, Queiroz SJ, Policena GM, Souza AC. Trend of fall-related mortality among the elderly. Rev Esc Enferm USP. 2021;55:e20200069. 51. Oliveira AE, Machado AB, Santos ED, Almeida EB. Alarm fatigue and the implications for patient safety. Rev Bras Enferm. 2018;71(6):3035–40. 34. Diniz JL, Sousa VF, Coutinho JF, Araújo IL, Andrade RM, Costa JS, et al. Internet of things gerontechnology for fall prevention in older adults: an integrative review. Acta Paul Enferm. 2022;35:eAPE003142. Review. 52. Teixeira TC, Cassiani SH. Conclusão Root cause analysis of falling acidentes and medication errors in hospital. Acta Paul Enferm. 2014;27(2):100–7. 53. Souza LE. Health, development and innovation: a contribution of the critical theory of technology to the discussion. Cad Saude Publica. 2016;32(32 Suppl 2):e00029615. 35. Brasil. Ministério da Saúde. Agência Nacional de Vigilância Sanitária (ANVISA). Fundação Oswaldo Cruz (FIOCRUZ). Protocolo prevenção 10 Acta Paul Enferm. 2023; 36:eAPE01462.
https://openalex.org/W2518867241	https://europepmc.org/articles/pmc5043589?pdf=render	English	null	The value of poly-urethane cuffed endotracheal tubes	Critical care	2,016	cc-by	744	Correspondence: Jan.poelaert@uzbrussel.be 1Department of Anesthesiology and Perioperative Medicine, University Hospital Brussels, Laarbeeklaan 101, 1090 Brussels, Belgium 2Faculty of Medicine and Pharmacy, VUB, Brussels, Belgium Competing interests Competing interests The author declares that he has no competing interests. Competing interests The author declares that he has no competing interests. Poelaert Critical Care (2016) 20:290 DOI 10.1186/s13054-016-1474-3 Poelaert Critical Care (2016) 20:290 DOI 10.1186/s13054-016-1474-3 Abbreviations: ET, Endotracheal tube; ID, Internal diameter; PU-ET, Poly-urethane cuffed endotracheal tube; PVC-ET, Poly-vinylchloride cuffed endotracheal tube Another important issue which has not been discussed is the proper choice of the size of the ET. There is a clear tendency in clinical anesthesia practice to decrease the size of the ET to 8 mm internal diameter (ID) for male and 7 mm ID for female patients. This allows a theoretical improved expansion of the cuff, with a consequent decline of pleas and channels in the PVC cuffs; whether this reduction in size leads to improved sealing has still to be demonstrated. I read with special interest the meta-analysis of Blot et al. in a recent issue of Critical Care [1]. In a systematic review of the available literature of both laboratory and clinical studies, comparing the sealing efficacy of poly- vinylchloride cuffed endotracheal tubes (PVC-ET) versus poly-urethane cuffed endotracheal tubes (PU-ET), they demonstrated a clear advantage with a demonstrated delay of chance of descent of oropharyngeal secretions when a PU-ET was used. However, prevention of ventilator- associated pneumonia could not be demonstrated in the overall data. Several open questions remain unanswered. Though there are some arguments in favor of the use of PU-ET, we still have to wait for the demonstration of a clear delay of onset of ventilator-associated pneumonia. The characteristics of the PU cuff allow better expansion of the cuff within the trachea, in comparison with the thicker material of PVC. In a recent re-analysis of the initial data [2], we demonstrated that a clear benefit of the PU-ET is obvious after 16.6 h if patients are treated in a comparable manner with tidal volume 6 ml/kg, PEEP 7–8 cmH2O, and positioned in 30–40° inclined position postoperatively [3]. Furthermore, in longer term ventilated ICU patients, intubated with PU-ET, condensation fluid is sometimes present within the cuff (unpublished data), abolishing proper assessment of cuff pressure. The latter could be one of the reasons why no clear advantage is found in mechanical ventilated ICU patients. In addition, we demonstrated that, indeed, micro-aspiration still occurs with PU-ET, but is retarded for several hours [4]. This phenomenon is also suggested in in-vitro studies [5]. Funding Only departmental funding is present. Funding Funding Only departmental funding is present. Author’s contribution JP wrote and conceived the manuscript/text. The value of poly-urethane cuffed endotracheal tubes Jan Poelaert1,2 e related Research by Blot et al., https://ccforum.biomedcentral.com/articles/10.1186/s13054-016-1380-8 Abbreviations: ET, Endotracheal tube; ID, Internal diameter; PU-ET, Poly-urethane cuffed endotracheal tube; PVC-ET, Poly-vinylchloride cuffed endotracheal tube 3. Poelaert J, Haentjens P, Blot S. Association among duration of mechanical ventilation, cuff material of endotracheal tube, and postoperative nosocomial pneumonia in cardiac surgical patients: a prospective study. J Thorac Cardiovasc Surg. 2014;148(4):1622–7. 4. D’Haese J, De Keukeleire T, Remory I, Van Rompaey K, Umbrain V, Poelaert J. Assessment of intraoperative microaspiration: does a modified cuff shape improve sealing? Acta Anaesthesiol Scand. 2013;57(7):873–80. 5. Zanella A, Scaravilli V, Isgrò S, Milan M, Cressoni M, Patroniti N, Fumagalli R, Pesenti A. Fluid leakage across tracheal tube cuff, effect of different cuff material, shape, and positive expiratory pressure: a bench-top study. Intensive Care Med. 2011;37(2):343–7. Poelaert Critical Care (2016) 20:290 References 1. Blot SI, Rello J, Koulenti D. The value of polyurethane-cuffed endotracheal tubes to reduce microaspiration and intubation-related pneumonia: a systematic review of laboratory and clinical studies. Crit Care. 2016;20(1):203. y y 2. Poelaert J, Depuydt P, De Wolf A, Van de Velde S, Herck I, Blot S. Polyurethane cuffed endotracheal tubes to prevent early postoperative pneumonia after cardiac surgery: a pilot study. J Thorac Cardiovasc Surg. 2008;135(4):771–6. © 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Page 2 of 2 Poelaert Critical Care (2016) 20:290
https://openalex.org/W4384694845	https://www.frontiersin.org/articles/10.3389/fnagi.2023.1213977/pdf	English	null	Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease	Frontiers in aging neuroscience	2,023	cc-by	6,362	Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease OPEN ACCESS EDITED BY Muthuraman Muthuraman, University Hospital Würzburg, Germany REVIEWED BY Wooyoung Jang, Gangneung Asan Hospital, Republic of Korea Steven Gunzler, University Hospitals Cleveland Medical Center, United States CORRESPONDENCE Ming-Hong Chang cmh500809@gmail.com RECEIVED 28 April 2023 ACCEPTED 04 July 2023 PUBLISHED 18 July 2023 CITATION Fang T-C, Tsai Y-S and Chang M-H (2023) Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease. Front. Aging Neurosci. 15:1213977. doi: 10.3389/fnagi.2023.1213977 COPYRIGHT © 2023 Fang, Tsai and Chang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. OPEN ACCESS EDITED BY Muthuraman Muthuraman, University Hospital Würzburg, Germany REVIEWED BY Wooyoung Jang, Gangneung Asan Hospital, Republic of Korea Steven Gunzler, University Hospitals Cleveland Medical Center, United States CORRESPONDENCE Ming-Hong Chang cmh500809@gmail.com RECEIVED 28 April 2023 ACCEPTED 04 July 2023 PUBLISHED 18 July 2023 CITATION Fang T-C, Tsai Y-S and Chang M-H (2023) Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s di OPEN ACCESS EDITED BY Muthuraman Muthuraman, University Hospital Würzburg, Germany REVIEWED BY Wooyoung Jang, Gangneung Asan Hospital, Republic of Korea Steven Gunzler, University Hospitals Cleveland Medical Center, United States CORRESPONDENCE Ming-Hong Chang cmh500809@gmail.com RECEIVED 28 April 2023 ACCEPTED 04 July 2023 PUBLISHED 18 July 2023 CITATION Fang T-C, Tsai Y-S and Chang M-H (2023) Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease. Front. Aging Neurosci. 15:1213977. doi: 10.3389/fnagi.2023.1213977 COPYRIGHT © 2023 Fang, Tsai and Chang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease Ting-Chun Fang1, Yu-Shan Tsai1 and Ming-Hong Chang1,2,3 1Department of Neurology, Taichung Veterans General Hospital, Neurological Institute, Taichung, Taiwan, 2Department of Post-Baccalaureate Medicine, College of Medicine, National Chung Hsing University, Taichung, Taiwan, 3Brain and Neuroscience Research Center, College of Medicine, National Chung Hsing University, Taichung, Taiwan CITATION Fang T-C, Tsai Y-S and Chang M-H (2023) Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease. Introduction: Hyposmia is a common prodrome in patients with Parkinson’s disease (PD). This study investigates whether olfactory changes in PD differ according to the degree of olfactory dysfunction and whether there are changes in motor and non-motor symptoms. Front. Aging Neurosci. 15:1213977. doi: 10.3389/fnagi.2023.1213977 Front. Aging Neurosci. 15:1213977. doi: 10.3389/fnagi.2023.1213977 Methods: The 129 subjects with PD were divided into two groups: anosmia and non-anosmia. All cases were reassessed within 1–3 years after the initial assessment. The assessment included the MDS-Uniﬁed PD Rating Scale (MDS- UPDRS), the University of Pennsylvania Smell Identiﬁcation Test (UPSIT), Beck’s Depression Inventory-II (BDI-II), Montreal Cognitive Assessment (MoCA), and equivalence dose of daily levodopa (LEDD). The generalized estimating equation (GEE) model with an exchangeable correlation structure was used to analyze the change in baseline and follow-up tracking and the disparity in change between these two groups. COPYRIGHT © 2023 Fang, Tsai and Chang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Results: The anosmia group was older and had a longer disease duration than the non-anosmia group. There was a signiﬁcant decrease in UPSIT after follow-up in the non-anosmia group (β = −3.62, p < 0.001) and a signiﬁcant difference in the change between the two groups (group-by-time effect, β = 4.03, p < 0.001). In the third part of the UPDRS motor scores, there was a tendency to increase the score in the non-anosmia group compared to the anosmia group (group- by-time effect, β = −4.2, p < 0.038). There was no signiﬁcant difference in the group-by-time effect for UPDRS total score, LEDD, BDI-II, and MoCA scores. TYPE Original Research PUBLISHED 18 July 2023 DOI 10.3389/fnagi.2023.1213977 TYPE Original Research PUBLISHED 18 July 2023 DOI 10.3389/fnagi.2023.1213977 Parkinson’s disease, olfactory dysfunction, UPSIT, MDS-UPDRS, equivalence dose of daily levodopa, cognition, depression KEYWORDS 2.2. Variables The olfactory function was evaluated with the validated Taiwanese version of UPSIT, an odor identiﬁcation (Jiang et al., 2010). The total score was 40 in this test and the cutoﬀvalue of total anosmia was less than 19. Considering that the mean UPSIT score is 17–20 in PD patients which is close to the cutoﬀvalue 19 of anomia in UPSIT, we divided subjects into two groups, anosmia and non-anosmia based on the UPSIT score at the ﬁrst visit to represent the characteristics of profound olfactory deﬁcit or milder symptom in PD, respectively (Doty, 2001, 2012; Picillo et al., 2014; Lawton et al., 2016). Non-motor symptoms of PD were also assessed. For cognition, we used MoCA due to its validation for assessing global cognitive abilities in PD (Litvan et al., 2012). BDI-II was used for mood investigation (Beck et al., 1996). Regarding the severity of motor symptoms in PD, the part 3 score of MDS-UPDRS (UPDRS 3) and the equivalent dose of daily levodopa (LEDD) were used to determine the severity of motor symptoms (Goetz et al., 2008). The total score of MDS-UPDRS (UPDRS T) was used to represent the disease burden of PD. Scores of MDS-UPDRS Part 1 and Part 2 were used to represent the non-motor and motor experiences of daily living, and Part 4 was used for motor complication. To determine the motor subtypes, we utilized 11 items (2.10, 3.15–3.18) for tremor and ﬁve items (2.12, 2.13, 3.10–3.12) for postural instability/gait diﬃculty (PIGD) from the MDS-UPDRS. The ratio of mean tremor scores to the mean PIGD scores was g However, hyposmia is still one of the common non- motor symptoms in PD related to Lewy body pathology in the olfactory system (Haehner et al., 2009; Rodríguez-Violante et al., 2017). As the disease progresses, Lewy body pathology increases in the olfactory system, but most studies show inconsistent results in the relationship between hyposmia and disease severity (Berendse et al., 2011; Yoo et al., 2020). Few studies discuss the association between the duration of the disease and olfactory dysfunction, and most of them did not show an obvious correlation, and even the results of some longitudinal studies are inconsistent (Ercoli et al., 2022). Due to the ambiguous relationship between olfactory dysfunction and disease duration, this study aims to investigate the longitudinal change of olfactory function in PD patients based on their degree of olfactory dysfunction. 2.1. Participants Parkinson’s disease (PD) is the second most common neurodegenerative disease, with a prevalence of around 1.4–3.0 per thousand in Taiwan which increases with aging (Liu et al., 2016a,b). In addition to motor symptoms such as bradykinesia, tremor, and rigidity, non-motor symptoms contribute to poor quality of life in patients with PD (Rodríguez-Violante et al., 2015; Tibar et al., 2018; Santos Garcia et al., 2019). Some non-motor symptoms appear before motor symptoms, known as prodromal non-motor symptoms of PD (Poewe et al., 2017). Olfactory dysfunction, constipation, depression, and rapid eye movement (REM) sleep behavior disorder (RBD) can represent prodromal symptoms. Participants were recruited from the outpatient clinic at Taichung Veteran General Hospital from 2017. Subjects were selected on the basis of International Parkinson and MDS Clinical Diagnostic Criteria for Parkinson’s disease. At the ﬁrst visit (T0), all subjects received a complete survey that included the MDS-Uniﬁed Parkinson’s Disease Rating Scale (MDS-UPDRS), the University of Pennsylvania Smell Identiﬁcation Test (UPSIT), Beck’s Depression Inventory-II (BDI-II), and Montreal’s Cognitive Assessment (MoCA). Follow-up (T1) was conducted for these patients with PD in 1–3 years after the ﬁrst visit, and a comprehensive survey was also conducted including MDS-UPDRS, UPDIT, BDI, and MoCA. The equivalent dose of daily levodopa was also calculated on the date of the ﬁrst visit and the follow- up, respectively. Individuals were excluded if they did not meet the MDS clinical diagnostic criteria for Parkinson’s disease during follow-up or could not complete the questionnaire. Those who had upper respiratory tract infection and sinonasal disease which might aﬀect olfaction were also excluded. Written informed consent was obtained from all participants. This study was approved by Taichung Veterans General Hospital Institutional Review Board/Ethics Committee (No. CE22189B). All methods were performed in accordance with the Declaration of Helsinki guidelines and hospital regulations. The Braak staging system explains prodromal symptoms because alpha-synuclein aggregates, a pathological hallmark of PD, are initially found in the olfactory bulb and the dorsal motor nucleus of the vagus (Braak et al., 2003). One route of propagation of alpha-synuclein inclusion in the dual-hit hypothesis starts from the enteric nervous system with the gut to brain spreading. This route is suggested to be associated with the involvement of the autonomic nervous system and premotor RBD, naming the body-ﬁrst subtype. 2.1. Participants The other route of alpha-synuclein pathology starts from the olfactory bulb and anterior olfactory nucleus and spreads to adjacent areas such as the olfactory tubercle, piriform cortex, periamygdaloid cortex, and entorhinal cortex. However, the evidence of entry via the olfactory pathway is still controversial because no advanced lesions are found in non-olfactory cortical areas (Braak et al., 2003; Horsager et al., 2020). In a PD mouse model, RBD- like behavior occurred earlier than hyposmia, which correlates with the ﬁnding in humans that PD patients with RBD were more hyposmic than PD patients without RBD (Taguchi et al., 2020). These ﬁndings suggest that the ascending pathway of the brainstem may predominate in the spread of alpha-synuclein, despite the initial deposition in the olfactory bulb (Braak et al., 2003; Horsager et al., 2020). Sequential change in olfaction and (non) motor symptoms: the difference between anosmia and non-anosmia in Parkinson’s disease Discussion: In conclusion, this study found that olfactory sensation may still regress in PD with a shorter disease course without anosmia, but it remains stable in the anosmia group. Such a decline in olfaction may not be related to cognitive status but may be associated with motor progression. Parkinson’s disease, olfactory dysfunction, UPSIT, MDS-UPDRS, equivalence dose of daily levodopa, cognition, depression 01 Frontiers in Aging Neuroscience frontiersin.org Fang et al. 10.3389/fnagi.2023.1213977 3.3. Change in olfactory identiﬁcation in anosmia/non-anosmia groups In the GEE analysis (Table 2), a signiﬁcant group eﬀect revealed a lower UPSIT score in the anosmia group (β = −10.58, p < 0.001). The time eﬀect was signiﬁcant in the non-anosmia group (β = −3.62, p < 0.001) but not in the anosmia group. The group-by-time eﬀect was also signiﬁcant (β = 4.03, p < 0.001), indicating that the UPSIT score remained stable in the anosmia group but decreased signiﬁcantly in the non-anosmia group (Figure 1A). These results remained signiﬁcant after adjusting for age, gender, and disease duration (Table 3). 3.1. Demographic data p = 0.007), after adjusting for age, gender, and disease duration. However, no signiﬁcant correlations were found between UPSIT and BDI or LEDD. A total of 129 participants were enrolled in this study. Table 1 shows that the anosmia and non-anosmia groups comprised 73 and 56 subjects, respectively. At baseline, the anosmia group was older than the non-anosmia group (66.65 vs. 63.21, p = 0.032) and had a longer disease duration (4.89 years vs. 3.27, p = 0.033). The group with anosmia also demonstrated higher scores on UPDRST and UPDRS3, but exhibited lower scores on the MoCA. However, no signiﬁcant diﬀerences were found between the two groups regarding gender, follow-up interval, motor subtypes, scores of UPDRS1, 2, and 4, LEDD and BDI scores. After the follow-up for UPSIT re-evaluation, it was observed that 20 patients from the non-anosmia group at the ﬁrst visit had developed anosmia, accounting for 35.7% of the non-anosmia group. Conversely, seven patients from the anosmia group had transitioned to non-anosmia. Eventually, the anosmia and non-anosmia groups comprised 86 and 43 subjects, respectively. 2.2. Variables Given this uncertainty, we also conducted a comprehensive analysis of disease severity, medication usage, cognitive function, and depression during the longitudinal follow-up period to provide a more comprehensive comparison. Frontiers in Aging Neuroscience 02 frontiersin.org Fang et al. 10.3389/fnagi.2023.1213977 TABLE 1 Characteristics of the participants at baseline. TABLE 1 Characteristics of the participants at baseline. Characteristic Anosmia (n = 73) Non- anosmia (n = 56) P-value Age, mean (SD), y 66.65 (8.63) 63.21 (9.33) 0.032* Gender (%) Male 39 (53.5) 37 (66.1) 0.148 Female 34 (46.5) 19 (33.9) Disease duration, mean (SD), y 4.89 (5.07) 3.27 (3.46) 0.033* Follow-up interval, mean (SD), y 1.52 (0.63) 1.65 (0.73) 0.288 UPDRST† 51.0 (37.0, 66.0) 43.0 (28.2, 53.0) 0.007* UPDRS1† 9.0 (5.0, 14.0) 8.0 (5.0, 11.0) 0.378 UPDRS2† 8.0 (3.0, 13.5) 7.0 (3.2, 10.7) 0.211 UPDRS3† 32.0 (23.0, 41.0) 25.5 (17.0, 34.7) 0.002* UPDRS4† 0 (0, 1.0) 0 (0, 0.7) 0.843 Motor subtype (%) 0.907 PIGD 34 (46.6) 24 (42.9) Tremor 25 (34.2) 21 (37.5) Indeterminate 14 (19.2) 11 (19.6) LEDD† 474.0 (201.8, 787.5) 377.5 (140.6, 637.5) 0.090 MoCA† 26.0 (21.5, 28.0) 27.0 (25.0, 29.0) 0.016* BDI† 10.0 (4.0, 17.0) 8.5 (4.0, 14.5) 0.447 y, years; SD, standard deviation; UPDRS, MDS-UPDRS; LEDD, equivalent dose of daily levodopa; MoCA, Montreal’s Cognitive Assessment; BDI-II, Beck’s Depression Inventory- II; PIGD, postural instability/gait diﬃculty. †Analyzed by non-parametric Mann–Whitney U-tests, and presented with median (1st and 3rd quartile). Signiﬁcance, p < 0.05. employed to deﬁne the following subtypes: (1) tremor subtype with a ratio ≥1.15 and (2) PIGD subtype with a ratio ≤0.90 (Stebbins et al., 2013). 3. Results y, years; SD, standard deviation; UPDRS, MDS-UPDRS; LEDD, equivalent dose of daily levodopa; MoCA, Montreal’s Cognitive Assessment; BDI-II, Beck’s Depression Inventory- II; PIGD, postural instability/gait diﬃculty. †Analyzed by non-parametric Mann–Whitney U-tests, and presented with median (1st and 3rd quartile). Signiﬁcance, p < 0.05. 2.3. Statistical analysis Baseline clinical characteristics between the anosmia and non-anosmia groups were compared by using chi-square test for binary variables. UPSIT, MoCA, BDI-II, MDS UPDRS scores, and LEDD scores were analyzed as continuous variables. For continuous variables that follow a normal distribution, Student t-tests were used for analysis. For variables that do not follow a normal distribution, non-parametric Mann–Whitney U tests were used for analysis. Multiple linear regression adjusted for age, gender and disease was carried out to analyze the relationship between UPSIT and each variable including MoCA, BDI-II, UPDRS 3, UPDRS T, and LEDD scores at baseline. Generalized estimating equation (GEE) model with an exchangeable correlation structure, was used to assess the change of longitudinal data, including MoCA, BDI-II, UPDRS 3, UPDRS T, and LEDD, between the anosmia and non-anosmia groups at T1 compared with T0. All tests were with a statistical signiﬁcance level of p < 0.05 and were reported with 95% conﬁdence intervals (CIs). Data analysis was performed with SPSS software (IBM Corporation, Armonk, New York, NY, USA). Frontiers in Aging Neuroscience 3.4. Change in motor symptoms in anosmia/non-anosmia groups The UPSIT scores of all participants at baseline were signiﬁcantly correlated with MoCA (β = 0.14, p = 0.015), UPDRS 3 (β = −0.67, p = 0.001), and UPDRS T (β = −0.84, Signiﬁcant group eﬀects were found for UPDRS 3 and UPDRS T (Table 3), indicating higher scores in the anosmia group at Frontiers in Aging Neuroscience 03 frontiersin.org me e P- < me e P- < me e P- < fect vs. mia) Time effect, anosmia (T1 vs. T0) Time effect, non-anosmia (T1 vs. T0) Group × tim P-value B P-value B P-value B <0.001* NA NA NA NA NA 0.41 0.55 −3.62 <0.001* 4.03 0.064 NA NA NA NA NA 152.01 <0.001* 158.79 <0.001* −6.78 0.007* NA NA NA NA NA −0.19 0.684 −0.46 0.183 0.27 0.416 NA NA NA NA NA −0.46 0.682 0.83 0.462 −1.29 <0.001* NA NA NA NA NA −1.24 0.425 3.03 0.022* −4.28 0.001* NA NA NA NA NA −1.15 0.683 6.37 0.014* −7.52 a Smell Identiﬁcation Test; LEDD, equivalent dose of daily levodopa; MoCA, Montreal’s Cognitive Assessment; BDI-II, Beck’s Depres Fang et al. 10.3389/fnagi.2023.1213977 FIGURE 1 Difference in changes between non-anosmia and anosmia group in UPSIT (A), UPDRST (B), UPDRS3 (C), LEDD (D), MoCA (E), and BDI-II (F) during the follow-up. T0, ﬁrst visit; T1, follow-up visit; UPSIT, University of Pennsylvania Smell Identiﬁcation Test; UPDRST, total score of MDS-UPDRS; UPDRS3, part 3 score of MDS-UPDRS; LEDD, equivalent dose of daily levodopa; MoCA, Montreal’s Cognitive Assessment; BDI-II, Beck’s Depression Inventory-II. FIGURE 1 Difference in changes between non-anosmia and anosmia group in UPSIT (A), UPDRST (B), UPDRS3 (C), LEDD (D), MoCA (E), and BDI-II (F) during the follow-up. T0, ﬁrst visit; T1, follow-up visit; UPSIT, University of Pennsylvania Smell Identiﬁcation Test; UPDRST, total score of MDS-UPDRS; UPDRS3, part 3 score of MDS-UPDRS; LEDD, equivalent dose of daily levodopa; MoCA, Montreal’s Cognitive Assessment; BDI-II, Beck’s Depression Inventory-II. those of other longitudinal studies. For instance, Lewis et al. (2020) analyzed PD patients annually and found that UPSIT signiﬁcantly decreased in early and middle-stage PD but not in later-stage PD with disease duration exceeding 5 years. Domellof et al. (2017) explored the UPSIT outcome with the interaction eﬀect between the group (hyposmic/normosmic) and time, revealing that UPSIT deteriorated over time in the normosmic group while remaining stable in the hyposmic group. Meusel et al. group (Figure 1C). Regarding LEDD, the time eﬀects for both the anosmia group (β = 116.90, p = 0.001) and the non-anosmia group (β = 122.28, p = 0.001) were signiﬁcant, but there was no signiﬁcance in either group eﬀect or group-by-time eﬀect (Figure 1D and Table 3). Our results support these ﬁndings by indicating that the rate of olfactory decline with disease progression is more pronounced in patients without severe initial olfactory deﬁcits, whereas the olfactory deﬁcit remains relatively stable in patients with profound olfactory deﬁcits. While olfactory impairment is considered a premotor feature of Parkinson’s disease (PD), it is important to note that the olfactory impairment may continue to progress even after motor symptoms have emerged until it reaches a point known as the “ﬂoor eﬀect” in the current olfactory test (Fullard et al., 2017). This corresponds to the hypothesis proposed by Huisman et al. (2004) suggesting that dopaminergic neurons in the olfactory bulb, which act as possible suppressors in olfactory transmission, increase as a compensatory mechanism to the dopamine deﬁcit in the basal ganglia. With disease progression, the decrease in olfactory bulb volume and the deposition of Lewy bodies in the olfactory bulb may neutralize such inhibitory changes, resulting in less signiﬁcant olfactory degeneration (Herting et al., 2008). However, olfactory loss in PD may not be simply explained by imbalance of dopamine projection because the olfactory function involves several neurotransmitters such as acetylcholine, norepinephrine, serotonin and GABA (Doty, 2017). As olfactory 3.5. Change in cognition and mood in anosmia/non-anosmia groups The MoCA score was lower in the anosmia group with a signiﬁcant group eﬀect (β = −2.13, p = 0.007) (Table 2), but the signiﬁcance disappeared after adjusting for age, gender, and disease duration (Table 3). No signiﬁcant eﬀects for MoCA were found for time or group-by-time eﬀects (Figure 1E and Table 3). Likewise, no signiﬁcant eﬀects were found for BDI in terms of group eﬀect, time eﬀect, or group-by-time eﬀect (Figure 1F and Table 3). Frontiers in Aging Neuroscience 3.4. Change in motor symptoms in anosmia/non-anosmia groups (2010) showed a larger olfactory decline in the subgroup of patients with no severe initial olfactory deﬁcit over 5 years of tracking. The patients with marked olfactory regression had an average disease duration of 2.3 years at the beginning of the visit. baseline. Although there were trends of increasing UPDRS T and UPDRS 3 scores in the non-anosmia group at follow-up (Figures 1B, C and Table 2), the time eﬀect lost its signiﬁcance after adjustment (Table 3). However, the group-by-time eﬀect for UPDRS 3 remained signiﬁcant (β = −4.2, p < 0.038) after adjusting for age, gender, and disease duration (Table 3). The increase in UPDRS 3 score was much more pronounced in the non-anosmia group (Figure 1C). ed for a smia v a) -value <0.001* 0.112 0.061 0.751 0.002* 0.015* mell Iden 4. Discussion The present study demonstrated that the UPSIT score regressed in the non-anosmia group while remaining stable in the anosmia group. Notably, the non-anosmia group had a relatively short course of the disease in this study. Our ﬁndings are consistent with frontiersin.org 05 ime P ressio for age, gender, and disease duration. mia vs. ) Time effect, anosmia (T1 vs. T0) Time effect, non-anosmia (T1 vs. T0) Group × t value B P-value B P-value B 0.001* NA NA NA NA NA 0.55 0.435 −3.48 <0.001* 4.03 .112 NA NA NA NA NA 116.9 0.001* 122.28 0.001* −5.37 .061 NA NA NA NA NA −0.01 0.978 −0.23 0.53 0.21 .751 NA NA NA NA NA −0.8 0.491 0.42 0.712 0 002* NA NA NA NA NA −1.9 0.239 2.32 0.095 −4.2 015* NA NA NA NA NA −3.31 0.265 4.01 0.138 −7.33 ell Identiﬁcation Test; LEDD, equivalent dose of daily levodopa; MoCA, Montreal’s Cognitive Assessment; BDI-II, Beck’s Depr 10.3389/fnagi.2023.1213977 Fang et al. 10.3389/fnagi.2023.1213977 model (Borghammer et al., 2021). These features of the body- ﬁrst model may reﬂect the association between olfactory Lewy pathology and the caudo-rostral progression of Lewy pathology. However, olfactory Lewy pathology is not only related to caudo- rostral progression. Kok et al. (2021) found two features of olfactory Lewy pathology in the Vantaa85 + cohort: caudo-rostral progression and amygdala-based progression, corresponding to the body-ﬁrst and brain-ﬁrst models, respectively. This may also explain why not all patients in the non-anosmia group in our study turned to anosmia during follow-up and indicates that the severity and pathological changes of olfaction in PD are not a single pattern of progression. Further research with larger, more deﬁnitive patient classiﬁcation, longer follow-up studies, and the inclusion of pathology and imaging is required to elucidate the relationship between olfaction and PD. dysfunction appears to be more closely associated with the body- ﬁrst type of alpha-synuclein propagation, the pathology primarily aﬀecting the dorsal motor nucleus of the vagus or brainstem may impact olfactory function through the development of alpha-synucleinopathy in the bilateral olfactory bulbs or other brainstem nuclei that project to the olfactory system (Borghammer, 2021). Some cross-sectional studies have shown that olfactory degeneration is unrelated to the disease’s course (Cavaco et al., 2015; Masala et al., 2018). 4. Discussion Other longitudinal studies have also shown no signiﬁcant change in olfaction over time in patients with PD (Doty et al., 1988; Muller et al., 2002; Herting et al., 2008; Campabadal et al., 2017; Fujio et al., 2020). Such diﬀerent results may be related to diﬀerent study designs, such as the number of patients enrolled, the characteristics of diﬀerent patient groups, and so on. In our study, patients were divided into two groups, anosmia and non- anosmia, and the course of the disease diﬀered between the two groups. Therefore, grouping patients according to the degree or duration of olfactory abnormalities may explain the discrepancies between the results of these studies. This study has some limitations. First, the follow-up period of 1–3 years and the single follow-up session may not have been suﬃcient to detect changes in clinical data over a longer period. However, changes in olfaction in patients with shorter disease duration and non-anosmia progressed within 3 years, while the severity of signiﬁcant motor symptoms and cognitive function may require a longer follow-up period to observe a diﬀerence. Second, although we tried to exclude the possibility that olfactory tests were aﬀected by diseases such as sinonasal disease or upper respiratory tract infection, which commonly aﬀect the sense of smell, there are many other causes of olfactory abnormalities, including idiopathic causes (which may account for 18% of patients with olfactory abnormalities), that may aﬀect test results (Temmel et al., 2002). Thirdly, for safety and the subjects’ preference, we used the On status UPDRS score for the assessment of motor symptoms and disease severity, and therefore, the assessment may be inﬂuenced by medication. Nevertheless, these patients are regularly followed up in the outpatient clinic, and the physician ensures that the patient’s medication dosage is adequate. We also analyzed LEDD, which showed that the non-anosmia group had a lower LEDD than the anosmia group, but there was no signiﬁcant diﬀerence between the two. This indirectly implies that the non-anosmia group was not using fewer medications despite having a lower UPDRS score. Therefore, the eﬀect of insuﬃcient dosage of medication on the increasing UPDRS score in the non-anosmia group in this study may be subtle. Olfactory deterioration in patients with PD is thought to be associated with cognitive decline, and in particular, the accuracy of olfactory identiﬁcation tests is often aﬀected by cognitive decline (Laing and Doty, 2003). 5. Conclusion In conclusion, this study shows that olfactory sensation may still regress in Parkinson’s patients with a shorter course of the disease without anosmia, while it remains stable in the anosmia group. Such a decline in olfaction may not be related to cognitive status but may be associated with disease progression. Larger, long-term follow-up studies incorporating pathology and imaging analysis are needed to elucidate the underlying mechanisms. 4. Discussion However, the results of this study showed that although the UPSIT scores of the non-anosmia group decreased after follow-up, there was no signiﬁcant diﬀerence in the MoCA scores for the cognitive function component. This may suggest that while there is a signiﬁcant association between hyposmia in PD patients and cognitive decline, the initial regression in olfactory identiﬁcation is not solely attributed to cognitive decline. Other factors, such as Lewy body-related pathology in the peripheral and central olfactory organs or change in the balance of neurotransmitters, may play a role. Regarding disease severity, although the association with olfactory abnormalities remains inconclusive, our study found a signiﬁcant association between UPSIT and UPDRS T score and UPDRS 3 scores, in line with the results of other studies (Roos et al., 2019). Unlike the longitudinal study by He et al. (2020) which showed that olfactory abnormalities were predictive of disease progression, our study found no change in UPDRS T and UPDRS 3 score in the anosmia group during short-term follow-up, but there was a tendency for symptoms to progress in the non-anosmia group. These diﬀerent results may be due to diﬀerences in the length of follow-up, patient subgroups, and analysis methods. In addition, the worsening of Parkinson’s symptoms and olfaction in the non-anosmia group during the follow-up period may indirectly support the theory of Lewy body pathology between the brainstem and olfactory organs, as well as the inﬂuence of neurotransmitters such as dopamine. In the Braak staging system, Lewy body pathology was initially found in the olfactory bulb, but this lesion did not progress further, suggesting that a cascade of pathological changes from the brainstem upward is the main pathway (Braak et al., 2003). Horsager et al. (2020) proposed a body-ﬁrst and brain-ﬁrst model for the progression of PD pathology based on the presence or absence of RBD and the results of 123I-metaiodobenzylguanidine (MIBG) scintigraphy. The body- ﬁrst model corresponds to the spreading pathway of the Braak staging system. In addition to autonomic-related prodrome and RBD, the body-ﬁrst model has a faster progression of motor symptoms and earlier olfactory abnormalities than the brain-ﬁrst Frontiers in Aging Neuroscience References Fujio, H., Inokuchi, G., Kuroki, S., Tatehara, S., Katsunuma, S., Kowa, H., et al. (2020). Three-year prospective study on olfaction of patients with Parkinson’s disease. Auris. Nasus Larynx 47, 899–904. doi: 10.1016/j.anl.2019.08.008 Beck, A., Steer, R., and Brown, G. (1996). BDI-II: Beck Depression Inventory Manual, 2nd Edn. San Antonio: Psychological Corporation. Berendse, H. W., Roos, D. S., Raijmakers, P., and Doty, R. L. (2011). Motor and non-motor correlates of olfactory dysfunction in Parkinson’s disease. J. Neurol. Sci. 310, 21–24. doi: 10.1016/j.jns.2011.06.020 Fullard, M. E., Morley, J. F., and Duda, J. E. (2017). Olfactory Dysfunction as an Early Biomarker in Parkinson’s Disease. Neurosci. Bull. 33, 515–525. doi: 10.1007/ s12264-017-0170-x Borghammer, P. (2021). The alpha-Synuclein Origin and Connectome Model (SOC Model) of Parkinson’s Disease: Explaining motor asymmetry, non-motor phenotypes, and cognitive decline. J. Parkinsons Dis. 11, 455–474. doi: 10.3233/JPD-202481 Goetz, C. G., Tilley, B. C., Shaftman, S. R., Stebbins, G. T., Fahn, S., Martinez- Martin, P., et al. (2008). Movement Disorder Society-sponsored revision of the Uniﬁed Parkinson’s Disease Rating Scale (MDS-UPDRS): scale presentation and clinimetric testing results. Mov. Disord. 23, 2129–2170. doi: 10.1002/mds. 22340 Borghammer, P., Horsager, J., Andersen, K., Van Den Berge, N., Raunio, A., Murayama, S., et al. (2021). Neuropathological evidence of body-ﬁrst vs. brain-ﬁrst Lewy body disease. Neurobiol. Dis. 161:105557. doi: 10.1016/j.nbd.2021.105557 Haehner, A., Boesveldt, S., Berendse, H. W., Mackay-Sim, A., Fleischmann, J., Silburn, P. A., et al. (2009). Prevalence of smell loss in Parkinson’s disease–a multicenter study. Parkinsonism Relat. Disord. 15, 490–494. doi: 10.1016/j.parkreldis. 2008.12.005 Braak, H., Del Tredici, K., Rub, U., de Vos, R. A., Jansen Steur, E. N., and Braak, E. (2003). Staging of brain pathology related to sporadic Parkinson’s disease. Neurobiol. Aging 24, 197–211. doi: 10.1016/s0197-4580(02)00065-9 Campabadal, A., Uribe, C., Segura, B., Baggio, H. C., Abos, A., Garcia-Diaz, A. I., et al. (2017). Brain correlates of progressive olfactory loss in Parkinson’s disease. Parkinsonism Relat. Disord. 41, 44–50. doi: 10.1016/j.parkreldis.2017.05.005 He, R., Zhao, Y., He, Y., Zhou, Y., Yang, J., Zhou, X., et al. (2020). Olfactory dysfunction predicts disease progression in Parkinson’s Disease: A longitudinal study. Front. Neurosci. 14:569777. doi: 10.3389/fnins.2020.569777 Cavaco, S., Goncalves, A., Mendes, A., Vila-Cha, N., Moreira, I., Fernandes, J., et al. (2015). Abnormal Olfaction in Parkinson’s Disease Is Related to Faster Disease Progression. Behav. Neurol. 2015:976589. doi: 10.1155/2015/976589 Herting, B., Schulze, S., Reichmann, H., Haehner, A., and Hummel, T. (2008). A longitudinal study of olfactory function in patients with idiopathic Parkinson’s disease. Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. frontiersin.org 07 Fang et al. 10.3389/fnagi.2023.1213977 Funding This study was funded by the Taichung Veterans General Hospital TCVGH 1123402C and 1123401C. Acknowledgments The studies involving human participants were reviewed and approved by the Taichung Veterans General Hospital Institutional Review Board/Ethics Committee (No. CE22189B). The patients/participants provided their written informed consent to participate in this study. We thank the staﬀs at the Department of Neurology in Taichung Veteran General Hospital for participant recruitment. Author contributions The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. T-CF and M-HC conceptualized the project. T-CF and Y-ST performed the data acquisition and analysis. T-CF wrote the ﬁrst draft of the manuscript. M-HC critically reviewed the manuscript. All authors contributed to writing and revising the manuscript. Publisher’s note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their aﬃliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. References doi: 10.1016/j.parkreldis.2016.09.023 Rodríguez-Violante, M., Camacho-Ordoñez, A., Cervantes-Arriaga, A., González- Latapí, P., and Velázquez-Osuna, S. (2015). Factors associated with the quality of life of subjects with Parkinson’s disease and burden on their caregivers. Neurología 30, 257–263. doi: 10.1016/j.nrleng.2014.01.002 Lewis, M. M., Harkins, E., Lee, E. Y., Stetter, C., Snyder, B., Corson, T., et al. (2020). Clinical Progression of Parkinson’s Disease: Insights from the NINDS Common Data Elements. J. Parkinsons Dis. 10, 1075–1085. doi: 10.3233/JPD-201932 Rodríguez-Violante, M., Ospina-García, N., Pérez-Lohman, C., and Cervantes- Arriaga, A. (2017). Spotlight on olfactory dysfunction in Parkinson’s disease. J. Parkinson. Restless Legs Syndr. 7, 33–41. doi: 10.2147/jprls.S125390 Roos, D. S., Twisk, J. W. R., Raijmakers, P., Doty, R. L., and Berendse, H. W. (2019). Hyposmia as a marker of (non-)motor disease severity in Parkinson’s disease. J. Neural Transm. 126, 1471–1478. doi: 10.1007/s00702-019-02074-0 Litvan, I., Goldman, J. G., Troster, A. I., Schmand, B. A., Weintraub, D., Petersen, R. C., et al. (2012). Diagnostic criteria for mild cognitive impairment in Parkinson’s disease: Movement Disorder Society Task Force guidelines. Mov. Disord. 27, 349–356. doi: 10.1002/mds.24893 Santos Garcia, D., de Deus Fonticoba, T., Suarez Castro, E., Borrue, C., Mata, M., Solano Vila, B., et al. (2019). Non-motor symptoms burden, mood, and gait problems are the most signiﬁcant factors contributing to a poor quality of life in non-demented Parkinson’s disease patients: Results from the COPPADIS Study Cohort. Parkinsonism Relat. Disord. 66, 151–157. doi: 10.1016/j.parkreldis.2019. 07.031 Liu, C. C., Li, C. Y., Lee, P. C., and Sun, Y. (2016a). Variations in incidence and prevalence of parkinson’s disease in Taiwan: A population-based nationwide study. Parkinsons. Dis. 2016:8756359. doi: 10.1155/2016/8756359 Liu, W. M., Wu, R. M., Lin, J. W., Liu, Y. C., Chang, C. H., and Lin, C. H. (2016b). Time trends in the prevalence and incidence of Parkinson’s disease in Taiwan: A nationwide, population-based study. J. Formos Med. Assoc. 115, 531–538. doi: 10.1016/ j.jfma.2015.05.014 Stebbins, G. T., Goetz, C. G., Burn, D. J., Jankovic, J., Khoo, T. K., and Tilley, B. C. (2013). How to identify tremor dominant and postural instability/gait diﬃculty groups with the movement disorder society uniﬁed Parkinson’s disease rating scale: comparison with the uniﬁed Parkinson’s disease rating scale. Mov. Disord. 28, 668–670. doi: 10.1002/mds.25383 Masala, C., Solla, P., Liscia, A., Defazio, G., Saba, L., Cannas, A., et al. (2018). Correlation among olfactory function, motors’ symptoms, cognitive impairment, apathy, and fatigue in patients with Parkinson’s disease. J. Neurol. 265, 1764–1771. References J. Neurol. 255, 367–370. doi: 10.1007/s00415-008-0665-5 Domellof, M. E., Lundin, K. F., Edstrom, M., and Forsgren, L. (2017). Olfactory dysfunction and dementia in newly diagnosed patients with Parkinson’s disease. Parkinsonism Relat. Disord. 38, 41–47. doi: 10.1016/j.parkreldis.2017.02.017 Horsager, J., Andersen, K. B., Knudsen, K., Skjaerbaek, C., Fedorova, T. D., Okkels, N., et al. (2020). Brain-ﬁrst versus body-ﬁrst Parkinson’s disease: a multimodal imaging case-control study. Brain 143, 3077–3088. doi: 10.1093/brain/awaa238 Doty, R. L. (2001). The Smell Identiﬁcation Test Administration Manual. Philadelphia, PA: Sensonics, Inc. Huisman, E., Uylings, H. B., and Hoogland, P. V. (2004). A 100% increase of dopaminergic cells in the olfactory bulb may explain hyposmia in Parkinson’s disease. Mov. Disord. 19, 687–692. doi: 10.1002/mds.10713 Doty, R. L. (2012). Olfactory dysfunction in Parkinson disease. Nat. Rev. Neurol. 8, 329–339. doi: 10.1038/nrneurol.2012.80 Jiang, R. S., Su, M. C., Liang, K. L., Shiao, J. Y., Wu, S. H., and Hsin, C. H. (2010). A pilot study of a traditional Chinese version of the University of Pennsylvania Smell Identiﬁcation Test for application in Taiwan. Am. J. Rhinol. Allergy 24, 45–50. doi: 10.2500/ajra.2010.24.3388 Doty, R. L. (2017). Olfactory dysfunction in neurodegenerative diseases: is there a common pathological substrate? Lancet Neurol. 16, 478–488. doi: 10.1016/S1474- 4422(17)30123-0 Doty, R. L., Deems, D. A., and Stellar, S. (1988). Olfactory dysfunction in parkinsonism: a general deﬁcit unrelated to neurologic signs, disease stage, or disease duration. Neurology 38, 1237–1244. doi: 10.1212/wnl.38.8.1237 Kok, E. H., Savola, S., Raunio, A., Oinas, M., Tuimala, J., Polvikoski, T., et al. (2021). Alpha-synuclein pathology of olfactory bulbs/peduncles in the Vantaa85+ cohort exhibit two divergent patterns: a population-based study. Acta Neuropathol. 142, 777–780. doi: 10.1007/s00401-021-02364-6 Ercoli, T., Masala, C., Cadeddu, G., Mascia, M. M., Oroﬁno, G., Gigante, A. F., et al. (2022). Does Olfactory Dysfunction Correlate with Disease Progression in Parkinson’s Disease? A Systematic Review of the Current Literature. Brain Sci. 12, 513. doi: 10. 3390/brainsci12050513 Laing, D. G. D. G., and Doty, R. L. (2003). Psychophysical measurement of human olfactory function, including odorant mixture assessment. New York City, NY: Marcel Dekker. Frontiers in Aging Neuroscience 08 frontiersin.org Fang et al. Fang et al. 10.3389/fnagi.2023.1213977 Lawton, M., Hu, M. T., Baig, F., Ruﬀmann, C., Barron, E., Swallow, D. M., et al. (2016). Equating scores of the University of Pennsylvania smell identiﬁcation test and sniﬃn’ sticks test in patients with Parkinson’s disease. Parkinsonism Relat. Disord. 33, 96–101. References doi: 10.1007/s00415-018-8913-9 Taguchi, T., Ikuno, M., Hondo, M., Parajuli, L. K., Taguchi, K., Ueda, J., et al. (2020). alpha-Synuclein BAC transgenic mice exhibit RBD-like behaviour and hyposmia: a prodromal Parkinson’s disease model. Brain 143, 249–265. doi: 10.1093/brain/ awz380 Meusel, T., Westermann, B., Fuhr, P., Hummel, T., and Welge-Lussen, A. (2010). The course of olfactory deﬁcits in patients with Parkinson’s disease–a study based on psychophysical and electrophysiological measures. Neurosci. Lett. 486, 166–170. doi: 10.1016/j.neulet.2010.09.044 Temmel, A. F., Quint, C., Schickinger-Fischer, B., Klimek, L., Stoller, E., and Hummel, T. (2002). Characteristics of olfactory disorders in relation to major causes of olfactory loss. Arch. Otolaryngol. Head Neck Surg. 128, 635–641. doi: 10.1001/archotol. 128.6.635 Muller, A., Reichmann, H., Livermore, A., and Hummel, T. (2002). Olfactory function in idiopathic Parkinson’s disease (IPD): results from cross-sectional studies in IPD patients and long-term follow-up of de-novo IPD patients. J. Neural Transm. 109, 805–811. doi: 10.1007/s007020200067 Tibar, H., El Bayad, K., Bouhouche, A., Ait Ben Haddou, E. H., Benomar, A., Yahyaoui, M., et al. (2018). Non-motor symptoms of Parkinson’s disease and their impact on quality of life in a cohort of moroccan patients. Front. Neurol. 9:170. doi: 10.3389/fneur.2018.00170 Picillo, M., Pellecchia, M. T., Erro, R., Amboni, M., Vitale, C., Iavarone, A., et al. (2014). The use of University of Pennsylvania smell identiﬁcation test in the diagnosis of Parkinson’s disease in Italy. Neurol. Sci. 35, 379–383. doi: 10.1007/s10072-013- 1522-6 Yoo, H. S., Chung, S. J., Lee, Y. H., Ye, B. S., Sohn, Y. H., and Lee, P. H. (2020). Association between olfactory deﬁcit and motor and cognitive function in Parkinson’s Disease. J. Mov. Disord. 13, 133–141. doi: 10.14802/jmd.19082 Poewe, W., Seppi, K., Tanner, C. M., Halliday, G. M., Brundin, P., Volkmann, J., et al. (2017). Parkinson disease. Nat. Rev. Dis. Primers 3:17013. doi: 10.1038/nrdp.2017.13 09 09 Frontiers in Aging Neuroscience frontiersin.org
https://openalex.org/W4381487452	https://atheneadigital.net/article/download/v23-n2-cazorla-reyes/3331-pdf-en	es		Corresponsabilidad en el cuidado informal de personas dependientes: construcción discursiva en la política pública chilena	Athenea digital	2,023	cc-by	8,128	Athenea Digital - 23(2): e3331 (julio 2023) -ARTÍCULOS- ISSN: 1578-8946 CORRESPONSABILIDAD EN EL CUIDADO INFORMAL DE PERSONAS DEPENDIENTES: CONSTRUCCIÓN DISCURSIVA EN LA POLÍTICA PÚBLICA CHILENA CO-RESPONSIBILITY IN THE INFORMAL CARE OF DEPENDENT PERSONS: DISCURSIVE CONSTRUCTION IN CHILEAN PUBLIC POLICY Ketty Cazorla-Becerra* ; María Isabel Reyes-Espejo* Universidad de Valparaíso; Millennium Institute for Care Research MICARE; **Pontificia Universidad Católica de Valparaíso; ketty.cazorla@uv.cl; maria.reyes@pucv.cl Historia editorial Resumen Recibido: 09-06-2022 En el escenario de una crisis global de cuidados este estudio aborda cómo se construye discursivamente la corresponsabilidad del cuidado y autocuidado de la cuidadora informal de personas dependientes en la política pública chilena. Para ello, desde un enfoque interpretativo, se optó por una metodología cualitativa de tipo documental para analizar textos del subsistema de apoyos y cuidados desde la perspectiva del discurso pragmáticos. Los resultados muestran un marcado discurso sobre una mujer cuidadora sobrecargada que autogestiona su autocuidado con escaso apoyo, para luego ver emerger una familia que, gestionada por la cuidadora, debe comenzar a hacerse cargo de la corresponsabilidad del cuidar. Tras ello, surge un Estado subsidiario que apoya a esta emprendedora del cuidado y a su familia cuando no hay recursos suficientes. Concluimos que este andamiaje se sostiene en los hombros de las mujeres, manteniendo la invisibilización de oportunidades que redistribuyan el cuidado con mayor justicia social. Aceptado: 10-05-2023 Publicado: 21-06-2023 Palabras clave Cuidado; Corresponsabilidad; Política pública; Discurso Abstract Keywords Care; Co-responsibility; Public policy; Discourse In the scenario of a global care crisis, this study addresses how the co-responsibility of care and self-care of the informal caregiver of dependent people is discursively constructed in Chilean public policy. To this end, an interpretative methodology, qualitative method and through a documentary approach text of the Support and Care Subsystem are analyzed from the perspective of pragmatic discourse. The results show a marked discourse about an overburdened female caregiver who self-manages her self-care with little support, and then sees the emergence of a family that, managed by the caregiver, must begin to take charge of the co-responsibility of caregiving. After that, a subsidiary State emerges that supports this care entrepreneur and her family when there are not enough resources. We conclude that this scaffolding is supported on the shoulders of women, maintaining the invisibility of opportunities that redistribute care with greater social justice. Cazorla-Becerra; Ketty & Reyes-Espejo, María Isabel (2023). Corresponsabilidad en el cuidado informal de personas dependientes: construcción discursiva en la política pública chilena. Athenea Digital, 23(2), e3331. https://doi.org/10.5565/rev/athenea.3331 Introducción La emblemática consigna feminista relativa a “poner la vida en el centro” se ha revitalizado en los últimos años debido a la situación de pandemia global (Arriagada, 2021). La alta demanda de labores de cuidado producidas por el cierre de instituciones y medidas de prevención de contagio como el distanciamiento o asilamiento social, visibili- 1 Corresponsabilidad en el cuidado informal de personas dependientes zaron la alta carga de trabajo que el cuidado implica en la vida cotidiana de las mujeres que asumen mayoritariamente esta tarea (Pérez, 2014). Este agravamiento de la crisis del cuidado se venía denunciando hace décadas, reclamando la ausencia de una corresponsabilidad social, entendida como una distribución compartida y justa de la práctica del cuidar (Rico, 2011). Siguiendo a Chiara Saraceno y Wolfgang Keck (2008), en Latinoamérica, los regímenes de bienestar en clave de cuidado presentan una alta variabilidad de corresponsabilidad entre países, aunque, en general se comparte —como soporte— un perfil familiarista. Este bienestar centrado en la familia puede producirse por defecto, ante la ausencia de Estado y mercado; por apoyo, cuando el Estado colabora con la familia ante la falta de acceso al mercado (con foco en población vulnerada); por opción, cuando hay alternativas estatales y de mercado, pero la familia elige no usarlos; y por desfamiliarismo, cuando el sujeto foco del bienestar ya no es la familia sino el ciudadano, colectivizando los efectos del cuidado (Batthyány, 2021; Genta et al., 2022). El cuidado en el caso de personas en situación de dependencia se enmarca en una organización familiarista que resuelve privadamente los costos del cuidar e invisibiliza el valor de relaciones colaborativas (Tronto, 2017). Así, el cuidado desde un contexto mercantilizado como el chileno1, sitúa a la dependencia como una carga y un gasto privado (Ortega, 2019). Esta lógica propicia desigualdades en el bienestar de las cuidadoras, quienes no cuentan con los recursos para desarrollar prácticas de (auto)cuidado en condiciones de calidad y justicia. Para María Durán (2018) este grupo emerge como un nuevo sujeto social que denomina “cuidatoriado”, aludiendo a la noción de precariado (Standing, 2011). Este refiere a una clase social formada principalmente por mujeres que ejercen el cuidado en condiciones de mínimos derechos garantizados y carecen de capacidad reivindicativa frente al Estado y la sociedad. De acuerdo con la literatura, las mujeres de este cuidatoriado podrían clasificarse en, al menos, dos perfiles. La cuidadora formal, que brinda apoyo con mayor especialización, siendo remunerada, y pudiendo estar —o no— enmarcada en una institución pública o privada (Ramacciotti, 2020). Y la cuidadora informal, que brinda apoyo de manera voluntaria, sin remuneración, ni institución u organización social pública o privada de por medio, enmarcándose en el contexto familiar (Batthyány, 2015). Tanto a nivel nacional como internacional, existe suficiente evidencia para denunciar la desvalorización de la práctica del cuidado informal que las mujeres 1 2 El origen se encuentra en las reformas neoliberales realizadas desde 1975 en Dictadura y que, sostenidas en el tiempo, impulsaron la constitución de un estado subsidiario que permanece hasta nuestros días (Sisto et al., 2021). En particular, las reformas en salud y previsión, orientadas a la privatización e individualización de las protecciones sociales, moldearon las políticas públicas con atención focalizada y en base a condiciones estrictas para su financiamiento y control (Morales, 2017; Olivares et al., 2018; Pávez, 2018). Ketty Cazorla-Becerra; María Isabel Reyes-Espejo brindan a personas dependientes, coincidiendo en precarias condiciones para ejercer tanto el cuidado hacia otro como en el cuidado personal (Vellone et al., 2020). Este último aspecto se hace visible cuando afecta la labor de cuidado, pero no cuando se trata de proteger la autonomía de la cuidadora, reproduciendo una injusta visión de cuerpos femeninos que no siempre son reconocidos más allá del cuidado que brindan (Butler, 2017). Paradójicamente, esta noción de autocuidado desvalorizada es erigida como una herramienta estratégica para transformar desigualdades de género (Pautassi, 2018). El cuidar, entonces, no es una acción unidireccional, sino que representa una práctica de corresponsabilidad multidireccional hacia otros y hacia uno mismo en un entorno situado (Pereira y Cárdenas, 2019). A nivel global, la literatura científica reciente señala que el perfil de las personas cuidadoras en el contexto informal corresponde a mujeres que rodean los 50 años o más, mostrando como tendencia un mayor envejecimiento a medida que aumenta la edad de la persona cuidada (Escandell-Rico y Pérez-Fernández, 2022). Se reconoce, además, como una práctica anclada en la cotidianidad familiar y en la que, por este perfil generizado, se requiere equilibrar el cuidado de personas dependientes con el cuidado familiar, propiciando un alto costo para el bienestar de la cuidadora (Ezquerra y De Eguía, 2020). Las prácticas de autocuidado de estas cuidadoras resultan, por lo general, insuficientes (Cazorla-Becerra y Gamboa-Morales, 2023; Dixie y Querido, 2020). Se argumenta que tanto la mujer que cuida como la misma política pública, tienden a priorizar el cuidado del otro dependiente y sus necesidades, quedando en segundo plano los costos biopsicosociales que ellas deben enfrentar (Pereira et al., 2019; Reyes et al., 2022). Si bien hay coincidencia en la severidad de la sobrecarga de trabajo, las investigaciones se concentran en describir efectos físicos en la cuidadora, como problemas musculares, óseos y enfermedades crónicas (Duarte et al., 2018) y, subsecuentemente, se señalan efectos psicológicos como depresión, ansiedad y estrés (Mosquera et al., 2020). Las consecuencias sociales, como el aislamiento, precariedad laboral y empobrecimiento, son dimensiones menos abordadas en las investigaciones (Rahimi et al., 2021). En Chile, la política de cuidado surge como una iniciativa que, desde el 2016, propone un subsistema de apoyos y cuidados en favor de personas en situación de dependencia, con énfasis en personas mayores y discapacidad (Hernández y Rodríguez, 2022). Sin embargo, inicialmente, se presenta con una limitada cobertura geográfica (6 % del total nacional), focalizada en familias de precarios quintiles de ingreso, brindando un soporte esporádico de relevo para cuidadoras, y situándolas como recurso de apoyo al sistema sanitario, sin considerarlas como sujeto de derecho propio (Grandón, 2021). 3 Corresponsabilidad en el cuidado informal de personas dependientes En este marco, el objetivo de este estudio se dirige a comprender cómo se construye discursivamente la corresponsabilidad del cuidado y autocuidado de la cuidadora informal de personas dependientes en la política pública chilena vigente. Para ello, utilizamos una metodología de perfil interpretativo cualitativo basado en el análisis documental de discurso pragmático, que examina tres documentos fundamentales del Subsistema Nacional de Apoyos y Cuidados durante el período 2022. En lo siguiente, el escrito abordará una revisión de la literatura que desarrolla los conceptos de cuidado y autocuidado para luego abordar las políticas públicas de cuidado. Ambos ejes resultan estratégicos para el análisis de la corresponsabilidad social, y permiten dar un contexto que justifica las decisiones metodológicas que levantan tres resultados relevantes: la naturalización de la sobrecarga de la cuidadora y su autogestión, la familia biparental como responsable principal del cuidado y el rol subsidiario del Estado en el cuidado informal. Finalmente, desarrollamos algunas conclusiones a partir de los hallazgos del estudio. Nociones de cuidado y autocuidado El cuidado, etimológicamente, se refiere a “preocuparse por” o “atender a”, un significado que tiene un claro contenido ético al demandar un compromiso con la protección y salvaguarda de la humanidad (Quezada y Bascones, 2020). Este cuidado será entendido como una actividad característica de la especie humana que incluye todo lo que hacemos con vistas a mantener, continuar o reparar el mundo, para así poder vivir en él de la mejor manera posible, asumiendo equitativamente —y sin excusa— las responsabilidades del cuidado en torno a las dimensiones material, económica y afectiva (Tronto, 2017). Así, el acto de cuidar, anclado al propósito del sostenimiento de la vida, implica que la compleja trama de cuidado se construye desde vínculos interdependientes y multidireccionales que confluyen tanto a nivel productivo como reproductivo (García y Urbina, 2021). En particular, el cuidado de personas dependientes se dirige a proteger la subsistencia considerando acciones básicas como comer o bañarse, llegando hasta el sostenimiento de vínculos sociales, como la participación ciudadana. Este cuidado se constituye en la acción de ayuda a una persona que, por razones de dependencia, requiere apoyo para su bienestar en la vida cotidiana, implicando un sustrato material (trabajo), económico (costo) y psicológico (afectos) (Batthyány et al., 2017). Al sostenerse sobre prácticas repetitivas y cotidianas, compromete a las mujeres a dedicar una gran cantidad de tiempo, teniendo dificultades para articular su quehacer con otros trabajos y dimensiones de la cotidianeidad. Así, la alta demanda y la multiposición de roles que im- 4 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo plica su ejercicio, intensifica desigualdades de género para las cuidadoras (Genta, 2021). En este contexto problemático y de alta demanda para las cuidadoras, el ejercicio de autocuidado aparece como estrategia relevante para la política pública a nivel global, constituyéndose, en el escenario actual Latinoamericano, como un componente clave del derecho al cuidado, junto a la tarea de cuidar y ser cuidado (Pautassi, 2018). Desde la contingencia institucional local, en Chile ha sido entendido como una actividad autónoma que realiza —y debe realizar— la cuidadora para proteger su propia salud y bienestar (Ministerio de Salud, 2013). Complementariamente, el autocuidado puede ser comprendido desde la noción foucaultiana del cuidado de sí (Foucault, 1982/2014), extendiendo su significado a la demanda por la elaboración de un arte de sí. Esta perspectiva permite alejarse de reflexiones binarias que refieren al uso de la noción de autocuidado, ayudando a complejizar su análisis. En este sentido, Foucault señala que este arte se revela en la ética de ser un sujeto que se construye en relación, con el propósito de ser feliz y colaborar por una comunidad mejor (1982/2014). De este modo, el “cuidado de sí” implicaría la acción de tres componentes relacionales: con uno mismo, con los otros y con el entorno, constituyéndose en los escenarios de aplicación en una lógica de corresponsabilidad que, mancomunadamente, se actualiza en la intersubjetividad de cada sujeto (Muñoz, 2012). No obstante, en cuanto al reconocimiento que se hace de estos aportes, es importante consignar el debate crítico que, desde una ética feminista del cuidado, diversas autoras realizan al planteamiento foucaultiano del cuidado de sí (Dilts, 2011; Myers, 2008). Al respecto, se señala que la propuesta reflexiva del sujeto sobre sí mismo, terminaría siendo un ejercicio individualizante y despolitizador que asimilaría fácilmente el proyecto ético-político neoliberal conformando subjetividades resilientes y autovalentes, desde la promoción de una autogestión centrada en las capacidades individuales, dejando poco espacio para los principios de reciprocidad e interdependencia a nivel cultural y comunitario (Allen, 2004). Esto, siguiendo a Verónica Schild (2007), trae el riesgo de reproducir un sujeto consumidor-ciudadano-empoderado que, a través de su potencial libertad para elegir, debe responsabilizarse de sí mismo. Reconociendo estos debates, nos quedaremos, al menos para efectos del presente estudio, con la visión más conciliadora del cuidado de sí de feministas foucaultianas que argumentan que esta noción no tiene por qué equivaler a nada más que individualismo, ya que ninguna práctica ética es posible sin la comunicación con otros y su entorno (McLaren, 1997). 5 Corresponsabilidad en el cuidado informal de personas dependientes Política pública y corresponsabilidad del cuidado En Chile, las diversas encuestas del uso del tiempo declaran que, pese al aumento de la participación laboral femenina en el trabajo remunerado, no ha aumentado la participación en las labores de cuidados de los hombres (Hernández y Rodríguez, 2022). La conciliación entre la vida laboral y familiar, con una organización de corresponsabilidad del cuidado, es uno de los principales puntos ciego de las políticas públicas (Grandón, 2021). Las mujeres cuidadoras asumen los efectos de esta ausencia de corresponsabilidad, ya que no existe una organización del cuidado que distribuya de manera justa estas responsabilidades y que reconozcan apoyo en dimensiones clave como el tiempo, la remuneración o los servicios (Fraser, 2016). Los estudios sobre políticas públicas de cuidado informal en Chile destacan su raigambre neoliberal, modelo desde el que se ha construido una organización de cuidados sostenida en una diada intra-hogar compuesta por una persona dependiente y una mujer cuidadora (Comelín y Leiva, 2018). Esta diada responde a un modelo de bienestar con apoyo estatal de tipo subsidiario que convive con una oferta privada, costosa y desregulada; combinación que carece de compatibilidad con las múltiples necesidades en las prácticas del cuidar (Arriagada, 2021). Desde la economía feminista se han propuesto otras formas de provisión de bienestar y distribución del cuidado, como la Teoría del diamante, que busca equilibrar el autocuidado y la corresponsabilidad por cuidar, involucrando otros actores: miembros del hogar (incluyendo hombres), políticas públicas de Estado, organizaciones ciudadanas y el mundo privado (Razavi, 2007). Las fronteras entre estos actores no son nítidas ni estáticas, por ejemplo, el papel del Estado es cualitativamente diferente, porque no es sólo un proveedor de servicios, sino un decisor significativo cuando se trata de los derechos sociales. En este modelo, es fundamental identificar cómo el Estado orienta y evalúa el acceso a cuidados de calidad y quién carga con sus costos. La creación, regulación y financiamiento de un sistema de cuidados con enfoque de derechos sociales, permite incrementar la disponibilidad y la calidad de los cuidados y reducir la carga de tiempo que se asigna a las cuidadoras no remuneradas, junto con mayores oportunidades para su autocuidado (Pautassi, 2021). Al respecto, es posible señalar que, en Chile, existe una invisibilización del cuidado informal y autocuidado de la cuidadora de personas dependientes, que incluye a la política pública. La disputa por visibilizar y poner en valor este trabajo es una primera batalla que hay que ganar para alcanzar mayor justicia social y de género. Y para colaborar en este sentido, en este estudio nos preguntamos: ¿Cómo se construye 6 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo discursivamente la corresponsabilidad del cuidado y autocuidado de la cuidadora informal de personas dependientes en la política pública chilena vigente? Decisiones metodológicas En este estudio, de tipo exploratorio-descriptivo (Denzin y Lincoln, 2003), asumimos una posición epistemológica interpretativa (Creswell, 2013). Recurrimos al método cualitativo (Flick, 2004) circunscrito a un enfoque de producción de información documental, comprendiendo al documento como un actor que prescribe relaciones y formas de interacción (Prior, 2008), orientada desde una perspectiva de la psicología discursiva pragmática que permite analizar los textos, entendiéndolos como formas de acción social (Potter y Wetherell, 1987). Así, el discurso se comprende como un grupo de prácticas lingüísticas que constituyen relaciones sociales (Íñiguez y Antaki, 1994). Para la selección de los documentos se realizó, durante el año 2022, un muestreo pragmático (Denzin y Lincoln, 2003), considerando los siguientes criterios: i) texto de autoría de un organismo estatal abocado al cuidado informal; ii) vigente y publicado en la página web del Subsistema de Apoyo y Cuidados chileno; y iii) que se dirija particularmente al/la cuidador/a informal de personas dependientes. Como criterios de exclusión se consideraron: i) texto de autoría de sector privado; ii) publicado en una página ajena al Subsistema de Apoyo y Cuidados de Chile; iii) que en su contenido no refiera /la cuidador/a de personas dependientes; y iv) que su destinatario no sea el/la cuidador/a. Este encuadre orientó la búsqueda en la página web del Subsistema de Apoyos y Cuidados, situándose dentro del Programa denominado Red local de Apoyos y Cuidados perteneciente al Ministerio de Desarrollo Social y Familia, seleccionando tres documentos (Ver Tabla 1). Nombre Objetivo Guía de Corresponsabilidad en el Cuidado (Mi- Entregar contenidos de género y corresponsabilidad dirigidos a equipos técnicos nisterio de Desarrollo Social y Familia, 2022a) y cuidadoras de personas con dependencia, sus familias y comunidad. Ficha educativa N.°1: Aprendiendo a cuidarme (Ministerio de Desarrollo Social y Familia, (2022b) Brindar orientaciones sobre autocuidado a cuidadoras de personas en situación de dependencia. Ficha educativa N.°2: ¿Cómo avanzar hacia la corresponsabilidad en el cuidado? (Ministerio de Desarrollo Social y Familia, 2022c) Indicar sugerencias sobre corresponsabilidad a cuidadoras de personas en situación de dependencia. Tabla 1. Documentos seleccionados para el estudio 7 Corresponsabilidad en el cuidado informal de personas dependientes En su conjunto, estos documentos entregan, a nivel nacional, uno de los lineamientos más importantes para comprender la retórica que prescribe el cuidado y el autocuidado de la cuidadora informal; constituyéndose en la política orientadora para cada región del país donde funciona el Subsistema de Apoyo y Cuidados de Chile, por lo que su análisis permitirá comprender cómo éstos prescriben discursos relativos a la corresponsabilidad. Analíticamente recurrimos a la perspectiva de Análisis de Discurso Pragmático que entiende el discurso como un conjunto de prácticas lingüísticas que actúan en el presente, promoviendo posiciones de sujeto y relaciones sociales (Potter y Wetherell, 1987). Se utiliza la herramienta de análisis denominada repertorios interpretativos, entendida como unidades lingüísticas que construyen versiones de acontecimientos sociales mediante recursos retóricos (Wetherell y Potter, 1996). El procedimiento contempla, en primer lugar, lecturas sucesivas para simplificar el corpus textual mediante la selección de categorizaciones ligadas al objetivo propuesto. En segundo lugar, se examinan estas categorizaciones y sus contextos en el documento, asignando nombres con el fin de identificar los repertorios interpretativos emergentes (Potter y Wetherell, 1987). En tercer lugar, cada repertorio es analizado a través de la identificación de sus recursos de factualización, es decir, de las estrategias lingüísticas específicas que se utilizan para hacer que algo aparezca como un hecho incuestionable, destacándose: descripción factual, acreditación de categorías, discurso epocal y modalización (Potter, 1998). Los repertorios interpretativos resultantes son tres: 1) La naturalización de la sobrecarga de la cuidadora y su autogestión, analizado a través de la descripción factual, entendida como el reconocimiento de mecanismos retóricos que propician un hecho social determinado; 2) La familia biparental como responsable principal del cuidado informal, analizado a través de la acreditación de categorías, comprendida como la intención del discurso para habilitar el pertenecer a una determinada condición a fin de reforzar una versión, y el discurso epocal, referido a la construcción de alternativas que se fundamentan entre un pasado y un presente; y 3) El rol subsidiario del Estado en el cuidado informal, analizado desde los recursos de acreditación de categorías (ya descrito) y de modalización, que busca modificar una afirmación que se utiliza para reforzarla o debilitarla (Hozven y Sisto, 2021). A continuación, describimos los repertorios reconstruidos con mayor profundidad: 8 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo Naturalización de la sobrecarga de la cuidadora informal y su autogestión Este primer repertorio interpretativo es identificado en todos los documentos, pero se focalizará en dos descripciones factuales: la sobrecarga de la cuidadora y su autogestión. Como primera descripción factual, el discurso reconoce que una de las principales problemáticas del cuidar está en lo que denomina síndrome de “sobrecarga del cuidador” (sic2, p. 41), entendiéndolo como algo esperable y normalizado (Ver Figura 1). Fuente: Ministerio de Desarrollo Social y Familia, 2022a, p. 41. Figura 1. Extracto de Guía de responsabilidad en el cuidado La frase presente en el título “es normal sentirse agotado/a” intenta naturalizar una sobrecarga que, además, es definida como aquellos efectos negativos que están circunscritos a dimensiones físicas y psicológicas del bienestar, dando a entender que sancionar este estrés es una práctica que se debe evitar. Si bien se considera la dimensión social, cuando se refiere al aislamiento y la pérdida de vínculos, el discurso invisibiliza otras dimensiones sociales como la pérdida de estatus social, la disminución de ingresos autónomos, daños previsionales futuros, entre otros aspectos generizados. Como segunda descripción factual el discurso coloca la resolución de la sobrecarga de cuidado en manos de la autogestión de la mujer, dejando pendiente el compro - 2 La guía comienza con un capítulo denominado “¿qué es el enfoque de género?” que releva la importancia de la igualdad de género en las actividades de cuidado. Sin embargo, se refieren únicamente al sexo masculino para hacer referencia al síndrome de quien cuida. 9 Corresponsabilidad en el cuidado informal de personas dependientes miso de corresponsabilidad que asume el documento desde su título. En las Figuras 2 y 3 ilustramos un fragmento que describe tareas autogestionadas. Fuente: Ministerio de Desarrollo Social y Familia, 2022b, p. 2. Fuente: Ministerio de Desarrollo Social y Familia, 2022b, p. 2. Figura 2. Extracto de Ficha N.°1 Aprendiendo a cuidarme Figura 3. Extracto de Ficha N.°1 Aprendiendo a cuidarme Las recomendaciones identificadas en el discurso utilizan verbos en mandato como “haga”, “busque” o “evite” que hablan a la cuidadora exclusivamente, buscando que autogestione recursos para (auto)cuidar su salud física y psicológica, sin aludir a redes públicas ni privadas para satisfacer lo que parece un listado de necesidades, más que derechos a proteger. Esta autogestión del autocuidado se construye discursivamente como una decisión autónoma y descontextualizada de categorías transversales al cuidado como la clase, el género, la raza, entre otros. Esto, a nuestro juicio, desdibu ja la imagen de un autocuidado enmarcado en una corresponsabilidad que redistribuya más justamente el cuidar. Ambas descripciones factuales en conjunto, y siguiendo la noción de cuidado de sí (Foucault, 1982/2014), limitan la noción de autocuidado a solo una de sus dimensiones, referida a la “relación consigo mismo”. Esta autogestión del autocuidado sostenida con los propios recursos de las cuidadoras evoca al sujeto neoliberal consumidor-ciudadano-empoderado, construyendo un discurso de responsabilización individual para invertir en sí mismo, administrando activos personales para aumentar la calidad del cuidado prestado (Schild, 2007). Esta construcción de cuidadora como sujeto neoliberal implica que debe gobernarse a sí misma gestionando sus propios riesgos en un marco de constante autorregulación, alejándose de la corresponsabilidad social. Esta cuidadora autorregulada se en- 10 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo marca en una política pública que se focaliza en mujeres que pertenecen a hogares del estrato de calificación socioeconómica más bajo declarado en el Registro Social de Hogares chileno, por lo que la autorregulación se hace más compleja. La familia biparental como responsable principal del cuidado informal Este repertorio interpretativo se presenta con fuerza y autoridad en los documentos. Mediante el uso de los recursos retóricos denominados acreditación de categorías y discurso epocalista, se obtienen dos resultados relevantes referidos al mandato dirigido a la mujer para implicar a la familia en el cuidado y sostener una figura de familia cuidadora de tipo biparental y heteronormada. En primer lugar, el discurso mediante el recurso de acreditación de categorías sitúa a la mujer cuidadora en la condición de gestora ideal de los recursos familiares. A continuación, un fragmento ilustrador (Ver Figura 4). Fuente: Ministerio de Desarrollo Social y Familia, 2022c, p. 2. Figura 4. Extracto de Ficha N°2 ¿Cómo avanzar hacia la corresponsabilidad en el cuidado Este mandato sitúa a la familia como una figura central de la organización del cuidado en reemplazo de la cuidadora individual. Por un lado, se dan indicaciones que se basan en la consigna de que “la familia está para ayudar”, por lo tanto, la mujer cuidadora debe ser competente en la tarea de involucrar a familiares en el cuidado. Solo cuando la familia es insuficiente, el discurso insta a buscar colaboración en la comunidad representada por el sujeto “vecino”, visto como un recurso afuerino e instrumental para ampliar la red de apoyo. A lo largo de los documentos se menciona en algunas ocasiones la relevancia de actores comunitarios en la trama de corresponsabilidad del cuidado, pero su presencia en el discurso es baja comparada con el actor familiar, invi- 11 Corresponsabilidad en el cuidado informal de personas dependientes sibilizando las variadas instancias de co-cuidado que alberga el territorio, y que han sido evidentes en los actuales tiempos de crisis. Por otro lado, la tarea de gestionar una red de apoyo familiar construye una figura de cuidadora que debe encargarse de “enseñar lo que sabe hacer a otras personas”, adjudicándole un rol de entrenadora de competencias del cuidar. Este discurso evoca nuevamente el imaginario del sujeto neoliberal consumidor-ciudadano-empoderado, construyendo un cuidado individualista en donde cada cuidadora autogestione sus recursos familiares, con el fin de proveer un cuidado de la mayor calidad y al menor costo posible. En segundo lugar, el discurso epocalista, como recurso discursivo, intenta transitar desde “la cuidadora” como centro de la organización del cuidado hacia “la familia”, pero con foco exclusivo en la diada “madre y padre”. Así se evidencia en el siguiente fragmento que invita a identificar cómo se distribuye, entre ellos, el trabajo de cuidado (Ver Figura 5). Fuente: Ministerio de Desarrollo Social y Familia, 2022a, p. 34. Figura 5. Extracto de Guía de corresponsabilidad en el cuidado El texto presenta en su párrafo introductorio la frase “vamos a comenzar a redistribuir las cargas”, ubicando un cambio epocal en la figura central de cuidado, pasando de la mujer a la diada parental como únicos actores relevados. El discurso promueve que las tareas de cuidado dejen de estar únicamente centradas en la madre, para ser compartidas con el padre, siendo un primer esfuerzo por abordar la corresponsabilidad familiar y en miras de minimizar el síndrome de sobrecarga de la cuidadora. Pero esta emergente categorización de cuidadores en dupla biparental heteronormada con hijos, 12 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo invisibiliza, por un lado, otras conformaciones de familia y hogar; y, por otro lado, también margina a otras personas que habitualmente emergen como cuidadoras de relevo como hijas/os o abuelas/os. Asimismo, el discurso da por sentado la factibilidad de negociación entre la pareja para redistribuir tareas de cuidado, invisibilizando contextos de desigualdad de género que complejizan la negociación. Si bien la guía en sus primeras páginas deja espacio para el abordaje de igualdad de género, observamos que en el discurso se invisibiliza las posibles desigualdades estructurales de género que transversalizan las dinámicas de arreglo y ajuste familiar, produciendo una inconsistencia factible de corregir. Cabe destacar que el discurso de redistribución del cuidado se circunscribe al espacio familiar y privado, sin hacer alusión a una redistribución en el espacio público no familiar, donde probablemente también se sostienen las complejidades para negociar debido a las desigualdades de género. Ambos recursos discursivos, tanto la acreditación de categorías como el discurso epocalista, considerando el marco de regímenes de bienestar en clave de cuidado (Saraceno y Keck, 2008), marcan una clara organización social de bienestar familiarista por defecto. El rol subsidiario del estado en el cuidado informal En el caso de este repertorio interpretativo analizado mediante los recursos de acreditación de categorías y modalización, lo primero que llama nuestra atención es que la definición de corresponsabilidad del cuidado deja, en un primer momento, al margen otros actores posibles, no mencionándolos. Esto es ilustrado en la siguiente cita (Ver Figura 6). Fuente: Ministerio de Desarrollo Social y Familia, 2022a, p. 29 Figura 6. Extracto de Guía de corresponsabilidad en el cuidado. 13 Corresponsabilidad en el cuidado informal de personas dependientes El actor acreditado para participar en la acción de corresponsabilidad citado en la definición es exclusivamente el hogar de tipo familiar. No se hace alusión a otros acto res como el Estado, el sector privado o la comunidad. Lo segundo que llama nuestra atención, y utilizando el recurso de modelización del discurso, es que esta definición de corresponsabilidad debilita la consideración de variados actores en la trama, pero, en un segundo momento del discurso, si se consigna al Estado en un rol de apoyo subsidiario, solo destinado a familias que no cuenten con propios recursos, propiciando una inconsistencia discursiva. Esto se aprecia en estos fragmentos comparados (Ver Figura 7 y 8). Fuente: Ministerio de Desarrollo Social y Familia, 2022a, p. 29 Fuente: Ministerio de Desarrollo Social y Familia, 2022a, p. 55 Figura 7. Extracto Guía de corresponsabilidad en el cuidado Figura 8. Extracto Guía de corresponsabilidad en el cuidado El primer discurso ilustrado es categórico en señalar que el hombre no sólo debe “ayudar”, ya que esto es insuficiente, sino que debe ser parte activa e igualitaria de la corresponsabilidad de cuidado, involucrando a la figura masculina exclusivamente en su rol de padre de familia, marginando otros roles como empleador o agente de servi cio público. Llama la atención está marcada prescripción de corresponsabilidad dada al padre que no se extiende a otros actores posibles, enquistando el cuidado al mundo privado familiar. Es así como, en el segundo párrafo ilustrado se señala que, en algunas situaciones de mayor vulnerabilidad, la corresponsabilidad del cuidado debería salir del hogar familiar, para acudir, en lógica clientelar estatal (Schild, 2007), a servicios otorgados por el Estado, en los que, por cierto, se invisibilizan otros tejidos de 14 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo apoyos posibles de apalancar como los recursos en contextos comunitarios, ciudadanos y/o privados. Al subrayar este recurso discursivo respecto a la categorización del Estado como parte subsidiaria de la trama de cuidado, no nos referimos a sancionar la lógica de apoyo en el cuidar, sino a poner en cuestión que el texto promociona en el hombre-familiar la corresponsabilidad de cuidar (superando el nivel de apoyo), pero no exige la misma lógica a otros actores de la organización social, como el Estado. Esta modelización vuelve a presentar inconsistencia cuando avanzado en el discurso, se consigna a la práctica de cuidado como un derecho social, lo que aparece in compatible con el Estado subsidiario neoliberal imperante. En términos conceptuales un derecho social promueve condiciones de vida digna en un marco de justicia e igualdad, poniendo al servicio del bien común los recursos dispuestos por el Estado (Serey y Giaconi, 2021), posición opuesta a la lógica de focalización de recursos en población acreditada de riesgo propia del sistema subsidiario. Integrando los resultados obtenidos tanto del recurso de categorización como de modelización, y siguiendo a Schild (2000; 2007), pareciera que, la política pública de cuidado, comandada por un Estado subsidiario, continúa produciendo una ciudadanía de mercado que recibe asistencia social solo en un grupo menor de personas, y apoyándose en los recursos de las mujeres cuidadoras y sus familias, a través de su autoformación y empoderamiento. Conclusiones El artículo se propone aportar a la comprensión acerca de cómo se construye discursivamente la corresponsabilidad del cuidado y autocuidado de la cuidadora informal de personas dependientes en la política pública chilena. Para ello hemos compartido un análisis del tipo discurso pragmático de documentos de política pública, los que, sin duda, son esfuerzos parciales que requieren ser complementados con otras estrategias investigativas. Por ello, este estudio es significado como una primera fase de trabajo que se hace cargo de una parte de la política de cuidados. Esta dificultad es aún mayor cuando se trata de una política relativamente nueva en Chile, que ofrece reciente material para el análisis. Podemos concluir que los resultados obtenidos permiten identificar una construcción de discurso de la política pública chilena sobre cuidado informal de personas dependientes y autocuidado de la cuidadora, que se sostiene en tres dimensiones situadas en lógica de andamio, que van de lo individual a lo colectivo y de lo privado a lo públi - 15 Corresponsabilidad en el cuidado informal de personas dependientes co. Los repertorios interpretativos identificados comienzan con un marcado discurso sobre una mujer cuidadora sobrecargada que autogestiona su autocuidado con escaso apoyo externo, para luego ver emerger una familia biparental que, gestionada y capacitada por la mujer cuidadora, debe transitar hacia el centro de la corresponsabilidad del cuidado, identificándose como la principal figura en esta organización social. Finalmente, propone una figura de Estado subsidiario que apoya a esta emprendedora del cuidado y a su familia cuando no hay recursos autónomos suficientes. Este andamiaje se sostiene en los hombros de las mujeres cuidadoras informales, manteniendo lógicas de invisibilización y desvalorización de su trabajo. Este discurso da cuenta de una política de cuidados que en Chile está caminando desde un lugar con énfasis liberal hacia otro que intenta aproximarse a un énfasis social. Así, se transita desde el reconocimiento de la diada cuidadora/a-persona dependiente como el agente clave de la política hacia el reconocimiento de la familia como un agente ordenador de las prácticas del cuidado informal. Asimismo, se aprecia un tránsito en la misma dirección cuando se deja de subrayar las necesidades de cuidar para adscribirse a un discurso ligado al cuidado como un derecho social. En este senti do, observamos que aún falta consistencia en el discurso para asumir la categoría social en el derecho al cuidado, pero, sin duda, es un significativo avance en las políticas públicas, al menos de tipo discursivo. Nuestros análisis coinciden con la literatura internacional al señalar el lento recorrido que se produce desde la diada del cuidado liberal hacia la corresponsabilidad dada en contextos familiares. Algunas investigaciones indican que la ruta de transformación no debiera finalizar en la figura familiar, más aún si se sustenta en un imaginario de familia conservadora. El desafío es superar los límites familiares y privados para alcanzar espacios colectivos y públicos, en donde la noción de Estado social, como aquel actor capaz de garantizar el derecho al cuidado y asumir la organización de la corresponsabilidad, sigue siendo una pista pertinente, pero aún aspiracional para el caso chileno analizado. Al respecto, cabría considerar la perspectiva denominada “cuidadanía”, la que concibe a ciudadanos y ciudadanas como cuidadores y receptores de cuidados a la vez, al margen de su género, edad o raza (Leiva-Gómez, 2017). Se releva el rol del Estado como el actor garante del cumplimiento de los derechos asociados al cuidar, debido a que se entiende que esta práctica emerge de un sistema de relaciones interdependientes que consignan el principio de situar el “cuidado en el centro”. Es así como, desde la cuidadanía está la posibilidad de que repensemos los derechos ligados al cuidar, al ser cuidado, al autocuidado e incluso al negarse a cuidar, porque entendido como derecho, 16 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo éste debe darse bajo condiciones dignas y autodeterminadas para todos los partícipes de la relación (Pautassi, 2018). Por último, podemos indicar que los tres repertorios interpretativos emergentes del análisis permitirían identificar factores claves a considerar para transformar, con fines de mejora, la política pública de cuidados vigente. Desde esta perspectiva, sería favorable una política que permita acercarse a una organización del cuidado que no se sostenga en un cuidatoriado, sino más bien, en una cuidadanía, lo que necesariamente contempla la consideración de factores estructurales de desigualdad social. Este tránsito no solo requiere políticas públicas ligadas a servicios, transferencias económicas y de tiempo, sino que, siguiendo a Karina Batthyány (2021), también requiere de dos políticas estratégicas adicionales: la de protección de las cuidadoras y la de transformación cultural del cuidado a nivel de sociedad, que incluya más presencia del actor polí tico o comunitario (Reyes y Olivares, 2020). En este potencial transformador la categoría de corresponsabilidad social del cuidado y el autocuidado de la cuidadora informal deben ser consideradas desde el diseño hasta la evaluación de las políticas públicas, para con ello brindar mayores oportunidades para que todas las personas, no solo las mujeres, puedan contar con tiempo, recursos y apoyos suficientes en la práctica de cuidar(nos). En este sentido, la alerta que im pone la crítica feminista a la implementación de políticas públicas, alejadas de espacios de participación y deliberación social, imprime a este desafío una paradoja difícil de abordar cuando se pretende construir una práctica social del cuidado contrahegemónica que se sustenta sobre un modelo de sociedad que sigue siendo neoliberal y patriarcal. El modelo de Estado subsidiario imperante en Chile se constituye en una barrera difícil de superar por lo que, sin duda, los resultados de este estudio también pueden ser leídos prospectivamente como una pequeña, pero constructiva crítica sobre las reales posibilidades que tenemos de asumir un discurso sobre corresponsabilidades e ideales sobre nuestros derechos de cuidado, cuando el discurso de la política pública sigue navegando bajo el inmenso océano de la precariedad y la desigualdad social. Agradecimiento Artículo derivado de la tesis denominada “Prácticas del cuidado de sí en cuidadoras informales de personas dependientes” desarrollada durante el periodo 2022-2023 en el Doctorado en Psicología de la Pontificia Universidad Católica de Valparaíso y Millennium Institute for Care Research (MICARE), Santiago, Chile. 17 Corresponsabilidad en el cuidado informal de personas dependientes Financiamiento Artículo realizado en el marco del Proyecto FONDECYT Regular Nº 1230530 “El enfoque comunitario en la política social de género: explorando la acción pública y la participación de las mujeres en el Chile actual.” Referencias Allen, Amy (2004). Foucault, feminism and the self: The politics of personal transformation. In Dianna Taylor & Karen Vintges (Eds.), Feminism and the final Foucault (pp. 235-257). University of Illinois. Arriagada, Irma (2021). Crisis social y de la organización social de los cuidados en Chile. Estudios Sociales Del Estado, 7(13), 6-41. https://doi.org/10.35305/ese.v7i13.250 Batthyány, Karina (2015). Las políticas y el cuidado en América Latina. Una mirada a las experiencias regionales. CEPAL - Serie Asuntos de Género (124). https://repositorio.cepal.org/handle/11362/37726 Batthyány, Karina (2021). Miradas latinoamericanas a los cuidados. Siglo XXI Editores México. https://www.clacso.org/miradas-latinoamericanas-a-los-cuidados/ Batthyány, Karina; Genta, Natalia & Perrotta, Valentina (2017). El aporte de las familias y las mujeres a los cuidados no remunerados en salud en Uruguay. Revista Estudios Feministas, 25(1), 187-213. https://doi.org/10.1590/18069584.2017v25n1p187 Butler, Judith (2017). Cuerpos aliados y lucha política. Hacia una teoría performativa de la asamblea. Paidós. Cazorla-Becerra, Ketty & Gamboa-Morales, Dahiana (2023). Cartografía participativa de cuidadoras informales: disputa de lugares y tiempos en la práctica del cuidar en Valparaíso, Chile. Prospectiva. Revista de Trabajo Social e intervención social, 36, 1-22. https://doi.org/10.25100/prts.v0i36.12614 Comelín, Andrea & Leiva, Sandra (2018). Necesidades emergentes del cuidado informal de mayores dependientes: develando las voces de los cuidadores familiares. Interciencia, 43(5),358-364. https://www.redalyc.org/articulo.oa? id=33957311010 Creswell, John (2013). Philosophical Assumptions and Interpretative Frameworks. En John Creswell (Ed.), Qualitative Inquiry and Research Design: Choosing Among Five Approaches (pp. 15-41). SAGE Publications. Denzin, Norman & Lincoln, Yvonna (2003). Introduction: The Discipline and Practice of Qualitative Research. En Norman Denzin, & Yvonna Lincoln (Eds.), The Sage handbook of qualitative research (3 ed., pp. 1–32). Sage Publications. Dilts, Andrew (2011). From ‘Entrepreneur of the Self’ to ‘Care of the Self’: Neo-liberal Governmentality and Foucault’s Ethics. Foucault Studies, 12, 1-17. https://ssrn.com/abstract=1580709 18 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo Dixe, María & Querido, Ana (2020). Informal caregiver of dependent person in selfcare: Burden-related factors. Revista Enfermagem, 5(3), 1-8. https://doi.org/10.12707/RV20013 Duarte, Irene; de Aquino, Jessica & Silva, Eliete (2018). Entre o Estado, a sociedade e a família: o care das mulheres cuidadoras. Revista Enfermagem, 71(6), 2882-2890. https:// doi.org/10.1590/0034-7167-2018-0111 Durán, María (2018). La riqueza invisible del cuidado. Universitat de València. Escandell-Rico, Francisco & Pérez-Fernández, Lucía (2022). Impacto en la salud y sus determinantes de las personas mayores cuidadoras. Atención Primaria, 54, 1-3. https://doi.org/10.1016/j.aprim.2022.102464 Ezquerra, Sandra & De Eguía, María (2020). ¿Redistribución de los cuidados? El papel de la familia, el mercado y las políticas públicas en Catalunya. Política y Sociedad, 57(3), 769-795. https://doi.org/10.5209/poso.60900 Flick, Uwe (2004). Introducción a la Investigación Cualitativa. Ediciones Morata S. L. Foucault, Michel (1982/2014). La hermenéutica del sujeto. Fondo de Cultura Económica. Fraser, Nancy (2016). Capitalism’s Crisis of Care. Dissent, 63(4), 30-37. https://doi.org/10.1353/dss.2016.0071 García, Manuela & Urbina, Carolina (2021). Ella es mi favorita: Perspectivas infantiles sobre el buen cuidado en residencias. Revista Latinoamericana de Ciencias Sociales, Niñez y Juventud, 19(3), 1-24. https://doi.org/10.11600/rlcsnj.19.3.4179 Genta, Natalia (2021). Pareja de doble ingreso y cuidado infantil: una mirada a las relaciones de género en los hogares. Cadernos de Pesquisa, 51.(1), 1-18. https://doi.org/10.1590/198053148559 Genta, Natalia; Batthyány, Karina; Perrota, Valentina; Scavino, Sol & Katzkowicz, Sharon (2022). ¿Cuál es el vínculo entre las estrategias de cuidado infantil y la inserción laboral de las cuidadoras? Revista Española de Sociología, 31(1),1-27. https://doi.org/10.22325/fes/res.2022.87 Grandón, Débora (2021). Lo personal es político: un análisis feminista de la experiencia cotidiana de cuidadoras informales de personas adultas en situación de dependencia, en Santiago de Chile. Cadernos Brasileiros de Terapia Ocupacional, 29, 1-14. https://doi.org/10.1590/2526-8910.ctoAO2161 Hernández, Macarena & Rodríguez, Claudio (2022). Política social y experiencias de cuidado en tiempos de pandemia, estallido social y proceso constituyente. Desafíos para la protección social en Chile. En Carolina Tetelboin, Daisy Iturrieta y José Noronha (Eds.), Pandemias, desigualdades y prácticas sociales en salud: miradas para la transformación de América Latina (pp. 135-156) CLACSO. https://doi.org/10.2307/j.ctv2v88fnd.8 Hozven, Raúl & Sisto, Vicente (2021). Repertorios interpretativos en el trabajo social: ¿disciplina, profesión o prestación de servicios? Cuaderno de Trabajo Social, 1(16), 12-34. https://cuadernots.utem.cl/?p=637 Íñiguez, Lupiciño & Antaki, Charles (1994). El análisis del discurso en Psicología. Boletín de Psicología, 44, 57-75. https://eprints.lancs.ac.uk/id/eprint/19036 19 Corresponsabilidad en el cuidado informal de personas dependientes Leiva-Gómez, Sandra (2017). Organización social del cuidado en Bolivia y Chile: Estado y ciudadanía. Revista Austral de Ciencias Sociales, 28, 61-81. https://doi.org/10.4206/rev.austral.cienc.soc.2015.n28-04 McLaren, Margaret (1997). Foucault y el sujeto de Feminismo. Teoría y práctica social, 23(1), 109-128. https://www.jstor.org/stable/23559121 Ministerio de Desarrollo Social y Familia, (2022a, 12 de marzo). Documentos. Guía de Corresponsabilidad en el Cuidado. https://apoyosycuidados.ministeriodesarrollosocial.gob.cl/guia-decorresponsabilidad-en-el-cuidado Ministerio de Desarrollo Social y Familia, (2022b, 12 de marzo). Fichas educativas. Aprendiendo a cuidarme. https://apoyosycuidados.ministeriodesarrollosocial.gob.cl/aprendiendo-acuidarme Ministerio de Desarrollo Social y Familia, (2022c, 12 de marzo). Fichas educativas. ¿Cómo avanzar hacia la corresponsabilidad en el cuidado? https://apoyosycuidados.ministeriodesarrollosocial.gob.cl/corresponsabilidad Ministerio de Salud (2013). Automanejo de enfermedades No transmisibles: una tarea compartida entre usuarios y equipo de salud. Autor. http://www.repositoriodigital.minsal.cl/handle/2015/878 Morales, Karol (2017). Prácticas de organización colectiva en trabajadores municipales a honorarios: Construyendo colectivos en modalidades de vinculación laboral flexible (Tesis de posgrado sin publicar). Pontificia Universidad Católica de Valparaíso. Mosquera, Isabel; Larrañaga, Isabel; Del Río, María; Calderón, Carlos; Machón, Mónica & García, María (2020). Desigualdades de género en los impactos del cuidado informal de mayores dependientes en Gipuzkoa: Estudio CUIDAR-SE. Revista Española de Salud Pública, 93, 1-14. https://www.scielosp. org/article/resp/2019.v93/e201901002/ Muñoz, Nora (2012). Aprendizajes de género y cuidado de sí en la salud masculina: entre lo universal y lo específico. Psicología, Conocimiento y Sociedad, 2(2), 626. https://www.colibri.udelar.edu.uy/jspui/handle/20.500.12008/28902 Myers, Ella (2008). Resisting Foucauldian Ethics: Associative Politics and the Limits of the Care of the Self. Contemporary Political Theory, 7, 125-146. https://link.springer.com/article/10.1057/palgrave.cpt.2007.25 Olivares, Barbara; Winkler, María Inés; Reyes, María Isabel; Berroeta, Héctor & Montero, Maritza (2018). ¿Y si pensamos la comunidad con derechos? Psicología Comunitaria, derechos y políticas públicas: Una relación compleja. Universitas Psychologica, 17(1),1-13. https://doi.org/10.11144/Javeriana.upsy172.pcdp Ortega, Julián (2019). Trabajos de(s)cuidados: reformas neoliberales, feminización y precarización de las condiciones de trabajo en enfermería. Athenea Digital, 19(3), 1-26. https://doi.org/10.5565/rev/athenea.2333 20 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo Pautassi, Laura (2018). El cuidado como derecho. Un camino virtuoso, un desafío inmediato. Revista de la Facultad de Derecho de México, 68(272), 717-742. https:// doi.org/10.22201/fder.24488933e.2018.272-2.67588 Pautassi, Laura (2021). A un año de la pandemia: Los cuidados en el centro y en los márgenes. Instituto Nacional de Ciência e Tecnologia em Políticas Públicas, Estratégias e Desenvolvimento; Desenvolvimento em Debate, 9(1), 213-229. https:// doi.org/10.51861/ded.dmvu.1.019 Pávez, Javiera (2018). Significados y prácticas de la precariedad en trabajadores y trabajadoras de un programa social de infancia en la Región de Valparaíso. (Tesis de posgrado sin publicar). Pontificia Universidad Católica de Valparaíso. Pereira, Magaly & Cárdenas, María (2019). Visión filosófica del cuidado humano en la mujer en la etapa del climaterio. Enfermería: Cuidados Humanizados, 8(1), 3445. https://doi.org/10.22235/ech.v8i1.1795 Pereira, Nildete; Chaves, Larissa; Pereira, Nadirlene; de Oliveira, Elaine; Araújo, Luciana & Andrade, Alice (2019). Consecuencias del cuidado para la salud de las ancianas cuidadoras de familiares dependientes. Revista da Escola de Enfermagem da USP, 53, 1-9. https://doi.org/10.1590/S1980220X2018002303446 Pérez, Amaia (2014). Subversión feminista de la economía. Aportes para un debate sobre el conflicto capital-vida. Traficantes de Sueños. https://doi.org/10.1387/pceic.15712 Potter, Jonathan (1998). La representación de la realidad: Discurso, retórica y construcción social. Paidós. Potter, Jonathan & Wetherell, Margaret (1987). Discourse and social psychology: Beyond attitudes and behaviour. Sage. Prior, Lindsay (2008). Repositioning documents in social research. Sociology, 42(5), 821836. https://doi.org/10.1177/0038038508094564 Quezada, Martha & Bascones, Luis (2020). El cuidado de las personas vulnerables a lo largo de la historia. Sobre Ruedas, 105, 6-10. https://siidon.guttmann.com/files/sr106_quezada_bascones_cuidado.pdf Rahimi, Tahereh; Dastyar, Neda & Rafati, Foozieh (2021). Experiences of family caregivers of patients with COVID-19. BMC Family Practice, 22(137), 2-10. https://doi.org/10.1186/s12875-021-01489-7 Ramacciotti, Karina (2020). Cuidar en tiempos de pandemia. Descentrada - Revista interdisciplinaria de feminismos y género, 4(2), 1-8. https://doi.org/10.24215/25457284e126 Razavi, Shahra (2007). The Political and Social Economy of Care in a Development Context Conceptual Issues, Research Questions and Policy Options. UN Women. https://socialprotection-humanrights.org/wp-content/uploads/2018/04/ Razavi-paper.pdf Reyes-Espejo, María Isabel; Cazorla-Becerra, Ketty; González-Seguel, Antonia; Bandim-Pedroza, Tacianna, & Cárcamo-Morales, Javiera (2022). Efectos performativos de las Políticas Sociales de Género en el campo comunitario. 21 Corresponsabilidad en el cuidado informal de personas dependientes Revista Liminales. Escritos Sobre Psicología Y Sociedad, 11(22), 57-88. https://doi.org/10.54255/lim.vol11.num22.688 Reyes, María Isabel & Olivares, Barbara (2020). La política de las políticas: La posición de los psicólogos/as comunitarios/as frente al quehacer gubernamental. En Jaime Alfaro, Bárbara Olivares, María Monreal, Paris Gamonal & Francisco Jeannerett (Eds.), Diálogos Contemporáneos en Psicología Comunitaria (pp. 146–170). Astrolabio Ediciones https://www.researchgate.net/publication/348233100_La_Politica_de_las_Poli ticas_La_Posicion_de_losas_Psicologosas_Comunitariosas_frente_al_Quehac er_Gubernamental Rico, Nieves (2011). “Crisis del Cuidado y Políticas Públicas.” Las familias latinoamericanas interrogadas: hacia la articulación del diagnóstico, la legislación y las políticas. CEPAL. http://hdl.handle.net/11362/6982 Saraceno, Chiara & Keck, Wolfgang (2008). The institutional framework of intergenerational family obligations in Europe: A conceptual and methodological overview. Wissenschaftszentrum Berlin für Sozialforschung. http://www.multilinks-project.eu/info/papers Schild, Verónica (2000). Neo-liberalism’s New Gendered Market Citizens: The ‘Civilizing’ Dimension of Social Programmes in Chile. Citizenship Studies, 4(3), 275-305. https://doi.org/10.1080/713658800 Schild, Verónica (2007). Empowering Consumer Citizens or Governing Poor Female Subjects? The Institutionalization of “Self-Development” in the Chilean Social Policy Field. Journal of Consumer Culture, 7(2), 179-203. https://doi.org/10.1177/1469540507077672 Serey, Mariela & Giaconi, Carolina (2021). ¿Por qué debe incorporarse el derecho al cuidado en la futura Constitución? Una reflexión desde el cuidado no remunerado de personas en situación de dependencia. Revista Intervención, 10(1), 111-118. https://doi.org/10.53689/int.v10i1.94 Sisto, Vicente; Ramírez-Casas del Valle, Lorena; Núñez, Lorena, & López, Andrea (2021). La ética de lo público y la impertinencia del managerialismo como modelo de organización del trabajo en tiempos de crisis. Psicoperspectivas, 20(3), 1-12. https://doi.org/10.5027/psicoperspectivas-vol20-issue3-fulltext2443 Standing, Guy (2011). The Precariat. The New Dangerous Class. Bloomsbury Academic. Tronto, Joan (2017). There is an alternative: hominess curans and the limits of neoliberalism. International Journal of Care and Caring, 1(1), 27-43. https://doi.org/10.1332/239788217X14866281687583 Vellone, Ercole; Barbaranelli, Claudio; Pucciarelli, Gianluca; Zeffiro, Valentina; Alvaro, Rosaría & Riegel, Barbara (2020). Validity and reliability of the caregiver contribution to self-care of heart failure index version 2. Journal of Cardiovascular Nursing, 35(3), 280-290. https://doi.org/10.1097/JCN.0000000000000655 22 Ketty Cazorla-Becerra; María Isabel Reyes-Espejo Wetherell, Margaret & Potter, Jonathan (1996). El análisis del discurso y la identificación de los repertorios interpretativos. En Ángel Gordo & José Lizana (Comps.), Psicología, discursos y poder (pp. 63-78). Visor. Este texto está protegido por una licencia Creative Commons 4.0. Usted es libre para Compartir —copiar y redistribuir el material en cualquier medio o formato— y Adaptar el documento —remezclar, transformar y crear a partir del material— para cualquier propósito, incluso comercialmente, siempre que cumpla la condición de: Atribución: Usted debe reconocer el crédito de una obra de manera adecuada, proporcionar un enlace a la licencia, e indicar si se han realizado cambios . Puede hacerlo en cualquier forma razonable, pero no de forma tal que sugiera que tiene el apoyo del licenciante o lo recibe por el uso que hace. Resumen de licencia - Texto completo de la licencia 23
https://openalex.org/W3099697876	https://www.frontiersin.org/articles/10.3389/fnhum.2020.565664/pdf	English	null	The Attentive Cursor Dataset	Frontiers in human neuroscience	2,020	cc-by	6,018	1. INTRODUCTION We introduce a large-scale dataset of mouse cursor movements that can be used to predict user attention, infer demographics information, and analyze ﬁne-grained movements. Attention is a ﬁnite resource, so people spend their time on things they ﬁnd valuable, especially when browsing online. Objective measurements of attentional processes are increasingly sought after by researchers, advertisers, and other key stakeholders from both academia and industry. With every click, digital footprints are created and logged, providing a detailed record of a person’s online activity. However, click data provide an incomplete picture of user interaction, as they inform mainly about a users’ end choice. A user click is often preceded by several valuable interactions, such as scrolling, hovers, aimed movements, etc. and thus having access to this kind of data can lead to an overall better understanding of the user’s cognitive processes. For example, previous work has evidenced that when the mouse cursor is motionless, the user is processing information (Hauger et al., 2011; Huang et al., 2011; Diriye et al., 2012; Boi et al., 2016), i.e., essentially “users ﬁrst focus and then execute actions” (Martín-Albo et al., 2016). We have collected mouse cursor tracking logs from near 3K subjects performing a transactional search task that together account for roughly 2 h worth of interaction data. Our dataset has associated attention labels and ﬁve demographics attributes that may help researchers to conduct several analysis, like the ones we discuss later in this section. 1 Department of Communications and Networking, Aalto University, Espoo, Finland, 2 Telefónica Research, Madrid, Spain Keywords: aimed movements, attention, demographics, web search, mouse cursor Edited by: Claudio De Stefano, University of Cassino, Italy University of Cassino, Italy Reviewed by: Hugo Gamboa, New University of Lisbon, Portugal Nicole Cilia, University of Cassino, Italy Correspondence: Luis A. Leiva ﬁrstname.lastname@aalto.ﬁ Reviewed by: Hugo Gamboa, New University of Lisbon, Portugal Nicole Cilia, Uni e sit of Cassino Ital Research in mouse cursor tracking has a long track record. Chen et al. (2001) were among the ﬁrst ones to note a relationship between gaze position and cursor position during web browsing. Mueller and Lockerd (2001) investigated the use of mouse tracking to create compelling visualizations and model the users’ interests. It has been argued that mouse movements can reveal subtle patterns like reading (Hauger et al., 2011) or hesitation (Martín-Albo et al., 2016), and can help the user regain context after an interruption (Leiva, 2011a). Others have also noted the utility of mouse cursor analysis as a low-cost and scalable proxy of eye tracking (Huang et al., 2012; Navalpakkam et al., 2013). Several works have investigated closely the utility of mouse cursor data in web search (Arapakis et al., 2015; Lagun and Agichtein, 2015; Liu et al., 2015; Arapakis and Leiva, 2016; Chen et al., 2017) and web page usability evaluation (Arroyo et al., 2006; Atterer et al., 2006; Leiva, 2011b), two of the most prominent use cases of this technology. Mouse biometrics is another active research area that has shown promise in controlled settings (Lu et al., 2017; Krátky and Chudá, 2018). Researchers have started to analyze mouse movements on websites for the detection of neurodegenerative disorders (White et al., 2018; Gajos et al., 2020). In practice, commercial web search engines often use mouse cursor tracking to improve search results (Huang et al., 2011, 2012), optimize page design (Leiva, 2012; Diaz et al., 2013), and oﬀer better recommendations to their users (Speicher et al., 2013). In what follows, we provide a brief survey of what others have accomplished by analyzing mouse cursor movements in web search tasks. These analyses highlight potential use cases of our dataset, thereby allowing researchers to investigate similar environments and behaviors. Correspondence: Luis A. Leiva ﬁrstname.lastname@aalto.ﬁ Specialty section: This article was submitted to Cognitive Neuroscience, a section of the journal Frontiers in Human Neuroscience Received: 25 May 2020 Accepted: 19 October 2020 Published: 16 November 2020 The Attentive Cursor Dataset Luis A. Leiva 1* and Ioannis Arapakis 2 DATA REPORT published: 16 November 2020 doi: 10.3389/fnhum.2020.565664 Luis A. Leiva 1* and Ioannis Arapakis 2 1 Department of Communications and Networking, Aalto University, Espoo, Finland, 2 Telefónica Research, Madrid, Spain Citation: Leiva LA and Arapakis I (2020) The Attentive Cursor Dataset. Front. Hum. Neurosci. 14:565664. doi: 10.3389/fnhum.2020.565664 November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org The Attentive Cursor Dataset Leiva and Arapakis 1https://www.ﬁgure-eight.com 1.2. Inferring Visual Attention 1.2. Inferring Visual Attention Mouse cursor tracking has been used to survey the visual focus of the user, thus revealing valuable information regarding the distribution of user attention over the various SERP components. Despite the technical challenges that may arise from this analysis, previous work has shown the utility of mouse movement patterns to measure within-content engagement (Arapakis et al., 2014a; Carlton et al., 2019) and predict reading experiences (Hauger et al., 2011; Arapakis et al., 2014b). Lagun et al. (2014a) introduced the concept of motifs, or frequent cursor subsequences, in the estimation of search result relevance. Similarly, Liu et al. (2015) applied the motifs concept to SERPs and predicted search result utility, searcher eﬀort, and satisfaction at the search task level. Boi et al. (2016) proposed a method for predicting whether the user is actually looking at the content pointed by the cursor, exploiting the mouse cursor data and a segmentation of the web page contents. Lastly, Arapakis and Leiva (2016) investigated user engagement with direct displays on SERPs and provided further evidence that supports the utility of mouse cursor data for measuring user attention at a display-level granularity (Arapakis and Leiva, 2020; Arapakis et al., 2020). Others have also examined the extent to which mouse cursor movements can help identify gender and age (Yamauchi and Bowman, 2014; Kratky and Chuda, 2016; Pentel, 2017), however the experimental settings have limited generalizability, either because the tasks are not well-connected to typical activities that users perform online, such as web search, because the data include multiple samples per participant, thereby increasing the risks of information leakage, or because researchers could not verify their ground-truth data. In our dataset, we limit the training samples to exactly one mouse cursor trajectory per participant, who are veriﬁed, high-quality crowdworkers. 1.3. Inferring Emotion We ran an online crowdsourcing study that reproduced the conditions of a transactional search task. Participants were presented with a simulated information need that explained that they were interested in purchasing some product for them or a friend. Overall, the study consisted of three parts, to be described later: (1) pre-task guidelines, (2) the web search task, and (3) a post-task questionnaire. The connection between mouse cursor movements and the underlying psychological states has been a topic of research since the early 90s (Card et al., 1987; Accot and Zhai, 1997). Some studies have investigated the utility of mouse cursor data for predicting the user’s emotional state. For example, Zimmermann et al. (2003) investigated the eﬀect of induced aﬀective states on the motor-behavior of online shoppers and found that the total duration of mouse cursor movements and the number of velocity changes were associated to the experienced arousal. Kaklauskas et al. (2009) created a system that extracts physiological and motor-control parameters from mouse cursor interactions and then triangulated those with psychological data taken from self-reports, to correlate the users’ emotional state and productivity. In a similar line, Azcarraga and Suarez (2012) combined electroencephalography signals and mouse cursor interactions to predict self-reported emotions like frustration, 1.1. Inferring Interest interest, conﬁdence and excitement. Yamauchi (2013) studied the relationship between mouse cursor trajectories and generalized anxiety in human subjects. Lastly, Kapoor et al. (2007) predicted whether a user experiences frustration, using an array of aﬀective- aware sensors. g For a long time, commercial search engines have been interested in how users interact with Search Engine Result Pages (SERPs), to anticipate better placement and allocation of ads in sponsored search or to optimize the content layout. Early work considered simple, coarse-grained features derived from mouse cursor data to be surrogate measurements of user interest (Goecks and Shavlik, 2000; Claypool et al., 2001; Shapira et al., 2006). Follow- up research transitioned to more ﬁne-grained mouse cursor features (Guo and Agichtein, 2008, 2010) that were shown to be more eﬀective. These approaches have been directed at predicting open-ended tasks like search success (Guo et al., 2012) or search satisfaction (Liu et al., 2015). Mouse cursor position is mostly aligned to eye gaze, especially on SERPs (Guo and Agichtein, 2012; Lagun et al., 2014a), and that can be used as a good proxy for predicting good and bad abandonment (Diriye et al., 2012; Brückner et al., 2020). 1.4. Inferring Demographics g g p Prior work has linked age with motor control and pointing performance in tasks that involve the use of a computer mouse (Walker et al., 1997; Bohan and Chaparro, 1998; Hsu et al., 1999; Smith et al., 1999; Jastrzembski et al., 2003; Lindberg et al., 2006). Overall, aging is marked by a decline in motor control abilities, therefore it is expected to aﬀect the users’ pointing performance and, by extension, how they move the computer mouse. For example, Smith et al. (1999) observed that older people incurred in longer mouse movement times, more sub- movements, and more pointing errors than the young. These ﬁndings underline potential age eﬀects on the way a mouse device is used in an online search task. Prior research has also noted sensory-motor diﬀerences due to gender (Landauer, 1981; Chen and Chen, 2008; Yamauchi et al., 2015), such as signiﬁcant variation in the cursor movement distance, pointing time, and cursor patterns. The cause of these variations has been attributed to gender-based diﬀerences in how users move a mouse cursor or to diﬀerent cognitive mechanisms (perceptual and spatial processes) involved in motor control. Frontiers in Human Neuroscience \| www.frontiersin.org 2.3. Pre-task Guidelines Participants were instructed to read carefully the terms and conditions of the study which, among other things, informed them that they should perform the task from a desktop or laptop computer using a computer mouse (and refrain from using a touchpad, tablet, or mobile device) and that their browsing activity would be logged. Moreover, participants consented to share their browsing data and their (anonymized) responses for later analysis. Using this ﬁnal selection of search queries, we produced the static version of the corresponding Google SERPs and injected custom JavaScript code that allowed us to capture all client-side user interactions. For this, we used EVTRACK3, an open source JavaScript event tracking library derived from the smt2ǫ mouse tracking system (Leiva and Vivó, 2013). EVTRACK can capture browser events either via event listeners (the event is captured as soon as it is ﬁred) or via event polling (the event is captured at ﬁxed-time intervals). We captured mousemove events via event polling, every 150 ms to avoid unnecessary data overhead (Leiva and Huang, 2015), and all the other browser events (e.g., load, click, scroll) via event listeners. Whenever an event was recorded, we logged the following information: mouse cursor position (x and y coordinates), timestamp, event name, XPath of the DOM element that relates to the event, and the DOM element attributes (if any). Participants were asked to act naturally and choose anything that would best answer a given search query, since all “clickable” elements (e.g., result links, images, etc.) on the SERP were considered valid answers. The instructions were followed by a brief search task description using this template: “You want to buy <noun> (for you or someone else as a gift) and you have submitted the search query <noun> to Google Search. Please browse the search results page and click on the element that you would normally select under this scenario.” The template was populated with the corresponding <noun> entities, based on the assigned query. Participants were allowed as much time as they needed to examine the SERP and proceed with the search task, which would conclude whenever they clicked on any SERP element. The payment for the participation was $0.20. Participants could also opt out at any moment, in which case they were not compensated. Each participant could take the study only once. 2https://trends.google.com/trends/ 3https://github.com/luileito/evtrack 2.1. Participants We recruited participants from the FIGURE EIGHT crowdsourcing platform1. They were of mixed nationality (e.g., American, Belgian, British, German) and had diverse educational backgrounds (see Table 1). All participants were proﬁcient in English and were experienced (Level 3) contributors, i.e., they had a proven track record of successfully completed November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org Frontiers in Human Neuroscience \| www.frontiersin.org 2 The Attentive Cursor Dataset Leiva and Arapakis TABLE 1 \| Demographics information from our dataset. Age Count Gender Count Nationality Count Education Count Income Count 18–23 380 Male 1,605 USA 1,755 High school 593 <25K 881 24–29 716 Female 1,118 VEN 251 College 472 25–34K 446 30–35 590 NA 14 GBR 209 Bachelor’s 704 35–49K 367 36–41 417 CAN 66 Graduate 499 50–74K 394 42–47 223 EGY 37 Master’s 399 75–99K 249 48–53 174 UKR 31 Doctorate 30 100–149K 145 54–59 132 IND 29 NA 40 150–249K 42 60–65 63 SRB 27 >250K 23 +66 24 RUS 25 NA 190 NA 18 ... TABLE 1 \| Demographics information from our dataset. tasks and of a diﬀerent variety, thus being considered very reliable contributors. code that removed all ads excepting one that would be selected for a given participant. In any case, native bottom-most ads were not shown to the participants. 2.2. Materials Starting from Google Trends2, we selected a subset of the Top Categories and Shopping Categories that were suitable representatives of transactional tasks. Then, we extracted the top search queries issued in the US during the last 12 months. Next, we narrowed down our search query collection to 150 representative popular queries. The ﬁnal collection of transactional queries was repeated as many times needed to produce the desired number of search sessions for the ﬁnal dataset. 2.3. Pre-task Guidelines All queries triggered some form of advertisements on the SERPs, according to three diﬀerent formats: “native” (organic ads) or “bundled” (direct display ads). All SERPs included one or more native ads together with one bundled ad. The native advertisements could appear either at the top or bottom position of the SERP, whereas the bundled ads could appear either at the top-left or top-right position. We ensured that only one ad was visible per condition and participant at a time. This was possible by instrumenting each downloaded SERP with custom JavaScript 2.5. Post-task Questionnaire Starting from a set of 3,223 participants who initially accessed the study, we ﬁltered automatically those who did not ﬁnish it (138 cases) as well as participants who did not move their mouse at all (176 cases). We concluded to a dataset with 2,909 observations comprising at least one mouse movement, together with their associated browser’s and user’s metadata. See Table 1 for a summary of the available demographics information. Upon concluding the search task, participants were asked to answer a series of questions. The questions were forced-choice type and allowed multi-point response options. The ﬁrst question asked the degree to which the user noticed the advertisements shown on the SERP: While performing the search task, to what extent did you pay attention to the advertisement? We used a 5-point Likert-type scale to collect the labels: 1 (“Not at all”), 2 (“Not much”), 3 (“I can’t decide”), 4 (“Somewhat”), and 5 (“Very much”). In practice, these scores should be collapsed to binary labels (true/false), but we felt it was necessary to use a 5-point Likert-type scale for several reasons. First, using 2-point scales often results in highly skewed data (Johnson et al., 1982). Second, it is important to leave room for neutral responses, because some users may not want to say one way or another, otherwise this can produce response biases. But 3-point scales can lead more users to stay neutral, because the remaining options can be seen as “too extreme.” Therefore, we opted for a 5-point scale, which leaves more room for “soft responses” and in addition is easy to understand. With this scoring scheme, therefore, we are conﬁdent that eventual binary labels would actually reﬂect positive and negative user votes. There are 92 unique combinations of query and ad style, each of which assessed by 32 users on average (SD = 17 users). There are 1,942 observations from the attended condition (self-reported Likert-type score ≥4), 776 observations from the non-attended condition (score ≤2), and 191 observations from the neutral condition (score of 3). The average mouse cursor trajectory has 15.78 coordinates (SD = 16.5, min = 1, max = 222), which is around the same order of magnitude as reported in similar studies (Huang et al., 2011; Leiva and Huang, 2015; Arapakis and Leiva, 2016). Excepting the automatic ﬁltering procedure explained above, our data is in raw form and therefore some columns require further processing. 3.1. Data Format 5. What is your current income? [25K, 35K,..., 250K, Prefer not to say] 5. What is your current income? [25K, 35K,..., 250K, Prefer not to say] The Attentive Cursor dataset includes the following resources: The Attentive Cursor dataset includes the following resources: 1. A folder with mouse tracking log ﬁles, as recorded by the EVTRACK software: 1. A folder with mouse tracking log ﬁles, as recorded by the EVTRACK software: 2.5. Post-task Questionnaire For example, most columns pertaining demographics information are stored as integers, therefore researchers should consult Table 1 to retrieve the corresponding categorical labels. We also recommend researchers to apply other ﬁltering methods, depending on the nature of their experiments, such as collapsing the ground-truth attention labels from the original 1–5 scale to a binary scale (Arapakis and Leiva, 2020; Arapakis et al., 2020) or ignoring cursor trajectories having <5 coordinates, which in most cases would correspond to 1 s of interaction data. The questionnaire also comprised the following demographics-related questions: 1. What is your gender? [Male, Female, Prefer not to say] 2. What is your age group? [18–23, 24–29,..., 60–65, +66, Prefer not to say] 3. What is your native language? [Pull-down list, Prefer not to say] 4. What is your education level? [High school, College,..., Doctorate, Prefer not to say] 4. What is your education level? [High school, College,..., Doctorate, Prefer not to say] 2.4. Task Procedure Post-task Questionnaire Upon concluding the search task, participants were asked to answer a series of questions. The questions were forced-choice type and allowed multi-point response options. Frontiers in Human Neuroscience \| www.frontiersin.org 2.4. Task Procedure Each participant was presented with a search task description, then provided with a predeﬁned search query (selected at random from our pool of queries) and the corresponding SERP, and they were asked to click on any element of the page that best solved the task. This way, we ensured that participants interacted with the same pool of web search queries and avoided any unaccounted systematic bias due to query quality variation. All possible combinations of query and ad style (i.e., format and position) November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org 3 The Attentive Cursor Dataset Leiva and Arapakis were pre-computed so that whenever a new user accessed the study, they were assigned one of these combinations at random. were pre-computed so that whenever a new user accessed the study, they were assigned one of these combinations at random. experiments. On the downside, experimenters have to account for potential threats to ecological validity, distractions in the physical environment of the participant, and privacy issues, to name a few. Still, crowdsourcing allows for exploring a wider range of parameters in a more controlled manner as compared to in-the-wild large-scale studies. Participants accessed the instrumented SERPs through a dedicated web server that did not alter the look and feel of the original SERPs. This allowed us to capture ﬁne-grained user interactions while ensuring that the content of the SERPs remained consistent with the original version. Each participant was allowed to perform the search task only once to avoid introducing possible carry over eﬀects and, thus, altering their browsing behavior in subsequent search tasks. In sum, each participant was exposed only to a single condition; i.e., a unique combination of query and ad style. Finally, at the end of the study participants had to copy a unique code and paste it on FIGURE EIGHT in order to have their job validated. We collected self-reported ground-truth labels in a similar vein to previous work (Feild et al., 2010; Lagun et al., 2014b; Liu et al., 2015; Arapakis and Leiva, 2016) which also administered post-task questionnaires. To mitigate and discount low-quality responses, several preventive measures were put into practice, such as introducing test (gold-standard) questions to our tasks, selecting experienced contributors with high accuracy rates, and monitoring their task completion time, thus ensuring the internal validity of our experiment. 2.5. AUTHOR CONTRIBUTIONS All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. ETHICS STATEMENT Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study. 3. VALIDATION AND FILTERING a. Browser events: space-delimited ﬁles (CSV) with information about each event type (8 columns). Crowdsourcing studies oﬀer several advantages over in-situ methods of experimentation (Mason and Suri, 2012), such as access at a larger and more diverse pool of participants with stable availability, collection of real usage data at a relatively large scale, and a low-cost alternative to the more expensive laboratory-based y b. Browser metadata: XML ﬁles with information about the user’s browser (e.g., viewport size). 2. A TSV ﬁle with ground-truth labels (4 columns). November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org 4 The Attentive Cursor Dataset Leiva and Arapakis FIGURE 1 \| File content samples (top) and SERP snapshots with mouse cursor trajectories (bottom). An ellipsis (...) denotes an intentional omission of some data, for brevity’s sake. The gray-colored rectangles in the bottommost ﬁgures denote the different ad types, from left to right: right-aligned bundled ad, left-aligned bundled ad, and native ad. FIGURE 1 \| File content samples (top) and SERP snapshots with mouse cursor trajectories (bottom). An ellipsis (...) denotes an intentional omission of some data, for brevity’s sake. The gray-colored rectangles in the bottommost ﬁgures denote the different ad types, from left to right: right-aligned bundled ad, left-aligned bundled ad, and native ad. FIGURE 1 \| File content samples (top) and SERP snapshots with mouse cursor trajectories (bottom). An ellipsis (...) denotes an intentional omission of some data, for brevity’s sake. The gray-colored rectangles in the bottommost ﬁgures denote the different ad types, from left to right: right-aligned bundled ad, left-aligned bundled d d ti d 4. CONCLUSION We have presented a large-scale, in-the-wild dataset of mouse cursor movements in web search, with associated ground-truth labels about user’s attention and demographics attributes. The dataset represents real-world behavior of individuals completing a transactional web search task. What makes this dataset both unique and challenging is the fact that there is only one observation per user. It is not possible to leak information from any data splits; e.g., training, validation, and testing splits typically used in machine learning studies. It is our hope that the dataset will foster research in several scientiﬁc domains, Including, e.g., information retrieval, movement science, and psychology. DATA AVAILABILITY STATEMENT 3. A tab-delimited ﬁle (TSV) with user’s demographics and stimulus condition (12 columns). 3. A tab-delimited ﬁle (TSV) with user’s demographics and stimulus condition (12 columns). The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. 4. A folder with all SERPs in HTML format. 4. A folder with all SERPs in HTML format. 5. A README ﬁle with a detailed explanation of each resource. Figure 1 provides some examples of the kind of data that researchers can ﬁnd in our dataset. We provide the URL to the repository in the “Data Availability Statement” section below. Figure 1 provides some examples of the kind of data that researchers can ﬁnd in our dataset. We provide the URL to the repository in the “Data Availability Statement” section below. Frontiers in Human Neuroscience \| www.frontiersin.org REFERENCES Gajos, K. Z., Reinecke, K., Donovan, M., Stephen, C. D., Hung, A. Y., Schmahmann, J. D., et al. (2020). Computer mouse use captures ataxia and parkinsonism, enabling accurate measurement and detection. Mov. Disord. 35, 354–358. doi: 10.1002/mds.27915 Accot, J., and Zhai, S. (1997). “Beyond ﬁtts’ law: Models for trajectory- based HCI tasks,” in Proceedings of CHI (Atlanta, GA), 295–302. doi: 10.1145/258549.258760 Goecks, J., and Shavlik, J. (2000). “Learning users’ interests by unobtrusively observing their normal behavior,” in Proceedings of IUI, 129–132. doi: 10.1145/325737.325806 Arapakis, I., Lalmas, M., Cambazoglu, B. B., Marcos, M. C., and Jose, J. M. (2014a). User engagement in online news: Under the scope of sentiment, interest, aﬀect, and gaze. J. Assoc. Inf. Sci. Technol. 65, 1988–2005. doi: 10.1002/asi.23096 Guo, Q., and Agichtein, E. (2008). “Exploring mouse movements for inferring query intent,” in Proceedings of SIGIR (Singapore), 707–708. doi: 10.1145/1390334.1390462 Arapakis, I., Lalmas,M., and Valkanas, G. (2014b). “Understanding within-content engagement through pattern analysis of mouse gestures,” in Proceedings of CIKM, 1439–1448. doi: 10.1145/2661829.2661909 Guo, Q., and Agichtein, E. (2010). “Ready to buy or just browsing? Detecting web searcher goals from interaction data,” in Proceedings of SIGIR, 130–137. doi: 10.1145/1835449.1835473 Arapakis, I., and Leiva, L. A. (2016). “Predicting user engagement with direct displays using mouse cursor information,” in Proceedings of SIGIR (Pisa), 599–608. doi: 10.1145/2911451.2911505 Guo, Q., and Agichtein, E. (2012). “Beyond dwell time: estimating document relevance from cursor movements and other post-click searcher behavior,” in Proceedings ofWWW, 569–578. doi: 10.1145/2187836.2187914 Arapakis, I., and Leiva, L. A. (2020). “Learning eﬃcient representations of mouse movements to predict user attention,” in Proceedings of SIGIR. doi: 10.1145/3397271.3401031 Guo, Q., Lagun, D., and Agichtein, E. (2012). “Predicting web search success with ﬁne-grained interaction data,” in Proceedings of CIKM, 2050–2054. doi: 10.1145/2396761.2398570 Arapakis, I., Leiva, L. A., and Cambazoglu, B. B. (2015). “Know your onions: understanding the user experience with the knowledge module in web search,” in Proceedings of CIKM (Melbourne, VIC), 1695–1698. doi: 10.1145/2806416.2806591 Hauger, D., Paramythis, A., and Weibelzahl, S. (2011). “Using browser interaction data to determine page reading behavior,” in Proceedings of UMAP (Berlin; Heidelberg: Springer), 147–158. doi: 10.1007/978-3-642-22362-4_13 Arapakis, I., Penta, A., Joho, H., and Leiva, L. A. (2020). A price-per-attention auction scheme using mouse cursor information. ACM Trans. Inf. Syst. 38:13. doi: 10.1145/3374210 Hsu, S. H., Huang, C. C., Tsuang, Y. H., and Sun, J. S. (1999). REFERENCES Eﬀects of age and gender on remote pointing performance and their design implications. Int. J. Ind. Ergon. 23, 461–471. doi: 10.1016/S0169-8141(98) 00013-4 Arroyo, E., Selker, T., and Wei, W. (2006). “Usability tool for analysis of web designs using mouse tracks,” in Proceedings of CHIEA (Montréal, QC), 357–362. doi: 10.1145/1125451.1125557 Huang, J., White, R., and Buscher, G. (2012). “User see, user point: gaze and cursor alignment in web search,” in Proceedings of CHI, 1341–1350. doi: 10.1145/2207676.2208591 Atterer, R., Wnuk, M., and Schmidt, A. (2006). “Knowing the user’s every move: user activity tracking for website usability evaluation and implicit interaction,” in Proceedings of WWW (Edinburgh), 203–212. doi: 10.1145/1135777.1135811 Huang, J., White, R. W., and Dumais, S. (2011). “No clicks, no problem: using cursor movements to understand and improve search,” in Proceedings of CHI, 1225–1234. doi: 10.1145/1978942.1979125 Azcarraga, J., and Suarez, M. T. (2012). “Predicting academic emotions based on brainwaves, mouse behaviour and personality proﬁle,” in Proceedings of PRICAI (Berlin; Heidelberg: Springer), 728–733. doi: 10.1007/978-3-642-32695-0_64 Jastrzembski, T., Charness, N., Holley, P., and Feddon, J. (2003). Input devices for web browsing: age and hand eﬀects. Universal Access Inf. 4, 39–45. doi: 10.1007/s10209-003-0083-5 Bohan, M., and Chaparro, A. (1998). Age-related diﬀerences in performance using a mouse and trackball. Hum. Factors 42, 152–155. doi: 10.1177/154193129804200202 Johnson, S., Smith, P., and Tucker, S. (1982). Response format of the job descriptive index: assessment of reliability and validity by the multitrait-multimethod matrix. J. Appl. Psychol. 67, 500–505. doi: 10.1037/0021-9010.67.4.500 Boi, P., Fenu, G., Spano, L. D., and Vargiu, V. (2016). Reconstructing user’s attention on the web through mouse movements and perception-based content identiﬁcation. ACM Trans. Appl. Percept. 13, 15:1–15:21. doi: 10.1145/2912124 Kaklauskas, A., Krutinis, M., and Seniut, M. (2009). “Biometric mouse intelligent system for student’s emotional and examination process analysis,” in Proceedings of ICALT (Riga), 189–193. doi: 10.1109/ICALT.2009.130 Brückner, L., Arapakis, I., and Leiva, L. A. (2020). “Query abandonment prediction with deep learning models of mouse cursor movements,” in Proceedings of CIKM. doi: 10.1145/3340531.3412126 Kapoor, A., Burleson, W., and Picard, R. W. (2007). Automatic prediction of frustration. Int. J. Hum. Comput. Stud. 65, 724–736. doi: 10.1016/j.ijhcs.2007.02.003 Card, S. K., English, W. K., and Burr, B. J. (1987). “Evaluation of mouse, rate controlled isometric joystick, step keys, and text keys, for text selection on a CRT,” in Human-Computer Interaction, eds R. M. Baecker and W. A. S. Buxton (Taylor & Francis), 386–392. Kratky, P., and Chuda, D. (2016). ACKNOWLEDGMENTS Two manuscripts using a post-processed version of this dataset have been recently published by the authors (Arapakis and Leiva, 2020; Arapakis et al., 2020). LAL acknowledged the support from the Finnish Center for Artiﬁcial Intelligence (FCAI). November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org 5 The Attentive Cursor Dataset Leiva and Arapakis REFERENCES “Estimating gender and age of web page visitors from the way they use their mouse,” in Proceedings of WWW Companion, 61-62. doi: 10.1145/2872518.2889384 Carlton, J., Brown, A., Jay, C., and Keane, J. (2019). “Inferring user engagement from interaction data,” in Proceedings of CHI EA, 1212, 1–6. doi: 10.1145/3290607.3313009 Krátky, P., and Chudá, D. (2018). Recognition of web users with the aid of biometric user model. J. Intell. Inf. Syst. 51, 621–646. doi: 10.1007/s10844-018-0500-0 Chen, M. C., Anderson, J. R., and Sohn, M. H. (2001). “What can a mouse cursor tell us more? Correlation of eye/mouse movements on web browsing,” in Proceedings of CHI EA, 281–282. doi: 10.1145/634067.634234 Lagun, D., Ageev, M., Guo, Q., and Agichtein, E. (2014a). “Discovering common motifs in cursor movement data for improving web search,” in Proceedings of WSDM, 183–192. doi: 10.1145/2556195.2556265 Chen, R. C. C., and Chen, T.-K. (2008). The eﬀect of gender-related diﬀerence on human-centred performance using a mass assessment method. IJCAT 32, 322–333. doi: 10.1504/IJCAT.2008.021387 Lagun, D., and Agichtein, E. (2015). “Inferring searcher attention by jointly modeling user interactions and content salience,” in Proceedings of SIGIR, 483–492. doi: 10.1145/2766462.2767745 Chen, Y., Liu, Y., Zhang, M., and Ma, S. (2017). User satisfaction prediction with mouse movement information in heterogeneous search environment. IEEE Trans. Knowl. Data. Eng. 29, 2470–2483. doi: 10.1109/TKDE.2017.2739151 Lagun, D., Hsieh, C.-H., Webster, D., and Navalpakkam, V. (2014b). “Towards better measurement of attention and satisfaction in mobile search,” in Proceedings of SIGIR, 113–122. doi: 10.1145/2600428.2609631 Claypool, M., Le, P., Wased, M., and Brown, D. (2001). “Implicit interest indicators,” in Proceedings of IUI, 33–40. doi: 10.1145/359784.359836 Landauer, A. A. (1981). Sex diﬀerences in decision and movement time. Percept. Mot. Skills 52, 90–90. doi: 10.2466/pms.1981.52.1.90 Diaz, F.,White, R., Buscher, G., and Liebling, D. (2013). “Robust models of mouse movement on dynamic web search results pages,” in Proceedings of CIKM (San Francisco, CA), 1451–1460. doi: 10.1145/2505515.2505717 Leiva, L. A. (2011a). “Mousehints: easing task switching in parallel browsing,” in Proceedings of CHI EA, 1957–1962. doi: 10.1145/1979742.1979861 Diriye, A., White, R., Buscher, G., and Dumais, S. (2012). “Leaving so soon? Understanding and predicting web search abandonment rationales,” in Proceedings of CIKM, 1025–1034. doi: 10.1145/2396761.2398399 Leiva, L. A. (2011b). “Restyling website design via touch-based interactions,” in Proceedings of Mobile HCI (Stockholm), 599–604. doi: 10.1145/2037373.2037467 Feild, H. A., Allan, J., and Jones, R. (2010). “Predicting searcher frustration,” in Proceedings of SIGIR, 34–41. doi: 10.1145/1835449.1835458 Leiva, L. A. (2012). REFERENCES “Automatic web design reﬁnements based on collective user behavior,” in Proceedings of CHI EA, 1607–1612. doi: 10.1145/2212776.2223680 November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org 6 The Attentive Cursor Dataset Leiva and Arapakis Leiva, L. A., and Huang, J. (2015). Building a better mousetrap: compressing mouse cursor activity for web analytics. Inf. Process. Manag. 51, 114–129. doi: 10.1016/j.ipm.2014.10.005 Speicher, M., Both, A., and Gaedke, M. (2013). “TellMyRelevance! Predicting the relevance of web search results from cursor interactions,” in Proceedings of CIKM (Burlingame, CA), 1281–1290. doi: 10.1145/2505515. 2505703 j Leiva, L. A., and Vivó, R. (2013). Web browsing behavior analysis and interactive hypervideo. ACM Trans. Web 7, 20:1–20:28. doi: 10.1145/2529995.2529996 Walker, N., Philbin, D. A., and Fisk, A. D. (1997). Age-related diﬀerences in movement control: adjusting submovement structure to optimize performance. Lindberg, T., Näsänen, R., and Müller, K. (2006). How age aﬀects the speed of perception of computer icons. Displays 27, 170–177. doi: 10.1016/j.displa.2006.06.002 J. Gerontol. A Biol. Sci. Med. Sci. 52, 389–396. doi: 10.1093/geronb/ 52B.1.P40 Liu, Y., Chen, Y., Tang, J., Sun, J., Zhang, M., Ma, S., et al. (2015). “Diﬀerent users, diﬀerent opinions: predicting search satisfaction with mouse movement information,” in Proceedings of SIGIR (Santiago), 493–502. doi: 10.1145/2766462.2767721 White, R., Doraiswamy, P., and Horvitz, E. (2018). Detecting neurodegenerative disorders from web search signals. NPJ Digit. Med. 1:8. doi: 10.1038/s41746-018-0016-6 Yamauchi, T. (2013). “Mouse trajectories and state anxiety: feature selection with random forest,” in Proceedings of ACII (Geneva), 399–404. doi: 10.1109/ACII.2013.72 Lu, H., Rose, J., Liu, Y., Awad, A., and Hou, L. (2017). “Combining mouse and eye movement biometrics for user authentication,” in Information Security Practices, eds I. Traoré, A. Awad, and I. Woungang (Cham: Springer), 55–71. doi: 10.1007/978-3-319-48947-6_5 Yamauchi, T., and Bowman, C. (2014). “Mining cursor motions to ﬁnd the gender, experience, and feelings of computer users,” in Proceedings of ICDMW (Shenzhen), 221–230. doi: 10.1109/ICDMW.2014.131 Martín-Albo, D., Leiva, L. A., Huang, J., and Plamondon, R. (2016). Strokes of insight: user intent detection and kinematic compression of mouse cursor trails. Inf. Process. Manag. 52, 989–1003. doi: 10.1016/j.ipm.2016.04.005 Yamauchi, T., Seo, J. H., Jett, N., Parks, G., and Bowman, C. (2015). Gender diﬀerences in mouse and cursor movements. Int. J. Hum. Comput. Interact. 31, 911–921. doi: 10.1080/10447318.2015.1072787 Mason, W., and Suri, S. (2012). Conducting behavioral research on Amazon’s Mechanical Turk. Behav. Res. Methods 44, 1–23. Frontiers in Human Neuroscience \| www.frontiersin.org November 2020 \| Volume 14 \| Article 565664 REFERENCES doi: 10.3758/s13428-011-0124-6 Zimmermann, P., Guttormsen, S., Danuser, B., and Gomez, P. (2003). Aﬀective computing-a rationale for measuring mood with mouse and keyboard. Int. J. Occup. Saf. Ergon. 9, 539–551. doi: 10.1080/10803548.2003. 11076589 Mueller, F., and Lockerd, A. (2001). “Cheese: tracking mouse movement activity on websites, a tool for user modeling,” in Proceedings of CHI EA, 279–280. doi: 10.1145/634067.634233 Conﬂict of Interest: IA was employed by the company Telefonica Research, though no payment or services from the institution has been received or requested for any aspect of the submitted work. Navalpakkam, V., Jentzsch, L., Sayres, R., Ravi, S., Ahmed, A., and Smola, A. (2013). “Measurement and modeling of eye-mouse behavior in the presence of nonlinear page layouts,” in Proceedings of WWWW, 953–964. doi: 10.1145/2488388.2488471 The remaining author declares that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Pentel, A. (2017). “Predicting age and gender by keystroke dynamics and mouse patterns,” in Adjunct Proceedings of UPMAP (Bratislava), 381–385. doi: 10.1145/3099023.3099105 Shapira, B., Taieb-Maimon,M., andMoskowitz, A. (2006). “Study of the usefulness of known and new implicit indicators and their optimal combination for accurate inference of users interests,” in Proceedings of SAC (Dijon), 1118–1119. doi: 10.1145/1141277.1141542 Copyright © 2020 Leiva and Arapakis. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Smith, M. W., Sharit, J., and Czaja, S. J. (1999). Aging, motor control, and the performance of computer mouse tasks. Hum. Factors 41, 389–396. doi: 10.1518/001872099779611102 November 2020 \| Volume 14 \| Article 565664 Frontiers in Human Neuroscience \| www.frontiersin.org 7
https://openalex.org/W2049089797	https://figshare.com/ndownloader/files/1848763	Zulu	null	Autophagy in osteoblasts is involved in mineralization and bone homeostasis	Autophagy	2,014	cc-by	256	Exon 3 Wild-type allele Deleted allele Flox allele Neo P1 P2 700 bp P2 P3 350 bp A loxP loxP B M Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ 700 bp 500 bp 300 bp Flox Deleted C 1 2 3 4 5 6 C Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ LC3-I LC3-II ACTB relative levels 25000 35000 45000 Exon 3 Wild-type allele Deleted allele Flox allele Neo P1 P2 700 bp P2 P3 350 bp A loxP loxP B M Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ 700 bp 500 bp 300 bp Flox Deleted C 1 2 3 4 5 6 C Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ LC3-I LC3-II ACTB relative levels 25000 35000 45000 Exon 3 Wild-type allele Deleted allele Flox allele Neo P1 P2 700 bp P2 P3 350 bp A loxP loxP Exon 3 Wild-type allele Deleted allele Flox allele Neo P1 P2 700 bp P2 P3 350 bp A loxP loxP B M Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ 700 bp 500 bp 300 bp Flox Deleted C 1 2 3 4 5 6 C Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ LC3-I LC3-II ACTB LC3-II to ACTB relative levels 5000 15000 25000 35000 45000 C M Exon 3 A B M Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ 700 bp 500 bp 300 bp Flox Deleted C 1 2 3 4 5 6 B B 1 Deleted C Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ LC3-I LC3-II ACTB LC3-II to ACTB relative levels 5000 15000 25000 35000 45000 C M C Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ LC3-I LC3-II ACTB C Atg5flox-flox Col1A-Cre- atg5flox-flox Col1A-Cre+ C M
https://openalex.org/W2903106546	https://revistas.marilia.unesp.br/index.php/dialogoseperspectivas/article/download/8339/5347	es		Desafíos en Educación Superior: el caso de las personas con discapacidad intelectual en el Perú	Revista Diálogos e Perspectivas em Educação Especial	2,018	cc-by	9,048	Dossiê Desafíos en Educación Superior Desafíos en Educación Superior: el caso de las personas con discapacidad intelectual en el Perú Desafios da Educação Superior: o caso das pessoas com deficiencia intelectal no Perú Estefani Aime Apaza QUISPE1 Resumen: El logro de la inclusión de las personas con discapacidad intelectual (DI) es uno de los mayores desafíos a los que se enfrentan las instituciones educativas y comunidad. Según la Organización Mundial de la Salud (OMS, 2011), un 15% de la población presenta situación de discapacidad, mientras que en el Perú existe alrededor de un 5.2%, siendo de mayor proporción la discapacidad intelectual. En respuesta a ello, se plantea el enfoque inclusivo o de Atención a la diversidad por medio del actual Currículo Nacional de la Educación Básica y desde Educación Superior, normativas que garanticen el ingreso y permanencia en esta. Sin embargo, todavía existe dificultades en las instituciones educativas de estos niveles educativos; ya que, en educación básica, prevalece un 25% de personas con discapacidad en edad escolar sin acceso al sistema educativo dentro del cual se incluye a las personas con DI y en educación superior, solo un 5,99% accede a institutos y universidades, del cual no se han encontrado reportes sobre el caso de las personas con DI. De tal modo que surge la importancia de explorar sobre la situación de estas personas; para ello, el presente estudio tiene como objetivo analizar las características sociodemográficas y de habilidades instrumentales de aprendizaje de 10 participantes (2 varones y 8 mujeres) con DI asistentes a un servicio de salud de un Hospital Público de Lima Metropolitana. El tipo y diseño de investigación corresponde a Empírico-no experimental con estrategia descriptiva mediante encuestas. La muestra fue recogida mediante un muestreo intencionado. Los instrumentos utilizados fueron una ficha sociodemográfica y una lista de cotejo AD HOC de las habilidades instrumentales de aprendizaje. En esta comunicación se fundamentan breves avances sobre el constructo de discapacidad intelectual, el enfoque de inclusión y las políticas educativas que inciden en el acceso y permanencia de las personas con DI en la educación superior. Los hallazgos buscan la visibilización de la situación de algunas personas con DI y la generación de evidencia que sirva de base para la elaboración de medidas oportunas y pertinentes que brinden iguales oportunidades para todos y todas. Palabras clave: Discapacidad intelectual; inclusión educativa; habilidades instrumentales de aprendizaje; educación básica y superior; políticas educativas. Abstract: The achievement of the inclusion of people with intellectual disability (ID) is one of the greatest challenges facing educational institutions and the community. According to World Health Organization (WHO, 2011), 15% of the population present disability, while in Peru there is roughly 5.2%, with intellectual disability being a greater proportion. In response to this, the inclusive or attention to diversity approach is proposed through the current National Curriculum of Basic Education, and Higher Education, regulations that guarantee entry and permanence in this. However, there are still difficulties in the educational institutions of these educational levels; in basic education, 25% of people with disabilities of school age prevail without access to the education system, which includes people with ID, and in higher education, only 5.99% access to institutes and colleges, which no reports have been found about the case of people with ID. In such a way that the importance of exploring about the situation of these people; thus, the present study aims to analyze the sociodemographic characteristics and instrumental learning skills of 10 participants (2 men and 8 women) with ID attending a health service of a Public Hospital in Lima Metropolitana. The type and design of research corresponds to Empirico-no experimental with descriptive strategy through surveys. The sample was collected by intentional sampling. The instruments used were a sociodemographic index and an AD HOC list of instrumental learning skills. This communication is based on brief advances on the construction of intellectual disability, the inclusion approach and educational policies that affect the access and permanence of people with ID in higher education. The findings seek the visibility of the situation of some people with ID and the generation of evidence that serves as the basis for the development of timely and relevant measures that provide equal opportunities for all. Keywords: Intellectual disability; educational inclusion; instrumental learning skills; basic and superior education; educational policies. 1 Estefani Aime Apaza Quispe. Bachiller en Psicología por la Universidad Nacional Mayor de San Marcos (UNMSM), Lima, Perú. Especializada en el “Estudio en Discapacidad” por la Facultad de Derecho de la Pontificia Universidad Católica del Perú (PUCP) y la Red Iberoamericana de Expertos sobre la Convención de los Derechos de las Personas con Discapacidad (Red CDPD). E-mail: estef.aime@gmail.com. https://doi.org/10.36311/2358-8845.2018.v5n1.11.p141 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 141 QUISPE, E. A. A. Introducción La educación es un derecho y el camino para garantizar el desarrollo integral de las personas. En el caso de las personas con discapacidad, el acceso y permanencia a una educación en todos los niveles educativos –desde lo inicial hasta educación superior– que brinde iguales oportunidades que los demás, potencia su desarrollo y salud integral aún más. En este sentido, Muntaner (2013) señala que lograr una educación de calidad para todos y con todos es un objetivo compartido por el mundo entero y al partir del marco de inclusión educativa se pretende acabar con la exclusión de cualquier estudiante, tanto para los que no tienen posibilidad de acceder o presentan una situación de desventaja en el sistema educativo. Al igual que otros países latinoamericanos, el Perú es uno de los Estados Partes de la Convención Internacional sobre los Derechos de las Personas con Discapacidad-CDPD (Organización de las Naciones Unidas, 2008), en este documento se reconoce el valor de las contribuciones que pueden brindar las personas con discapacidad al bienestar general y a la diversidad de sus comunidades, asimismo, se ratifica en la Ley General de las Personas con Discapacidad (Ley Nº 29973). En ambos documentos se define discapacidad como el resultado de la interacción de la persona con deficiencia de manera permanente -física, sensorial, mental o intelectual- y las barreras actitudinales y del entorno, que pueda impedir su participación plena en la sociedad en las mismas condiciones que los demás (De Asis Roig, 2013). Pese a contar con un marco sobre discapacidad y ser un país alineado a las normativas internacionales como la CDPD, todavía no hay un consenso sobre los tipos de discapacidades que se utilizan en los censos y las encuestas; así como, una escasa información actualizada y literatura sobre la situación educativa de las personas con DI destacándose la escasa información en educación superior. Esto conlleva a pensar que los servicios educativos que se brindan en el Perú todavía no están encaminados a superar la transición a la vida adulta y el desarrollo integral de una vida autónoma en las personas con discapacidad. Semejante panorama revela el estudio realizado por Neves-Silva & Álvarez-Martín (2014) sobre la situación en general de las personas con discapacidad en el que señalan la gran dificultad en encontrar datos sobre discapacidad, asimismo, al analizar la situación en 13 países (Salvador, Guatemala, Honduras, Paraguay, Argentina, Brasil, Chile, Ecuador, Nicaragua, Brasil, Colombia, Perú y Uruguay) encontraron que existe mayor prevalencia de discapacidad en población en situación de pobreza, envejecimiento, bajas oportunidades educativas, baja asistencia perinatal, etc. Por lo que ven limitado el ejercicio de muchos derechos como a la vivienda, trabajo y educación. De acuerdo con el modelo social de la discapacidad, el Perú fundamenta la educación bajo un enfoque de atención a la diversidad tal como se subscribe en la Ley General de Educación (Ley N° 28044) y se garantiza en todos sus niveles educativos; por ejemplo, en el actual Currículo Nacional de Educación Básica se incide en (i) el reconocimiento al valor inherente de cada persona y de sus derechos, por encima de cualquier diferencia;(ii) la disposición a enseñar ofreciendo a los estudiantes las condiciones y oportunidades que cada uno necesita para lograr los mismos resultados; y (iii) disposición a depositar expectativas en una persona, creyendo sinceramente en su capacidad de superación y crecimiento por sobre cualquier circunstancia (Ministerio de Educación [MINEDU], 2016). Estos son principios que se relacionan con la justicia, equidad, calidad e igualdad de género y apoyan la participación de las personas con discapacidad; así también, son concordantes con el objetivo cuatro para el desarrollo sostenible 142 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Desafíos en Educación Superior Dossiê “Garantizar una educación inclusiva, equitativa y de calidad, y promover oportunidades de aprendizaje durante toda la vida para todos”. Al respecto, la inclusión educativa se define a partir de tres principios fundamentales: presencia, participación y progreso (Muntaner, 2013). Esto implica que todos los estudiantes estén presentes participando en todas las actividades, experiencias y situaciones que se les brinda en el aula y el centro para que alcancen aprendizajes. Según la Organización de la Naciones Unidas para la Educación, la Ciencia y la Cultura (Unesco, 2008) es “un proceso orientado a responder a la diversidad de los estudiantes incrementando su participación y reduciendo la exclusión en y desde la educación. Está relacionada con la presencia, la participación y los logros de todos los alumnos, con especial énfasis en aquellos que, por diferentes razones, están excluidos o en riesgo de ser marginados, constituyendo un impulso fundamental para avanzar en la agenda de la Educación para todos y todas” (p.7). Para las personas con discapacidad, el proceso de inclusión educativa se ha venido dando de manera paulatina; ya que desde el 2012 con la promulgación de una nueva Ley General de la Persona con Discapacidad (Ley N° 29973) y el reglamento de la Ley General de Educación se comienza a ejecutar medidas para facilitar este proceso en todos los servicios educativos. Por su parte, el Ministerio de la Mujer y Poblaciones Vulnerables crea el Consejo Nacional para la integración de la persona con Discapacidad (CONADIS), instancia que promueve que las instituciones incorporen en sus planes un enfoque inclusivo. Adicionalmente, el estado peruano califica como “El decenio de las personas con Discapacidad en el Perú” al periodo de los años 2007 al 2016. La inclusión de estudiantes con discapacidad en educación básica se ha venido realizando con la colaboración de los Servicios de Apoyo y Asesoramiento a las Necesidades Educativas Especiales (SAANEE). Además, se dieron normas directivas de matrícula que aseguren un mínimo de dos vacantes para estudiantes con necesidades educativas especiales asociadas a discapacidad – Leve y moderado. Sin embargo, los hallazgos de Cueto, Miranda & Vasquéz (2016) al analizar la Reforma Magisterial 29944 del 2012 señalan que sufrir de alguna discapacidad es un obstáculo para acceder y recibir una educación de calidad y que en dicha normativa para docentes solo en 3 artículos del documento se hace referencia a disposiciones sobre discapacidad. Así también, acceden a menores niveles educativos durante su educación y se encuentran con barreras como infraestructura inadecuada, materiales no especializados y docentes poco preparados para atenderlos (Defensoría del Pueblo, 2000). Otra problemática se encuentra en el Servicio de Apoyo y Asesoramiento para la Atención a las Necesidades Educativas Especiales (SAANEE) que tiene como función apoyar la inclusión de los estudiantes con discapacidad a escuelas regulares, estos servicios aún son pequeños y no tienen correspondencia ni respaldo consistente en las instancias descentralizadas del sistema educativo (Tovar, 2013). Adicionalmente, Fernández (2007) entrevistó a directores, docentes, estudiantes y padres de familia de escuelas regulares evidenciando que los profesores no son muy positivos cuando se les pregunta por las facilidades para aprender con las que cuentan los niños o niñas con discapacidad a los que enseñan: el 56,3% se muestra inseguro y el resto señala que existe poca capacidad de aprender en ellos. Respecto a la inclusión de personas con discapacidad en educación superior, se establece que las instituciones deben realizarse los ajustes razonables para garantizar el acceso y Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 143 QUISPE, E. A. A. permanencia de las personas con discapacidad, así pues, existe una reserva del 5% de vacantes o en caso incurran en discapacidad y se vean obligados a interrumpir sus estudios cuentan con una reserva de matrícula hasta por 5 años. También existen disposiciones que mencionan la inclusión de cursos sobre discapacidad en los currículos y programas de estudios. Desde el 2014 se reglamentó la norma técnica de Infraestructura para Locales de Educación Superior y en su versión más reciente se indica que se debe brindar las condiciones de accesibilidad y seguridad para personas con diferentes tipos de limitaciones en concordancia con los criterios de Diseño Universal como uso de rampas o pisos podo táctiles para personas con limitación visual, etc. Estas disposiciones se ratifican en la nueva Ley universitaria (2014) y la Ley de institutos y escuelas de Educación Superior y de la Carrera Pública de sus Docentes (2017). A pesar de lo dispuesto, el panorama se torna más alarmante en este nivel educativo puesto que todavía no se han encontrado procedimientos de licenciamiento en la Superintendencia de Educación Universitaria (SUNEDU) para asegurar una educación de calidad para las personas con discapacidad; así pues, no existe un protocolo para admisión y solo algunas universidades como la Pontificia Católica de Perú (PUCP) realizan las adaptaciones a su criterio e incluso genera cartillas explicativas de conceptos para los docentes que cuenten con estudiantes con discapacidad (Constantino, 2013). Otra problemática es que no existen disposiciones vinculadas a los ajustes razonables del currículo y programas, ni a la disponibilidad de recursos de aprendizaje y bibliográficos adecuados a las necesidades de la población, tampoco existen servicios de acompañamiento y tutoría (Vega-Ganoza, 2016). Por lo que queda a disposición de cada centro de estudios realizar los ajustes razonables al programa de estudios como el caso de la Universidad San Ignacio de Loyola (USIL), la cual reporta el caso de un estudiante con síndrome de Down que concluyó sus estudios superiores en el 2014 a través del programa de inclusión educativa. De esto, aún continúan siendo los aspectos más problemáticos la adecuación metodológica y la cuestión de no haberse superado el modelo integracionista (Bregaglio, Constantino y Cruzado, 2014). En esta línea de ideas, partir de un marco referencial sobre el estado de la inclusión en el nivel superior resulta difícil pese a que algunas universidades tienen el interés en dar un procedimiento diferente a este grupo con base al modelo social en sus políticas institucionales son escasos los reportes. Son pocas las investigaciones más cercanas a visibilizar la situación de los estudiantes con discapacidad en una universidad. Tal como muestra un estudio que realizó Bregaglio, et al. (2014) en la Pontificia Universidad Católica de Perú (PUCP) desde el 2005 hasta el 2013 se registró el ingreso de 31 estudiantes con discapacidad, no obstante, no es un registro definitivo y oficial. Otros resultados encontrados, al explorar la percepción de la comunidad educativa en relación a los tipos de discapacidad, evidenciaron que existe actitudes más positivas hacia las personas con discapacidad física, seguida de la intelectual, psicosocial y, en el último lugar, la sensorial. En lo referido a accesibilidad, si bien no cumple en su totalidad con edificios con accesibilidad vertical y con un sistema de audio en sus elevadores; existen prácticas destacables como reubicar la clase en el primer piso cuando se detecta que un estudiante se encuentra en situación de discapacidad física. Por su parte, en lo referido a instituciones no universitarias, no se ha encontrado reportes similares. En definitiva, los datos estadísticos permiten afirmar que asegurar el aprendizaje y la participación plena de estas personas es una tarea pendiente en el Perú: De un total del alumnado con discapacidad, la cifra más resaltante se ubica en los estudiantes con Discapacidad In144 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Desafíos en Educación Superior Dossiê telectual, según el Censo escolar del 2016 existe 10671, ello equivale al 77%. Asimismo, según la Encuesta Nacional de Hogares prevalece 37.846 personas con discapacidad en edad escolar sin acceso al sistema educativo, ello equivale al 25% (Instituto Nacional de Estadística [INEI], 2015). Lo más alarmante es la exclusión educativa en el nivel secundario equivalente al 42%. (Constantino, Bregaglio & Cueva, 2016). Mientras que en Educación Superior, la situación es más incierta debido a que solo un 5,99% accede a institutos y universidades; y el registro más cercano sobre el tema se encuentra en el censo educativo del 2010, en el que se muestra que los estudiantes con limitaciones visuales son los que registran mayores matriculas en comparación a otras limitaciones como motoras, de lenguaje, múltiples y auditivas. Cabe mencionar que no se refiere a la situación de las personas con discapacidad intelectual (Vega-Ganoza, 2016). De manera general, desde la educación básica hasta la superior las personas con discapacidad se encuentran con dificultades para acceder y permanecer en el sistema educativo tal como lo señala la Organización para la Cooperación y Desarrollo Económico (OCDE, 2016), los desafíos pendientes que existen están en relación al acceso, específicamente en niveles superiores y técnicos, así también en relación a la desigualdad en el acceso entre distintos grupos socioeconómicos. Se resalta que “uno de los principales factores que determinan las diferencias en acceso al nivel educativo se relaciona con el ingreso, cuyo impacto se hace mayor a medida que se avanza en el ciclo educativo” (p.11). Sobre la base de esto, se destaca que el logro de la inclusión de las personas con Discapacidad Intelectual (en adelante DI), es uno de los mayores desafíos a los que se enfrentan las instituciones educativas peruanas –escuelas, institutos y universidades– debido a la gran proporción poblacional de personas con esta discapacidad en el Perú y la escasa investigación acerca de la situación de las mismas en niveles educativos superiores. Para fines de este estudio, se busca analizar los desafíos que se planean en educación superior a partir del análisis de algunas características sociodemográficas y de habilidades instrumentales de aprendizaje de un grupo de personas con DI. Entendiéndose que las características sociodemográficas son el conjunto de indicadores que muestran la situación y evolución de un grupo poblacional, y por habilidades instrumentales de aprendizaje como las destrezas necesarias o esenciales aprendidas de forma incidental o intencional que tienen una aplicación directa en la vida personal. Por lo que se busca responder a la pregunta ¿Cuáles son los desafíos en educación superior según las características sociodemográficas y de habilidades instrumentales de aprendizaje que presentan un grupo de jóvenes y adultos con DI asistentes a un servicio de salud de un Hospital público de Lima Metropolitana? Para responder a dicha pregunta, el análisis de la información recolectada proviene de dos cuestionarios que abordan dichas áreas y se explican de manera descriptiva. Consideraciones sobre la Discapacidad intelectual Para la comprensión de lo que implica la Discapacidad Intelectual (antes conocido como Retraso Mental), la Asociación Americana de Discapacidades Intelectuales y Del Desarrollo (AAIDD, por sus siglas en inglés) es el principal organismo que define y estudia esta condición. Por su parte, la Organización Mundial de la Salud (2016) a través de la Clasificación Internacional del Funcionamiento, de la Discapacidad y de la Salud describe lo que se entiende Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 145 QUISPE, E. A. A. por DI, así pues, menciona que requieren mayor asistencia sanitaria y cuentan con más necesidades insatisfechas en esta esfera, señalando que quedan excluidos más a menudo. En tal sentido, la discapacidad intelectual se explica sobre la base de tres conceptos significativos: a) el modelo socio-ecológico de discapacidad; b) un enfoque multifactorial de la etiología; y c) la distinción entre una definición operativa y otra constitutiva de la condición (Verdugo & Shalock, 2010). Por lo que la DI se define como las “limitaciones significativas tanto en funcionamiento intelectual como en conducta adaptativa tal y como se ha manifestado en habilidades adaptativas conceptuales, sociales y prácticas. Esta discapacidad aparece antes de los 18 años” (Schalock et al., 2010, p. 1). En esta nueva concepción, que surge desde la psicología, se vuelve relevante entender el concepto de “Funcionamiento humano”, este consta de dos componentes: cinco dimensiones (habilidades intelectuales, conducta adaptativa, Salud, Participación y Contexto) y una representación del rol de los apoyos. Figura 1. Esquema conceptual del Funcionamiento Humano. a. Habilidades intelectuales: capacidad mental general que considera al razonamiento, la planificación, la solución de problemas, el pensamiento abstracto, la comprensión de ideas complejas, el aprender con rapidez y el aprender de la experiencia. b. Conducta adaptativa: conjunto de habilidades conceptuales, sociales y prácticas que se han aprendido y se practican por las personas en su vida cotidiana. c. Salud: estado de completo bienestar físico, mental y social. d. Participación: se considera al desempeño de la persona en actividades reales en ámbitos de la vida social que se relaciona con su funcionamiento en la sociedad; la participación se refiere a los roles e interacciones en el hogar, trabajo, ocio, vida espiritual, y actividades culturales. e. Contexto: se refleja en las condiciones interrelacionadas en las que viven las personas su vida cotidiana; el contexto incluye factores ambientales (por ejemplo, físico, social, actitudinal) y personales (por ejemplo, motivación, estilos de afrontamiento, estilos 146 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Dossiê Desafíos en Educación Superior de aprendizaje, estilos de vida) que representan el ambiente completo de la vida de un individuo (Verdugo & Shalock, 2010). El otro componente es el rol de los apoyos, estos se entienden como los recursos y estrategias destinados a promover el desarrollo, la educación, los intereses y el bienestar personal, y a incrementar el funcionamiento individual (Luckasson et al, 2002). Del mismo derivan las necesidades de apoyo constituyéndose un “constructo psicológico que hace referencia al patrón e intensidad de los apoyos necesarios para que una persona participe en actividades relacionadas con un funcionamiento humano estándar” (Thompson et al, 2009). Este es un nuevo paradigma que permitirá repensar en cómo se deben orientar las medidas de accesibilidad y ajustes razonables para brindar los apoyos educativos necesarios que potencialicen a los estudiantes con DI. Ello conllevará a mejorar la vida y salud integral de las personas con DI. Desarrollo Este estudio es de tipo Empírico-no experimental (Ato, López y Benavente, 2013), de Estrategia descriptiva mediante encuestas – transversal (León y Montero, 2007); ya que permite recoger información sobre una o más variables en un tiempo específico. A continuación se describen las variables de estudio. Cuadro 1. Definición de Variables Áreas Variables Sexo Características sociodemográficas Habilidades instrumentales de aprendizaje Definición Esta variable indica si el sujeto es hombre o mujer. Nivel educativo Esta variable indica el tipo de nivel de instrucción que obtuvo el sujeto: Sin nivel, asistieron a EBR, EBA o EBE. Preparación laboral Esta variable indica si el sujeto tuvo presencia o ausencia de preparación laboral. Ocupación Laboral Esta variable indica si el sujeto tuvo presencia o ausencia de ocupación laboral. Participación ciudadana Esta variable indica si el sujeto tuvo presencia o ausencia de participación en organizaciones. Condiciones de vida Esta variable indica las condiciones en las que vive el sujeto: vive independiente o dependientes de su familia y si cuenta con servicios básicos. Comunicación y lenguaje Esta variable indica si el sujeto tiene presencia o ausencia de uso de gestos o palabras para comunicar sus mensajes y necesidades a otros. Lectura de palabras Esta variable indica si el sujeto tiene ausencia o presencia de decodificación palabras. Escritura de palabras Esta variable indica si el sujeto tiene presencia o ausencia de transcripción de palabras al dictado y copia. Uso de operaciones matemáticas Esta variable indica si el sujeto tiene presencia o ausencia de resolución de sumas y restas con números de dos unidades. Elaboración propia. Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 147 QUISPE, E. A. A. Los participantes se seleccionaron utilizando un tipo de muestreo no probabilístico propositivo (Kerlinger y Lee, 2002); es decir, el proceso de selección fue deliberado para obtener una muestra de la población de estudio. Dicha muestra estuvo constituida por 10 participantes con DI, 2 hombres y 8 mujeres cuyas edades fluctúan entre 18 a 50 años (ME=28.7; DE=10.7), los cuales son asistentes a un servicio de salud de un Hospital público de Lima Metropolitana. De los participantes, seis presentan comorbilidad con el Síndrome de Down. Para ello, se procedió presentando a los participantes y sus cuidadores el objetivo del estudio, en función de la autorización y conformidad de la información brindada, se procedió con la firma de consentimientos informados. Luego, se recogió la información a través de las evaluaciones individuales tanto al participante como al cuidador principal. Los instrumentos utilizados fueron: a) Ficha Sociodemográfica: El protocolo de evaluación fue diseñado para la evaluación de datos personales del participante, nivel educativo, nivel de ocupación, composición de hogar y nivel de ingresos. La interpretación de resultados se realiza de manera descriptiva. b) Lista de cotejo AD HOC de las habilidades instrumentales del aprendizaje: El protocolo de evaluación fue elaborado para identificar un conjunto de habilidades instrumentales para actividades como comunicación y lenguaje, la lectura, escritura y uso de operaciones básicas (Apaza, Huamán, Ramírez, Segovia y Suárez, 2016), en donde el evaluador trata de generar estímulos en función de los reactivos para que evoquen la conducta requerida. Se registran las conductas con los indicadores de logro (logrado o no logrado), adicionalmente, en caso que no se haya logrado una conducta se anota el tipo de apoyo que requiere para lograrlo, hay un acápite para consignar observaciones. La interpretación de resultados se realiza de manera descriptiva. Resultados y discusión El objetivo de estudio fue analizar las características sociodemográficas y de habilidades instrumentales de aprendizaje de un grupo de personas con DI. En cuanto a las características sociodemográficas se muestra que en la variable Nivel educativo existe mayor presencia de participantes con educación incompleta y menor presencia de personas no escolarizadas. De las personas escolarizadas, la mayoría asistió a Educación Básica Especial (EBE) y solo una accedió a educación superior. En lo referido al análisis de las Habilidades instrumentales de aprendizaje, se encuentra que existe mayor presencia de la variable Comunicación y lenguaje, y menor presencia del uso de operaciones matemáticas. A continuación se presenta la cantidad de ocasiones en las cuales se han presentado las variables descritas. De lo encontrado en el nivel educativo, se muestra que del grupo, ocho de diez asistieron a Educación Básica Especial y del resto, uno asistió a Educación Básica Alternativa y otro a Educación Básica Regular. En lo referido a educación superior, se destaca que nueve de diez no alcanzaron este nivel educativo. 148 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Dossiê Desafíos en Educación Superior Tabla 1. Incidencia de variable Nivel educativo en las personas con DI Variable Nivel educativo Indicadores Frecuencia Sin nivel 1 Nivel primaria incompleto 6 Nivel primaria completo 2 Nivel secundaria incompleto 1 Nivel secundaria completo 1 Educación superior completo 0 Educación superior incompleto 1 Elaboración propia. En detalle como muestra la tabla 1., en la variable Nivel educativo, uno de diez no ha contado con instrucción educativa y de los otros casos, nueve de diez han tenido algún tipo de instrucción educativa, de los cuales seis de diez participantes con DI presentan el nivel primario incompleto y solo dos lo han concluido. Cabe mencionar que los que no concluyeron el nivel de primaria asistían a Centros de Educación Básica Regular y Especial. Se observa que solo un participante ha concluido secundaria. Adicionalmente, la información vertida por los cuidadores principales sobre las razones de no concluir sus estudios se encontraron argumentos de pocas expectativas educativas tanto de los padres como maestros y por parte de las escuelas, signos de segregación, fracaso escolar y/o barreras para el acceso. De los participantes que permanecieron en EBE, se mantuvieron en esas escuelas ya que no fueron aceptados ni evaluados para acceder a otras, según los informantes. En lo referido a educación superior, solo un participante accedió a este nivel aunque no lo concluyó. Cabe destacar, que accedió a este nivel debido a que concluyó su nivel secundario y tuvo posibilidades de seguir una carrera universitaria. Las razones señaladas por las que no permaneció estudiando se relacionan con la falta de ajustes razonables por parte de la institución. Lo evidenciado en esta variable conlleva a repensar nuevamente la diversidad y en que no es suficiente el acceso a la educación sino que hace falta la permanencia en un sistema educativo con calidad y equidad. Como se muestra en los resultados, existen inequidades educativas y gran retraso en el cumplimiento de sus niveles educativos, ya que no han logrado completar grados lo cual implica que no han logrado adquirir alguna de las competencias descritas en el currículo nacional. Tal como lo mencionan Cueto, Miranda & Vásquez (2016) y el informe de la Defensoría del pueblo sobre Educación Especial (2000) son grandes las brechas educativas que tienen las personas con discapacidad para acceder a una educación con iguales oportunidades de aprendizaje que los demás. De igual modo, es concordante con los hallazgos de Neves-Silva & Álvarez-Martín (2014) donde mencionan que en la mayoría de países existe alrededor del 42% a más personas con discapacidad que no han completado la enseñanza básica. El plano educativo de las personas con DI es crítico, pese a que se encuentra vigente la política que refiere al derecho a una educación inclusiva (Ley general de educación 28044) no se ha pensado de qué forma se va a apoyar a las personas con DI que no pudieron tener una educación de calidad, puesto que esto continúa generando desigualdades sociales y no asegura el bienestar general de estas personas. Además, genera barreras de acceso a niveles superiores para optar por una carrera universitaria o técnica. Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 149 QUISPE, E. A. A. La mayoría de casos analizados pertenecen a jóvenes y adultos que tuvieron que asistir a educación especial; ya que los segregaron de escuelas regulares y solo el participante que concluyó su educación básica tuvo posibilidades de acceder a educación superior. Esto refuerza la idea de reconvertir los Centros de Educación Básica Especial como lo recomiendan desde un análisis realizado por el Consejo Nacional de Educación (2017), esto implica formular un plan oportuno y pertinente que comience a tomar medidas para reducir las brechas entre la educación especial y regular. Este contexto no se puede generalizar puesto que a nivel latinoamericano todavía prevalecen estos servicios como Centros de Educación Especial y están en la búsqueda por la superación de la dimensión médica y de discapacidad presentes históricamente en la Educación Especial y en el Servicio Educativo Especializado (Casagrande & Cruz, 2016). No obstante, para otros autores como Florián (2014) estos centros responden a un enfoque tradicional basado en las necesidades educativas especiales (NEE), que proporciona servicios “diferentes” o “adicionales” a los proporcionados a otros niños de edad similar, generando de algún modo segregación y perpetuando la discriminación. Otro punto que se debe analizar es el número reducido de participantes que acceden a educación superior. Si bien, en la actualidad, más peruanos acceden al sistema educativo, las desigualdades de acceso relacionadas con el nivel socioeconómico, el género, discapacidad, lengua, entre otras, son profundas y aumentan a medida que se avanza hacia los niveles educativos más altos como educación superior como lo señaló la OCDE (2016). No obstante, no basta con establecer medidas para acceso sino que resulta necesario tomar acciones adecuadas para que no resulte imposible la permanencia del estudiante con discapacidad (Bregaglio, et al. 2014). De lo encontrado en preparación laboral, se destaca que la mayoría ha logrado tener algún tipo de preparación. Tabla 2. Incidencia de variable Preparación Laboral en las personas con DI Variable Preparación Laboral Indicadores Frecuencia Presencia de Preparación 6 Ausencia de Preparación 4 Elaboración propia. En detalle como muestra la tabla 2., en la variable Preparación Laboral, seis de diez tuvo preparación laboral mientras que el resto no tuvo preparación alguna. El tipo de preparación que presenta está relacionado a talleres de cocina, repostería y cerámica y talleres técnicos de carpintería, jardinería y costura. Cabe resaltar que la mayoría de los talleres recibidos fueron dados por instituciones privadas. Estos resultados evidencian la falta de intervención de políticas de estado para garantizar la capacitación en cursos u oficios técnicos de las personas con DI. Esto es concordante con la Encuesta Nacional Especializada sobre Discapacidad (ENEDIS) del 2012, en la cual se refleja que el 97,3% no se capacita, de esta forma se ve limitada su posibilidad de ocuparse en alguna actividad laboral. De lo encontrado acerca de la ocupación laboral, se destaca que la mayoría de participantes no presenta ocupación alguna. Pese a que según lo encontrado en la anterior variable analizada la mayoría cuenta con preparación laboral. 150 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Dossiê Desafíos en Educación Superior Tabla 3. Incidencia de variable Ocupación Laboral en las personas con DI Variable Ocupación Laboral Indicadores Frecuencia Presencia de ocupación 1 Ausencia / No se ocupa 9 Elaboración propia. En detalle como muestra el Tabla 3., en la variable Ocupación Laboral, uno de diez presenta ocupación laboral mientras que el resto muestra ausencia de ocupación. El tipo de actividades en la que suelen ocuparse está relacionado con trabajos de limpieza, estampados, cerámica y repostería. Los resultados evidencian que las personas con DI a pesar de tener preparación no han podido ocuparse laboralmente y esto se relaciona con lo reportado en la Encuesta Nacional Especializada sobre Discapacidad (ENEDIS) que indica que existe un 76,8% de población que se encuentra Económicamente Inactiva. Aunque no se hayan encontrado reportes en DI, los hallazgos de la ENEDIS no se alejan de estos resultados. Por ello, pese a que existen normas y leyes que están, en teoría, posibilitando que se garantice que las empresas incluyan personal con DI según la Ley General para la Persona con Discapacidad aparentemente en la práctica esto no se refleja. De lo encontrado en participación ciudadana, se resalta que la mayoría participa de algún tipo de organización. Tabla 4. Incidencia de variable Participación Ciudadana en las personas con DI Variable Participación Ciudadana Indicadores Frecuencia Presencia de Participación 7 Ausencia de Participación 3 Elaboración propia. En detalle como muestra el tabla 4., en la variable Participación Ciudadana, siete de diez participan de alguna organización para personas con Discapacidad y el resto no participa. Entre las organizaciones a las que participan se destaca la Oficina Municipal de Atención a la persona con Discapacidad (OMAPED). No obstante, participan de actividades culturales y en su mayoría son matriculados a estas por sus padres. Cabe mencionar que esta variable refleja la participación, en términos de ingreso a estas organizaciones; ya que no se planteó explorar sobre su conocimiento, ejercicio de ciudadanía y participación política en dichas instituciones. De lo encontrado acerca de las condiciones de vida, se destaca que, en su totalidad, los participantes viven de manera dependiente a sus familias. Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 151 QUISPE, E. A. A. Tabla 5. Incidencia de variable Condiciones de Vida en las personas con DI Variable Condiciones de Vida Indicadores Frecuencia Independiente 0 Dependiente 10 Elaboración propia. En detalle como muestra tabla 5., en la variable Condiciones de vida, todos viven de manera dependiente a su familia o cuidadores. La composición de los hogares de los participantes varía en un promedio de 6 personas y se destaca que todos cuentan con los servicios básicos de agua, luz y desagüe. De lo mostrado, se debe resaltar la idea de que las personas con DI acceden a servicios a partir de la presencia de un familiar puesto que ellos representa una figura de cuidador, esto de alguna forma refuerza nociones de verlos como “niños eternos”. Sin embargo, desde lo encontrado no se puede reflejar las condiciones de pobreza en la viven ciertas personas con discapacidad como lo menciona Tovar & Fernández (2005) quiénes afirman que este grupo de personas viven en situación de pobreza y resulta un gasto adicional para la familias tener un familiar en esta condición puesto que la mayoría de sus ingresos se destinan a las rehabilitaciones porque no cuentan con servicios de salud gratuitos. En cuanto a las Habilidades instrumentales de aprendizaje, se muestra que ocho de diez han logrado la habilidad de Comunicación y lenguaje, esto implica que hacen uso de gestos o palabras para comunicar sus mensajes y necesidades a otros; mientras que las otras dos personas requieren apoyo de interpretación comunicativa. Tabla 6. Incidencia de variable Habilidades instrumentales de aprendizaje en las personas con DI Variable Habilidades instrumentales de aprendizaje Indicadores Frecuencia Comunicación y lenguaje 8 Lectura de palabras 6 Escritura de palabras 5 Uso de operaciones matemáticas 2 Elaboración propia. En detalle como se muestra en la tabla 6., en cuanto a lectura de palabras, se observa que seis de diez personas han logrado la habilidad de decodificar palabras y la otra parte que representa cuatro del total requieren apoyo de Sistemas Aumentativos y Alternativos de Comunicación. Estos apoyos de tipo visual como pictogramas, ayudas visuales y horarios visuales sirvieron para contrastar que si bien no han logrado la habilidad de leer palabras son capaces de nombrar el significado de las imágenes que se les muestra. En cuanto a la escritura de palabras, se muestra que cinco de diez personas han logrado la habilidad de transcribir palabras al dictado y copia, y la otra parte que representa la mitad del total requiere apoyo. Las palabras que fueron evaluadas fueron sus nombres, monosílabos y palabras de dos y tres sílabas. En cuanto a las operaciones matemáticas, se observa que sólo dos de diez han logrado 152 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Desafíos en Educación Superior Dossiê la habilidad de resolver sumas y restas de números con dos unidades, y de los otro ocho no han logrado dicha habilidad por lo cual requieren de apoyos de manejo y gestión de dinero. De los resultados encontrados en esta área, se observa que no todas las personas con DI han logrado habilidades básicas que le permitirán realizar actividades como comunicar las necesidades, leer anuncios, escribir sus nombres y manejar dinero. Sin embargo, no se ha podido recoger información de la forma en que esto limita su participación en la sociedad a pesar que se considera como una barrera de aprendizaje que no hayan logrado habilidades porque los sitúa en desventaja en comparación a los demás. Así también, existe escasa información sobre el aprendizaje de personas con DI que permita contrastar lo obtenido en este estudio. Al igual que en la Evaluación Censal de Estudiantes (ECE) que se realiza anualmente para recoger el rendimiento de todos los estudiantes peruanos, todavía no se ha hecho el esfuerzo suficiente para incluir a los estudiantes con discapacidad intelectual pese a que desde el 2015 la Oficina de la Medición de la Calidad de los Aprendizajes (UMC) ha podido realizar las acomodaciones para los estudiantes con discapacidad sensorial (MINEDU, 2016). Cabe mencionar que de los dos participantes que lograron estás habilidades, uno de ellos fue el que tuvo más posibilidades de acceder a estudios superiores. En este sentido, pese a que en este estudio no se ha podido analizar el rol de los diferentes actores del entorno educativo como los docentes, auxiliares y directivas y su relación con estas habilidades; se promueve la visión de que para potenciar el desarrollo de los estudiantes más vulnerables se requieren a los mejores profesionales puesto que son los docentes los constructores de aprendizaje (Cueto, Miranda & Vásquez, 2016; Díaz y Ñopo, 2016; Ames & Uccelli, 2008) y estos profesionales deben enseñar habilidades teniendo en cuenta que varios de esos aprendizajes le servirán al estudiante con DI para desarrollar su autonomía, ejercer su ciudadanía y tener un bienestar integral (psicológico, físico y mental). De esta forma, también se refuerzan ideas relacionadas a que la educación recibida a pesar de brindarles habilidades instrumentales, no ha devenido en una educación que dote a personas de herramientas para un desarrollo posterior hacia otros niveles educativos como la educación superior y desemboca muchas veces en que estas personas tengan que pasar al finalizar su paso en la escuela a sus casas o muchas veces no concluir el nivel educativo que alcanzaron. Pocos de ellos acceden a educación superior y en caso lo hagan encuentran dificultades para mantenerse estudiando tal como lo señalo Bregaglio et al. (2014). En definitiva, los esfuerzos para facilitar la inclusión de las personas con DI en el Perú no han sido suficientes porque se requiere contar con mayor repertorio y cantidad de recursos pedagógicos para la inclusión, la necesidad de involucrar a equipos y no solo docentes de manera individual, la mejora de la infraestructura, y mayor información y desarrollo profesional docente para la inclusión tal como lo señala el Consejo Nacional de Educación (CNE, 2017). Desde esta perspectiva, con este estudio se plantea generar políticas públicas con base al paradigma de apoyos (Verdugo & Shalock, 2010; Luckacson y cols, 2002) debido a que como los resultados evidencian existe la necesidad, de brindar apoyos educativos y mayores oportunidades, lo cual está relacionado con las cinco dimensiones del funcionamiento humano. Así pues, tener un marco común implica que todas instancias entiendan e intervengan hacia un mismo objetivo. Así también, acrecentar la investigación en materia de discapacidad con datos estadísticos claros como se planteó en el presente puesto que existe poca literatura y hallazgos sobre Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 153 QUISPE, E. A. A. los logros educativos de las personas con discapacidad intelectual; esto resulta crítico frente a la recomendación de la División de Estadística de las Naciones Unidas y la OMS (2011), los cuales destacan la importancia de tener información fidedigna sobre el número de personas con discapacidad, los tipos de discapacidad, su edad y su ubicación como clave para la elaboración de políticas justas y efectivas. En esta línea de ideas, lo encontrado nos sitúa en un panorama que permite plantearnos desafíos en la educación básica pero sobre todo en Superior. No obstante, no se puede generalizar a nivel nacional estos hallazgos por el tamaño de muestra utilizado por lo que se recomienda para próximos estudios con diseño similar ampliarlo. Lo primero que se encuentra es hacer frente al ingreso a educación superior debido a la falta de cumplimiento de niveles educativos y la escasa intervención por parte del estado para asegurar la educación postsecundaria de las personas con DI no ha permitido que este grupo pueda participar o por lo menos tener la posibilidad de acceso a este nivel. Además, el tipo de educación recibida como la especial no conlleva a que puedan seguir adquiriendo conocimientos en otros niveles, ya sean técnicos o universitarios; tal como lo señaló Florían (2014) son centros diferenciados. Como segundo punto se debe hacer frente a la permanencia en educación superior, como se muestra en este estudio las características de habilidades instrumentales de aprendizaje se sitúan como una barrera de aprendizaje porque están en desventaja educativa en comparación a los demás; esto no permite que se cumplan tres cuestiones claves para el desarrollo de las personas con DI: (a) la capacidad individual resultante del aprendizaje y desarrollo, (b) las oportunidades como el resultado de la influencia del ambiente y las experiencias, y (c) los apoyos y acomodaciones (Zacarías, Saad, Fiorentini & Nava, nd). Así también, el tipo de programa de estudios que se brindan en las distintas instituciones de educación superior no cuenta con lineamientos claros sobre la accesibilidad y los ajustes razonables que se darán para el currículo. De tal modo que como lo menciona Bregaglio et al. (2014) y Vega-Ganoza (2016), los estudiantes con discapacidad encuentran dificultades para mantenerse y concluir sus carreras pese a que existen normativas que buscan garantizar ello en la práctica no se da en su totalidad todavía aunque existen intentos de generar programas inclusivos en la universidad (Usil, 2014). Conclusión En el Perú, la discapacidad intelectual es un tema poco estudiado y por mucho tiempo postergado, lo que ha conllevado a incrementar las brechas entre lo real y lo esperado. Por lo que las condiciones en las que viven las personas con DI necesitan visibilidad para poder establecer políticas justas y consecuentemente mejorar su calidad de vida. Desde este estudio se muestra un panorama sobre las características sociodemográficas y habilidades instrumentales de aprendizaje en el que se plantea hacer frente a distintas problemáticas relacionadas al ingreso, permanencia pero sobre todo a asegurar en cada nivel educativo oportunidades de aprendizaje que le permita a las personas con discapacidad intelectual acceder a niveles educativos superiores. El primer desafío es reducir la brecha entre educación especial y regular. En este estudio se encontró que la mayoría asistió a Educación Especial, solo uno accedió a Educación Superior porque concluyó los niveles anteriores. Esto de algún modo limita el desarrollo de la autonomía de las personas con DI, y por ende, su desarrollo integral. 154 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Desafíos en Educación Superior Dossiê Otro desafío es generar posibilidades para insertarlos al ámbito laboral mediante mayores posibilidades de capacitación en niveles educativos superiores que certifiquen sus capacidades y les brinde iguales ofertas de empleo que otras personas. El tercer desafío está relacionado con la participación a organizaciones como OMAPED; ello no ha devenido en que puedan tener mayores posibilidades para conectarse con instituciones de educación superior puesto que su asistencia a esos espacios está más relacionada con actividades culturales pese a que en teoría dicha instancia tiene como objetivo promover su inclusión social. Un cuarto desafío está relacionado con las condiciones en la que viven, todos de forma dependiente a su familia o cuidadores desembocando muchas veces en que sean estas personas quiénes decidan por ellos y mantengan sus bajas expectativas hacia otras posibilidades educativas. Un quinto desafío está relacionado con sus logros de aprendizaje; en este estudio no se ha podido recoger ello, no obstante, lo recolectado sitúa que la mayoría de participantes con DI pueden comunicarse y transmitir sus necesidades pero, en su mayoría, frente a actividades que implican lectura, escritura y uso de operaciones matemáticas ven limitada su participación y requieren el aprendizaje de apoyos de interpretación comunicativa; Sistemas Aumentativos y Alternativos de Comunicación (Apoyos visuales) y apoyos de manejo y gestión de dinero. Apoyos que debieran considerarse al momento de realizar medidas de accesibilidad y ajustes razonables. El sexto desafío está relacionado con la permanencia en educación, en su mayoría logran acceder pero no culminan sus niveles educativos como el caso de uno de ellos que tuvo la posibilidad de acceder a educación superior pero en el camino tuvo dificultades por la falta de ajustes al programa de estudios que limitaron su continuación. Otro desafío que se torna importante es tener una herramienta capaz de recolectar datos estadísticos con adecuadas propiedades psicométricas e información fiable en aras de conocer la situación educativa y social de las personas con discapacidad intelectual. Finalmente, el desafío principal de esta investigación apunta a visibilizar desde otro paradigma a la discapacidad intelectual y contagiar de ello a otros grupos organizados de docentes, investigadores, programas nacionales y proyectos de cooperación internacional. Parte de ello no solo implica eliminar estereotipos sociales y dejar de utilizar la denominación de “retraso mental” sino conlleva a que los generadores de políticas, investigadores y otros actores busquen cómo mejorar la eficacia de las escuelas, la formación del profesorado, la calidad de la educación, la participación de las familias, el aseguramiento del acceso y permanencia en las instituciones educativas, la generación de los sistemas de apoyos para la inclusión de las personas con DI a la escuela y comunidad. De tal forma que respeten el cumplimiento de las normas peruanas e internacionales como la Convención Internacional sobre los Derechos de las Personas con Discapacidad (Organización de las Naciones Unidas, 2006) porque no se trata de un ideal sino de derechos reconocidos internacionalmente. Referências AMES, P. & UCCELLI, F. Formando futuros maestros: observando las aulas de institutos superiores pedagógicos públicos. En M. Benavides (Ed.), Análisis de programas, procesos y resultados educativos en el Perú. Contribuciones empíricas para el debate, 2008, pp. 131-175. Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 155 QUISPE, E. A. A. APAZA, A., HUAMÁN, J., RAMIREZ, G., SUÁREZ, M. Y SEGOVIA, S. Estudio de casos en niños con Trastornos del Neurodesarrollo. 2016. 1ra edición. ATO, M., LÓPEZ, J. & BENAVENTE, A. Un sistema de clasificación de los diseños de investigación en Psicología. Anales de Psicología, 2013, v. 29, n. 3, pp. 1038-1059. CASAGRANDE, R. & CRUZ, G. Educação Especial e Atendimento Educacional Especializado: Contribuições Científicas de 2000 a 2010. Revista Diálogos e Perspectivas em Educação Especial, 2016, v.3, n.2, pp. 95-110. CONSTANTINO, R. Cartilla para docentes que tengan a cargo estudiantes con discapacidad (Cartillas) (p. 9). Instituto de Democracia y Derechos Humanos PUCP. Lima, Perú. 2013. Disponible en http:// idehpucp.pucp.edu.pe/wp-content/uploads/2013/08/Cartilla.pdf BREGAGLIO, R., CONSTANTINO, R. & CRUZADO, P. Hacia una universidad para todos: propuestas para una PUCP inclusiva. Lima: Pontificia Universidad católica del Perú. 2014. Disponible en: http://idehpucp.pucp.edu.pe/wp-content/uploads/2014/03/Diagn%C3%B3stico-Personas-conDiscapacidad-PUCP.pdf CONSTANTINO, R., BREGAGLIO, R. & CUEVA, S. Al final del salón: Un diagnóstico de la situación de la educación de personas con Discapacidad en el Perú. Lima: Paz y Esperanza. 2016. CUETO, S., MIRANDA, A. & VÁSQUEZ, M. Inequidades en la Educación. Lima: Grupo de Análisis para el Desarrollo. 2016. CONSEJO NACIONAL DE EDUCACIÓN. Calidad de la Educación inclusiva. 2017. Disponible en: https://goo.gl/ceLG8N Defensoría del Pueblo. Situación de la educación especial en el Perú: hacia una educación de calidad. Lima: Defensoría del Pueblo. 2000. DE ASIS ROIG, R. El modelo social de la discapacidad. Críticas y éxito. Papeles el tiempo de los derechos. 2013, v. 1, n. 1. Disponible en: https://goo.gl/KV8M6S DÍAZ, J. J. Y ÑOPO, H. La carrera docente en el Perú. En Investigación para el desarrollo en el Perú. Lima: GRADE. 2016. Disponible en: http://www.grade.org.pe/wp-content/uploads/carreradocente_ JJDHN_35.pdf INSTITUTO NACIONAL DE ESTADÍSTICA E INFORMÁTICA DEL PERÚ (INEI). Encuesta Nacional Especializada sobre Discapacidad. 2012. Disponible en: https://www.inei.gob.pe/media/ MenuRecursivo/publicaciones_digitales/Est/Lib1171/ENEDIS%202012%20-%20COMPLETO.pdf INSTITUTO NACIONAL DE ESTADÍSTICA E INFORMÁTICA [ENAHO]. Caracterización de las condiciones de vida de la población con discapacidad. 2015. Disponible en: https://www.inei.gob.pe/ media/MenuRecursivo/publicaciones_digitales/Est/Lib1 369/index.html Florian, L. La educación especial en la era de la inclusión: ¿El fin de la educación especial o un nuevo comienzo? Revista Latinoamérica de Inclusión Educativa. 2014, v. 7, n. 2, pp. 27-36. Disponible en: http://www.rinace.net/rlei/numeros/vol7-num2/art1.pdf PERU, Ley General De La Persona Con Discapacidad N° 29973. Diario Oficial El Peruano, Lima, Perú, 2014. Disponible en: http://www.leyes.congreso.gob.pe/Documentos/Leyes/29973.pdf PERU, Ley General de Educación N° 28044. Diario Oficial El Peruano, Lima, Perú, 2003. Disponible en: http://www.minedu.gob.pe/p/ley_general_de_educacion_28044.pdf PERU, Ley universitaria N° 30220. Diario Oficial El Peruano, Lima, Perú, 2014. Disponible en: http://www.minedu.gob.pe/reforma-universitaria/pdf/ley_universitaria.pdf 156 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 Desafíos en Educación Superior Dossiê PERU, Ley de institutos y escuelas de Educación Superior y de la Carrera Pública de sus Docentes N° 30512. Diario Oficial El Peruano, Lima, Perú, 2017. Disponible en: http://www.minedu.gob.pe/leyde-institutos/ LUCKASSON, R. et al. Retraso Mental: Definición, clasificación y sistemas de apoyo. 10ª edición. Madrid: American Association on Mental Retardation. 2002. THOMPSON, J. et al. Conceptualizando los apoyos y las necesidades de apyo de las personas con discapacidad Intelectual. Revista Española sobre la Discapacidad Intelectual, 2010, v. 41, n. 1, pp. 7-22. MINISTERIO DE EDUCACIÓN. Currículo Nacional de Educación Básica del Perú. 2016. Disponible en: https://goo.gl/wyCUIG MINEDU. Marco de fundamentación de las pruebas de la evaluación censal de estudiantes. 2016. Disponible en: https://goo.gl/mfK1LX MINISTERIO DE EDUCACIÓN. Educación Básica Especial y Educación Inclusiva-Balances y perspectivas. 2012. Disponible en: https://goo.gl/bDrZJm MONTERO, I. & LEÓN, O. A guide for naming research studies in Psychology [Una guía para denominar estudios de investigación en Psicología]. International Journal of Clinical and Health Psychology, 2007, v. 7, n. 3, pp. 43-52. MUNTANER, J. J. Calidad de vida en la escuela inclusiva. Revista Iberoamericana de Educación, 2013, v. 63, n. 1, pp. 35-49. Disponible en http://www.rieoei.org/rie63a02.pdf NEVES-SILVA, P. & ÁLVAREZ-MARTÍN, E. Estudio descriptivo de las características sociodemográficas de la discapacidad en américa latina. Universidad Rey Juan, España. 2014. Disponible en: DOI: 10.1590/1413-812320141912.17142013 ORGANIZACIÓN DE LAS NACIONES UNIDAS. Convención Internacional de las Personas con Discapacidad. México. 2008. ORGANIZACIÓN DE LAS NACIONES UNIDAS. Objetivos de Desarrollo Sostenible. 2017. Disponible en: https://goo.gl/hELoqD ORGANIZACIÓN MUNDIAL DE LA SALUD. Informe Mundial sobre Discapacidad. 2011. Disponible en: https://goo.gl/snBy9d ORGANIZACIÓN PARA LA COOPERACIÓN Y DESARROLLO ECÓNOMICO. Avanzando hacia una mejor educación para Perú. Lima: OECD. 2016. Disponible en: https://www.oecd.org/dev/ Avanzando-hacia-una-mejor-educacion-en-Peru.pdf ORGANIZACIÓN DE LA NACIONES UNIDAS PARA LA EDUCACIÓN, LA CIENCIA Y LA CULTURA. Conferencia Internacional de educación. La educación inclusiva: el camino hacia el futuro. Ginebra: UNESCO. 2008. Disponible en: http://www.ibe.unesco.org/fileadmin/user_upload/Policy_ Dialogue/48th_ICE/CONFINTED_48-3_Spanish.pdf PORTAL ARAGÓNES DE LA COMUNICACIÓN AUMENTATIVA Y ALTERNATIVA. Sistemas Aumentativos y Alternativos de comunicación (SAAC). 2017. Disponible en: https://goo.gl/svWUPU UNIVERSIDAD SAN IGNACIO DE LOYOLA. Alumno con habilidades diferentes culmina con éxito programa de inclusión educativa en la USIL. 2014. Disponible en: http://english.usil.edu.pe/0/modulos/ NOT/NOT_DetallarNoticia.aspx?PFL=8&NOT=2902 VEGA-GANOZA, J. Educación Superior en Iberoamérica. Informe Nacional: Perú. Lima: Perú. 2016. Disponible en: http://www.cinda.cl/wp-content/uploads/2013/12/Per%C3%BA-Informe-porpa%C3%ADs.pdf Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018 157 QUISPE, E. A. A. VERDUGO, M. & SHALOCK, R. Últimos avances en el enfoque y concepción de las personas con discpacidad intelectual. Revista Española sobre la Discapacidad Intelectual, 2010, v. 41, n. 4, pp. 7-21. VERDUGO et al. Escala ARC-INICO de Evaluación de la autodeterminación. Salamanca: INICO. 2014. TOVAR, T. La década de la Educación Inclusiva 2003-2012 para niños con discapacidad. Lima, Perú. 2013. TOVAR, T. & FERNÁNDEZ, P. Aprender vida: La educación de las personas con Discapacidad. 2005. Disponible en: https://www.conadisperu.gob.pe/observatorio/images/PDFs/Estudio-Educa.pd ZACARÍAS, J., SAAD, E., FIORENTINI, Z. & NAVA, E. Desarrollo de la autodeterminación en jóvenes con discapacidad intelectual que participan en un entorno universitario. Disponible en: https:// goo.gl/Mw3AmW Recebido em: 02 de janeiro 2018 Modificado em: 29 de março de 2018 Aceito em: 30 de junho de 2018 158 Revista Diálogos e Perspectivas em Educação Especial , v. 5, n.1, p. 141-158, Jan.-Jun., 2018
https://openalex.org/W4298080337	https://zenodo.org/record/5912471/files/671-688.pdf	English	null	Isoindolinones (Phthalimidines)	Zenodo (CERN European Organization for Nuclear Research)	2,001	cc-by	11,267	J. Indian Chern. Soc., Vol. 78, October-December 2001, pp. 671-688 J. Indian Chern. Soc., Vol. 78, October-December 2001, pp. 671-688 J. Indian Chern. Soc., Vol. 78, October-December 2001, pp. 671-688 I. Introduction lsoindole 1 is isomeric with indole 2 which comprises a benzene ring fused with a pyrrole nucleus 13. The parent com- pound and the 2-unsubstituted derivatives can tautomerize with the I H isomer, i.e. isoindolenine or (I H-isoindole) 3la,b. lsoindole is much more unstable compared with indole and undergoes rapid oxidation in air to form polymers Ia. lsoindole is thermodynamically more stable than its isoindolenine isomer at room temperature Ia. The next stable reduction state ofisoindole is isoindoline 4. Isoindolinone (phthalimidine) 5 is the more stable derivative of isoindole. Contents Contents I. Introduction 2. Structure of isoindole 3. Naturally occurring isoindole and (phthalimidine) derivatives isoindolinone 4. 5. 6. 7. 7A.I. 7A.2. 7A.3. 7A.4. 7A.5. 7A.6. Biologically important isoindole and isoindolinonc (phthalimidine) derivatives lsoiudolinone in the synthesis of dmgs Isoindolinone in the synthesis of natural products Classical methods Synthesis of isoindolinones (phthalimidines) through Gabriel's procedure Alkylidenc isoindolinones through Grignard procedure Synthesis of isoindolinones through Iithiation procedure 3-Aikylidene isoindolinones through Wittig reaction Synthesis of isoindolinones through Diels-Alder reaction Synthesis of isoindolinones through reduction processes 7A.7. Synthesis of isoindolinones through condensation reactions 7A.8. Synthesis of isoindolinones through rearrangement reac- tions 7 A.9. Synthesis of isoindolinones through photochemical reactions 7 A.IO. Synthesis of isoindolinonrs through miscellaneous processes 7B. Synthesis of isoindolinones through metal mediated reac- tions 8. Synthesis of isoindolinones through palladium-copper cata- lyzed reactions Contents I. Introduction 2. Structure of isoindole 3. Naturally occurring isoindole and (phthalimidine) derivatives isoindolinone 4. 5. 6. 7. 7A.I. 7A.2. 7A.3. 7A.4. 7A.5. 7A.6. Biologically important isoindole and isoindolinonc (phthalimidine) derivatives lsoiudolinone in the synthesis of dmgs Isoindolinone in the synthesis of natural products Classical methods Synthesis of isoindolinones (phthalimidines) through Gabriel's procedure Alkylidenc isoindolinones through Grignard procedure Synthesis of isoindolinones through Iithiation procedure 3-Aikylidene isoindolinones through Wittig reaction Synthesis of isoindolinones through Diels-Alder reaction Synthesis of isoindolinones through reduction processes 7A.7. Synthesis of isoindolinones through condensation reactions 7A.8. Synthesis of isoindolinones through rearrangement reac- tions 7 A.9. Synthesis of isoindolinones through photochemical reactions 7 A.IO. Synthesis of isoindolinonrs through miscellaneous processes 7B. Synthesis of isoindolinones through metal mediated reac- tions 8. Synthesis of isoindolinones through palladium-copper cata- lyzed reactions lsoindole (CNH 4 lsoindoline (JCN 3 3 lsoindoline I H-lsoindole (lsoindolenine) 5 ;.rNH 6~2 7 I 0 s I H-lsoindole (lsoindolenine) NH s lsoindolin-1-one/ Phthalimidine/ 2,3-Dihydroisoindol-1-one Phthalimidine/ 2,3-Dihydroisoindol-1-one To generate isoindole, phthalimidine ring requires nu- cleophilic addition or reduction at the carbonyl group fol- lowed by elimination ofwater1c. The chemical instability of isoindole 1 is well documented lb which prevented its isola- tion and detailed characterization until 1972 and 19733. The preparation of the first isoindole derivative, i.e. N- methylisoindole 6 in 1951 2 and the unsubstituted parent isoindole I in 1972 and 19733 demonstrated that the ring system was stable enough for isolation. aPresent address : Department of Chemistry, Bangladesh University of Engineering and Technology, Dhaka. Bangladesh. tThc relevant literature on isoindolinones has been reviewed through December 2000. aPresent address : Department of Chemistry, Bangladesh University of Engineering and Technology, Dhaka. Bangladesh. tThc relevant literature on isoindolinones has been reviewed through December 2000. Isoindolinones (Phthalimidines)t Nitya G. Kundu, M. Wahab Khana and Rupa Mukhopadhyay Department of Organic Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700 032, India Fax : 91-33-4732805 Manuscript received 30 April 2001 Nitya G. Kundu, M. Wahab Khana and Rupa Mukhopadhyay Nitya G. Kundu, M. Wahab Khana and Rupa Mukhopadhyay Department of Organic Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700 032, India Fax : 91-33-4732805 Manuscript received 30 April 2001 4 50::;;..-o] NH 6 ;:::.., =- 2 7 I lsoindole (JCN 3 I H-lsoindole (lsoindolenine) 5 ;.rNH 6~2 7 I 0 s lsoindolin-1-one/ Phthalimidine/ 2,3-Dihydroisoindol-1-one 4 5~]2 6~N)l 7 H I 2 Indole (CNH 4 lsoindoline 4 5~]2 6~N)l 7 H I 2 Indole 2. Structure of isoindole lsoindole I is a bicyclic I On electron array and com- plies with the Hi.ickel (4n + 2) rule for aromatic stabi- lization4. There have been several calculations of the elec- tronic structure of isoindoles 5- 8. The distribution of charge density around the isoindole nucleus was calculated based 671 J. Indian Chern. Soc., Vol. 78, October-December 200 I October-December 200 I 0 OMe <0 0 OMe- OMe 0 < 0 OMe OMe MeO 0 o (o~~ N _oMe OMe 0 -h A , " /) O~e 0 15 MeO OMe 14 (t) Pa!marune Magallanesme on the LA CO-MO method or the 'frontier electron concept', and the relatively high electron density found at position- I. Therefore, the expectation is that electrophilic substitution on carbon will occur most readily at this position. The semiempirical calculations of Dewar, and Polansky and Derflinger estimate a substantial degree of resonance stabi- lization for isoindole with a value of about 56 k.cal mol- 1 which is significantly larger than the value of pyrrole and is close to that ascribed for indole. Isoindole I should be fa- vored over its tautomer, isoindolenine 3 by about 8 kcal mol- 1 according to a molecular orbital calculation ofVeber and Lwowski9. Theoretical studies by Dewar et a/. 10 are consistent with structure I OMe MeO 0 o (o~~ N _oMe OMe 0 -h A , " /) O~e 0 15 MeO OMe 14 (t) Pa!marune Magallanesme MeO 0 o (o~~ N _oMe OMe 0 -h A , " /) O~e 0 15 MeO OMe 14 (t) Pa!marune Magallanesme 0 15 Magallanesme 0 15 Magallanesme (t) Pa!marune A series of cytostatically active metabolites have been isolated from microorganisms in which a highly substituted hydrogenated isoindolone unit is fused to an II- to 14-mem- 3. Naturally occurring isoindole and isoindolinone (phthalimidine) derivatives The first isoindolobenzazepine alkaloid (±)-chilenine 7 has been found in Berbens empetrifolw Lam. (Berberidaceae) by Fajardo et a/ 11 . Valencia et at. 12 iso- lated nuevamine 8, the first known isoindoloisoquinoline alkaloid and lennoxamine 9, an isoindolobenzazepine struc- turally related to (±)-chilenine from Berberzs darwmu Hook (Berberidaceae ). H I N ~ M .... yo'[ .... H MeO~ 16 NHMe Staurosponne Me 0 I : NMe MeO 0 17 Me* 0 I : NMe MeO 0 17 Staurosponne Staurosponne OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe OMe bered macrocycle. The isolation of the first two substances (cytochalasin B 18 and cytochalasin D 19) of the cytochala- sin series was achieved by Rothweiler and Tamm~ 8 and Aldridge et al. 19 . The cytochalasins are a group of about two dozen structurally related fungal metabolites having a wide range ofbiological activities20a,b_ ~~~ 0 0 18 Cytochalasm B 4. Biologically important isoindole and isoindolinone (phthalimidine) derivatives ~ ~ ~~~ 0 0 18 Cytochalasm B Valencia et a/. 13 also isolated a series of novel isoindolobenzazepines including (±)-13-deoxychilenine 10, pictonamine 11, chileninone 12, (±)-chilenamine 13, and (±)-palmanine 14 from three Chilean Berbens species, namely, B actmacantha Mart. ex Schult., B darwmu Hook and B valdzvzana Phil. The isolation of the first known isoindolobenzazocine alkaloid magallanesine 15 from Berberzs darwznn Hook was also achieved by Valencia et az.I4_ 18 Cytochalasm B 4. Biologically important isoindole and isoindolinone (phthalimidine) derivatives 4. Biologically important isoindole and isoindolinone (phthalimidine) derivatives Heterocyclic compounds containing phthalimidine(2,3- dihydroisoindol-1-one) skeleton have outstanding biologi- cal and physicochemical activities, few of these are being cited here. The I ,2,3,4-tetrahydropyrazino[2, l-a ]isoindol- Staurosporine 16 containing an isoindoli-none moiety was isolated from Saccharothrix Sp. AM-2282 15, and has very interesting biological activities, such as antimicrobiatl6a, hypotensive 16b, cytotoxic 16c activities. It is also an inhibi- tor of protein kinase 16c, and platelet aggregation 16d. ~_)~ rAJ' ~~CI ~N- ~N---o-CHMeCO,H ~C-: N zo o 21 o rJ U~ 22 (ON 2327) O The naturally occurring isoindole(2,5-dimethyl-6- methoxy-4, 7-dihydroisoindole-4, 7-dione) 17 was isolated from sponge Remera Sp. 17 . 672 Kundu et al. : Isoindolinones (Phthalimidines) 6-one 20 was found to lower blood pressure in spontane- ously hypertensive rats21 . A new nonsteroidal antiinflam- matory agent indoprofen 21 was synthesized by Li eta/. 22. Achinami et a/.28 prepared alkylidene isoindolinone de- rivatives 29 as vasodilators which exhibited IC 50 of5.6 f.1M in inhibition of40 rnM K+-induced coronary contraction in vitro and inhibitory effect to the thrornboxan A2 analog (U- 46619)-induced vasoconstriction. The isoindolo[2, 1- a]quinolines of type 30 showed protective effect against N2 induced hypoxia29. Lippmann 30 prepared I ,3-dihydro-3-(2- hydroxy-2-methylpropyi]-2H-isoindoi-I-one 31 and found it to be useful in the treatment of ulcers. A novel promising anxiolytic drug is DN-2327, a non- benzodiazepine isoindoline derivative 22, which has shown in animals to have anxiolytic, taming, antiaggressive and anticonvulsive effects without relevant sedative properties, or signs of dependence23 . DN-2327 showed a higher affin- ity for the BZ 1-GABA receptor in comparison to diazepam or flunitrazepam. 29a, R ~OCH3 29b, R • OH ~H,CMe20H ~NH 11 0 ~ 0 30 29a, R ~OCH3 29b, R • OH ~H,CMe20H ~NH 11 0 ~ 0 30 Bellioti et a/.24 discovered a series of isoindolinone de- rivatives 23 having affinity for the dopamine 0 4 receptor. Compounds containing substituents in 3- and 4-positions of the phenyl ring have the highest 0 4 receptor affinity. ~{~)-·· ¥,-{X . X • Hor Me Y~NorCH Ar ~ C6Hs e1c The study of cytotoxic effects31 of isoindol-1-one de- rivatives 32 and 33 on leukemia P388 cells showed that nearly half of the compounds inhibited uridine incorpora- tion, but only three compounds affected incorporation of thymidine and L-valine. The cytotoxic effects of the com- pounds could be related not only to the types and position of the substituents but to stereoisomerism as well. 4. Biologically important isoindole and isoindolinone (phthalimidine) derivatives The synthesis of a group of cyclic amide analogs, e.g. 24 and. 25 of 4-(2'-methoxyphenyl)-1-[2' -[N-(2" -pyridyl)- p-iodobenzamido ]ethyl]piperazine (p-MPPI 26) as 5-HT lA receptor ligands was achieved by Zhuang et a/.25. Some of these compounds displayed very high in vitro binding atlin- ity for 5-HT lA receptors, comparable to or exceeding that of the parent compound 26. ~ Jl 0 CH,O P·MPPI26 6. Isoindolinones in the synthesis of natural products Vedejs et a/.42 reported the synthesis ofisoindolone 49 having part of the structure of cytochalasin 0 19. in spontaneously hypertensive rats. in spontaneously hypertensive rats. (DDATHF) 37a, in which the glutamate moiety is tied back to the benzoyl ring through an isoindolinone moiety. The preliminary biological evaluation of 37b revealed that it to be an excellent inhibitor of human CCRF-CEM lympho- blastic leukemic cells and a non-competitive inhibitor of mammalian glycinamide ribonucleotide formyltransferase. Scheme 2. Reagents : (i) MeOH, l N HC1, Jci2, Pd/C, (ii) THF, 1 M diborane in THF. Hexahydroisoindole 38 was found to be a potent herbi- cida!35. Kim and Ryu36 developed a synthetic method for the herbicidal2,3-dihydro-3-methylene-2-substituted-phe- nyl-l H-isoindol-1-one derivatives. Compound 39 showed good herbicidal activity against various weed species. Abramovitch et a/.40 obtained the ring expanded 2- benzazepin-1-one 46 through the reaction of3-substituted- ( I H)-1-isoindolones 45 (Scheme 3). THF, or CH3CN ~O - - - - - NE1, CH3C:CNEt2 CH, R ~ Ph . R ..OEt R 46 THF, or CH3CN ~O - - - - - NE1, CH3C:CNEt2 CH, R ~ Ph . R ..OEt R 46 P·MPPI26 Mappicine ketone (MPK) 34, an analog of mappic;ine 35, a naturally derived alkaloid isolated from Mapiafoetida Miers. (Olacaceae)32a has been identified as an antiviral lead compound with selective activity against herpes viruses HSV-1; HSV-2 and human cytomegalovirus (HCMV). MPK appears t-o be herpes virus selective in that it does not in- hibit other DNA 'Or RNA viruses32b. Norman et a/.26 prepared a series of isoindolinone de- rivatives 27 bridged to 4-( I ,2-benzisothiazol-3-yl)-1- piperazinyl moiety with a variety of different bridging units. Norman et a/.26 prepared a series of isoindolinone de- rivatives 27 bridged to 4-( I ,2-benzisothiazol-3-yl)-1- piperazinyl moiety with a variety of different bridging units. The compounds were evaluated il1 vitro at dopamine 0 2 and serotonin 5-HT Ia and 5-HT 2 receptors and in vivo for their ability to antagonize the apomorphine-induced climb- ing response in mice and found as potential antipsychotic agents. A four-carbon spacer provided optimal activity within the homologous series. The compounds were evaluated il1 vitro at dopamine 0 2 and serotonin 5-HT Ia and 5-HT 2 receptors and in vivo for their ability to antagonize the apomorphine-induced climb- ing response in mice and found as potential antipsychotic agents. A four-carbon spacer provided optimal activity within the homologous series. Dihydrothiazoloisoindolones 36 were found to be po- tent non-nucleosidic HIV reverse transcriptase inhibitors33. N~s rAI'N-Z-Nf""\N_j,_ A ~ \_/ 'LJ 0 27 Z; ·(CH2)2, -(CH2) 1 - etc. R-<d ~) 36 0 R • H. CHJ 3-( 4-Acetoxyphenyl )methylene-2-(2-diethylamino )ethyl- 2,3-dihydro-1 H-isoindol-1-one 28 was found to exhibit lo- cal anesthe.tic activity superior to that of procaine27. Taylor eta/.34 synthesizeda confonnationally constrained analog 37b of 5,1 0-dideaza-5,6,7 ,8-tetrahydrofol ic acid 673 J. Indian Chern. Soc., Vol. 78, October-December 200 I in spontaneously hypertensive rats. Scheme I A synthesis of pyrazino[2, 1-a ]isoindole derivative 44 via isoindolinone 43 was developed by Ferland et a/. 21 (Scheme 2). The analog, e.g I ,2,3,4-tetrahydropyrazino[2, 1-a]isoindol-6-one 20 was found to lower blood pressure 5. Isoindolinone in the synthesis of drugs lsoindolinones (phthalimidines) have been widely used as intermediates for the synthesis of various important drugs. Shihunine, a popular Chinese drug, sold in Hong Kong, was isolated about 30 years ago by lnubushi et a/.38 from Dendrobium iohonense as a phthalide pyrrolidine alkaloid which they named shihunine as it was a component ofShi- Hu. Breuer and Zbaida39 synthesized shihunine 41 from the lactam, spiro[( 1-methylpyrrolidine )-2,3' -(2' -methyl-1'- isoindolinone)] 42 (Scheme 1). C.~:~~ ~-~CH=CH-CH~Ac C02Et 0 N I Ph COPh 49 Marsili et a/.43 synthesized isoindolobenzazepinone de- rivative 51 which are structurally related to the alkaloid pictonamine ll, from isoindolinone 50 (Scheme 5). rtf.i:-"' 41 0 Scheme I Ph Br . Ph~ ~N-CHP>"" _M_eo_H_. h_v_ ~NJ so 0 51 0 Ph Br . Ph~ ~N-CHP>"" _M_eo_H_. h_v_ ~NJ so 0 51 0 Scheme 3 Recently, Pigeon and Decroix41 developed a new ap- proach to isoindolo[2, 1-b ]benzodiazepines 48 through an intramolecular cyclization of a N-substituted-3-hydroxy- isoindolinone 47 (Scheme 4). Other important derivatives of isoindole are the metal complexes 40 of phthalocyanines, used as dyes and pigments37a,b. rAr-Z I ..<: i.S002, CH,CI2; ii.H,O. iii. RNH2 0~"" ~ iv. PTSA. toluene, reftux 47 0 O OH R ~ H, Me, Bu Scheme 4 0 c<N-Q-a Nffi2CH,Ol1 38 ·~:rn 39 7A.I. Synthesis of isoindolinones (phthalimidines) through Gabriel's procedure In 1885 Gabriel developed48 a general method for the synthesis ofylidenephthalides by condensation of phthalic anhydride 59 with arylacetic acid in the presence of sodium or potassium acetate at 230-250° (a type of Perkin reac- tion). He also reported that benzylidene phthalide 60 on treatment with ammonia in alcohol at I 00° for 8-10 h yielded benzylidene phthalimidine 61 (Scheme 9). ~0 + CH2CO,H Ph Ph ~ CH,CONa ~0 ~H /EtOH ©Q 59 0 ~ 230-2500C • ~0 1oOoc • NH 2h 6() 0 8-12h 61 0 ~0 + CH2CO,H Ph Ph ~ CH,CONa ~0 ~H /EtOH ©Q 59 0 ~ 230-2500C • ~0 1oOoc • NH 2h 6() 0 8-12h 61 0 Scheme 6 Scheme 6 COOB CH, EIOH. NaOH(aq) N I 0 frcH, renux. 2h 53 C<iOH Acetone. Et3wo --= CH, CH ethylchloro ~ N-CH I J ' N-yN....,_ lblmate 0 N u CH, ' 0 0 ~ Scheme 7 Scheme 7 Othman et a/. 46 also developed a synthesis of benzo[a]thieno[g]indolizinones 56, polycyclic compounds analogous to the alkaloid nuevamine 8, via an intramolecu- lar a-amidoalkylation reaction ofisoindolinone derivative 55 with thionyl chloride (Scheme 8a). Scheme I 0 Perjessy et a/.50 reported the synthesis of3-arylmethylene phthalimidines 66 through a Gabriel modification of Perkin condensation where phthalimide 64 was reacted with the corresponding arylacetic acids 65 (Scheme II). ~N-0)~~ 0 N R OH R 55 R R H. a. Ph. Y= H. a 56 s @J r:s V o-a s . b Scheme 8a «~ 64 0 CH2C02H + h CH,C01K y 260-2700C 65 X 2-3h ~X ©:$" 66 0 X= 4-N02, 4-0CH,, 3-(;H,, 4-F, 3-0CH, 4-1, 4-(;1, 4-Br, 3-F, l·l, 4-NH, Scheme II Scheme 8b 0 j i. Acetylchloride, CH;.CI2 ii Zncl:z in 1HF iii. PI • CIH;.. Acetic acid Scheme 6 COOB CH, EIOH. NaOH(aq) N I 0 frcH, renux. 2h 53 C<iOH Acetone. Et3wo --= CH, CH ethylchloro ~ N-CH I J ' N-yN....,_ lblmate 0 N u CH, ' 0 0 ~ Scheme 7 0 j i. Acetylchloride, CH;.CI2 ii Zncl:z in 1HF iii. PI • CIH;.. Acetic acid Scheme 6 7A. Classical methods Srheme 5 From isoindolinone derivatives 52 Napolitano et al.44 synthesized the isoindolobenzazepine alkaloids lennoxamine 674 Kundu et al. : lsoindolinones (Phthalimidines) Kundu et al. : lsoindolinones (Phthalimidines) OMe «N'-© 57 ° ~ 58 0 Scheme 8b 9 and chilenamine 13 (Scheme 6). Recently, Othman and Decroix45 reported a synthesis of benzo[ a ]pyrrolo[2,3-j]indolizine 54, a new heterocyclic ana- log of chilenine alkaloid 7, from the isoindolinone deriva- tive 53 (Scheme 7). Scheme 9 Honzl49 reported the synthesis of methylene isoindo- linone 63 by heating aniline and phthalylidene acetic acid 62 in acetic acid (Scheme I 0). 53 a-tCO,H NH2 ©Qo +6 62 0 rn, ~H ~N-\Q) 63 0 Scheme 7 Scheme 14 Recently, Achinami eta! 28 reported the synthesis of a series ofalkylideneisoindolinone derivatives 77 by follow- ing the above Grignard procedure (Scheme 15). 17: 1) Ar • Arr • Ph; b) Ar • 4-BrC,H., Arr ... Ph; c) Ar"" 2-Br-4-MeC1H1, Ar1 • Pfl; d) Ar • Ph, Ar1 • 3.4-(0CH,),C.H,. 7A.3. Synthesis of isoindolinones through lithiation procedures 7A.3. Synthesis of isoindolinones through lithiation procedures Watanabe et al. 54 reported a synthesis of3-substituted- phthalimidines 80 from benzanilide 78 through lithiation with n-butyllithium and subsequent reaction with benzonitrile (Scheme 16). 1-n-Buli \"'_ Ph N=01Ph ~ THF-hexane r()r'C-NH ~ r()"0NR CONHR 2·C,H,CN, ~OJNHR ~- 78 J· "H;~'"'"' 79 80 0 Scheme 18 ~ Alkylmagnesium Mt:~ ~R, ~N-R, halide ~N--R, H,so, ~N--R, no ?Jo . uo R, (R, )R, - H (CO,EI or H) H or CO,El ;Me(CO,EI.,.. H) H or CO,E; Ph (Me or H) H or Me ; Ph (Me) Me; Ph (H) H; Me (Me or H) H or Me; Me (Me) Me; Me (H) H; H (Me or H) H ot Me. An effective reaction of Schiff bases with certain functionalized organolithium reagents was described by Bradsher and Hunt57 for the synthesis of I ,2-diarylisoin- dolines 87 (Scheme 19). R, (R, )R, - H (CO,EI or H) H or CO,El ;Me(CO,EI.,.. H) H or CO,E; Ph (Me or H) H or Me ; Ph (Me) Me; Ph (H) H; Me (Me or H) H or Me; Me (Me) Me; Me (H) H; H (Me or H) H ot Me. R, (R, )R, - H (CO,EI or H) H or CO,El ;Me(CO,EI.,.. H) H or CO,E; Ph (Me or H) H or Me ; Ph (Me) Me; Ph (H) H; Me (Me or H) H or Me; Me (Me) Me; Me (H) H; H (Me or H) H ot Me. rn,a . + Ar-N=CHAr1 [i -©Q-Ar 82 B1 Ar1 Scheme 16 Parham et at. 55 prepared N-phenylphthalimidine 83 by selective halogen-lithium exchange in bromophenyl- alkylhalides 81 with n-butyllithium followed by treatment with phenyl isocyanate (Scheme 17). rArrn,a rna ~ -~= 1)Y 1 ~ ©riNPh Br ~u 81 82 D 0 rArrn,a rna ~ -~= 1)Y 1 ~ ©riNPh Br ~u 81 82 D 0 Scheme 13 Scheme 12 A series of benzylidenes phthalimidines 71 were pre- pared starting from phthalic anhydride27 59. A typical ex- ample is given in Scheme 13. Scheme 17 Alkylidene isoindolinones (alkylidenephthalimidines) were synthesized by the action of a Grignard reagent on a N-substituted-phthalimide and subsequent dehydration 52. Ang and Halton 53 applied the Grignard procedure reported by Heidenbluth eta/. 52 to prepare 2-substituted 3-alkyl-3- hydroxyphthalimidines 73 through the reaction of alkyl magnesium halide in ether on phthalimide 72 in benzene at 55° for 3 h. Dehydration of the alcohols 73 in concentrated sulfuric acid afforded the alkylidenephthalimidines 74 in good yields (Scheme 14). Parham and Jones 56 also synthesized N-phenylphtha- Jimidine derivative 86 from 2-bromobenzonitrile 84 and n- butyllithium at -78° followed by the reaction with phenyl isocyanate (Scheme 18). CN NH fAr n-BuLi ©(CN ~ ~ ~' THF-Hexane NPh Br -780C u ~ ~ u 0 Scheme II Scheme 8a Scartoni et al. 51 reported the synthesis of 3-benzyl-3- hydroxyphthalimidin-2-ylacetic acid 67 which yielded 3- ben-zylidene phthalimidin-2-ylacetic acid 68. This was subsequently converted to isoindolobenzazepine 69 (Scheme 12). Recently, Heaney and Shuhaibar47 reported a synthesis of the alkaloid nuevamine skeleton, e.g. isoindoloisoquino- line derivative 58 from an a-methoxyisoindolone deriva- tive 57 (Scheme 8b ). 675 J. Indian Chern. Soc., Vol. 78, October-December 2001 J. Indian Chern. Soc., Vol. 78, October-December 2001 Ph ,.glycine NaOH, EtOH t:.. 30 mln 60 0 OOC, 12h l· 2N HCI Scheme 23 Reagents and conditions : (i) KHMDS (2.2 cquiv.), THF, -78° toRT 2 h; (ii) R3CHO, THF, -30°; (iii) -30° toRT 0.5 h; (iv) aqueous HCI. Scheme 25. Reagents and conditions : (i) KHMDS (2.2 cquiv.), THF, -78° toRT 2 h; (ii) R3CHO, THF, -30°; (iii) -30° toRT 0.5 h; (iv) aqueous HCI. Scheme 19 0 0 ,(0!,)1~ R R,a ·~' ~:~ R~OI ~ NH- ~ NR,~ R. NR1 K sak, DMF TifF, rl ?5 o Rellu• 76 o n o 0 0 ,(0!,)1~ R R,a ·~' ~:~ R~OI ~ NH- ~ NR,~ R. NR1 K sak, DMF TifF, rl ?5 o Rellu• 76 o n o R1 -·(wb.,iluled)alkyi,Ph, PhSO,; R,• H, alko•r. OH. halo, (subsbluted)alkyl, acyl; R,, R. = H, alkyl,aryl, alk0.1ty, OH; n- Q, I, 2. Scheme 15 0 0 ,(0!,)1~ R R,a ·~' ~:~ R~OI ~ NH- ~ NR,~ R. NR1 K sak, DMF TifF, rl ?5 o Rellu• 76 o n o Hendi et al. 58 reported a Parham-type cyclization59 in which N-acyl-2-bromobenzamides 88 participated in metal- halogen exchange with n-BuLi to form N-acyl-2-lithioben- zamides 89. The resulting ortho-lithio intermediates 89 then underwent cyclization to afford 3-alkylidenephthalimidines 90 (Scheme 20). R1 -·(wb.,iluled)alkyi,Ph, PhSO,; R,• H, alko•r. OH. halo, (subsbluted)alkyl, acyl; R,, R. = H, alkyl,aryl, alk0.1ty, OH; n- Q, I, 2. Scheme 15 676 Kundu et al. : lsoindolinones (Phthalimidines) 0 0 0 0 ~ ~R2 1.NaH,251': o. ~ II R ~Br~' __ 2_._,._-B_u_u_. _-7_e_~,;•QJ__Li ('~ 2 ~ ~ + H30 -H,O Scheme 20 ~NR 99 0 I. LOA, THF -780C, IS min 2. e+. THF -7SOC ton. 2h 3. H3o+ ~ BHJTHF ~ ~NR Ren~x. 12h ~NR 100 0 101 R =Me. Y;H40Me(p) E= Me, Bn 7A.4. 3-Alkylidene isoindolinones through Wittig reac- tion Scheme 22. Reagents and conditions : (i) I. BuLi, THF, -78 to oo; 2. methyl benzoate, in THF, -78° to RT; (ii) diethylmalonate, methanesulfonic acid, (Ac0h0, I 00°; (iii) 20% HCl solution, heated upto boiling; (iv) I. oxalylchloride, C2H4Cl2, reflux, 0.2 h; 2. AlC13, C2H4Cl2, reflux, 2 h. Scheme 22. Reagents and conditions : (i) I. BuLi, THF, -78 to oo; 2. methyl benzoate, in THF, -78° to RT; (ii) diethylmalonate, methanesulfonic acid, (Ac0h0, I 00°; (iii) 20% HCl solution, heated upto boiling; (iv) I. oxalylchloride, C2H4Cl2, reflux, 0.2 h; 2. AlC13, C2H4Cl2, reflux, 2 h. Scheme 22. Reagents and conditions : (i) I. BuLi, THF, -78 to oo; 2. methyl benzoate, in THF, -78° to RT; (ii) diethylmalonate, methanesulfonic acid, (Ac0h0, I 00°; (iii) 20% HCl solution, heated upto boiling; (iv) I. oxalylchloride, C2H4Cl2, reflux, 0.2 h; 2. AlC13, C2H4Cl2, reflux, 2 h. Scheme 23 A series of3-substituted-isoindolinones 23 having affini- ty for the dopamine 0 4 receptor was prepared by Belliotti et a/. 24 starting from isoindolinoneacetic acid 102 (Scheme 24). Scheme 20 Campbell et a/.60 developed a synthesis ofisoindolinones 94 through a lithiation procedure (Scheme 21 ). A< 0 0 c) 0 ~NH ~ rarNMe K,~o,, rarNMe ~CD,H ~ CH1CN.820C ~ 1\ y- N, CH OTs N\._/YAr Ar • 3,4-dimcthylphenyl,2-naphthyl, 4-methylphcnyl, 4,!i-dunethyl-2-thiazolyl, phenyl, 4-melhyl-2-pyridyl, 4-chlorophc:nyl and 4-methoxyphenyl Scheme 24. Reagents: (i) NaH, DMF; (ii) Mel; (iii) NaBH 4, MeOH; (iv) p-TsCl, TEA in CH2Cl2. ~~li·~~: 93 a, R I = Et, R2 ~ Ph b,R]=Pr,R2~Ph c,RJ =Pb,R2=Ph d,RJ ~Bu.R2~Pr Scheme 21. Reagents and conditions: (i) BF3.0Et2, -78°; (ii) PhLi,- 78°; (iii) H20, wann to 50°, overnight. Scheme 24. Reagents: (i) NaH, DMF; (ii) Mel; (iii) NaBH 4, MeOH; (iv) p-TsCl, TEA in CH2Cl2. Couture eta/. 63 developed the synthesis of a series of3- (alkyl and aryl)methylene-2,3-dihydro-1 H-isoindol-1-one derivatives. Treatment of phosphorylated amides 104 with KHMDS (2.2 equiv.) in THF at -78° followed by quench- ing with selected aromatic and aliphatic caraboxaldehydes affords a mixture of E- and Z-isomers of2-alkyl-3-(aryl and alkyl)methylene-2,3-dihydro-1 H-isoindol-1-ones 105 (Scheme 25). Scheme 21. Reagents and conditions: (i) BF3.0Et2, -78°; (ii) PhLi,- 78°; (iii) H20, wann to 50°, overnight. 0 Rz~N-R, i-iv ~~ llt ... Ph- x ti'> 104 0 Ph R,~ vO N-R~ 105 R, R, = 4-0MeC,H,CH2, Me. R,-H,OPr. X= 2- or 3- Cl. Br, F Y -CH, N. R, = Bn, Ph. i-Pr. Scheme 25. Reagents and conditions : (i) KHMDS (2.2 cquiv.), THF, -78° toRT 2 h; (ii) R3CHO, THF, -30°; (iii) -30° toRT 0.5 h; (iv) aqueous HCI. 0 Rz~N-R, i-iv ~~ llt ... Ph- x ti'> 104 0 Ph R,~ vO N-R~ 105 R, C,H,CH2, Me. X= 2- or 3- Cl. Br, F 0 Rz~N-R, i-iv ~~ llt ... Ph- x ti'> 104 0 Ph R,~ vO N-R~ 105 R, C,H,CH2, Me. X= 2- or 3- Cl Br F Recently, Jozwiak and Sadokierska61 reported the syn- thesis ofhydroxyazaisoindolinones and their successful con- version to azaisoindoleacetic acids as effective precursors for the corresponding 4b-phenyldihydroazaisoindol [2, 1- a]quinolinediones 98 (Scheme 22). R, = 4-0MeC,H,CH2, Me. R, = 4-0MeC,H,CH2, Me. R,-H,OPr. Y -CH, N. R, = Bn, Ph. i-Pr. R, = Bn, Ph. i-Pr. 0 ""oH ""oooH~ ©yM!Ph ~©(N-1'1>1-~Ph ~lQli-Y 950 960 970 098 Scheme 25. Scheme 32 Milewska et al. 70 reported a synthesis of N-methyl- isoindolinones 131 by regioselective reduction of72 using Lawesson's reagent (LR) to yield monothioimides 130 fol- lowed by desulfurisation of 130 with Raney nickel (Scheme 33). J. Indian Chern. Soc., Vol. 78, October-December 200 I Mali and Yeola65 described a novel synthesis of N-sub- stituted-3-carbethoxymethylphthalimidines 110 from sec- ondary benzamides through a combination oflithiation and Wittig reaction (Scheme 27). An isoindolinone ring was synthesised through a series of steps which started with a Diels-Aider reaction of the Danishefsky diene 119 with 4,4-diethoxybut-2-ynal 120 by Taylor eta!. 34 (Scheme 30). f: I 010 HCJ CHO ""+II- I I 0 OMe CHO [ OMe f: TM~ CH(OEI), TMS~CH(OEI), I~CHO I 19 JlO Ill Ill diethyl L-glutamate 0~~~~00,~ DMF ~ CO,EI 123 0 Scheme 30 R, 0 I. n-C4ifgLii R, 0 ~NHR;, TMIDArrHF,~~Colls~aco,c,lls ~ 2. DMF/ether, ~ ~ ISO.IIIO"C 107 O"C tort 108 H Oil 107-110 a-e) R 1= H. f) R 1 = OCH3; a) R2 = CH3, b) R2 =Ph, c) R C6H40Me-o, d) R2 = C6H40Me-m, e) R2 = C6H40Me-p, f) R2 = P Scheme 30 Scheme 27 Epsztajn et a/. 66 described the synthesis of the 3-hy- droxy-5-azaisoindolin-1-ones 112 and their conversion to 2-aryl-2,3-dihydro-3-oxo-1 H-azaisoindole-1-acetic acids 113 and subsequent cyclization to the corresponding dihydroazaisoindolo[2, 1-a ]quinolinediones 114 (Scheme 28). 7A.6. Synthesis of isoindolinones through reduction pro- cesses 107-110 a-e) R 1= H. f) R 1 = OCH3; a) R2 = CH3, b) R2 =Ph, c) R2 = C6H40Me-o, d) R2 = C6H40Me-m, e) R2 = C6H40Me-p, f) R2 = Ph 107-110 a-e) R 1= H. f) R 1 = OCH3; a) R2 = CH3, b) R2 =Ph, c) R2 = C6H40Me-o, d) R2 = C6H40Me-m, e) R2 = C6H40Me-p, f) R2 = Ph Brewster et al. 68 reduced N-alkylphthalimides 72 to phthalimidines 125 under Clemmensen condition (Scheme 31). tion Flitsch and Peters 64 synthesized 3-ethoxycarbonyl- methyleneisoindolin-1-ones (phthalimidines) 106 from the reaction of phthalimides 72 with Wittig reagents at 140° (Scheme 26). Couture et a/. 62 recently reported that 2-alkyl-2,3- dihydroisoindol-1-ones 99 are easily and regioselectively deprotonated at the 3-position ofthe heterocyclic nucleus with LOA, thus providing ready access to 3-substituted and/ or functionalized isoindolinones 100 and consequently to the corresponding isoindolines 101 (Scheme 23). (C.H5)]P=CH<XJOC,H, 140"C R•H,CHJ Scheme 26 677 Scheme 31 ©Jylg-. "--©II---o-·-4 Ill 0 112 0 ©5~·-'-~' Scheme 28. Reagents and conditions :(a) i. BuLi, THF, -78 to 0°; ii. DMF, -78 to 0°; (b) Wittig reagent, 170°, 0.7 h; (c) i. oxalyl chloride, CI2C2H4; ii. AICI3, CI2C2H4. Ciganek69 prepared a mixture of N-4-pentenylisoindole 127, isoindoline 128 and isoindolinone 129 by the reduc- tion of N-4-pentenylphthalimide 126 with sodium bis(2- methoxyethoxy)aluminium hydride; N-4-pentenylisoindole 127 was the major product (Scheme 32). Scheme 32 Scheme 28. Reagents and conditions :(a) i. BuLi, THF, -78 to 0°; ii. DMF, -78 to 0°; (b) Wittig reagent, 170°, 0.7 h; (c) i. oxalyl chloride, CI2C2H4; ii. AICI3, CI2C2H4. Scheme 28. Reagents and conditions :(a) i. BuLi, THF, -78 to 0°; ii. DMF, -78 to 0°; (b) Wittig reagent, 170°, 0.7 h; (c) i. oxalyl chloride, CI2C2H4; ii. AICI3, CI2C2H4. 7A.7. Synthesis of isoindolinones through condensation reactions using phthalaldehyde as starting material 7A.7. Synthesis of isoindolinones through condensation reactions using phthalaldehyde as starting material 27 .-(' 27 .-(' Scheme 34 oReagents : (i) Ethanolamine or 3-amino-1-propanol, 205- 210, (u) lrans-4-aminocyclohexanol hydrochloride, KzC03, toluene, H20, reflux; (iii) SOCI2, toluene, 60°; (iv) Et3N, CH2Cl2, MsCJ, 0°; (v) 3-( 1-piperazinyl)-1 ,2-benzisothiazole, Et3N, CH3CN, reflux Scheme 34 oReagents : (i) Ethanolamine or 3-amino-1-propanol, 205- 210, (u) lrans-4-aminocyclohexanol hydrochloride, KzC03, toluene, H20, reflux; (iii) SOCI2, toluene, 60°; (iv) Et3N, CH2Cl2, MsCJ, 0°; (v) 3-( 1-piperazinyl)-1 ,2-benzisothiazole, Et3N, CH3CN, reflux Yamamoto showed 73 that an equimolar mixture of phe- nyl isocyanate and phthalaldehyde 151 on heating at 170° for 4 h afforded N-phenylphthalimidine 152 (Scheme 39). rAY'QlO ~CHO 151 ~ r()'I'N-Ar + CO; 1700C ~ 152 0 Scheme 39 rAY'QlO ~CHO 151 ~ r()'I'N-Ar + CO; 1700C ~ 152 0 Scheme 39 hydroxyisoindolinone 136 and a coupled product 137 was performed71 by a low-valent-titanium reagent in an aromatic solvent or in the absence of solvent (Scheme 35). ©ISNH 64 0 I. TiCI4-Zn IIOOC, 20h Neat or xylene 2. liP!. K2(l)J Scheme 35 DoMinh et al. 74 described a reaction ofphthalaldehyde 151 with ammonia and amines in Me2S04 (Scheme 40). Major products were phthalimidines 153 and 3-(2- cyanophenyl)isoquinoline 154. OH ~ Q-10 RNH2 ©< · H20 rQq 0 '<: ____. · NR __... NR + OfO Me2S04 N lSI 152 OH 153 0 154 O R- H, M<, Ph, 2,6-Me,C6H4 Scheme 40 OH ~ Q-10 RNH2 ©< · H20 rQq 0 '<: ____. · NR __... NR + OfO Me2S04 N lSI 152 OH 153 0 154 O R- H, M<, Ph, 2,6-Me,C6H4 Scheme 40 The coupled product 137 (23%) and phthalimidine 138 (44%) were obtained in the absence of solvent. In p-xylene, three monomolecular reductive products were produced (phthalimidine 138 24%, 3-hydroxyisoindolinone 136 19% and phthalide 132 II%); no coupling product was isolated. Grigg et a/.75 investigated the reaction of o-phthalal- dehyde 151 with a-amino acids 155 in boiling acetic acid and observed that a rapid reaction (5-1 0 min) occurred to give the N-substituted-isoindolin-1-ones J 57 (Scheme 41) in good yield. Luzzio and Zacherl 72 described the synthesis of isoindolinone (Scheme 36). « O Aluminium H OH Thiophene! H S4.H.s NPMB amalgam ~NPMB dichloromethane . Scheme 41 Scheme 41 A new route to isoindolinone derivatives was developed by Nefkens and Zwanenburg76 as outlined in Scheme 42. 7A.5. Synthesis of isoindolinones through Diels-Aider reaction An efficient approach for the synthesis of hexahydro- isoindolones 118 was reported by Gutierrez et a/.67 (Scheme 29). 0 ~ LR, Toluene ~t<--01, /:, 72 0 Scheme 33 ~oa );R II .. HN~R 6a ,., - II .. R-H,Ph Scheme 29 '~' C_HJ )k R lilt Scheme 33 A series of isoindolinone derivatives26 27a-c attached to the 4-(1 ,2-benzisothiazol-3-yl)-1-piperazinyl moiety were prepared by a variety of methods (Scheme 34). R-H,Ph Scheme 29 A method for the reduction of phthalimide to the 3- 678 Kundu et al. : Isoindolinones (Phthalimidines) 133 0 0 ~ _....OH iiioriv ~ 0 X ~N-Z -~J.-./~-z...- 134 a~c X C1 C1 OMs 135 •-c 27 .-(' Scheme 34 oReagents : (i) Ethanolamine or 3-amino-1-propanol, 205- 210, (u) lrans-4-aminocyclohexanol hydrochloride, KzC03, toluene, H20, reflux; (iii) SOCI2, toluene, 60°; (iv) Et3N, CH2Cl2, MsCJ, 0°; (v) 3-( 1-piperazinyl)-1 ,2-benzisothiazole, Et3N, CH3CN, reflux 0 ~ COOH ISO OMe r\ Sp'N N N..~ OMe _ 147 \_./ ·--v -NH, 0 I. HAc, MeOH tQr~r·\~ r--.. 0 2. NaCNBH3 \___/ '-' N 24a 0 Scheme 38 0 _....OH iiioriv ~ 0 X -~J.-./~-z...- 135 •-c 27 .-(' 135 •-c Scheme 38 7A.7. Synthesis of isoindolinones through condensation reactions using phthalaldehyde as starting material ~PMB THF/Water ~· p·toluene sulfonic ac1d ~· 139 0 140 0 141 0 R H ~NPMB 144 0 m-chloroperbenzoici ac1d, methylene Raney R SO,C,H, chloride H SO,C,H, mckel ~ I NaH, THF/DMSO ~ ~ ~NPMB orTHFIHMPA ~NPMB 143 0 2 CH]I, . 142 0 PMB "'p-methoxybenzyl or~ ph.ene1hylbromrdc, etc R • CH], CHzCH2C6il5. CH2CH2COCH3, CH2CHzCOOCHJ Scheme 36 ~· rAY-'010 H,~~H ~CHO in boiling ace1ic acid 151 [©r}:t-C~H]-- ©GN-f~CO,H oo,~ •no R1 • H, Ph, Me, iPr, CH1Ph, CH10H Scheme 41 Scheme 44 Recently, Allin et a/. 83 have studied the reaction of a- amino alcohols 176 with 2-formylbenzoic acid 175 as part of an on-going investigation into the preparation and re- activity of the isoindolinone ring system. They have re- ported the extremely high diastereoselectivity of the reac- tion to produce the tricyclic y-lactam products 177 (Scheme 45) through condensation of a-amino alcohols 176 with 2-formylbenzoic acid 175. This class of heterocycles can act as N-acyliminium ion precursors in the synthesis of substituted-isoindolinone derivatives 179 (Scheme 46). The treatment ofthe (+/-)-valinol derived tricyclic lactam 178 with titanium tetrachloride at room temperature in dichloromethane prior to addition of allyl trimethylsilane afforded the desired substituted-isoindolinone 179 in 96% isolated yield. 00 0 184 ©:Qo O ~H~N .. ) (i) SOOz or PAC. (QJ_ (ii)MeCH 185 Scheme 48 Scheme 48 7A.9. Synthesis of isoindolinones through photochemi- cal reaction Scheme 43 Allin et a/.82 subsequently discovered that the reaction of a-amino acids I 71 and amino alcohol172 with o-phthal- aldehyde 151 produced the corresponding phthalimidines 173 and 174, respectively, in good yields (Scheme 44). H,NOC CN R~£~a ,.Jo ~ 0 ·~ ~ 181 Ill! R .. H,OCHl Scheme 47. Reagents and conditions : (i) KOH (3.2 equiv.), ethanol, reflux; (ii) KOH (3.2 equiv.), ethyleneglycol or methoxyethanol, reflux. H,NOC CN R~£~a ,.Jo ~ 0 ·~ ~ 181 Ill! R .. H,OCHl Rt,_~.t .0 H,..-·~· I( rAI'N-(R, ~~~ (YOlO+ l7l 0 ffil\.--f,H ~ )-OH alcohol ~010 CH,CN ~ CO,H 174 0 R, 172 151 ""' 173 0 171 : L-valine, L-alanine, L-phenylalanine, L-serine, DL-Ieucine, DL- isoleucine; J 73 R = CH(CH3h, CI-1 3, CH 2Ph, CH 20H, CH2CH(CH3)2, CHCH3CH2CH3 respectively; 172 : 2-amino-3- methylbutanol, 2-amino-1-butanol, 2-amino-1-phenylpropanol, 2- amino-1-propanol; 174 : R 1/R2 = CH(CH3)z/H, CH2CH/H, CI-13/ Ph, CH3/H respectively. Scheme 47. Reagents and conditions : (i) KOH (3.2 equiv.), ethanol, reflux; (ii) KOH (3.2 equiv.), ethyleneglycol or methoxyethanol, reflux. Scheme 47. Reagents and conditions : (i) KOH (3.2 equiv.), ethanol, reflux; (ii) KOH (3.2 equiv.), ethyleneglycol or methoxyethanol, reflux. Heaney and Shuhaibar47 developed a route to 3-methoxy- 2-phenylethylisoindolin-1-one 57 which functions as a pre- cursor to synthetically useful acyliminium ions (Scheme 48). 7A.8. Synthesis of isoindolinones through rearrange- ment reactions A new route to phenylmethyleneisoindolinones was reported by Guillaume! eta/. 84 . The alkaline hydrolysis of 2-(2-benzofuranyl)benzonitriles I 80 led to Z-2,3- dihydro-3 -(2-hydroxyphenyl methylene)-( I H)-isoindolin- 1-ones 183 by rearrangement (Scheme 4 7). The forma- tion of the amide 181 or the acid 182 depended on the solvent used. 7A.9. Synthesis of isoindolinones through photochemi- cal reaction Freccero et al. 85 reported a synthesis of 3-hydroxy- isoindolinones 187 by photochemical reactions of phthalimides 72 (Scheme 49). H CO,H 0 H,X--011 ©: ~~0 176 CHO ~ PhCH,, /:::,. 17~ 011 ·HzO ~:f·,~· 177• 177b 176:(+/-) vahnol, (S} phenylglycmol. (R) phenyi&Jycinol. (S} ph<nylalaninol. (R) phenylalaninol; 177a : R • Ph, CH,Ph; I T7b: R ~Ph, CH,Ph; Scheme 45 + PhCHXY 186 X 3 H, SiMeJ, CHPh2 v~ H,"" S h 49 ~y ~~NK 187 0 R • H. Me + PhCHXY 186 X 3 H, SiMeJ, CHPh2 v~ H,"" ~y ~~NK 187 0 R • H. Me X 3 H, SiMeJ, CHPh2 v~ H,"" Scheme 45 Scheme 36 Zhuang et at.25 developed the synthesis of a group of isoindol-1-ones 24, cyclic amide analogs of p-MPPI 26 as 5-HT 1 A receptor I igands (Schemes 3 7 and 3 8). CICH2CN ©t~ ~ ~OMe NH NH N 1"-\ ~/',...........,. 1 ~I46 ~OMe CN 147 \.._/ + Q...,/"~NH, 0 MeOH. H 147 '--1 OMe -0 O l.p-ToOH,PnH •©C r\N"""vk COOMc 2. NaSH, N\.._/ 0 149 M 0 Scheme 37 "O~ CICH2CN ©t~ ~ ~OMe NH N NH N 1"-\ ~/',...........,. 1 \..___/ ~I46 ~OMe CN 147 \.._/ + Q...,/"~NH, 0 MeOH. H 147 '--1 OMe 0 -0 O l.p-ToOH,PnH •©C r\N"""vk COOH COOMc 2. NaSH, N\.._/ 0 149 M 0 Scheme 37 Scheme -t2 Scheme -t2 Scheme 37 679 J. Indian Chern. Soc., Vol. 78, October-December 200 I ~TMS 200C Scheme 46 Simons, Stobaugh and Takahashi, reported the quanti- tative formation of 2H-isoindole derivatives I 68-170 (Scheme 43) 77- 81 . Scheme 46 7A.8. Synthesis of isoindolinones through rearrange- ment reactions Kundu et al : Isoindolinones (Phthalimidines) ( ) 194 ... 194, 195, 196 X Y a NO;! H b NO;! CH3 c OMe NO;! d Cl H H NQz Scheme 52 ©(~ 197 ° _An_iline- rry-t'--fi Reflux ~"~ 6h 0 198 R- H,CH, Scheme 53 Weidner-Wells et a/. 86 showed that N-methylphthali- mide (NMP) 72 undergoes a photolysis reaction with phenylcyclopropane in acetonitrile to give two isomeric spirotetrahydrofuranyllactams ( 188 and 189). When there- action was carried out in the nucleophilic solvent MeOH, hydroxylactam 190 and 1-methoxy-3-phenylpropane were obtained (Scheme 50). ©rSNMe C>--© hv ICH3CN 12 0 PC (Phcnylcyclo propane) l bv IMeOH ~OMe" ~NMe - Ph~OMe 190 0 Scheme 50 lished by Whelton and Huitric90 involving the reaction of the nitrolactone 199 with bases. The reaction of 199 with aniline led to the phthalimidine 200 (Scheme 54). Griesbeck et al. 87 photochemically transformed N- phthaloylcysteine derivative 191 by elimination, decarboxy- lation and via electron transfer cyclization to isoindolone derivatives 192 (Scheme 51). The excited singlets were prone to elimination and y-H abstractions to yield compound 193 whereas the triplets on cyclization afforded I ,3,4, I Ob- tetrahydro-1 Ob-hydroxy-6H-[ I ,4 ]thiazino[3,4-a ]isoindol-6- ones 192. ~~ 6 N~ ~0 ~ ©G----© 199 0 reflux, 18h 0 200 Scheme 54 Scheme 54 ~N--(.:HSMe ""\ 0 _....c:.hv'---+- ~HO N S)R ~ OX>H Solvent 191 ° 192 ° R•H.COOH Scheme 51 ~({X)H ~z 0 193 Camilleri and Cassola91 reported a reaction in which the 2-(1-aminoethyl)benzoate anion cyclised to the lactam 3-methylisoindolinone 203 at high pH. 2-(1-Aminoethyl)- benzoic acid 201 (prepared by catalytic hydrogenation of the oxime of 2-acetyl benzoic acid) was cyclised in pres- ence of 0.1 M NaOH to yield lactam 203. They proposed the mechanism shown in Scheme 55. Scheme 51 Scheme 49 680 Kundu et al : Isoindolinones (Phthalimidines) 78. Synthesis of isoindolinones through metal mediated reactions Several methods of synthesis of isoindolinones (phthalimidines) under metal catalysis were discussed by Wilkinson et a/. 94 where various starting compounds in- cluding Schiff bases, aromatic nitriles, ketoximes, phenylhydrazones, semicarbazones and azines have been carbonylated. A series ofisoindolinones 220 were produced by the carbonylation of Schiff bases 219 of aromatic and alipathic amines95•96 . The reaction was catalyzed by metal carbonyls like pentacarbonyl iron and octacarbonyl dicobalt. The best results were obtained from cobalt-catalyzed carabonylation which was performed in the temperature range 200-230° under a CO pressure of 100-200 bar (Scheme 60). The reaction was also catalyzed by Pd 11-ac- etate under milder condition (Scheme 61). Polar solvents were found to inhibit the reaction. Scheme 56 Scheme 56 ©{u:;· '"~ ~: R • C(;Hs, 3-CFJC61i4. CH2Q;Hs 201l 0 Scheme 57 ~ R F'l.(COl.J I (Co)z(C0)8 __rAy-\ X NR, +CO - X~NR, 219 R~oR2 ~ alky~aryl 220 o X • CHO, N0 2, NH2 Scheme 60 3-carboxylate was reported by Othman and Decroix~ 5 (Scheme 58). rn, I ()()()fl a; aJOEt H,C-.yN'-"a;,NH,rAY-< ©(~a ~©(~a Um ~N\-~' a; 209 210 aJOEt 0 l_)- ] rAr--Z BBr1 00C 1h 212 ~NH CH2C1 2 ' 213 0 Scheme 58 a; aJOEt (~a ~ 7A.IO. Synthesis of isoindolinones through miscella- neous processes 7A.IO. Synthesis of isoindolinones through miscella- neous processes Rowe et a/. 88 reported that the treatment of benzo-2'- nitrophenylhydrazide-2-,8-acrylic acid 194a with boiling di- lute Na2C03 solution, H20 or nitrobenzene afforded 2-(2'- nitrophenylamino )isoindolinone-3-acetic acid 195a. When isoindolinone derivative 195a was boiled with acetic anhy- dride with or without addition of pyridine or retluxed with toluene in the presence of phosphorus trichloride, 1 mol- ecule of water was eliminated, yielding 2,5-diketo-3-(2'- nitrophenyl)isoindolinopyrazolidocoline 196a which was readily hydrolyzed to 195a preferably by acids (Scheme 52). Similarly, other compounds 195b-e and 196b-e were syn- thesized. Scheme 55 Scheme 55 Kimura et a/.92 synthesized a new macrocyclic polyamine 207 containing a phthalimidine ring (Scheme 56). A general procedure for 2,3-dihydro-3-methylene-2-sub- stituted-1 H-isoindol-1-ones (3-methylenephthalimidines) 208 was developed by Howe and Shelton93 (Scheme 57). Pojer et al. 89 prepared methyleneisoindolinone 198 by retluxing acetylbenzoic acid or its methyl ester 197 with aniline (Scheme 53). An improved method for the synthesis ofphthalimidine- An interesting approach to isoindolinone 200 was pub- 681 J. Indian Chern. Soc., Vol. 78, October-December 200 I & Br Br OMe 204 Ts....NH ~Ts ( 2NaH T -N T DMF s l__,.~-J s Ts 205 (~) ~ l___..~J 'H H 207 Scheme 56 TsC,&:d~Ts 0 OMe Ts-N Ts l__,.~-J Ts 206 _j (i) 48% HBr I AcOH (ii) anion exchange ©{u:;· '"~ ~: R • C(;Hs, 3-CFJC61i4. CH2Q;Hs 201l 0 Scheme 57 Scheme 58 Phthalimidine derivative 223 was synthesized from the carbonylation of aromatic nitrile 222 under a pressure of CO containing variable amounts of hydrogen in the pres- ence of a cobalt catalyst; the best yields were observed when a small quantity of pyridine was added to the reaction97 (Scheme 62). A facile synthetic method for the herbicidal2,3-dihydro- 3-methylene-2-aryl-l H-isoindol-1-one system 218a-c was described by Kim and Ryu36 (Scheme 59). ::«:~d R, 0 K2CO, H H CJR.. ~1f?(-1Y-o R.X R,~:~ 217 R, 0 F 216 218 a-c 214,216,218 RJ Rz R3 a H H F b H H Cl H OCH:J H H CH:J CH3 e CJ H Cl 217, J4X: a) propargyl bromide, b) ethyl iodide, c) 2-chloropropionitrile 2PhCN +CO+ Hz 222 Coz(CO)s, Py 2300C, 270 bar ~N-CH,Ph + NH:J 223 0 222 Scheme 61 Scheme 58 Scheme 62 3-Substituted-phthalimidines 225 were prepared from the carbonylation of aromatic ketoximes 224 in the pres- ence of hydrogen and the catalyst octacarbonyl dicobalt. The reaction was more selective with diary! ketoximes than with arylalkyl ketoximes97 (Scheme 63). 216 214,216,218 RJ Rz R3 H H F 214,216,218 RJ Rz R3 a H H F b H H Cl H OCH:J H H CH:J CH3 e CJ H Cl R ~NOH + 00 + H Coz(CO,)s X~ 2 2500C, 285 bar 224 X=H,R=Ph X=benro,R=Me R ~NH +H20 225 0 X=H,R=Ph X=benro,R=Me e CJ H Cl 217, J4X: a) propargyl bromide, b) ethyl iodide, c) 2-chloropropionitrile Scheme 69 Mori et al. 100 developed a synthesis of benzolactams, i.e. isoindolinone, isoquinoline and benzazepinone deriva- tives by palladium catalyzed amidation. In a typical example, N-benzyl-o-bromobenzylamine 242 was treated with n- Bu3N, PPh3 and a catalytic amount of Pd(OAch in pres- ence of CO at I 00° for 26 h to afford N-benzylisoindolin-1- one 243 (Scheme 70). Scheme 65 Phthalimidine 233 was also prepared from acetophenone- !, 1-dimethylhydrazone 232 through carbonylation reaction mediated by Pdll-acetate under much milder conditions98 (Scheme 66). CH, )j1NHR Pd(OAc), (0.02 equiv) «H \ PPh, (0.'64 equi.<) 0 ~ N-R Br n-BuJN, CO 242 243 0 CH, )j1NHR Pd(OAc), (0.02 equiv) «H \ PPh, (0.'64 equi.<) 0 ~ N-R Br n-BuJN, CO 242 243 0 Scheme 66 3-Phenylphthalimidines 227 and 235 were obtained when benzophenone semicarbazone 234 reacted with CO in the presence of cobalt catalyst at higher temperature (245°)97-98 (Scheme 67). Scheme 68 Thompson and Heck99 described the carbonylation of the palladium acetate complex 237 to yield 3-acetoxy-2- phenylphthalimidine 240. The carbonylation of237 in pres- ence of a nucleophile, e.g. aniline, led to 3-anilino-2-phe- nylphthalimidines 241 rather than the acetate (Scheme 69). ~ ~" 227 0 N-NHPb • CO C CO), 190-2000C Ph 250bar ~ 226 2J0-2400C ~N-cO-NHPh 228 0 Scheme 64 In the case of unsymmetrically substituted diary I ketone phenyl hydrazones, e.g. 229, a mixture of isomeric phthalimidines 230 and 231 were obtained (Scheme 65). Scheme 70 ~N-NMe2 232 p~ CO. 1.5 bar IOOOC CH2 «N-NMe2 2J] 0 ~N-NMe2 232 p~ CO. 1.5 bar IOOOC CH2 «N-NMe2 2J] 0 p~ CO. 1.5 bar IOOOC CH2 «N-NMe2 2J] 0 ·However, o-bromo-N-( o' -bromobenzyl)pheny !ethyl- amine 244, was cyclized to yield isoindolinone 245 and isoquinolinone 246 in a 2.6 : I ratio under the same condi- tions (Scheme 71). R = CH2Ph, CONH:H2Ph It was reported that benzophenone phenyl hydrazones 226 gave rise to 3-phenylphthalimidine 227 in the presence of octacarbonyl di-cobalt without hydrogen gas at 190-200°, but at higher temperature (230-240°), 3-phenyi-N-(N-phe- nylcarbamoyl)phthalimidine 228 was obtained via insertion of CO into the nitrogen-nitrogen bond96·98 (Scheme 64). Scheme 71 Ph >=N-NHCONH, CD2(CO\_ Ph C0,24SOC 234 ©4" 227 0 Ph • ~NCONHCHPh, 235 0 Brunet eta!. 101 developed cobalt carbonyl catalyzed carbonylation of aryl halides 247 under phase-transfer-ca- talysis conditions for the synthesis of benzolactam 249 (Scheme 72). ©? { 234 Scheme 63 Scheme 63 Scheme 59 682 Kundu et al. : Isoindolinones (Phthalimidines) Kundu et al. : Isoindolinones (Phthalimidines) The same reaction applied to the o-m ethyl ether of ben- zophenone oxime resulted in 3-phenylphthalimidines. Scheme 73 ©e CHO + Br Z67 ¥ ~) 0 Scheme 78. Reagents and conditions : (i) 2-Bromobenzaldehyde (2 mmol), ethylenediamine (3 mmol), (PPh 3)zPdCI 2 (0.03 mmol), PPh3 (0.08 mmol), Et3N (5 mmol), CH3CN (10 ml), CO ( 13 atm), 100°, 20 h. ©e CHO + Br Z67 ¥ ~) 0 Larock et al. 103 carried palladium-promoted olefination reaction ofthallated N-methylbenzamide 254 with methyl acrylate to yield an isoindolinone 255 (Scheme 74). rATCONHCH, H,C=CHCO,CH, ~0 0- HPdX ~0 ~ U,PdClo(l<qllN) ~CH, -- 0 NCH1 25-4 WO,CCF,), CH,CN . H CHCO,CH, NaH (U eqliY) I 255 CHCO,CH SOOC, S b PdX J Scheme 74 Scheme 78. Reagents and conditions : (i) 2-Bromobenzaldehyde (2 mmol), ethylenediamine (3 mmol), (PPh 3)zPdCI 2 (0.03 mmol), PPh3 (0.08 mmol), Et3N (5 mmol), CH3CN (10 ml), CO ( 13 atm), 100°, 20 h. Scheme 74 Scheme 67 ~--H ©?.;{n ~b'R -o NR 248 11 ~o(CO)• 249 ~ 0 Scheme 72 Benzaldehyde azine 236 also gave different N-substi- tuted-phthalimidines 138 and 223 under the reaction with CO in the presence of cobalt catalyst (Scheme 68). Scheme 72 683 Dikshit 1 dimethoxyph carboxylic ac 2-benzopyran 3-arylmethyl # 250a,b 0 2S0a b Larock et reaction oftha acrylate to yie rATCONHCH, ~ 25-4 WO,CCF,), Grigg et a for the prepar 260-262 (Sch ,y N---.._ Ph 256 0 Bums et at. cyclization-a enamide 263 u Dikshit 102 repo dimethoxyphenyl)-3 carboxylic acid 250a 2-benzopyran-4-carb 3-arylmethylene-1 H # 250a,b 0 R 2S0a R"' OM~ bR•H Larock et al. 103 c reaction ofthallated acrylate to yield an i rATCONHCH, H,C=CHC ~ U,PdClo(l< 25-4 WO,CCF,), CH,CN NaH (U eq SOOC, S b Grigg et al. 104 re for the preparation o 260-262 (Scheme 75 ,y_ N---.._ Ph 256 0 Bums et at. 105 dev cyclization-anion ca J. Indian Chem. Soc., Vol. 78, October-December 2001 Scheme 76. Reagents and conditions : (i) Pd(OAch. PPh3, Bu4NC!, sodium fonnate, DMF or CH3CN. 80°/l 00°,6/16 h. Dikshit 102 reported a novel reaction of 3-(3 ',4'- dimethoxyphenyl)-3,4-dihydro-1-oxo-1 H-2-benzopyran-4- carboxylic acid 250a and 3-phenyl-3,4-dihydro-1-oxo-1 H- 2-benzopyran-4-carboxylic acid 250b to yield 2,3-dihydro- 3-arylmethylene-1 H-isoindol-1-ones 253a,b (Scheme 73). Scheme 76. Reagents and conditions : (i) Pd(OAch. PPh3, Bu4NC!, sodium fonnate, DMF or CH3CN. 80°/l 00°,6/16 h. # R co,H R~o 0 l-&£)H, rctiu., 0,~ _ 4h COX 'P(OPh), 250a,b 0 151•,b R 2S0a R"' OM~ lSI a X,. Nl bR•H bX•N 2.5Ja R=CHJ b R•H R R Cho et a/. 106 synthesised isoindolinones 266 via carbo- nylative cyclization of2-(2-bromophenyl)-2-oxazolines 265 by a bimetallic palladium nickel catalyst system with car- bon monoxide (3 atm) in alcohol as shown in Scheme 77. + Scheme 77. Reagents and conditions : PdC12(PPh3)2, NiCJ2, 6H20, PPh3, Et3N, 3 atm, I 00°. + Scheme 77. Reagents and conditions : PdC12(PPh3)2, NiCJ2, 6H20, PPh3, Et3N, 3 atm, I 00°. A synthesis of tricyclic isoindolinone 269 by the palla- dium-catalyzed carbonylative cyclization of2-bromobenzal- dehyde 267 with diamines 268 was carried out by Cho et al. 107 (Scheme 78). 2.5Ja R=CHJ b R•H lSI a X,. Nl bX•N 8. Synthesis of isoindolinones through palladium-cop- per catalyzed reactions Grigg et al. 104 reported intramolecular Heck reactions for the preparation of isoindolinone derivatives 257, 258, 260-262 (Scheme 75). Recently, Kundu and Khan 108 have reported a new strat- egy for the regio- and stereoselective synthesis ofisoindolin- 1-ones 273,274 through the palladium-catalyzed condensa- tion of o-iodobenzamides 270 with terminal alkynes 271 and subsequent cyclization (Scheme 79). In most of the cases a number of (Z)-3-alkyl(aryl)ideneisoindolin-1~ones 274 were formed without any formation ofthe corresponding isoquinolinones. Only in a few cases, a mixture ofisoindo- linones 274 and isoquinolinones 275 were obtained. The configuration of3-alkyl(aryl)ideneisoindolinones as Z was established by NMR studies and confirmed by X-ray dif- fraction studies 109. ,y_ N---.._ Ph 256 0 Scheme 75 Scheme 75 Conclusion lsoindolinones are an important group of heterocyclic compounds which occur naturally and are also of biological importance. There have been many procedures available for the synthesis of this group of compounds. However, no re- cent review of this field is available in literature 1. In this report we have attempted to bring together the information available on the natural occurrence, biological importance and synthetic methodologies avai !able for this group of com- pounds. We hope this will be of interest to many organic and medicinal chemists active in this area and stimulate more activities in this promising area. Kundu et al. 110 have also described another novel and facile procedure to 3-alkylideneisoindolinones under mild conditions. In this case, a palladium-catalyzed reaction was followed by Friedel-Crafts acylation and simultaneous cy- clization to yield 3-alkylideneisoindolinones in good to ex- cellent yields. N-Aikyl- or N-aryl-2-iodobenzamides 270 underwent facile reaction with trimethyl silyl acetylene, in the presence of(PPh3hPdCI 2 and Cui at room temperature to yield 2-(trimethylsilyl)ethynylbenzamides 278 in excel- lent yields (76-89%). 2-Trimethylsilyl ethynylbenzamides were then subjected to Friedel-Crafts reaction with acid chlo- rides or acetic anhydride to afford the 3-acylmethyl- ideneisoindolinones 279 in good yields (Scheme 81 ). Scheme 75 Bums et at. 105 developed a palladium-catalyzed tandem cyclization-anion capture process. They reported that enamide 263 undergoes regiospecific 5-exo trig-biscycliza- tion (DMF, 80°) followed by hydride capture to afford the tricyclic products 264 (Scheme 76). Hydrogenation of 3-methyleneisoindolinones 273 af- forded 3-methylisoindolinones 276 (Scheme 80). Hydroge- nation of3-arylideneisoindolinones 274 gave 3-benzyliso- indolinones 277. 684 Kundu et al. : Isoindolinones (Phthalimidines) Rl'rA-r'J HC~Rl <.)..8J..._• _..NHR.l (PPh,),J'dC Z70 g Cui, El,N, D H orPd(OA<h 17J 0 H UC\ K,Co, DMF -R2 Rl'rA-r'J HC~Rl R<.)..8J..._• _..NHR.l (PPh,),J'dC Z70 g Cui, El,N, D H orPd(OA<h 17J 0 H UC\ K,Co, DMF -R2 lyzed heteroannulation reaction of 2-iodobenzamides 270 with acetylenic carbinols 280 containing a terminal acety- lenic group with a carbinol group next to it to afford 3- acylmethyl-isoindolinones 282 instead of the expected 3- alkylidene-isoindolinone 281 (Scheme 82). ~C(OH)HAr ~N-R 281 0 · I (W~~Q C<l, flJN + IIC...,C-CH(OH)Ar:.::::.:.=:::.:......j NHR ,280 DM:vJ\800C CH2COAr 270 O CondtiooA ~ l (PPh;J)2PdC1z, Cui ~N-R ElJN, DMF, rt, 24h 282 o Cordi:ion B l emc-c~_o_H>_h __ ~~~~~~~--~ NaDEll ElOH, rcfux. 4 h _ NHR J 0 Scheme 82 R1 = Ph, C6H40Me(p), C 6H4CI(m), !-naphthyl, CH201-l, SiMe3• n-hexyne, H, CM~OH, 2,4-dimethoxypyrimidin-5-yl y , , , ypy y R2 = H, CH3, Ph, CH2Ph, C61-1 40Me(p), C 6H4Me(p), C6H4CI(m) R3 = H,OCH3 . R3 = H,OCH3 R4 = 1-1, N02, OCH3, 01-1 R4 = 1-1, N02, OCH3, 01-1 Scheme 79 R': H, Ph, C,H,OMe(p), CMe20H Pd-C(IO%) Hz, EtOH or AcOH n, 24h R' = H, CH2Ph, Ph, C.H.Me(p), C.H.OMe(p). C,H,CI(m) Scheme 80 MRI ~N-R2 276,277 0 Scheme 79 Acknowledgement Thanks are due to Dr. John Boulton of the University of East Anglia for reading the manuscript and helpful sugges- tions. One of the authors (R.M.) is grateful to C.S.I.R., New Delhi, for a Fellowship. 2. G. Wittig, H. Tenhaeff, W. Schoch and G. Koenig,Juslus Lie bigs Ann. Chem., 1951, 572, I. 4. E. ~Hickel, Z. Phys., 1931, 70, 204. I. (a) D. Barton and W. D. Ollis, "Comprehensive Organic Chem- istry", Pergamon, Oxford, 1979, Vol. 4, p. 477; (b) J.D. White and M. F. Mann, Adv. Heterocyc/. Chem., 1969, 10, 113; (c) A. R. Katritzky and C. W. Rees, "Comprehensive Heterocyclic Chemistry", Pergamon, Oxford, 1984, Vol. 4, pp. 35, 36. 3. (a) R. Bonnett and R. F. C. Brown,J. Chem. Soc., Chem. Commun., 1972, 393; (b) E. J. Bomstein, D. E. Remy and J. E. Shields, J. Chem. Soc., Chem. Commun., 1972, 1149; (c) R. Bonnett, R. F. C. Brown and R. G. Smith,J. Chem. Soc., Perkin 7i·ans. I. 1973, 1432. J. Indian Chern. Soc., Vol. 78, October-December 2001 27. S. P. A. Laboratori Baldacci, Jap. Pat. JP 5946268/1984 Chern Abstr., 1984, 101, 54922) 5. M. J. S. Dewar, Trans. Faraday Soc., 1946, 42, 764. 27. S. P. A 6. H. C. Longuet-Higgins and C. A. Coulson, Trans. Faraday Soc., 1947, 43, 87. 28. K. Achinami, N. Ashizawa and F. Kobayasui, Jap. Pat. JP 3133955/1991 (Chem. Abstr., 1991, 115, 255977). 7. K. Fukui, T. Yonezawa, C. Nagata and H. Shingu, J. Chem. Phys., 1954, 22, 1433. 29. Y. Ishihara, Y. Kiyoto and G. Goto, Chem. Pharm. Bull., 1990, 38, 3024. 8. 0. E. Polansky and G. Derflinger, Monatsh. Chem., 1961, 92, 1114. 30. W. Lippmann, US Pat. 4267189/1981 ((:hem. Abstr., 1981,95, 61988) 9. D. F. Veber and W Lwowski, J. Am. Chem. Soc., 1964, 86, 4152. 31. J. Fuska, A. F.uskova and B. Proksa, Zb. Pr. Chemickotechnol Fak, SVST, 1979-1981 (pub. 1986), 285 (Chem Abstr., 1987, 106, 95582) 10. M. J. S. Dewar, A. J. Harget, N. Trinajstic and S. D. Worley, Tetrahedron, 1970, 26. 4505; see also, E'Chacko, J. Bornstein and D. J. Sardella, Tetrahedron Le/t., 1977, 1095. 32. (a) T R. Govindachari, K. R. Ravindranath and N. Viswanathan, J. Chem. Soc., Perkin Trans. I, 1974, 1215; (b) I. Pendrak, S. Barney, R. Wittrock, D. M. Lambert and W D. Kingsbury,]. Org. Chem., 1994, 59, 2623. 11. V. Fajardo, V. Elango, B. K. Cassels and M. Shamma, Tetrahe- dron Lett., 1982, 23, 39. 12. E. Valencia, A. J. Freyer. M. Shamma and V. Fajardo, Tetrahe- dron Lett., 1984, 25, 599. 33. E. De Clercq, J. Med. Chern., 1995, 38, 2491. 34. E. C. Taylor, P. Zhou, L. D. Jennings, Z. Mao, B. Hu and J.-0. Jun, Tetrahedron Lett., 1997, 38, 521. 13. E. Valencia, I. Weiss, S. Firdous, A. J. Freyer and M. Shamma, Tetrahedron, 1984, 40, 3957. 35. Takeda Chemical Industries Ltd., Jap. Pat. JP 58174362 (Chern Abstr., 1984, 100, 51452). 14. E. Valencia, V Fajardo, A. J. Freyer and M. Shamma, Tetrahe- dron Lett., 1985, 26, 993. 36. J. N. Kim and E. K. Ryu, Synth. Commun., 1996, 26, 67. 15. S. Omura, Y. lwai, A. Hirano, A. Nakagawa, J. Awaya, H. Tsuchiya, Y. Takahashi and R. Masuma,J. Antibiot .. 1977, 30, 275. 37. (a) R. 0. C. Norman and J. M. Coxon, "Principles of Organic Synthesis", 3rd. ed., Chapman & Hall, London, 1993, p. 691; (b) D. Barton and W. References R2cOCior fl gR2 IIC==C-SiMe1 ~""'C-SiMe3 (CII1CO),O ~ NllRI (PPh,),PdCI,, Cui 0 NHRI AICI,. 'qCHCHq 0 N-Rl 0 Et3N, DNF, r,t., 24h 178 0 OOC-n, 1-2 h or 3-4 h 279 0 R' ~II, Me, C,H,Mc(p), C.H40Mc(p), C.H.CI(m) R' • Me, Et, n-Pr, Ph, c.H.Me(p), c.fl.OMe(p) Scheme 81 Scheme 81 We have found that some of the isoindolinones 279 were in (E)-configuration while the rest were in the (Z)-configu- ration. The configuration of the 3-acylmethylidene- isoindol inones was con tinned by X-ray crystallography Ill. Kundu et a/. 112 have also reported the palladium-cata- 685 J. Indian Chern. Soc., Vol. 78, October-December 2001 J. Indian Chern. Soc., Vol. 78, October-December 2001 D: Ollis, "Comprehensive Organic Chem- istry", Pergamon, 1979, Vol. 4, p. 483. 16. (a) Y. Takahashi, M. Shinose, Y. Iwai and S. Omura, "II Oth. Annual Meeting of the Pharmaceutical Society of Japan", 1990, Abst. 3, p. 125; (b) S. Omura, Y. Iwai and A. Hirano, Jap. Pat. JP 7873501 (Chem. Abstr., 1978,89, 178086); (c) T. Tamaoki, H. Nomoto, I. Takahashi, Y. Kato, M. Morimoto and F. Tomita, Biochem. Biophys. Res. Commun., 1986, 135, 397; (d) S. Oka, M. Kodama, H. Takeda, N. Tomizuka and H. Suzuki,Agric. Bioi. Chern., 1986, 50, 2723. 38. Y. lnubushi, Y. Tsuda, T. Konita and S. Matsumoto, Chem. Pharm. Bull. Tokyo, 1964, 12, 749; 1968, 16, 1014 (Chern. Abstr., 1964, 61, 9541; 1968,69, 87256). 39. E. Breuer and S. Zbaida, Tetrahedron, 1975, 31, 499. 40. R. A. Abnm10vitch, I. Shinkai, B. J. Mavunkel, K. M. More, S. O'Connor, G. H. Ooi, W. T. Pennington, P. C. Srinivasan and J. R Stowers, Tetrahedron, 1996, 52, 3339. 17. J. M. Frincke and D. J. Faulkner,]_ Am. Chern. Soc., 1982, 104. 265. 41. P. Pigeon and B. Decroix, Tetrahedron Lett., 1997, 38, 2985. I 8. W Rothweiler and Ch. Tamm, Experientia, 1966, 22, 750. 42. E. Vedejs, J. B. Campbell, Jr., R. C, Gadwood, J. D. Rodgers, 19. D. C. Aldridge, J. 1. Armstrong, R. N. Speake and W. B. Turner, J. Chern. Soc., Chern. Commun., 1967, 26. K. L. Spear andY. Watanabe, f. Org. Chern., 1982, 47, 1534. 43. A. Marsili, V. Scartoni, I. Morelli and P. Pierangeli,J. Chem. Soc., Perkin Trans. I, 1977, 959. 20. (a) M. Binder and C. Tamm,Angew. Chern., Int. Ed. Engl., 1973, 12, 370; (b) "Frontiers of Biology", ed. S. W. Tanenbaum, North- Holland, New York, 1980, Vol. 46. 44. E. Napolitano, G. Spinelli, R Fiaschi and A. Marsili,]. Chern. Soc., Perkin Trans. 1, 1986, 785. 21. J. -M. Ferland, C. A. Demers on and L. G. Humber,Can.J. Chern., 1985,63, 361. 45. M. Othman and B. Decroix, Synth. Commun., 1996, 2803. 46. M. Othman, P. Pigeon and B. Decroix, Tetrahedron, 1997, 53, 2495. 22. S. Li, X. Wong, H. Guo and L. Chern, Yiyano Gongye, 1985, 16, 543 (Chern. Abstr., 1986, I 056378). 47. H. Heaney and K. F. Shuhaibar,Synlett, 1995, 47. 23. M. Linden, D. Hadler and S. Hofmann, Human Psychophar- macol., 1997, 12, 445. 48. S. Gabriel, Chern. Ber., 1885, 18, 3470. 49. J. Honzl, Collect. C=ech. Chem. Kundu et al. : lsoindolinones (Phthalimidines) Cornrnun .. 1978, 807. 65. R. S. Mali and S. N. Yeola, Synthesis, 1986, 755. 66. J. Epsztajn, R. Grzelak and A. Jozwiak, Synthesis, 1996, 1212. 92. E. Kimura, Y. Yoshiyama, M. Shionoya and M. Shiro,J. Org. Chern., 1990, 55, 764. 67. A. J. Gutierrez, K. J. Shea and J. J. Svoboda,!. Org. Chern., 1989, 54, 4335. 93. R. K. Howe and B. R. Shelton,}. Org. Chern., 1990, 55, 4603. 68. J. H. Brewster, A. M. Fusco, L. E. Carosino and B. G. Com1an, J. Org. Chern., 1963, 28, 498. 94. G. Wilkinson, F. G. A Stone and E. W. Abel, "Comprehensive Organometallic Chemistry", Pergamon, 1982, Vol. 8, pp. 181- 183. 69. E. Ciganek, J. Org. Chem., 1980, 45, 1512. 70. M . .1. Milewska, T. Bytner and T. Polonski, Synthesis, 1996, 1485. 95. A. Mullen in "New Syntheses with Carbon Monoxide", cd. J. Falbe, Springer-Verlag, Berlin, 1980, p. 414. 71. J. Li, J. Chen, H.-bo. Chen and W.-X. Chcn,Synth. Cornmun., 1998, 28, 1281. 96. J. Tsuji, "Organic Syntheses with Palladium Compounds", Springer-Verlag, Berlin, 1980, p. 67. 72. F. A. Luzzio and D. P. Zacher!, Tetrahedron Le/1., 1998, 39, 2285. 97. A. Rosenthal and I. Wender in "Organic Synthesis via Metal Carbonyls", eds. I. Wender and P. Pino, Wiley, New York, 1968, Vol. I, p. 405. 73. I. Yamamoto, Y. Tabo and H. Gotoh, Tetrahedron Lei/., 1971, 2295. 98. J. Fabe, "Carbon Monoxide in Organic Synthesis", Springer- Verlag, Berlin, 1970. 74. T. DoMinh, A L. Johnson, J. E. Jones and P. P. Senise, Jr.,J. Org. Chern., 1977, 42, 4217. 99. J. M. Thompson and R F.l-leck.J. Org. Chem., 1975,40,2667. 75. R. Grigg, H. Q. N. Gunaratne and V. Sridharan,J. Chern. Soc., Chern. Cornrnun., 1985, 1183. I 00. M. Mori, K. Chiba and Y. Ban,J. Org. Chem., 1978, 43, 1684. 101. J.-J. Brunet, C. Sidot and P. Caubere,J. Org. Chem., 1983, ~8. 1166. 76. G. H. L. Nefkens and B. Zwanenburg, Tetrahedron, 1985, 41. 6063. 102. D. K. Dikshit, Indian J. Chem., Sect. B, 1983, 22, 1243. 77. S. S. Simons, Jr. and D. F. Johnson, J. Am. Chern. Soc., 1976, 98, 7098. I 03. R. C. Larock, C.-L. Liu, H. H. Lau and S. Varaprath, Tetrahe- dron Leu., 1984, 25, 4459. 78. S. S. Simons, Jr. and D. F. Johnson,J. Org. Chern., 1978, 43, 2886. I 04. R. Grigg, V. Loganathan, V. Santhakurnar, V Sridharan and A. Kundu et al. : lsoindolinones (Phthalimidines) 54. H. Watanabe, C.-L. Mao, I. T. Barnish and C. R. Hauser,J. Org. Chern., 1969, 34, 919. 83. S. M. Allin, C. J. Northfield, M. I. Page and A. M. Z. Slawin, Tetrahedron Lett., 1997, 38, 3627. 83. S. M. Allin, C. J. Northfield, M. I. Page and A. M. Z. Slawin, Tetrahedron Lett., 1997, 38, 3627. 55. W. E. Parham, L. D. Jones andY. A. Sayed,J. Org. Chem., 1976, 41, 1184. 55. W. E. Parham, L. D. Jones andY. A. Sayed,J. Org. Chem., 1976, 41, 1184. 84. J. Guillaume!, N. Boccara, P. Dcmerseman and R. Royer,J Chern. Soc., Chem. Commun., 1988, 1604. 84. J. Guillaume!, N. Boccara, P. Dcmerseman and R. Royer,J Chern. Soc., Chem. Commun., 1988, 1604. 56. W. E. Parham and L. D. Jones,J. Org. Chern., 1976, 41, 1187. 85. M. Freccero, E. Fasani and A Albini,J. Org. Chem., 1993. 58. 1740. 57. C. K. Bradsher and D. A. Hunt,J. Org. Chern., 1981, 46, 327. 86. M. A Weidner-Wells, K. Oda and P. H. Mazzocchi, Tetrahe- dron, 1997, 53, 3475, and references cited therein. 58. M.S. Hendi, K. J. Natalie, Jr. S. B. 1-Iendi, J. A. Campbell, T. D. Greenwood and J. F. Wolfe, Tetrahedron Lett., 1989, 30, 275. 87. A. G. Griesbeck, J. Hirt, W. Kramer and P. Dallakian. Tetrahe- dron, 1998, 54, 3 169. 59. W. E. Parham and C. K. Bradsher,Acc. Chem. Res., 1982, 15, 300. 60. J. B. Campbell, R. F. Dedinas and S A. Trumbower-Walsh, J. Org. Chern., 1996, 61, 6205. 88. (a) F. M. Rowe, W C. Dovey, B. Garforth, E. Levin, .1. D. Pask and A. T. Peters, J. Chem. Soc., 1935, 1796; (b) F. M. Rowe, A. S. Haigh and AT. Peters,J. Chern. Soc., 1936, 1098; (c) F. M. Rowe and E. J. Cross,J. Chem. Soc., 1947, 461; (d) F. M. Rowe, W T. McFadyen and A T. Peters,J. Chern. Soc., 1947, 468. 61. A. Jozwiak and I. Sadokierska, Synth. Commun., 1998, 28, 451. 62. A. Couture, E. Deniau, D. lonescu and P. Grandclaudon,Tetra- hedron Lett., 1998, 39, 2319. 89. P.M. Pojer, E. Ritchie and W. C. Taylor,Aust. J. Chern., 1968, 21, 1375. 63. A. Couture, E. Deniau and P. Grandclaudon, Tetrahedron, 1997, 53, 10313. 90. B. D. Whellon and A C. Huitric,J. Org. Chern., 1970, 35, 3143. 64. W. Flitsch and H. Peters, Tetrahedron Lett .. 1969, 1161. 91. P. Camilleri and A Cassola, J. Chern. Soc., Chem. J. Indian Chern. Soc., Vol. 78, October-December 2001 Commun., 1956, 21, 725. 24. T. R. Belliotti, W. A. Brink, S. R. Kesten, J. R Rubin, D. 1. Wustrow, K. T Zoski, S. Z. Whetzel, A. E. Corbin, T A. Pugsley, T. G. Heffner and L. D Wise, Bioorg. Med. Chern. Lei/., 1998, 8, 1499. 50. A. Perjessy, M. Lacava and P. 1-lmciar, Collect. C=ech. Chem. Commun., 1971, 35, 2775. 51. V. Scartoni, R. Fiaschi, S.Catalano, I. Morelli and A. Marsili,J. Chern. Soc., Perkin Trans. 1, 1979, 1547. 25. Z.-P. Zhuang, M.-P. Kung, M. Mu and H. F. Kung.J. Med. Chern., 1998,41, 157 52. K. Heidenbluth, H. Tonjes and R. Scheffler, J. Prakt. Chem., 1965, 30, 204. 26. M. H. Norman, D. J. Minick and G. C. Rigdon,]. Med. Chem., 1996, 39, 149. 53. W. S. Ang and B. Halton,Aust. J. Chem., 1971,24,851. 686 Kundu et al. : lsoindolinones (Phthalimidines) Kundu et al. : lsoindolinones (Phthalimidines) Teasdale, Tetrahedron Leu., 1991, 32, 687. 79. J. F. Stobaugh, A Repta, L.A. Sternson and K. W. Garren,Anal. Biochem., 1983, 135, 495. I 05. 79. J. F. Stobaugh, A Repta, L.A. Sternson and K. W. Garren,Anal. Biochem., 1983, 135, 495. Teasdale, Tetrahedron Leu., 1991, 32, 687. I 05. B. Bums, R. Grigg, Y. Santhakumar, V Sridharan, P Stevenson I 05. B. Bums, R. Grigg, Y. Santhakumar, V Sridharan, P Stevenson and T. Worakun, Tetrahedron. 1992, 48, 7297. 80. I. Takahashi, M. Tsuzuki and M. Kitajima,Chem. Express, 1992, 7, 70 I, and references cited therein. and T. Worakun, Tetrahedron. 1992, 48, 7297. 106. C. S. Cho, .1. W. Lee, D. Y. Lee, S.C. Shim and T J. Kim.J. 80. I. Takahashi, M. Tsuzuki and M. Kitajima,Chem. Express, 1992, 7, 70 I, and references cited therein. an 106. C. 80 , j ,C xp ess, 99 , 7, 70 I, and references cited therein. , , , 106. C. S. Cho, .1. W. Lee, D. Y. Lee, S.C. Shim and T J. Kim.J. 7, 70 I, and references cited therein. 106. C. S. Cho, .1. W. Lee, D. Y. Lee, S.C. Shim and T J. Kim.J. 106. C. S. Cho, .1. W. Lee, D. Y. Lee, S.C. Shim and T J. Kim.J. 81. I. Takahashi, T. Kawakami, E. Hirano, H. Yokota and H. Kitajima, Synle/1, 1996, 353. Chern. Soc., Chem. Commun., 1996, 2115. I 07. C. S. Cho, L. H. Jiang and S. C. Shim,Synth. Commun., 1998, 81. I. Takahashi, T. Kawakami, E. Hirano, H. Yokota and H. Kitajima, Synle/1, 1996, 353. Chern. Soc., Chem. Commun., 1996, 2115. I 07. C. S. Cho, L. H. Jiang and S. C. Shim,Synth. Commun., 1998, Kitajima, Synle/1, 1996, 353. 82 S Alli C C dki d T j S l 1996 781 I 07. C. S. Cho, L. H. Jiang and S. C. Shim,Synth. Commun., 1998, 28 849 Kitajima, Synle/1, 1996, 353. 82. S.M. Allin, C. C. Hodkinson and N. Taj,Synlett, 1996, 781. I 07. C. S. Cho, L. H. Jiang and S. C. Shim,Synth. Commun., 1998, 28, 849. 82. S.M. Allin, C. C. Hodkinson and N. Taj,Synlett, 1996, 781. 82. S.M. Allin, C. C. Hodkinson and N. Taj,Synlett, 1996, 781. 28, 849. 687 J. Indian Chern. Soc., Vol. 78, October-December 2001 108. (a) M. W. Khan and N. G. Kundu,Synlett, 1997, 1435; (b) N. G. Kundu and M. W. Khan, Tetrahedron, 2000, 56, 4777. 6937; (b) N. G. Kundu, M. W. Khan Utld R. Mukhopadhyay, Tetrahellron, 1999, 55, 12361. I 09. (a) M. W. Khan, S. Guha, A. K. Mukherjee and N. G. Kundu, ActaCrysta/log~. 1998, C54. I 19;(b)N.G. Kundu, M. W. Khan, S. Guha and A. K. Mukherjee, Acta Crystallogr., 1999, C55, 239; (c) A. K. Mukherjee, S. Guha, M. W. Khan, N. G. Kundu and M. Helliwell, Acta Crystallogr., 2000, C56, 85. I I I. (a) S Guha, A. K. Mukherjee, M. W. Khan and N. G. Kundu, Acta Crystallogr., 1999, C55, 818; (b) K. Chowdhury, R. Mukhopadhyay, M. Mukherjee, C. K. Broder and N. G. Kundu, Acta Crystal/ogr., 2001, E57, 0421. 112. (a) N. G. Kundu, M. W. Khan and J. S Mahanty,J. Chem. Res. (S), 1999, 460; J. Chem. Res. (M), 1999, 190 I. II 0. (a) N. G. Kundu and M. W. Khan, Tetrahedron Lett., I 997, 38, 688
https://openalex.org/W4200003258	https://jrssem.publikasiindonesia.id/index.php/jrssem/article/download/44/88	English	null	The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at Panahmas Dwitama Distrindo Ltd. Jember	Journal Research of Social Science, Economics, and Management/Journal Research of Social Science, Economics and Management	2,021	cc-by-sa	5,297	Keywords: compensation; work discipline; employee performance. THE EFFECT OF COMPENSATION ON EMPLOYEE PERFORMANCE THROUGH WORK DISCIPLINE AS AN INTERVENING VARIABLE AT PANAHMAS DWITAMA DISTRINDO LTD. JEMBER Vika Dhelia Pristiawati 1* Rahmat Agus Santoso 2 1, 2 Universitas Muhammadiyah Gresik, Indonesia e-mail: vikadhelia122@gmail.com 1, ra_santoso@umg.ac.id 2 *Correspondence: vikadhelia122@gmail.com1 DOI: 10.36418/jrssem.v1i5.44 INTRODUCTION INTRODUCTION The establishment of a company and factors business capital, human resources are also included in one of the most critical factors. Therefore good management and development of human resources aim to achieve company goals (Hyväri, 2016). To achieve a company's success, there needs to be an encouragement that allows employees to develop and improve their abilities and skills optimally. One way that the company can provide is with compensation. Achieving good company performance and company progress Contribute to employee performance in their respective jobs (Ardiansyah & Bachri, 2021). According to (Priyono & Darma, 2016), compensation is income received by employees in the form of goods directly or indirectly by the company in return for services that have been provided to the company. Panahmas Dwitama Distrindo Ltd. is a distributor company Unilever with the status of a private company, where employees are required to have a good performance for the company to survive and compete with other companies. Therefore, companies must consider factors - factor labor that determines the company's success in achieving its goals by increasing employee performance. Based on observations made by the author, the fact that employees of Panahmas Dwitama Distrindo Ltd. have different performances is influenced by compensation and productivity. In providing compensation to employees following the level of education, period, and career period at the company Panahmas Dwitama Distrindo Ltd. The provision of compensation based on Law No. 13 of 2003 concerning Manpower Article 1 paragraph 30 requires companies to provide salaries and wages in which there are also allowances. The contents of Law No. 13 of 2003 (Undang-Undang, 2003) concerning Manpower Article 1 Paragraph 30 is that. Person obedience in actual behavior at the time of achieving the goals set and supported by high work discipline, then the success of a job carried out by employees of a company to obtain the maximum result. Mediation can be interpreted as a process of third party participation (third element) in solving a problem (Armengol, 2013), and here work discipline is an intermediary to determine the effect of compensation on employee performance. Submitted: 19 November 2021, Revised: 11 December 2021, Accepted: 14 December 2021 Submitted: 19 November 2021, Revised: 11 December 2021, Accepted: 14 December 2021 Abstract. Panahmas Dwitama Distrindo Jember is a product distributor company of Unilever Indonesia Ltd. has low employee discipline and decreased productivity of employee performance is one of the problems this company faces. This study aims to see the effect of compensation on employee performance through work discipline as an intervening variable at Panahmas Dwitama Distrindo Ltd. Jember. This research method uses quantitative research methods by taking 60 respondents. Answer the problem formulation with the help of Smart Partial Least Square (PLS) 3 and analyze the data using the Partial Least Square (PLS) analysis technique. The study results show that compensation has a negative and insignificant effect on employee performance, compensation has a positive and significant effect on work discipline. Work discipline has a positive and significant effect on employee performance, and compensation has a positive and significant effect on performance. Employees through work discipline as an intervening variable. Keywords: compensation; work discipline; employee performance. DOI: 10.36418/jrssem.v1i5.44 DOI: 10.36418/jrssem.v1i5.44 Vika Dhelia Pristiawati, Rahmat Agus Santoso\| 4 443 et al., 2021). This is also following previous research conducted by (Patmarina & Erisna, 2012) that work discipline affects the performance of employees' CV. Laut Selatan Jaya Bandar Lampung. INTRODUCTION An employee's performance or an employee is an evaluation of the employee's work in performing their duties and responsibilities and following the procedures determined by the organization within a certain period: profitable or non-profitable (Rachmawati "Wages are the workers' rights or workers received and expressed in the form of money as a rewarding entrepreneur or employer to the worker or laborer who is determined and paid according to an employment agreement, agreements, or statutory regulations, including allowances for DOI: 10.36418/jrssem.v1i5.44 444 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember workers or laborers and their families for a job and for services that have been or, will ‘be performed.”(Undang- Undang, 2003). indirectly by the company in return for services rendered to the company. Compensation is divided into him that compensation direct and indirectly. Direct compensation is in salary, wages, intensive while compensation is indirect to inform welfare employees. In his book, Priyono & Darma (2016) explains in detail the objectives of compensation include: In addition to compensation, companies can also improve employee performance by creating work discipline. Work discipline can affect employee performance, and this statement is reinforced by the opinion of (Hasibuan et al., 2017) that The better one's work discipline, the higher the performance will be achieved. The work discipline of Panahmas Dwitama Distrindo Ltd. can be seen from the attendance and punctuality when employees enter the office. This can be seen from the number of attendance achieved. The employee attendance method of Panahmas Dwitama Distrindo Ltd. Jember uses fingerprints to be precise and accurate in evaluating employee discipline. a. Cooperation Bonds, by providing compensation, a formal cooperative bond is established between employers and workers, where employees (laborers) must do their jobs well while employers are obliged to pay agreed compensation. b. Job Satisfaction, with remuneration employees, will be able to meet their life needs so that job satisfaction will be obtained from the position they hold. c. Motivation, If the remuneration provided is adequate, then managers will find it easier to motivate employees, which will have an impact on increasing productivity. In solving performance problems, poor work discipline is also a factor that reduces productivity, affecting performance because many employees are subject to disciplinary action. INTRODUCTION From the description above, the formulation of the problem proposed is whether compensation affects employee performance through work discipline as an intervening variable. This study aimed to determine the effect of compensation on employee performance through work discipline as an intervening variable. d. Employee Stability, with a compensation program based on fair and proper principles, guarantees employee stability because turnover is relatively small. e. Discipline Improvement, Compensation. Following procedures will have an impact on increasing employee discipline. Compensation According to Parwita (2015), work discipline is the willingness of an employee to comply with and carry out various According to Priyono & Darma (2016), compensation is all income received by employees in the form of goods directly or Vika Dhelia Pristiawati, Rahmat Agus Santoso\| 445 guide employees to perform their duties correctly following decisions made. provisions that must be obeyed. (Supomo & Nurhayati, 2018) also argue that work discipline is an attitude or behavior of employees in an organization or institution always to obey, respect, and respect all rules and norms that have been determined by the institution so that organizational goals are achieved. From the description above, it can be concluded that Work Discipline is an act of one's awareness to carry out all the regulations determined by the company so that organizational goals are achieved. According to (Farid et al., 2019), some factors that influence employee discipline, namely: 6. Whether you care about them or not, employees have different personalities. 7. Development habits that promote discipline. Performance According to (Mangkunegara, 2005) Performance is the result of work in quality and quantity achieved by an employee in carrying out his duties following the responsibilities given to him. According to (Simamora, 2004) there are several characteristics of employees who have high performance, including: 1. The amount of compensation may affect discipline. Employees will comply with all applicable rules. a. Achievement-oriented employees who have high performance have a strong desire to build the dreams they dream of. 2. Whether the company has exemplary leadership or not, an example of leadership is significant. Because in a corporate environment, all employees will always be vigilant about how they can control themselves with words, actions, and attitudes that they have control over, which may violate disciplinary rules. b. Confidence, employees with high performance have a positive attitude that leads to acting with a high level of confidence. c. Self-control, employees with high performance have high self-discipline. d. Competition, employees with high performance, can develop specific abilities. 3. Whether or not there are specific rules used as a guideline, a company cannot develop discipline without specific document rules that can be used as general guidelines. The Influence of Work Discipline on Employee Performance The company must create productivity in improving employee performance, starting with the first step, namely Work Discipline. By creating work discipline, employees will have high performance. (Liyas & Primadi, 2017) in their research, work discipline has a significant effect on Employee Performance at Rural Banks. H1: Compensation has a positive effect on the performance of employees of Panahmas Dwitama Distrindo Ltd. Jember. The Hypothesis of Research Effect of Compensation on Employee Performance 4. The courage of leadership to act. If employees have committed misconduct, management needs the courage to act according to their misconduct level. Argue by (Purnama Dewi & Harjoyo, 2019) that the purpose of compensation, in general, is to help the company achieve the goals of the company's strategy success and ensure the creation of internal and external justice. In addition, compensation 5. All activities conducted by the company, whether or not supervised by management, must be supervised to 446 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember 446 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember Discipline at Ltd. Son of the Hutama Board. is intended for the benefit of employees, the government, and the community. On the other hand, the compensation program must also be determined based on fair and reasonable principles to provide satisfaction for all parties, and the purpose of compensation is achieved. A previous study researched by (Junaedi & Lestari, 2017) stated that partial compensation significantly affects employee performance at the Brilliant Education Center Probolinggo Educational Institution. At the same time, the results of research (Asmayana, 2018) Compensation has a positive effect on employee performance at the City Trade Office Makassar. Based on the statement above, the following hypothesis can be formulated: is intended for the benefit of employees, the government, and the community. On the other hand, the compensation program must also be determined based on fair and reasonable principles to provide satisfaction for all parties, and the purpose of compensation is achieved. A previous study researched by (Junaedi & Lestari, 2017) stated that partial compensation significantly affects employee performance at the Brilliant Education Center Probolinggo Educational Institution. At the same time, the results of research (Asmayana, 2018) Compensation has a positive effect on employee performance at the City Trade Office Makassar. Based on the statement above, the following hypothesis can be formulated: From the description above, the following hypothesis can be drawn: H2: Compensation positively affects Employee Work Discipline at Panahmas Dwitama Distrindo Ltd. Jember. The Influence of Work Discipline on Employee Performance The Influence of Work Discipline on Employee Performance The Influence of Work Discipline on Employee Performance H3: Work Discipline Affects Employee Performance. The Effect of Compensation on Employee Performance through Work Discipline as an Intervening Variable. The Effect of Compensation on Work Discipline The Effect of Compensation on Work Discipline The company always expects employees to have high performance. Employees with high performance improve the company's overall performance to survive and compete in global competition. The high and low level of employee performance can be seen from the compensation they get from the company. Therefore the company must provide compensation following the responsibilities given. In addition to compensation, work discipline is also a factor causing employee performance that is less than optimal in completing work on time. (Isvandiari & Fuadah, 2017) in his research shows that compensation and Compensation is one factor that influences the enforcement of employee discipline in a company. According to (Reni Yesi, 2017), discipline is an attitude of respect for the company's rules and regulations that exist in employees, which causes employees to conform to the rules and regulations company voluntarily. The results of (Husain 2020) state a positive and significant influence between Compensation on Work Discipline of Ltd. Strategic Pest control Tebet, South Jakarta. A similar study was also conducted by (Suwarna & Juliansyah, 2018), which stated a significant influence between Compensation on Employee Work Vika Dhelia Pristiawati, Rahmat Agus Santoso\| 447 work discipline variables significantly affect the performance of Ltd. Meritjan Kediri. Quantitative research is systematic scientific research on parts and phenomena and their relationships. The purpose of quantitative research is to develop and use mathematical models, theories, and hypotheses related to natural phenomena. The measurement process is a crucial part of quantitative research. This provides an overview or answer to the fundamental relationship of quantitative relationships (Siyoto & Sodik 2015). Research with a quantitative approach is usually carried out with the number of samples determined based on the existing population. Calculation of the number of samples is done using a specific formula. The selection of the formula to be used is then adjusted to the type of research and population homogeneity (Priyono & Darma, 2016). H4: Compensation positively affects employee performance through work discipline as an Intervening variable. The research shows that updates and differences lie in the variables and population and how to process the data. The author's interest in taking the title "the effect of compensation on employee performance through work discipline as an intervening variable at panahmas dwitama distrindo Ltd. Jember” because the author wants to know and learn about compensation work discipline and employee performance. The Effect of Compensation on Work Discipline In addition, the author will also test whether there is an influence between these variables. Data Collection Tools Quantitative data collection methods rely heavily on random sampling and instruments structured data collection according to predetermined response categories so that the resulting data is easy to summarize, compare and generalize. Quantitative research is closely related to testing hypotheses made from previous theories or an occurrence of phenomena. The research used is approach descriptive and quantitative analysis so that the data are numbers (numeric). According to (Sugiyono, 2016) the Quantitative approach is a research method based on the philosophy of positivism used to examine populations specific or samples, data collection using instruments research, data analysis is quantitative and statistical, intending to test predefined hypotheses so that the results can find out what has been suspected. The purpose of this research using the descriptive method is to describe the object of research, or the research results can be seen clearly. (Sugiyono, 2017) also states that this descriptive research method is carried out to determine the existence of independent variable, either on one or more variables, without making comparisons of variables The approach in Data Analysis This study uses a systematic descriptive approach and inferential analysis of the relationship between variables by testing hypotheses. Where: The nature of the data in this study is quantitative data. According to (Siyoto S. & Sodik, 2015), quantitative data is data in the form of numbers or numbers. Following the form, quantitative data can be processed or analyzed using mathematical or statistical calculation techniques. Where: n: Sample N: Population D: Degrees of freedom Diet: N: 150 Employees D: 10% 𝑛= 150 150(10%2) + 1 𝑛= 150 150(0,01) + 1 𝑛= 150 2,5 = 60 Where: n: Sample N: Population D: Degrees of freedom Diet: N: 150 Employees D: 10% 𝑛= 150 150(10%2) + 1 𝑛= 150 150(0,01) + 1 𝑛= 150 2,5 = 60 Where: n: Sample N: Population D: Degrees of freedom Diet: N: 150 Employees D: 10% 𝑛= 150 150(10%2) + 1 𝑛= 150 150(0,01) + 1 𝑛= 150 2,5 = 60 D: Degrees of freedom Diet: Data Analysis Techniques The data analysis technique is carried out chronologically after data collection and processed and analyzed computerized. The data analysis technique used validity test, test reliability, simple linear regression, correlation coefficient, coefficient of determination, and significant test (t-test). Types of Research This research uses descriptive research development with survey descriptive research methods. The purpose of this research is to find out the effect of compensation on employee performance through work discipline as an intervening variable at Panahmas Dwitama Distrindo Ltd. Jember. specified in a study. data was obtained from a questionnaire survey on employees of Panahmas Dwitama Distrindo Ltd. Jember. At the same time, the secondary data from this study were obtained from other data sources related to the research title. Most members' sample is of the population taken using the sampling method. This study uses sampling developed by Isaac and Michael using a statistical approach for an error rate of 10%, which can be done with the formula: Types of Research Data This study uses primary and secondary data. According to (Sarwono, 2006), primary data is data in the form of text from interviews and obtained through interviews with informants sampled in his research. In connection with this research, the primary Data Collection Tool Data collection was done through questionnaires or questionnaires as a data collection tool chosen by researchers to collect data. So the sample collection of this study only used 60 of the 150 employees of Panahmas Dwitama Distrindo Ltd. Jember. Population and Sample The population used in this study were employees of Panahmas Dwitama Distrindo Ltd. Jember. The number of employees of Panahmas Dwitama Distrindo Ltd. Jember totals 150 employees. According to (Margono 2005), the population is the entire object of research consisting of a human object, animals, plant, symptoms, test scores of the event as sources of funds that have characteristics 448 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember 448 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember RESULTS AND DISCUSSION Data were managed using the partial least square (PLS) (Ghozali & Latan, 2015). method using the Smart PLS 3.0 program. The research results can be seen as follows: Vika Dhelia Pristiawati, Rahmat Agus Santoso\| 449 variables > 0.5, and it can be stated that all variables have good discriminant validity. Table 3. Composite Validity Composite Reliability X 0863 Y 0848 Z 0874 Source: Data processed (2021). variables > 0.5, and it can be stated that all variables have good discriminant validity. Table 3. Composite Validity Composite Reliability X 0863 Y 0848 Z 0874 Source: Data processed (2021). variables > 0.5, and it can be stated that all variables have good discriminant validity. Table 1. Convergent Validity Factor Loading P-Value X3 0.887 0.000 X4 0.874 0.000 X5 0.700 0.000 Y3 0.884 0.000 Y4 0.831 0.000 Z3 0.823 0.000 Z4 0.837 0.000 Z5 0.846 0.000 Source: Processed data (2021) Table 4. shows The Cronbach's Alpha value >0.6. Performance has the lowest value of 0.645. Evaluation of the Measurement Model or Outer Model In testing the validity and reliability, the research model can use latent constructs with indicators reflective and normative. a. Convergent validity is determined based on the principle that the measurement of a construct should be highly correlated (Jabbour et al., 2015). The convergent validity with indicators was reflectively evaluated by Average Variance Extracted (AVE). a. Convergent validity is determined based on the principle that the measurement of a construct should be highly correlated (Jabbour et al., 2015). The convergent validity with indicators was reflectively evaluated by Average Variance Extracted (AVE). Based on table 3 above, the composite value reliability of all research variables > 0.7, thus indicating that all variables have met the composite standard. So that all variables are considered reliable. Table 4. Cronbach's Alpha Cronbach's Alpha Compensation 0.764 Performance 0.645 Work Discipline 0.785 Source: Processed data (2021) Table 4. Cronbach's Alpha Table 4. Cronbach's Alpha Table 1. Convergent Validity Performance Table 6 shows that the relationship between compensation and employee performance is not significant with a P- Value of 0.264 > 0.05 with a beta value of -0.007 and a T-statistic value of 0.631 < t-table 1, 96 it can be concluded that compensation has a negative and insignificant effect on employee performance. This explains that compensation has a negative and insignificant effect on the performance of employee Panahmas Dwitama Distrindo Ltd. Jember. However, employees work to meet their needs and improve their lives where compensation given by large or small companies does not improve employees' performance. This is different from previous research conducted by (Junaedi & Lestari, 2017), which states that partial compensation significantly affects employee performance at the Brilliant Education Center Probolinggo. 3. 3. Testing the effect of work discipline on employee performance The p-value of the work discipline variable on employee performance is 0.000 <significantly 0.05 with a positive beta value of 0.788 and the t statistic has a value of 10.480>1.96, it can be concluded that work discipline has a positive and significant effect on performance employee. This explains that work discipline affects the performance of employees of Panahmas Dwitama Distrindo Ltd. Jember. Employees who have high work discipline certainly have high performance as well. (Liyas & Primadi, 2017) in previous research also produced a statement what work Discussion Based on table 6, obtained p-value and t-statistics for each variable are described as follows: Source: Processed data (2021) Source: Processed data (2021) positive and significant effect on work discipline. This explains that compensation has an effect on employee discipline at Panahmas Dwitama Distrindo Ltd. Jember. The factor that affect the enforcement of employee discipline is compensation, so to improve employee work discipline the company must provide appropriate compensation. This research is supported by previous research conducted by (Husain, 2020) which states that there is a positive and significant influent between compensation on the work discipline of Ltd. Strategic Pest control Tebet, South Jakarta. (Suwarna & Juliansyah, 2018) also argued that there is a significant influence between compensation on employee work discipline at Ltd. Putra Papan Hutama. Evaluation of the Structural Model (Inner Model) Based on table 1. It can be proven if the data in the study is valid because the value factor loading is more significant than 0.7. After measuring (outer model), then testing the model structural (inner model). R Square can show the strength and weakness of the influence caused by endogenous variables (Maryani et al., 2020). The following is the R-Square value on the latent variable: b. Validity Discriminant aims to ensure whether an indicator reflective is correct to measure good. Table 2. Discriminant Validity Table 2. Discriminant Validity (AVE) Compensation 0.680 Company Performance 0.737 Work Discipline 0.698 Source: Processed data (2021) Table 2 shows the AVE value of all Table 5. R Square R Square Compensation (X) Performance (Y) 0.571 Work Discipline (Z) 0.212 Source: Processed data (2021) Based on table 5, the latent variable of 450 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember 450 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember performance has a value of 0.571 which means that performance can explain the variance of Work Discipline of 57.1%. The value of R Square of Work Discipline is also influenced by compensation of 0.212, testing the hypothesis as follows: 2. The effect of compensation (X) on work discipline (Z). 3. The effect of work discipline (Z) on employee performance (Y). 4. The effect of compensation on employee performance through work discipline as an intervening variable. 1. Effect of compensation (X) on employee Figure 1. Path Diagram Figure 1. Path Diagram Figure 1 is the result of processing data carried out with the SmartPLS 3.0 program, then in hypothesis testing, display the results of data processing relationships between variables using SmartPLS as follows: Table 6. Results of Inter-Variable Influence Test Table 6. Results of Inter-Variable Influence Test Table 6. Results of Inter-Variable Influence Test Original Sample (O) Sample Mean (M) Standard Deviation (STDEV) T Statistics (\|O/STDEV\|) P Values Compensation (X) -> Employee Performance (Y) -0.077 -0.083 0.122 0.631 0.264 Compensation (X) -> Work Discipline (Z) 0.461 0.493 0.084 5.473 0.000 Work Discipline (Z) -> Employee Performance (Y) 0.788 0.801 0.075 10.480 0.000 Vika Dhelia Pristiawati, Rahmat Agus Santoso\| 451 REFERENCES Andryane, R., & Mukaffi, Z. (2019). Pengaruh kompensasi terhadap kinerja karyawan melalui disiplin kerja sebagai variabel interverning: studi kasus pada karyawan PT Telkomunikasi Indonesia TBK Witel Kediri. Jurnal Sosial Ekonomi Dan Humaniora, 2, 216–225. 2. Effect of compensation on work discipline The p-value of the compensation variable on work discipline is 0.000 < significant 0.05 with a positive beta value of 0.461, and the t statistic has a value of 5.473 > 1.96. It can be concluded that compensation has a 452 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember discipline had a significant effect on employee performance at Bank Kreditan Rakyat. 4. Testing the effect of compensation on performance through work discipline as an intervening variable Table 7. indirect effect test results Original Sample (O) Sample Mean (M) Standard Deviation (STDEV) T Statistics (\|O/STDEV\|) P Values Compensation (X) -> Work Discipline (Z) -> Employee Performance (Y) 0.363 0.396 0.083 4.397 0.000 Source: Processed data (2021) 452 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember Table 7. indirect effect test results prove and answer this research problem. The conclusions are compensation has a negative and insignificant effect on employee performance at Panahmas Dwitama Distrindo Ltd. Jember. Compensation has a positive effect on Work Discipline Panahmas Dwitama Distrindo Ltd. Jember. Work Discipline has a positive and significant effect on Employee Performance of Panahmas Dwitama Distrindo Ltd. Jember. Compensation has a positive and significant influence on employee performance through Work Discipline as an intervening variable. Work Discipline is able to mediate the effect of compensation on the performance of employees of Panahmas Dwitama Distrindo Ltd. Jember. Tables The above shows a P-value of 0.000 <0.05, with a beta value of 0.363 and a t-statistic value of 4.397 > at-value of 1.96, it can be concluded that compensation has a positive and significant effect on employee performance through work discipline as an intervening variable. This explains that the effect of compensation on company performance through employee discipline at Panahmas Dwitama, DIstrindo Jember Ltd. The first step that must be taken by the company to improve employee performance is to enforce good work discipline, for that the company must provide appropriate compensation so that employee performance increases. Previous research conducted by (Andryane & Mukaffi, 2019) stated that compensation has an indirect effect on employee performance through work discipline at PT Telekomunikasi Indonesia. CONCLUSIONS Based on the research results and the hypotheses that have been formulated, the results obtained are research that is able to Vika Dhelia Pristiawati, Rahmat Agus Santoso\| 453 Hyväri, I. (2016). Roles of top management and organizational project management in the effective company strategy implementation. Procedia- Social and Behavioral Sciences, 226, 108–115. Hyväri, I. (2016). Roles of top management and organizational project management in the effective company strategy implementation. Procedia- Social and Behavioral Sciences, 226, 108–115. Ardiansyah, M. F., & Bachri, A. A. (2021). Pengaruh Kepemimpinan Partisipatif, Kompensasi Dan Kepuasan Kerja Terhadap Kinerja Karyawan Divisi Coal Processing And Barge Loading–Pt Adaro Indonesia. Jurnal Indonesia Sosial Sains, 2(11), 1933–1952. Isvandiari, A., & Fuadah, L. (2017). Pengaruh Kompensasi dan Disiplin Kerja Terhadap Kinerja Karyawan Bagian Produksi PG. Meritjan Kediri. Jurnal Ilmiah Bisnis Dan Ekonomi Asia, 11(2), 1–8. Armengol, V. F. (2013). The principles of mediation and the role of third parties in peace processes. Norwegian Peacebuilding Resource Centre Report, Sep Tember, 5. Asmayana, H. A. (2018). Pengaruh Kompensasi Terhadap Kinerja Pegawai Pada Kantor Dinas Perdagangan Kota Makassar. Jurnal Profitability Fakultas Ekonomi Dan Bisnis, 2(2). Jabbour, C. J. C., Jugend, D., de Sousa Jabbour, A. B. L., Gunasekaran, A., & Latan, H. (2015). Green product development and performance of Brazilian firms: measuring the role of human and technical aspects. Journal of Cleaner Production, 87, 442–451. Farid, A., Haerani, S., Hamid, N., & Reni, A. (2019). The Effect of Organizational Culture, Organizational Commitment and Work Satisfaction Toward Teacher Performance of High School at Makassar. Hasanuddin Journal of Applied Business and Entrepreneurship, 2(2), 89–99. Junaedi, J., & Lestari, I. W. (2017). Pengaruh Disiplin Kerja dan Kompensasi Terhadap Kinerja yawan Pada Lembaga Pendidikan Brilliant Education Centre Probolinggo. Jurnal Ilmiah Ecobuss, 5(1), 21–32. Liyas, J. N., & Primadi, R. (2017). Pengaruh disiplin kerja terhadap kinerja karyawan pada bank perkreditan rakyat. Al-Masraf: Jurnal Lembaga Keuangan Dan Perbankan, 2(1), 17–26. Ghozali, I., & Latan, H. (2015). Partial least squares konsep, teknik dan aplikasi menggunakan program smartpls 3.0 untuk penelitian empiris. Semarang: Badan Penerbit UNDIP. Mangkunegara, A. A. A. P. (2005). Evaluasi kinerja SDM. Tiga Serangkai. Hasibuan, S., Mawarni, R., & Hendriandy, R. (2017). Respon Pemberian Pupuk Bokashi Ampas Tebu dan Pupuk Bokashi Eceng Gondok terhadap Pertumbuhan dan Produksi Tanaman Kedelai (Glycine max (L) Merril). Bernas: Jurnal Penelitian Pertanian, 13(2), 59–64. Margono, S. (2005). Metodologi penelitian pendidikan. Maryani, N. L. K. S., Widyani, A. A. D., & Saraswati, N. P. CONCLUSIONS A. S. (2020). Pengaruh Kompensasi Finansial Terhadap Kinerja Karyawan Dengan Motivasi Sebagai Variabel Intervening Pada Pt Arta Sedana Retailindo Cabang Hardys Malls Sanur Pengaruh Kompensasi Finansial Terhadap Kinerja Karyawan Dengan Motivasi Husain, B. A. (2020). Pengaruh Kompensasi Terhadap Disiplin Kerja Karyawan Pada PT. Strategic Pestcontrol Tebet Jakarta Selatan. JENIUS (Jurnal Ilmiah Manajemen Sumber Daya Manusia), 3(3), 277–285. 454 \| The Effect of Compensation on Employee Performance Through Work Discipline as an Intervening Variable at LTD. Panahmas Dwitama Distrindo Jember Metodologi Penelitian. Literasi Media. Sebagai Variabel Inter. VALUES, 1(2), 53–67. https://doi.org/10.35314/inovbiz.v4i2. 75 Siyoto, S., & Sodik, M. A. (2015). Dasar metodologi penelitian. Literasi Media Publishing. Parwita, G. B. S. (2015). Disiplin Kerja Karyawan (Suatu Kajian Teori). Jurnal Ilmu Manajemen Mahasaraswati, 5(2), 101040. Sugiyono. (2016). Metode Penelitian Kuantitatif, Kualitatif dan R&D. PT Alfabet. Sugiyono, P. D. (2017). Metode Penelitian Bisnis: Pendekatan Kuantitatif, Kualitatif, Kombinasi, dan R&D. Penerbit CV. Alfabeta: Bandung. Patmarina, H., & Erisna, N. (2012). Pengaruh Disiplin Kerja Terhadap Kinerja Karyawan Yang Dimediasi Oleh Produktivitas Kerja Perusahaan CV. Laut Selatan Jaya Di Bandar Lampung. Jurnal Manajemen Dan Bisnis, 3(1). http://dx.doi.org/10.36448/jmb.v3i1.4 93 Supomo, R., & Nurhayati, E. (2018). Manajemen Sumber Daya Manusia. Suwarna, A., & Juliansyah, R. (2018). Pengaruh Kompensasi Terhadap Disiplin Karyawan Di PT. Putera Papan Hutama. Jurnal Ilmiah MEA (Manajemen, Ekonomi, & Akuntansi), 2(1), 30–41. Priyono, P., & Darma, U. B. (2016). Buku manajemen sumber daya manusia (2). Surabaya: Penerbit Zifatama. Purnama Dewi, D., & Harjoyo, H. (2019). Manajemen Sumber Daya Manusia. Purnama Dewi, D., & Harjoyo, H. (2019). Manajemen Sumber Daya Manusia. Undang-Undang, R. I. (2003). Nomor 13 Tahun 2003. Tentang Ketenagakerjaan. Rachmawati, S. N., Lumbanraja, P., & Siahaan, E. (2021). The Effect of Adaptive Ability, Communication Skills, and Work Environment on Performance of Medan Mayor’s Office with Teamwork as Intervening Variables. Journal Research of Social Science, Economics, and Management, 1(4), 406–417. https://doi.org/10.36418/jrssem.v1i4. 37 © 2021 by the authors. Submitted for possible open-access publication under the terms and conditions of the Creative Commons Attribution (CC BY SA) license (https://creativecommons.org/licenses/by-sa/4.0/). Reni Yesi, S. (2017). Analisis Pengaruh Kepemimpinan Dan Disiplin Kerja Terhadap Kinerja Pegawai PT. Karisnamas Di Bekasi. Jurnal Administrasi FIA Unija, 4(2), 28–38. Sarwono, J. (2006). Metode penelitian kuantitatif dan kualitatif. Simamora, H. (2004). Manajemen sumber daya manusia. Siyoto, S. & Sodik, M. A. (2015). Dasar Siyoto, S. & Sodik, M. A. (2015). Dasar Siyoto, S. & Sodik, M. A. (2015). Dasar
https://openalex.org/W4283688699	https://link.springer.com/content/pdf/10.1007/s11230-022-10049-1.pdf	English	null	A new parasitic copepod (Copepoda; Cyclopoida; Chondracanthidae) from two pomacentrid fishes caught on the Great Barrier Reef, Queensland, Australia	Systematic parasitology	2,022	cc-by	5,214	Syst Parasitol (2022) 99:601–610 https://doi.org/10.1007/s11230-022-10049-1 (0123456789().,-volV)( 01234567 89().,-volV) RESEARCH A new parasitic copepod (Copepoda; Cyclopoida; Chondracanthidae) from two pomacentrid ﬁshes caught on the Great Barrier Reef, Queensland, Australia Geoff A. Boxshall Received: 11 April 2022 / Accepted: 31 May 2022 / Published online: 29 June 2022 The Author(s) 2022 The Chondracanthidae of Australian marine ﬁshes are reasonably well known, with a total of 17 named species recorded to date (Table 1). The ﬁrst two species reported were described by Heegaard (1940) under the names Acanthochondria platycephali Hee- gaard, 1940 and Acanthochondria platycephali forma alata-longicollis Heegaard, 1940. Both species were subsequently redescribed from the type material by Ho (1973) who accepted the former as valid and treated the latter as the type species of a new genus under the name Pterochondria alatalongicollis (Heegaard, 1940). In that paper, Ho (1973) considered that Acanthochondria diastema Kabata, 1965 described from waters off Tasmania was a junior synonym of Pterochondria alatalongicollis, but A. diastema was later resurrected as a valid species by Ho & Dojiri (1988). More recently, Tang et al. (2010) reassessed the validity of the genus Pterochondria Ho, 1973 and concluded that it did not differ signiﬁcantly from Acanthochondria. They transferred its only species back to Acanthochondria as A. alatalongicollis. Abstract A new species of the copepod genus Pseudacanthocanthopsis Yamaguti & Yamasu, 1959 (family Chondracanthidae) is described based on material of both sexes collected from two pomacentrid host ﬁshes caught off Lizard Island, Queensland. The type host is Neopomacentrus azysron (Bleeker) and the additional host is N. cyanomos (Bleeker). The new species is distinguishable from all congeners by the form of the antennule of the female, which is dorso- ventrally ﬂattened and extends out anteriorly to the front of the cephalothorax margin. G. A. Boxshall (&) Department of Life Sciences, Natural History Museum, Cromwell Road, London SW7 5BD, UK e-mail: g.boxshall@nhm.ac.uk Introduction The copepod family Chondracanthidae comprises 191 valid species currently classiﬁed in 50 genera (Walter & Boxshall, 2021). Adult female chondracanthids are highly transformed parasites that live on marine ﬁsh hosts. The males are reduced in size and have been traditionally been referred to as dwarf males (Øster- gaard et al., 2005). Male chondracanthids are typically found attached to specialized secretory organs located in the genital region of the females from which they obtain nutrients (Østergaard & Boxshall, 2004). Heegaard (1962) added Pseudoblias lyrifera Hee- gaard, 1962 to the Australian fauna and this species was redescribed in better detail by Kabata (1969). In the same paper Kabata established Neobrachiochon- dria quadrata Kabata, 1969, a new monotypic genus from southern Australia. Ho & Dojiri (1976) described Pseudacanthocanthopsis rohdei Ho & Dojiri, 1976 from the Great Barrier Reef, and this is the only species reported from Australian waters on a G. A. Boxshall (&) Department of Life Sciences, Natural History Museum, Cromwell Road, London SW7 5BD, UK e-mail: g.boxshall@nhm.ac.uk 12 3 123 Syst Parasitol (2022) 99:601–610 602 pomacentrid host the same host family as the new (1992) who recorded six species new to Australia Table 1 Species of Chondracanthidae known from Australian waters and their recorded hosts (listed in chronological order) Copepod species Host in Australian waters References Acanthochondria platycephali Heegaard, 1940 Platycephalus bassensis Cuvier Heegaard (1940) [as A. gemina Heegaard, 1962] Platycephalus richardsoni Castlenau [as Neoplatycephalus richardsoni] Heegaard (1962) Acanthochondria alatalongicollis Heegaard, 1940 Platycephalus bassensis Cuvier Heegaard (1940) [as Pterochondria alatalongicollis] Platycephalus bassensis Cuvier Ho (1973) Acanthochondria tasmaniae Heegaard, 1962 ‘‘sea perch’’ Heegaard (1962) Pseudoblias lyrifera Heegaard, 1962 Rhombosolea tapirina Gu¨nther Heegaard (1962) Pseudorhombus dupliciocellatus Regan Kabata (1969) Acanthochondria diastema Kabata, 1965 Platycephalus bassensis Cuvier Kabata (1965) Platycephalus sp. Ho & Dojiri (1988) Neobrachiochondria quadrata Kabata, 1969 Hypoplectrodes nigroruber (Cuvier) Kabata (1969) Pseudacanthocanthopsis rohdei Ho & Dojiri, 1976 Dascyllus reticulatus (Richardson) Ho & Dojiri (1976) Pomacentrus chrysurus (Cuvier) [as P. rhodonotus] Ho & Dojiri (1976) Lagochondria nana Ho & Dojiri, 1988 Callionymus sp. Introduction Ho & Dojiri (1988) Apodochondria medusa Ho & Dojiri, 1988 Neosebastes pandus (Richardson) Ho & Dojiri (1988) Acanthochondria incisa Shiino, 1955 Scorpaena papillosa (Schneider & Forster) [as Helicolenus papillosus] Kabata (1992) Chondracanthus genypteri Thomson, 1899 Genypterus blacodes (Forster) Kabata (1992) Chondracanthus neali Leigh-Sharpe, 1930 Malacocephalus laevis (Lowe) Kabata (1992) Chondracanthus polymixiae Yamaguti, 1939 Polymixia japonica Gu¨nther Kabata (1992) Rohdea cryptopoda Kabata, 1992 Genypterus blacodes (Forster) Kabata (1992) Lateracanthus novus Kabata, 1992 Cetonurus sp. Kabata (1992) Acanthocanthopsis quadrata Heegaard, 1945 Dicotylichthys punctulatus Kaup Tang & Ho (2005) Contusus brevicaudus Hardy Tang & Ho (2005) Chondracanthus goldsmidi Tang, Andrews & Cobcroft, 2007 Latris lineata (Forster) Tang et al. (2007) Chondracanthidae gen .sp. Paraulopus nigripinnis (Gu¨nther) [as Chlorophthalmus nigripinnis] Kabata (1992) Table 1 Species of Chondracanthidae known from Australian waters and their recorded hosts (listed in chronological order) Copepod species Host in Australian waters References Acanthochondria platycephali Heegaard, 1940 Platycephalus bassensis Cuvier Heegaard (1940) [as A. gemina Heegaard, 1962] Platycephalus richardsoni Castlenau [as Neoplatycephalus richardsoni] Heegaard (1962) Acanthochondria alatalongicollis Heegaard, 1940 Platycephalus bassensis Cuvier Heegaard (1940) [as Pterochondria alatalongicollis] Platycephalus bassensis Cuvier Ho (1973) Acanthochondria tasmaniae Heegaard, 1962 ‘‘sea perch’’ Heegaard (1962) Pseudoblias lyrifera Heegaard, 1962 Rhombosolea tapirina Gu¨nther Heegaard (1962) Pseudorhombus dupliciocellatus Regan Kabata (1969) Acanthochondria diastema Kabata, 1965 Platycephalus bassensis Cuvier Kabata (1965) Platycephalus sp. Ho & Dojiri (1988) Neobrachiochondria quadrata Kabata, 1969 Hypoplectrodes nigroruber (Cuvier) Kabata (1969) Pseudacanthocanthopsis rohdei Ho & Dojiri, 1976 Dascyllus reticulatus (Richardson) Ho & Dojiri (1976) Pomacentrus chrysurus (Cuvier) [as P. rhodonotus] Ho & Dojiri (1976) Lagochondria nana Ho & Dojiri, 1988 Callionymus sp. Ho & Dojiri (1988) Apodochondria medusa Ho & Dojiri, 1988 Neosebastes pandus (Richardson) Ho & Dojiri (1988) Acanthochondria incisa Shiino, 1955 Scorpaena papillosa (Schneider & Forster) [as Helicolenus papillosus] Kabata (1992) Chondracanthus genypteri Thomson, 1899 Genypterus blacodes (Forster) Kabata (1992) Chondracanthus neali Leigh-Sharpe, 1930 Malacocephalus laevis (Lowe) Kabata (1992) Chondracanthus polymixiae Yamaguti, 1939 Polymixia japonica Gu¨nther Kabata (1992) Rohdea cryptopoda Kabata, 1992 Genypterus blacodes (Forster) Kabata (1992) Lateracanthus novus Kabata, 1992 Cetonurus sp. Kabata (1992) Acanthocanthopsis quadrata Heegaard, 1945 Dicotylichthys punctulatus Kaup Tang & Ho (2005) Contusus brevicaudus Hardy Tang & Ho (2005) Chondracanthus goldsmidi Tang, Andrews & Cobcroft, 2007 Latris lineata (Forster) Tang et al. (2007) Chondracanthidae gen .sp. Materials and methods p g p Type Material: Holotype female from gills of Neopo- macentrus azysron (Fish No. 3023: 35.8 mm SL, 44.4 mm TL); Queensland Museum Registration No. W29609. Paratype female with male attached, from gills of N. azysron (Fish No. 3430: 42.9 mm SL, 51 mm TL); Queensland Museum Registration No. W29610. Paratype female (without male), from gills of N. azysron (Fish No. 3114: 55.2 mm SL, 73.8 mm TL); Natural History Museum, London Registration number NHMUK 2022.172. The material was collected by Alexandra S. Grutter as part of a ﬁeld experiment testing the effect of cleaner ﬁsh Labroides dimidiatus Valenciennes presence/ removal on ectoparasites (Grutter, 1996). Fish were collected by scuba divers using barrier nets and hand nets, and immediately placed in quick-sealing plastic bags to retain the parasites. Fish died when placed in an ice slurry during transport to the laboratory. The ﬁsh and the contents of the bag were ﬁxed in 10% formalin in seawater. Parasites were found by exam- ining the body of the ﬁsh, oral and branchial cavities, inside of operculum, and detached pectoral and pelvic ﬁns and gills spread out on a petri dish, all scanned under a microscope at a magniﬁcation of 35x. Collected parasites were transferred to individual vials containing 10% formalin in seawater. Fish standard length (SL) and total length (TL) were measured. Additional non-type material: one female with male attached, from gills of Neopomacentrus cyanomos (Fish 3082: 42.6 mm SL, 52 mm TL); Queensland Museum Registration No. W29611. One female with male attached, from gills of N. cyanomos (Fish 3065: 34.4 mm SL, 42.6 mm TL); Queensland Museum Registration No. W29612. One female with male attached, from gills of N. cyanomos (Fish 3009: 39.8 mm SL, 49.2 mm TL); Natural History Museum, London Registration number NHMUK 2022.173. Three females of Pseudacanthocanthopsis were collected from the gills of Neopomacentrus azysron (Bleeker, 1877) caught between October 18 and 26, 1993. A further three females of Pseudacanthocan- thopsis were collected from the gills of Neopomacen- trus cyanomos (Bleeker, 1856) caught between October 18 and 19, 1993. Specimens were cleared in lactic acid and observed whole on a Leitz dissecting microscope. Dissected appendages were examined on an Olympus BH2 microscope using differential inter- ference contrast. Drawings were made using a drawing tube and measurements were made using a stage micrometer. Morphological terminology conforms to Huys & Boxshall (1991). Materials and methods The structure referred to in older chondracanthid literature as the ‘‘accessory antennule’’ was shown to be the atrophied (and Abundances (range, median, 25th/75th quantile, prevalence, number of ﬁsh hosts sampled) were: Neopomacentrus azysron (0 – 4, 0, 0/0, 17.8%, 135); Neopomacentrus cyanomos (0-6, 0, 0/2, 49%, 104) (Unpublished data, Alexandra S. Grutter, personal communication). Introduction Paraulopus nigripinnis (Gu¨nther) [as Chlorophthalmus nigripinnis] Kabata (1992) (1992) who recorded six species new to Australia, including the new monotypic genus Rohdea Kabata, 1992, collected from ﬁshes caught in deeper waters off the coast of New South Wales. These existing records are summarized in Table 1. (1992) who recorded six species new to Australia, including the new monotypic genus Rohdea Kabata, 1992, collected from ﬁshes caught in deeper waters off the coast of New South Wales. These existing records are summarized in Table 1. pomacentrid host, the same host family as the new species described below. Two major contributions to our knowledge came from Ho & Dojiri (1988) who established two new monotypic genera, Lagochondria Ho & Dojiri, 1988 and Apodochondra Ho & Dojiri, 1988, from Australian marine ﬁshes, and Kabata 12 123 Syst Parasitol (2022) 99:601–610 603 Here, we describe a new species of Pseudacantho- canthopsis Yamaguti & Yamasu, 1959 collected from two pomacentrid hosts caught from small reefs in the lagoon and in front of Casuarina Beach off Lizard Island, Queensland (Great Barrier Reef, 144000S, 1452800E). For map see Grutter (1996). A third pomacentrid species, Pomacentrus moluccen- sis Bleeker, 1853, collected at the same time, in the same location, and using the same methods was uninfected by the parasite described here (Grutter, 1996). laterally displaced) terminal segment of the antenna (Ho, 1984). Names of hosts follow FishBase (Froese & Pauly, 2021). Type material and voucher specimens are deposited in the collections of the Queensland Museum and in the Natural History Museum, London. Systematics Family Chondracanthidae Milne Edwards, 1840 Genus Pseudacanthocanthopsis Yamaguti & Yamasu, 1959 Description of Female Adult female body transformed, consisting of head, trunk and genitoabdomen (Figs. 1, 2A, 3A); total body length 868 lm (excluding antennules and caudal setae). Head comprising fused cephalosome and ﬁrst pedigerous somite; slightly wider than long (347 x 362 lm); with median cuticular thickening along dorsal 12 3 Syst Parasitol (2022) 99:601–610 604 Fig. 1 Photomicrograph of adult female of Pseudacanthocan- thopsis grutterae n. sp. (ventral view) from Neopomacentrus cyanomos apical seta plus 1 inner distal seta, 1 dorsal seta and 2 lateral setae. Antennule (Fig. 2C, D) 2-segmented, comprising proximal segment forming enlarged dorso-ventrally ﬂattened plate and cylindrical distal segment. Proxi- mal segment expanded anteriorly, supplied with extrinsic muscles entering via lumen at base of limb. Three stubby elements (modiﬁed setae?) present on ventral surface of proximal segment (Fig. 2A). Distal segment (Fig. 2D) displaced posteriorly, armed with 5 setae apically. Antenna (Fig. 2E) comprising broad compound basal segment and strongly curved distal claw; atro- phied terminal segment (‘‘accessory antennule’’) located laterally at base of claw; armed with single apical seta. Labrum (Fig. 2F) with median indentation and expanded into paired lateral lobes. Mandible (Fig. 2G) forming tapering blade armed with about 22 teeth along convex margin and 8 on concave margin. Maxillule (Fig. 2H) lobate with 2 apical setae and 1 reduced inner seta. Maxilla (Fig. 2I) 2-segmented, proximal segment robust, unarmed; distal segment forming stout curved claw bearing strong accessory claw on concave margin; and armed with curved spiniform seta plus slender seta proximally. Maxil- liped Fig. 3B) 3-segmented; ﬁrst segment longest, unarmed; second segment with inner distal margin produced into swelling ornamented with ﬁne spinules; terminal segment forming curved claw bearing small tooth on concave margin. Fig. 1 Photomicrograph of adult female of Pseudacanthocan- thopsis grutterae n. sp. (ventral view) from Neopomacentrus cyanomos midline extending posteriorly from frontal margin to beyond middle of head (Fig. 3A). Proximal segment of antennules ﬂattened and extending anteriorly to form bipartite frontal plate, conspicuous in dorsal view. Head without obvious dorsal or lateral processes. Paired processes associated with antennae present; processes and antennae can together be displaced anteriorly (Fig. 1) or posteriorly (Fig. 2A) in preserved material. Small posterolateral rounded lobes present ventrally either side of labrum. Trunk about 1.5 times wider than long (370 x 574 lm), subrectangular, lacking processes. Genitoabdomen (Fig. 2B) small, comprising genital complex bearing paired genital openings dorsolaterally, and unsegmented abdomen. Description of Female Genital complex about 2.2 times wider than abdomen; ornamented with paired setules on ventral surface. Abdomen about as wide as long, fused to genital complex, bearing paired caudal rami on posterior margin. Caudal rami each armed with large fused Legs 1 (Fig. 3C) and 2 (Fig. 3D) located on walls of groove separating head and trunk; difﬁcult to observe. Both leg pairs biramous but very reduced, each comprising unsegmented protopodal part, unseg- mented, lobate endopod and indistinctly 2-segmented exopod. Endopod with single apical seta in both legs. Exopod armed with outer element on proximal segment and 1 lateral plus 3 apical elements on distal segment in both legs. Description of Male Body less transformed than female; consisting of cephalothorax (comprising fused cephalosome plus ﬁrst pedigerous somite), second pedigerous somite, 2 limbless somites (expressed in dorsal view), genital complex and 1-segmented abdomen (Fig. 3E). Cephalothorax slightly wider than long (91 x 100 lm); surfaced ornamented with 5 pairs of 12 123 605 Syst Parasitol (2022) 99:601–610 Fig. 2 Pseudacanthocanthopsis grutterae n. sp. adult female. A, habitus, ventral; B, genitoabdomen, dorsal; C, antennule showing musculature; D, distal segment of antennule; E, antenna; F, labrum in situ, ventral; G, mandible; H, maxillule; I, maxilla. Scale bars: A, C, 100 lm; B, E, F, 50 lm; H, I, 20 lm; G, 10 lm Fig. 2 Pseudacanthocanthopsis grutterae n. sp. adult female. A, habitus, ventral; B, genitoabdomen, dorsal; C, antennule showing musculature; D, distal segment of antennule; E, antenna; F, labrum in situ, ventral; G, mandible; H, maxillule; I, maxilla. Scale bars: A, C, 100 lm; B, E, F, 50 lm; H, I, 20 lm; G, 10 lm integumental pores. Second pedigerous somite nar- rowing strongly; posterior part ornamented with pair of pores dorsally. Third somite limbless; wider than long (14 x 42 lm). Fourth somite clearly expressed dorsally (Fig. 3E) but fully fused to genital complex ventrally (Fig. 3F); dimensions in dorsal view (13 x 39 lm). Genital complex (Fig. 3E) with slight lateral indentation marking plane of fusion of ﬁfth somite with genital somite; produced posterolaterally into rounded lobes carrying genital apertures ventrally. Abdomen as long as wide (17 x 17 lm); bearing paired caudal rami posteriorly. Caudal rami (Fig. 3F) about 12 3 606 Syst Parasitol (2022) 99:601–610 Fig. 3 Pseudacanthocanthopsis grutterae n. sp. adult female. A, habitus, dorsal; B, maxilliped; C, leg 1; D, leg 2. Adult ma habitus, dorsal; F, genital complex and abdomen, ventral; G, antennule. Scale bars: A, 200 lm, B-D, 20 lm; E, 50 lm; F, 25 lm; lm Fig. 3 Pseudacanthocanthopsis grutterae n. sp. adult female. A, habitus, dorsal; B, maxilliped; C, leg 1; D, leg 2. Adult male, E, habitus, dorsal; F, genital complex and abdomen, ventral; G, antennule. Scale bars: A, 200 lm, B-D, 20 lm; E, 50 lm; F, 25 lm; G, 10 lm Fig. 3 Pseudacanthocanthopsis grutterae n. sp. adult female. A, habitus, dorsal; B, maxilliped; C, leg 1; D, leg 2. Adult male, E, habitus, dorsal; F, genital complex and abdomen, ventral; G, antennule. 123 Description of Male 4G) slender; ﬁrst segment elongate, unarmed; second segment unarmed; third segment forming simple, tapering distal claw. Description of Male Scale bars: A, 200 lm, B-D, 20 lm; E, 50 lm; F, 25 lm; G, 10 lm twice as long as wide (13 x 6 lm); armed with large fused apical seta plus 1 inner distal seta, 1 dorsal seta and 2 lateral setae. surface;apicalsegmentshort,armedwith8setaeintotal, one of which fused basally with aesthetasc. Antenna (Fig. 4A, B) comprising large, unarmed basal segment bearing strongly recurved terminal claw; atrophied apical segment (‘‘accessory antennule’’) carried on Antennule (Fig. 3G) cylindrical, 2-segmented; large basal segment armed with 7 setae along anteroventral 123 123 Syst Parasitol (2022) 99:601–610 607 Fig. 4 Pseudacanthocanthopsis grutterae n. sp. adult male. A, antenna, with position of atrophied distal segment marked; B, atrophied apical segment of antenna, detached; C, labrum in situ, ventral; D, mandible; E, maxillule; F, maxilla; G, maxilliped; H, leg 1; I, leg 2. Scale bars: A, B, 20 lm; C-I, 10 lm Fig. 4 Pseudacanthocanthopsis grutterae n. sp. adult male. A, antenna, with position of atrophied distal segment marked; B, atrophied apical segment of antenna, detached; C, labrum in situ, ventral; D, mandible; E, maxillule; F, maxilla; G, maxilliped; H, leg 1; I, leg 2. Scale bars: A, B, 20 lm; C-I, 10 lm Reef ﬁshes and the impact of cleaner ﬁsh on these communities. lateral surface; apical segment (Fig. 4B) cylindrical, armed with single apical seta. Labrum (Fig. 4C) without median indentation on posterior margin; with rounded lateral margins. Mandible (Fig. 4D) forming tapering blade armed with 9 teeth along convex margin and 4 on concave margin. Maxillule (Fig.4E) lobate with 2 apical setae and 1 reduced inner seta. Maxilla (Fig. 4F) 2-segmented; proximal segment robust, unarmed; distal segment forming weakly curved claw bearing strong toothdistallyonconcavemarginand5denticlesonouter margin; segment armed with 2 slender setae proximally. Maxilliped (Fig. 4G) slender; ﬁrst segment elongate, unarmed; second segment unarmed; third segment forming simple, tapering distal claw. lateral surface; apical segment (Fig. 4B) cylindrical, armed with single apical seta. Labrum (Fig. 4C) without median indentation on posterior margin; with rounded lateral margins. Mandible (Fig. 4D) forming tapering blade armed with 9 teeth along convex margin and 4 on concave margin. Maxillule (Fig.4E) lobate with 2 apical setae and 1 reduced inner seta. Maxilla (Fig. 4F) 2-segmented; proximal segment robust, unarmed; distal segment forming weakly curved claw bearing strong toothdistallyonconcavemarginand5denticlesonouter margin; segment armed with 2 slender setae proximally. Maxilliped (Fig. Discussion The form of the female antennules in the new species is so unusual that investigation of the musculature was necessary in order to conﬁrm the interpretation. Two extrinsic muscles pass into the antennule from the head (Fig. 2C). The larger, more anteriorly located muscle has a tendinous origin inside the head and passes into the proximal segment of the antennule and on towards a broad insertion on the lateral wall of the segment. The smaller muscle also originates inside the head and passes into the limb, inserting directly onto the lateral wall of the segment posterior to the anterior muscle. The presence of musculature indicates that the ﬂattened structure is derived from paired limbs and the presence of the deﬁned, setose, distal segment, albeit displaced posteriorly, conﬁrms that this structure comprises the proximal part of the antennule. These muscles may represent an opposing pair and appear to function to adduct and abduct the ﬂattened proximal segment relative to the head, possibly moving the anterior expansions in to meet at the midline and to separate them. The new species can be readily distinguished from all four of its congeners by the form of the antennule in females, which is dorso-ventrally ﬂattened and extends out anteriorly to the front of the cephalotho- rax. Together the antennules are visible as a biﬁd plate in dorsal view (Fig. 3A). In contrast, in the three Japanese species, P. apogonis, P. secunda and P. bicornutus, the antennules of the female are cylindri- cal and more-or-less directed laterally. In P. rohdei the antennules are modiﬁed, with the large, ﬂeshy prox- imal segment forming a curved structure with a ventromedial lobe and a large protrusion at its posterodistal corner (Ho & Dojiri, 1976). However it is ﬂeshy and cylindrical rather than dorso-ventrally ﬂattened. The unique form of the antennule of the new species is observable in undissected females and supports the establishment of the new species. The morphology of the new species is most similar to that of P. rohdei and P. bicornutus; all three species have very similar gross morphology in the female and the appendages share numerous character states. Remarks The genus Pseudacanthocanthopsis was established by Yamaguti & Yamasu (1959) to accommodate their new species P. apogonis Yamaguti & Yamasu, 1959, collected from Ostorhinchus semilineatus Temminck & Schlegel [as Apogon semilineatus] in Japanese waters. The type species was subsequently reported from a second apogonid host, Ostorhinchus doeder- leini Jordan & Snyder [as Apogon doederleini] (Izawa, 1975). Two additional species were subsequently reported from Japan. The ﬁrst was P. secunda Yamaguti & Yamasu, 1960 from Apogon lineatus Temminck & Schlegel. This species was described under the name Pseudacanthopsis secunda: however, this generic name has never been proposed and is obviously an error since Yamaguti and Yamasu (1960) only made comparisons with Pseudacanthocanthopsis apogonis. Venmathi Maran et al. (2013) redescribed P. secunda and recorded its presence on two new hosts: the tetraodontid Takifugu poecilonotus Legs 1 and 2 (Fig. 4H, I) biramous, reduced; each comprising unsegmented protopod, unsegmented, lobate endopod and indistinctly segmented exopod. Endopod unarmed in both legs. Exopod armed with 5 outer and distal margin elements in leg 1 (Fig. 4H) and 4 outer and distal margin elements in leg 2 (Fig. 4I). Etymology. The new species honours its discoverer Alexandra S. Grutter for her important contributions to our knowledge of the ectoparasites of Great Barrier 12 3 3 Syst Parasitol (2022) 99:601–610 608 (Temminck & Schlegel) caught off Hiroshima Pre- fecture, Japan, and the sparid Pagrus major (Tem- minck & Schlegel) from the Seto Inland Sea, Japan. They also extended its known distribution range on the type host, A. lineatus, to include the East China Sea off Japan and Korea. The third Japanese species was P. bicornutus (Shiino, 1960) from a cepolid host, Owstonia totomiensis Tanaka. Ho & Kim (1995) redescribed P. bicornutus based on material from a second host, the pomacentrid Chromis notatus (Tem- minck & Schlegel), caught in the Sea of Japan. The fourth species in the genus is P. rohdei which was based on material collected from two hosts belonging to the family Pomacentridae, Dascyllus reticulatus (Richardson) and Pomacentrus chrysurus (Cuvier) [as P. rhodonotus Bleeker] in Australian waters (Ho & Dojiri, 1976). seta (vs. none) on the atrophied tip of the antenna, no tooth on the maxillary basis (vs. one small tooth – see Fig. 5D in Ho & Dojiri (1976) for P. rohdei and Fig. 6A in Ho & Kim (1995) for P. bicornutus), four setae (vs. Remarks ﬁve setae) on the distal exopodal segment of leg 1, one seta (vs. two setae) on leg 1 endopod and four setae (vs. 3 or 5 setae) on the distal exopodal segment of leg 2 in the female; and by having one seta (vs. 3 or 5 setae) on the atrophied tip of the antenna, and no setae (vs. 2 or 4 setae) on the endopod of legs 1 and 2 in the male. Conﬂict of interest The author has no competing interests. Kabata, Z. (1992). Copepoda parasitic on Australian ﬁshes, XIV. An assemblage of bathypelagic species. Journal of Natural History, 26, 9–45. Open Access This article is licensed under a Creative Com- mons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any med- ium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Østergaard, P., & Boxshall, G. A. (2004). Giant females and dwarf males: a comparative study of nuptial organs in female Chondracanthidae (Crustacea: Copepoda). Zoolo- gischer Anzeiger, 243, 65–74. Østergaard, P., Boxshall, G. A., & Quicke, D. L. J. (2005). Dwarfs or giants? Sexual size dimorphism in Chondra- canthidae (Poecilostomatoida, Copepoda). Crustaceana, 78, 397–408. Shiino, S. M. (1960). Two new parasitic copepods belonging to a new genus Prochondracanthopsis (Chondracanthidae). Report of the Faculty of Fisheries, Prefectural University of Mie, 3(3), 518–526. Tang, D., & Ho, J.-S. (2005). Resurrection of Acanthocan- thopsis Heegaard, 1945 (Copepoda, Poecilostomatoida, Chondracanthidae), parasitic on tetraodontiform ﬁshes. Crustaceana, 78, 809–818. Author contributions GAB is sole author Author contributions GAB is sole author Funding The Natural History Museum, London allowed the author access to laboratory facilities as a Scientiﬁc Associate. y Huys, R., & Boxshall, G. A. (1991). Copepod Evolution. Lon- don: The Ray Society, 468 pp. Izawa, K. (1975). A new and a known Chondracanthid copepods parasitic on ﬁshes from Tanabe Bay. Annotationes Zoo- logicae Japonenses, 48, 108–118. Data availability Type material is stored in the Queensland Museum and Natural History Museum, London and is available for study. Kabata, Z. (1965). Parasitic Copepods of ﬁshes. Report of the B.A.N.Z. Antarctic Research Expedition, 8(6), 1–16. Discussion Johnson for assistance in the ﬁeldwork, Bryony Dixon and Libby Liggins for conducting some of the parasite surveys, and Olivia Cenne for data compilation and producing the photographs. ASG is grateful for the support of the staff of the Lizard Island Research Station. ASG and the author acknowledge the Dingaal people, the traditional owners and custodians of the land and country upon which this research was carried out and pay respects to elders past and present of this nation. ASG’s research was funded by the Australian Museum in the form of a Post Graduate Award and a Lizard Island Doctoral fellowship and by James Cook University and The University of Queensland. Ho, J.-S. (1984). Accessory antennule and the origin of the Chondracanthidae (Poecilostomatoida). Crustaceana Sup- plement 7, 242–248. Ho, J.-S. & Dojiri, M. (1976). Parasitic copepods on the ﬁshes of the Great Barrier Reef, Australia. Part 1. Cyclopoida. Publications of the Seto Marine Laboratory, XXIII, 257–273. Ho, J.-S. & Dojiri, M. (1988). Copepods of the Family Chon- dracanthidae parasitic on Australian marine ﬁshes. Aus- tralian Journal of Zoology, 36, 273–291. Ho, J.-S. & Kim, I. H. (1995). Chondracanthid copepods (Poe- cilostomatoida) parasitic on ﬁshes of the Sea of Japan. Reports of the Sado Marine Biological Station, Niigata University, 25, 31–44. Author contributions GAB is sole author Discussion For example, all three species have a reduced number of setae on the atrophied tip of the antenna, an accessory claw on the maxillary basis, a spinulose lobe on the basis of the maxilliped, subequal legs 1 and 2, and a reduced number of setae on the endopod of both legs in the adult female; and lack a strong claw on the atrophied tip of the antenna and have a reduced number of setae on the endopod of leg 2 in the adult male. In addition to the antennules, the new species differs from P. rohdei and P. bicornutus by having a large ﬂeshy process near the base of each antenna, one Ho & Dojiri (1976) noted that the ‘‘pygmy’’ male of P. rohdei attached to one of a pair of small processes found at the junction of the trunk and the genital complex of the adult female. No such processes were observed in the new species but it is clear that the structures reported by Ho & Dojiri (1976) represent the paired nuptial organs. Østergaard & Boxshall (2004) interpreted these specialized structures as secretory organs which provide nutrient secretion that sustains the attached male. Acknowledgements I would like to thank Alexandra S. Grutter (ASG) for collecting the material and making it 123 123 Syst Parasitol (2022) 99:601–610 609 Ho, J.-S. (1973). Chondracanthid copepods parasitic on platy- cephalid ﬁshes of Australia, with discussion of known species occurring on ﬂatheads. Parasitology, 67, 123–131. available to me for study. ASG would like to thank Mark A. Johnson for assistance in the ﬁeldwork, Bryony Dixon and Libby Liggins for conducting some of the parasite surveys, and Olivia Cenne for data compilation and producing the photographs. ASG is grateful for the support of the staff of the Lizard Island Research Station. ASG and the author acknowledge the Dingaal people, the traditional owners and custodians of the land and country upon which this research was carried out and pay respects to elders past and present of this nation. ASG’s research was funded by the Australian Museum in the form of a Post Graduate Award and a Lizard Island Doctoral fellowship and by James Cook University and The University of Queensland. available to me for study. ASG would like to thank Mark A. Declarations Kabata, Z. (1969). Copepoda parasitic on Australian ﬁshes XI. Family Chondracanthidae. Journal of Natural History, 3, 497–507. Conﬂict of interest The author has no competing interests. Conﬂict of interest The author has no competing interests. Conﬂict of interest The author has no competing interests. References Tang, D., Andrews, M., & Cobcroft. J. M. (2007). The ﬁrst chondracanthid (Copepoda: Cyclopoida) reported from cultured ﬁnﬁsh, with a revised key to the species of Chondracanthus. Journal of Parasitology, 93, 788–795. Froese, R., & Pauly, D. (Eds) (2021). FishBase. World Wide Web electronic publication, www.ﬁshbase.org. Grutter, A. S. (1996). Experimental demonstration of no effect by the cleaner wrasse Labroides dimidiatus (Cuvier and Valenciennes) on the host ﬁsh Pomacentrus moluccensis (Bleeker). Journal of Experimental Marine Biololgy and Ecology, 196, 285–298. Tang, D., Kalman, J. E., & Ho, J.-S. (2010). Acanthochondria cyclopsetta Pearse, 1952 and A. alleni n. sp. (Copepoda; Cyclopoida; Chondracanthidae) from ﬂatﬁsh hosts of the U.S.A., with comments on the taxonomic position of A. zebriae Ho, Kim & Kumar, 2000 and A. bicornis Shiino, 1955 and the validity of Pterochondria Ho, 1973. Zootaxa, 2657, 18–32. Heegaard, P. (1940). Some new parasitic copepods (Chondra- canthidae and Lernaeopodidae) from Western Australia. Videnskabelige Meddelelser fra Dansk Naturhistoriske Forening, Copenhagen, 104, 87–101. Venmathi Maran, B. A., Tang, D., Madinabeitia, I., Izawa, K., Ohtsuka, S., Jang, D. J., & Nagasawa, K. (2013). Redescription of Pseudacanthocanthopsis secunda (Yam- aguti and Yamasu, 1960) (Copepoda: Chondracanthidae) Heegaard, P. (1962). Parasitic copepods from Australian waters. Records of the Australian Museum, 25, 149–234. 12 3 3 Syst Parasitol (2022) 99:601–610 610 remarks on two known species. Biological Journal of Okayama University, 5(3-4), 89–165. parasitic on marine ﬁshes from the Seto Inland Sea, Japan and the East China Sea off Japan and Korea. Journal of Natural History, 47, 553–562. Yamaguti, S., & Yamasu, T. (1960). Two new species of copepods parasitic on Japanese ﬁshes. Publications of the Seto Marine Laboratory, 8, 137–141. Walter, T. C., & Boxshall, G. A. (2021). World of Copepods Database. Chondracanthidae Milne Edwards, 1840. Accessed through: World Register of Marine Species at: https://www.marinespecies.org/aphia.php?p= taxdetails&id=128566 on 2021-11-10 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Yamaguti, S., & Yamasu, T. (1959). Parasitic copepods from ﬁshes of Japan, with descriptions of 26 new species and 123 12
https://openalex.org/W4386417202	https://lipidworld.biomedcentral.com/counter/pdf/10.1186/s12944-023-01910-3	English	null	Correlation between serum soluble ASGR1 concentration and low-density lipoprotein cholesterol levels: a cross-sectional study	Lipids in health and disease	2,023	cc-by	7,368	Open Access Open Access © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecom- mons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Abstract Background Recent studies have shown that loss-of-function mutations in hepatic asialoglycoprotein receptor 1 (ASGR1) are associated with low levels of circulating cholesterol and a reduced risk of coronary artery disease (CAD). In contrast to ASGR1 on the hepatocyte membrane, serum soluble ASGR1 (sASGR1) is a secreted form that has been detected in circulation. However, the functions of serum sASGR1 are unclear. This study aims to investigate the rela- tionship between human serum sASGR1 concentration and low-density lipoprotein cholesterol (LDL-C) levels. Methods In a cohort of 134 participants who underwent coronary angiography examination, basic information was recorded, and blood samples were collected for biochemical testing. The serum sASGR1 concentration was deter- mined by ELISA kits. The relationship between sASGR1 concentration and LDL-C levels was examined using linear regression models and interaction tests. Univariate and multivariate analyses were used to identify clinical variables that affect sASGR1 levels. Results After adjusting for potential confounders such as age, sex, BMI, and statin use, the serum sASGR1 concentra- tion was positively correlated with LDL-C levels (β = 0.093, 95% CI: 0.04 to 0.14, P < 0.001). Subgroup analysis and inter- action tests showed that the effect of serum sASGR1 concentration on LDL-C levels was significantly influenced by hypertension status (P for interaction = 0.0067). The results of a multivariate linear regression analysis incorporat- ing age, serum TG, LDL-C, nonesterified fatty acid (NEFA), white blood cell counts (WBCC), and fibrinogen revealed that LDL-C (β = 1.005, 95% CI: 0.35 to 1.66, P = 0.003) and WBCC (β = 0.787, 95% CI: 0.41 to 1.16, P < 0.0001) were inde- pendent influencing factors for serum sASGR1 levels. Conclusions The serum sASGR1 concentration was positively correlated with LDL-C levels. In addition, hyperten- sion status significantly affected the effect of serum sASGR1 on LDL-C levels. This study provides some research ideas for clinical doctors and researchers, as well as some references for additional research on serum sASGR1. Conclusions The serum sASGR1 concentration was positively correlated with LDL-C levels. In addition, hyperten- sion status significantly affected the effect of serum sASGR1 on LDL-C levels. This study provides some research ideas for clinical doctors and researchers, as well as some references for additional research on serum sASGR1. Conclusions The serum sASGR1 concentration was positively correlated with LDL-C levels. In addition, hyperten- sion status significantly affected the effect of serum sASGR1 on LDL-C levels. Abstract This study provides some research ideas for clinical doctors and researchers, as well as some references for additional research on serum sASGR1. Keywords ASGR1, Cholesterol metabolism, Low-density lipoprotein cholesterol, Atherosclerotic cardiovascular disease, Lipid metabolism Keywords ASGR1, Cholesterol metabolism, Low-density lipoprotein cholesterol, Atherosclerotic cardiovascular disease, Lipid metabolism Keywords ASGR1, Cholesterol metabolism, Low-density lipoprotein cholesterol, Atherosclerotic cardiovascular disease, Lipid metabolism luofei0058@csu.edu.cn Full list of author information is available at the end of the article Lipids in Health and Disease Lipids in Health and Disease Luo et al. Lipids in Health and Disease (2023) 22:142 https://doi.org/10.1186/s12944-023-01910-3 Introduction ASGR1 was identified as two splice variants [14, 15]. In addition to the hASGR1 located on the hepatocyte membrane mentioned above, Liu et al. detected solu- ble ASGR1 (sASGR1) in human serum and HepG2 cell supernatant by western blot analysis, indicating that it was secreted by hepatocytes [15]. Structurally, compared to hASGR1, sASGR1 lacks 117 nucleotides, making it unable to encode a transmembrane domain [16, 17]. Interestingly, they further found that sASGR1 competi- tively binds circulating asialoglycoproteins with hASGR1, and the sASGR1-asialoglycoprotein complex is still cap- tured by hASGR1 on the surface of hepatocytes and enters the cell [15], as shown in Fig. 1. Atherosclerotic cardiovascular disease (ASCVD) is still one of the diseases with the highest mortality and dis- ability rates in the world. Many studies have shown that abnormal lipid metabolism, especially high levels of low- density lipoprotein cholesterol (LDL-C), is a causal risk factor for ASCVD [1, 2]. Cholesterol-lowering therapy can significantly reduce the risk of cardiovascular events [3–5]. A meta-analysis showed that for every 1 mmol/l reduction in LDL-C, the risk of major vascular events decreased by 21% [6]. However, even with the use of high-intensity statins for lipid-lowering, certain patients are still at risk for cardiovascular events, so called resid- ual cardiovascular risk [7, 8]. Proprotein convertase sub- tilisin/kexin type 9 (PCSK9) inhibitors and statins, which are the most effective cholesterol-lowering medica- tions, increase LDL receptor (LDLR) expression to lower LDL-C levels [9, 10]. However, they are ineffective in patients with homozygous familial hypercholesterolemia [11]. Therefore, it is of great significance to find new lipid-lowering targets and develop new drugs. However, current research mainly focuses on the identity of ASGR1 as a liver surface receptor. The func- tions of sASGR1 are largely unknown. As an alterna- tive splice form of ASGR1, sASGR1 may have a similar function to hASGR1 or be able to reflect the expression level of hASGR1. However, sASGR1 may also regu- late lipid metabolism directly. The relationship between serum sASGR1 concentration and circulating cholesterol is unclear. Our study aims to investigate the relation- ship between human serum sASGR1 concentration and LDL-C levels. p g g p g In 2016, a human genetics study by Nioi et al. found that loss-of-function mutations in hepatic asialoglyco- protein receptor 1 (ASGR1) were associated with low levels of circulating cholesterol and a reduced risk of coronary artery disease (CAD) [12]. Recently, Wang et al. Introduction found that inhibiting the expression of hepatic ASGR1 (hASGR1) or its binding to circulating asialoglycopro- teins promotes cholesterol efflux into bile, resulting in a decrease of approximately 25% in circulating total choles- terol (TC) and LDL-C levels [13]. These studies suggest that hASGR1 is expected to be a potential new lipid-low- ering target. © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecom- mons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Luo et al. Lipids in Health and Disease (2023) 22:142 Page 2 of 8 Luo et al. Lipids in Health and Disease (2023) 22:142 Page 2 of 8 Clinical characteristics and biochemical measurementsh Clinical characteristics and biochemical measurements The general clinical data of patients were obtained through the hospital’s electronic medical record system. Basic information such as the name, sex, age, BMI, and disease history of each patient was recorded. After fast- ing overnight, peripheral blood samples were collected. Serum was isolated from the blood samples and stored at -80 ℃ for subsequent analysis. Serum triglycerides (TG), TC, LDL-C, high-density lipoprotein cholesterol (HDL-C), lipoprotein (a) [Lp (a)], nonesterified fatty acid (NEFA), apolipoprotein-A1 (APO-A1), apolipoprotein-B (APO-B), alanine aminotransferase (ALT), aspartate ami- notransferase (AST), albumin, creatinine, uric acid (UA), calcium, D-dimer, fibrinogen, high sensitivity troponin T (hsTNT), and N-terminal pro-brain natriuretic peptide (NT-proBNP) were determined using Hitachi 7600 series automatic analyzers in the Department of Laboratory Medicine of our hospital. Remanent cholesterol (RC) was estimated as TC minus LDL-C minus HDL-C. An automatic blood cell counter was used to measure white blood cell counts (WBCC). sASGR1 sandwich enzyme- linked immunosorbent assay (ELISA) kits (JL41965; Jianglai Biology, Shanghai) were used to measure the concentrations of serum sASGR1. Diagnostic criteria Clinical symptoms, ischemic electrocardiographic abnor- malities, and coronary angiography demonstrating at least one main coronary artery stenosis of at least 50% were diagnostic criteria for CAD. The definition of acute myocardial infarction (AMI) is based on the fourth uni- versal definition of myocardial infarction published by the ESC/ACC/AHA/WHF Task Force [18]. The diag- nostic criteria for type 2 diabetes mellitus (T2DM) were fasting blood glucose ≥ 126 mg/dl (7.0 mmol/l) or 2-h postprandial blood glucose ≥ 200 mg/dl (11.1 mmol/l). Hypertension was diagnosed when resting systolic blood pressure was > 140 mmHg or diastolic blood pressure was > 90 mmHg on a different day. Univariate analysis We utilized univariate linear regression analysis to assess whether the concentration of serum sASGR1 affects the level of LDL-C. As shown in Table 2, there was a signifi- cant difference in the effect of sASGR1 concentration on LDL-C (β = 0.10, P < 0.0001), which means that for every 1 ng/mL increase in sASGR1 concentration, the level of LDL-C rose by 0.1 mmol/L. In addition, TG, TC, RC, APO-B, NEFA, and the occurrence of CAD and AMI had statistically significant effects on LDL-C. Sex, age, BMI, statin use, smoking, HDL-C, Lp(a), APO-A1, and the occurrence of T2DM and hypertension were not associ- ated with LDL-C. Materials and methods Study subjectsh The study subjects were 134 patients who had undergone coronary angiography in the Department of Cardiovas- cular Medicine at the Second Xiangya Hospital of Cen- tral South University from September 2022 to April 2023. Patients with liver or kidney diseases, infections, autoim- mune diseases, malignancies, and other severe diseases Fig. 1 Relationship between serum sASGR1 and hASGR1. Circulating asialoglycoproteins bind hASGR1 on the surface of hepatocytes, ultimately resulting in reduced intracellular cholesterol efflux (red arrow). Circulating sASGR1 secreted by the liver competitively binds asialoglycoproteins with hASGR1. The sASGR1-asialoglycoprotein complex also enters the liver through hASGR1 on the surface of hepatocytes (black arrow) Fig. 1 Relationship between serum sASGR1 and hASGR1. Circulating asialoglycoproteins bind hASGR1 on the surface of hepatocytes, ultimately resulting in reduced intracellular cholesterol efflux (red arrow). Circulating sASGR1 secreted by the liver competitively binds asialoglycoproteins with hASGR1. The sASGR1-asialoglycoprotein complex also enters the liver through hASGR1 on the surface of hepatocytes (black arrow) Fig. 1 Relationship between serum sASGR1 and hASGR1. Circulating asialoglycoproteins bind hASGR1 on the surface of hepatocytes, ultimately resulting in reduced intracellular cholesterol efflux (red arrow). Circulating sASGR1 secreted by the liver competitively binds asialoglycoproteins with hASGR1. The sASGR1-asialoglycoprotein complex also enters the liver through hASGR1 on the surface of hepatocytes (black arrow) Luo et al. Lipids in Health and Disease (2023) 22:142 Page 3 of 8 Page 3 of 8 based on the data distribution. Categorical variables are expressed as frequencies or percentages. We used univar- iate and multivariate linear regression models to evaluate the relationship between serum sASGR1 concentration and LDL-C levels as well as to identify clinical variables that affect sASGR1 levels. Both the crude and adjusted results from our models were provided in accordance with the STROBE statement’s standards [19]. Stratified linear regression models were utilized to conduct sub- group analyses, and likelihood ratio tests were utilized to identify changes and interactions between subgroups. were excluded. The Second Xiangya Hospital’s ethics committee approved the study and determined that it complied with the ethical principles of the Declaration of Helsinki. Each patient gave informed consent. Relationship between sASGR1 and LDL‑C We further used multivariate analysis to evaluate the relationship between serum sASGR1 and LDL-C. Table 3 shows the crude model and three models after incorpo- rating confounding factors. In the crude model, there was a significant difference in the impact of serum sASGR1 on LDL-C [β = 0.1, 95% confidence interval (CI): 0.06 to 0.14, P < 0.001]. Similarly, in the three models (model I, model II, and model III), after incorporating other covar- iates, such as age, sex, and BMI, serum sASGR1 was still the influencing factor of LDL-C. Baseline characteristics of study participantsh The baseline characteristics of the participants are shown in Table 1. There were 134 participants in total, with a mean age of 54.6 ± 13.8 years and a male participation rate of 47.76%. The number of patients with CAD was 84 (62.69%), and among them, 21 were diagnosed with AMI. Sixty patients (44.78%) used statins for lipid lowering. The proportion of smokers was 30.6%. The proportions of patients with T2DM and hypertension were 23.88% and 47.01%, respectively. The serum sASGR1 concentration ranged from 0.61 to 19.13 ng/ml (median: 2.49 ng/ml). Statistical analysesh The data were analyzed using EmpowerStats, statistical software (version 4.1, http://www.empowerstats.com, X&Y Solutions, Inc., Boston, MA), and R language (ver- sion 4.2.0, http://www.R-project.org, The R Foundation). In this study, the statistical threshold was set at P = 0.05. Continuous variables are expressed as the mean ± standard deviation or as medians and quartiles Luo et al. Lipids in Health and Disease (2023) 22:142 Page 4 of 8 Table 2 The results of univariate analysis Abbreviations: CI Confidence interval Other abbreviations are as shown in Table 1 LDL-C β (95% CI) P value Sex Male ref Female -0.14 (-0.46, 0.18) 0.3987 Age (years) 0.01 (-0.00, 0.02) 0.1842 BMI (kg/m2) 0.02 (-0.02, 0.07) 0.3270 Statin use No 0 Yes 0.01 (-0.32, 0.33) 0.9622 Smoke No 0 Yes 0.09 (-0.26, 0.44) 0.6116 AMI No 0 Yes 0.53 (0.10, 0.97) 0.0184 TG (mmol/L) 0.21 (0.10, 0.33) 0.0003 TC (mmol/L) 0.78 (0.73, 0.83) < 0.0001 HDL-C (mmol/L) 0.23 (-0.36, 0.81) 0.4511 RC (mmol/L) 0.76 (0.33, 1.19) 0.0007 Lp (a) (mg/L) 0.00 (-0.00, 0.00) 0.1280 NEFA (mmol/L) 0.95 (0.23, 1.67) 0.0113 APO-A1 (g/L) 0.70 (-0.41, 1.81) 0.2167 APO-B (g/L) 3.47 (3.25, 3.68) < 0.0001 WBCC (109/L) 0.10 (0.01, 0.19) 0.0319 CAD No 0 Yes 0.36 (0.03, 0.69) 0.0340 T2DM No 0 Yes 0.27 (-0.10, 0.65) 0.1574 Hypertension No 0 Yes 0.17 (-0.16, 0.49) 0.3183 sASGR1 (ng/mL) 0.10 (0.06, 0.14) < 0.0001 Subgroup analysis and interaction tests As shown in Table 4, the interaction test showed that the effect of sASGR1 on LDL-C was significantly affected by hypertension status (P for interaction = 0.0067), while age, sex, BMI, and the status of CAD and T2DM were not statistically significant in the interaction test. Statistical analysesh Our results showed that the effect size of sASGR1 on LDL-C showed differences between hypertensive and nonhypertensive Table 1 Characteristics of study participants The data displayed are the mean ± SD, median (interquartile range), or n (%) Abbreviations: BMI body mass index, TG triglyceride, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, i Patients (n = 134) Age (years) 54.59 ± 13.78 BMI (kg/m2) 23.70 ± 3.29 TG (mmol/L) 1.46 (1.03, 2.06) TC (mmol/L) 4.42 (3.88, 5.27) LDL-C (mmol/L) 2.83 (2.36, 3.55) HDL-C (mmol/L) 1.12 (0.96, 1.29) RC (mmol/L) 0.43 (0.32, 0.56) Lp (a) (mg/L) 174.70 (76.33, 382.10) NEFA (mmol/L) 0.35 (0.25, 0.52) APO-A1 (g/L) 1.02 ± 0.17 APO-B (g/L) 0.77 (0.65, 0.97) WBCC (109/L) 6.45 (5.34, 7.72) ALT (μ/L) 17.20 (12.60, 25.38) AST (μ/L) 17.70 (15.40, 25.10) Albumin (g/L) 38.60 (37.05, 40.93) Creatinine (μmol/L) 70.00 (59.00, 85.23) UA (μmol/L) 316.00 (260.50, 382.50) Calcium (mmol/L) 2.19 (2.12, 2.28) Fibrinogen (g/L) 3.35 (2.90, 4.00) D-dimer (μg/mL) 0.26 (0.20, 0.42) Hs-TNT (pg/mL) 10.44 (5.33, 29.21) NT-proBNP (pg/mL) 174.00 (52.45, 1008.00) sASGR1 (ng/mL) 2.49 (1.71, 4.11) Sex N (%) Male 64 (47.76%) Female 70 (52.24%) Statin use No 74 (55.22%) Yes 60 (44.78%) Smoke No 93 (69.40%) Yes 41 (30.60%) AMI No 113 (84.33%) Yes 21 (15.67%) CAD No 50 (37.31%) Yes 84 (62.69%) T2DM No 102 (76.12%) Yes 32 (23.88%) Hypertension No 71 (52.99%) Yes 63 (47.01%) Table 2 The results of univariate analysis Abbreviations: CI Confidence interval Other abbreviations are as shown in Table 1 LDL-C β (95% CI) P value Sex Male ref Female -0.14 (-0.46, 0.18) 0.3987 Age (years) 0.01 (-0.00, 0.02) 0.1842 BMI (kg/m2) 0.02 (-0.02, 0.07) 0.3270 Statin use No 0 Yes 0.01 (-0.32, 0.33) 0.9622 Smoke No 0 Yes 0.09 (-0.26, 0.44) 0.6116 AMI No 0 Yes 0.53 (0.10, 0.97) 0.0184 TG (mmol/L) 0.21 (0.10, 0.33) 0.0003 TC (mmol/L) 0.78 (0.73, 0.83) < 0.0001 HDL-C (mmol/L) 0.23 (-0.36, 0.81) 0.4511 RC (mmol/L) 0.76 (0.33, 1.19) 0.0007 Lp (a) (mg/L) 0.00 (-0.00, 0.00) 0.1280 NEFA (mmol/L) 0.95 (0.23, 1.67) 0.0113 APO-A1 (g/L) 0.70 (-0.41, 1.81) 0.2167 APO-B (g/L) 3.47 (3.25, 3.68) < 0.0001 WBCC (109/L) 0.10 (0.01, 0.19) 0.0319 CAD No 0 Yes 0.36 (0.03, 0.69) 0.0340 T2DM No 0 Yes 0.27 (-0.10, 0.65) 0.1574 Hypertension No 0 Yes 0.17 (-0.16, 0.49) 0.3183 sASGR1 (ng/mL) 0.10 (0.06, 0.14) < 0.0001 Table 1 Characteristics of study participants Other abbreviations are as shown in Table 1 The data displayed are the mean ± SD, median (interquartile range), or n (%) Abbreviations: BMI body mass index, TG triglyceride, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, RC remnant cholesterol, Lp (a) lipoprotein (a), NEFA nonesterified fatty acid, APO-A1 apolipoprotein A1, APO-B apolipoprotein B, WBCC white blood cell counts, sASGR1 soluble ASGR1, ALT alanine aminotransferase, AST aspartate aminotransferase, UA uric acid, hs-TNT high sensitivity troponin T, NT-proBNP N-terminal pro-brain natriuretic peptide, AMI acute myocardial infarction, CAD coronary artery disease, T2DM type 2 diabetes mellitus The data displayed are the mean ± SD, median (interquartile range), or n (%) Discussion In this study, we used a linear regression model to explore the relationship between serum sASGR1 and LDL-C lev- els. Univariate analysis showed a positive relationship between serum sASGR1 concentration and LDL-C levels (β = 0.1, P < 0.001). Similar results were still obtained in the multivariable linear regression models incorporating factors, such as age, sex, BMI, and smoking status. n each case, the model is not adjusted for the stratification variable g g Previous animal studies have shown that the bind- ing of hASGR1 to asialoglycoproteins in the circula- tion determines the flow of hepatic cholesterol efflux into bile [13]. Inhibiting hASGR1 or its binding to cir- culating asialoglycoproteins will lead to increased cho- lesterol efflux, thereby reducing circulating TC and LDL-C levels [13]. Our study demonstrated a positive correlation between serum sASGR1 and LDL-C, which is similar to the relationship between hASGR1 and LDL-C, suggesting that the concentration of sASGR1 may reflect the protein expression level of hASGR1 to some extent. However, whether sASGR1 is directly involved in the metabolism of LDL-C remains unclear. showed differences in CAD status and age, although the interaction test was not statistically significant. For non- CAD patients and patients younger than 60 years, serum sASGR1 was not significantly associated with LDL-C. showed differences in CAD status and age, although the interaction test was not statistically significant. For non- CAD patients and patients younger than 60 years, serum sASGR1 was not significantly associated with LDL-C. Subgroup analysis and interaction tests As shown in Table 4, the interaction test showed that the effect of sASGR1 on LDL-C was significantly affected by hypertension status (P for interaction = 0.0067), while age, sex, BMI, and the status of CAD and T2DM were not statistically significant in the interaction test. Our results showed that the effect size of sASGR1 on LDL-C showed differences between hypertensive and nonhypertensive patients. In hypertensive patients, serum sASGR1 was positively correlated with LDL-C levels [β = 0.15, 95% CI (0.08, 0.23), P < 0.001], while this relationship could not be observed in nonhypertensive patients. In addi- tion, we noticed that the effect of sASGR1 on LDL-C The data displayed are the mean ± SD, median (interquartile range), or n (%) Abbreviations: BMI body mass index, TG triglyceride, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, RC remnant cholesterol, Lp (a) lipoprotein (a), NEFA nonesterified fatty acid, APO-A1 apolipoprotein A1, APO-B apolipoprotein B, WBCC white blood cell counts, sASGR1 soluble ASGR1, ALT alanine aminotransferase, AST aspartate aminotransferase, UA uric acid, hs-TNT high sensitivity troponin T, NT-proBNP N-terminal pro-brain natriuretic peptide, AMI acute myocardial infarction, CAD coronary artery disease, T2DM type 2 diabetes mellitus Luo et al. Lipids in Health and Disease (2023) 22:142 Page 5 of 8 Table 3 Multivariate analysis for the association between sASGR1 and LDL-C in different models Model I adjusted for age and sex Model II adjusted for model I plus BMI, statin use, smoking status, and hypertension Model III adjusted for models I and II plus AMI, NEFA, WBCC, CAD, and T2DM Outcome Crude Model Model I Model II Model III β P β P β P β P (95%CI) (95%CI) (95%CI) (95%CI) sASGR1 (ng/mL) 0.100 < 0.001 0.098 < 0.001 0.096 < 0.001 0.093 < 0.001 (0.06, 0.14) (0.06, 0.14) (0.06, 0.14) (0.04, 0.14) Table 3 Multivariate analysis for the association between sASGR1 and LDL-C in different models LDL-C, NEFA, APO-A1, APO-B, fibrinogen, WBCC, and AST. Sex, BMI, statin use, smoking, serum HDL-C, Lp(a), ALT, albumin, creatinine, UA, calcium, D-dimer, hs-TNT, NT-proBNP, and the presence of T2DM, AMI, and hypertension were not found to be associated with serum sASGR1 concentration. We then performed a multivariate linear regression analysis (Table 6). Based on sample size, collinearity between independent vari- ables, and univariate analysis results, we set age, LDL-C, TG, NEFA, WBCC, and fibrinogen as independent vari- ables. Subgroup analysis and interaction tests The results showed that the effect of LDL-C level on the concentration of sASGR1 was statistically signifi- cant [β = 1.005, 95% CI (0.35, 1.66), P = 0.003]. Notably, WBCC was also an independent factor affecting sASGR1 concentrations [β = 0.787, 95% CI (0.41, 1.16), P < 0.0001]. Table 4 Subgroup analysis for the association between sASGR1 and LDL-C V i bl Eff t i (95%CI) P l P f i t ti Table 4 Subgroup analysis for the association between sASGR1 and LDL-C The above model adjusted for age, sex, BMI, statin use, smoking status, hypertension, AMI, NEFA, WBCC, CAD, and T2DM In each case, the model is not adjusted for the stratification variable Variables Effect size (95%CI) P value P for interaction Age (years) 0.39 Below 60 0.06 (-0.01, 0.13) 0.09 Over 60 0.09 (0.02,0.17) 0.02 Sex 0.73 Male 0.08 (0.01, 0.14) 0.02 Female 0.11 (0.02, 0.21) 0.02 BMI (kg/m2) 0.39 < 24 0.07 (0.01, 0.13) 0.02 > = 24 0.11 (0.01, 0.21) 0.04 CAD 0.14 No -0.03 (-0.17, 0.11) 0.66 Yes 0.10 (0.04, 0.15) < 0.001 Hypertension 0.0067 No 0.04 (-0.04, 0.11) 0.33 Yes 0.15 (0.08, 0.23) < 0.001 T2DM 0.53 No 0.10 (0.03, 0.16) 0.004 Yes 0.13 (0.06, 0.19) 0.002 The relationship between clinical variables and sASGR1 concentrationf To identify other clinical variables that affect the con- centration of sASGR1, we conducted a univariate linear regression analysis. As shown in Table 5, the concentra- tion of serum sASGR1 is affected by age, serum TG, TC, Luo et al. Lipids in Health and Disease (2023) 22:142 Luo et al. Lipids in Health and Disease Page 6 of 8 PubMed literature search, only one study on ASGR1 and hypertension was obtained. This study showed that the single nucleotide polymorphism rs34870220 of ASGR1 was associated with the occurrence of essen- tial hypertension [20]. The authors speculated that ASGR1 may be involved in vascular endothelial injury [20]. Because there are few similar reports at present, we cannot explain why the linear correlation between sASGR1 and LDL-C can only be observed in hyperten- sive patients. Additionally, we noticed that the effect of serum sASGR1 on LDL-C showed differences in CAD status, although the P for the interaction test did not reach statistical significance. The interaction P value is related to the overall sample size and the number of stratified individuals. Therefore, the insufficient num- ber of non-CAD patients may be one of the reasons. In addition, the degree of dispersion of LDL-C levels in non-CAD patients was less than that in CAD patients (0.65 vs. 1.08 mmol/l), while serum sASGR1 may be affected by other unknown factors, showing a greater degree of dispersion in our results. This may also be one of the reasons for the lack of correlation between serum sASGR1 and LDL-C in non-CAD and nonhy- pertensive patients. However, we cannot rule out the possibility that there is no correlation between serum sASGR1 and LDL-C under healthy conditions. Furthermore, we attempted to identify clinical vari- ables that affect sASGR1 levels. Although univariate analysis suggests that a number of clinical variables are associated with sASGR1 levels, multivariate linear regression analysis shows that LDL-C and WBCC are independent influencing factors for sASGR1 levels. The relationship between clinical variables and sASGR1 concentrationf Table 5 Univariable correlations of the clinical variables with the serum sASGR1 levels Abbreviations are as shown in Table 1 Variables β (95%CI) P value Sex Male ref Female -0.31 (-1.63, 1.02) 0.6511 Age (years) 0.05 (0.00, 0.10) 0.0466 BMI (kg/m2) 0.07 (-0.13, 0.27) 0.4985 Statin use No 0 Yes 0.42 (-0.90, 1.75) 0.5337 Smoke No 0 Yes 0.40 (-1.03, 1.83) 0.5827 AMI No 0 Yes 1.84 (-0.36, 4.05) 0.1051 T2DM No 0 Yes 1.52 (-0.01, 3.04) 0.0535 Hypertension No 0 Yes 0.76 (-0.56, 2.07) 0.2629 TG (mmol/L) 0.65 (0.17, 1.12) 0.0082 TC (mmol/L) 1.55 (1.04, 2.07) < 0.0001 LDL-C (mmol/L) 1.66 (1.03, 2.30) < 0.0001 HDL-C (mmol/L) 1.28 (-1.10, 3.66) 0.2928 Lp(a) (mg/L) -0.00 (-0.00, 0.00) 0.8501 NEFA (mmol/L) 6.26 (3.43, 9.08) < 0.0001 APO-A1 (g/L) 6.73 (2.31, 11.15) 0.0035 APO-B (g/L) 7.01 (4.55, 9.48) < 0.0001 Fibrinogen (g/L) 0.96 (0.38, 1.55) 0.0016 WBCC (109/L) 0.86 (0.51, 1.20) < 0.0001 ALT (μ/L) 0.03 (-0.01, 0.07) 0.1048 AST (μ/L) 0.03 (0.00, 0.05) 0.0242 Albumin (g/L) 0.16 (-0.03, 0.34) 0.0951 Creatinine (μmol/L) 0.01 (-0.03, 0.04) 0.7542 UA (μmol/L) 0.00 (-0.00, 0.01) 0.4062 Calcium (mmol/L) 0.97 (-3.13, 5.06) 0.6447 D-dimer (μg/mL) -0.10 (-0.62, 0.41) 0.6959 Hs-TNT (pg/mL) 0.00 (-0.00, 0.00) 0.2316 NT-proBNP (pg/mL) -0.00 (-0.00, 0.00) 0.8771 Table 6 Multiple regression analysis detecting independent contributors to sASGR1 levels Abbreviations are as shown in Table 1 Variables β (95%CI) P value Vif Age (years) 0.009 (-0.03, 0.05) 0.676 1.185 TG (mmol/L) 0.274 (-0.16, 0.71) 0.211 1.149 LDL-C (mmol/L) 1.005 (0.35, 1.66) 0.003 1.242 NEFA (mmol/L) 2.405 (-0.20, 5.01) 0.07 1.229 WBCC (109/L) 0.787 (0.41, 1.16) < 0.0001 1.324 Fibrinogen (g/L) 0.005 (-0.66, 0.67) 0.989 1.465 Table 5 Univariable correlations of the clinical variables with the serum sASGR1 levels Abbreviations are as shown in Table 1 Variables β (95%CI) P value Sex Male ref Female -0.31 (-1.63, 1.02) 0.6511 Age (years) 0.05 (0.00, 0.10) 0.0466 BMI (kg/m2) 0.07 (-0.13, 0.27) 0.4985 Statin use No 0 Yes 0.42 (-0.90, 1.75) 0.5337 Smoke No 0 Yes 0.40 (-1.03, 1.83) 0.5827 AMI No 0 Yes 1.84 (-0.36, 4.05) 0.1051 T2DM No 0 Yes 1.52 (-0.01, 3.04) 0.0535 Hypertension No 0 Yes 0.76 (-0.56, 2.07) 0.2629 TG (mmol/L) 0.65 (0.17, 1.12) 0.0082 TC (mmol/L) 1.55 (1.04, 2.07) < 0.0001 LDL-C (mmol/L) 1.66 (1.03, 2.30) < 0.0001 HDL-C (mmol/L) 1.28 (-1.10, 3.66) 0.2928 Lp(a) (mg/L) -0.00 (-0.00, 0.00) 0.8501 NEFA (mmol/L) 6.26 (3.43, 9.08) < 0.0001 APO-A1 (g/L) 6.73 (2.31, 11.15) 0.0035 APO-B (g/L) 7.01 (4.55, 9.48) < 0.0001 Fibrinogen (g/L) 0.96 (0.38, 1.55) 0.0016 WBCC (109/L) 0.86 (0.51, 1.20) < 0.0001 ALT (μ/L) 0.03 (-0.01, 0.07) 0.1048 AST (μ/L) 0.03 (0.00, 0.05) 0.0242 Albumin (g/L) 0.16 (-0.03, 0.34) 0.0951 Creatinine (μmol/L) 0.01 (-0.03, 0.04) 0.7542 UA (μmol/L) 0.00 (-0.00, 0.01) 0.4062 Calcium (mmol/L) 0.97 (-3.13, 5.06) 0.6447 D-dimer (μg/mL) -0.10 (-0.62, 0.41) 0.6959 Hs-TNT (pg/mL) 0.00 (-0.00, 0.00) 0.2316 NT-proBNP (pg/mL) -0.00 (-0.00, 0.00) 0.8771 Table 6 Multiple regression analysis detecting independent contributors to sASGR1 levels Table 6 Multiple regression analysis detecting independent contributors to sASGR1 levels Abbreviations are as shown in Table 1 Variables β (95%CI) P value Vif Age (years) 0.009 (-0.03, 0.05) 0.676 1.185 TG (mmol/L) 0.274 (-0.16, 0.71) 0.211 1.149 LDL-C (mmol/L) 1.005 (0.35, 1.66) 0.003 1.242 NEFA (mmol/L) 2.405 (-0.20, 5.01) 0.07 1.229 WBCC (109/L) 0.787 (0.41, 1.16) < 0.0001 1.324 Fibrinogen (g/L) 0.005 (-0.66, 0.67) 0.989 1.465 PubMed literature search, only one study on ASGR1 and hypertension was obtained. Conclusion Since the relationship between hASGR1 and serum cholesterol was discovered in 2016, ASGR1 has attracted researchers’ attention mainly as a liver surface receptor. Although the presence of sASGR1 has been confirmed, there are few studies on serum sASGR1, especially inves- tigating its relationship with blood lipids. This is the first investigation of the correlation between serum sASGR1 and LDL-C levels that we are aware of. In the follow-up study, the relationship between serum sASGR1 and CAD risk needs further investigation. The current study sug- gests that hASGR1 may become a future lipid-lowering target. Inhibiting hASGR1 or its binding to circulating asialoglycoproteins may reduce circulating cholesterol by promoting cholesterol efflux. Therefore, the application of ASGR1 inhibitors in populations with high expres- sion of ASGR1 in the liver may be beneficial. However, there is still no simple and effective method to measure the expression level of hASGR1. Whether the concen- tration of serum sASGR1 reflects the expression level of hASGR1 deserves further study. In addition, the function of serum sASGR1 is unknown. Whether serum sASGR1 is also directly involved in the metabolism of LDL-C needs further study. Based on the research results of Liu et al. [15], it is intriguing to see whether the intervention of sASGR1 can directly achieve a lipid-lowering effect if entry of the sASGR1-asialoglycoprotein complex into the liver results in the same downstream effects as circulat- ing asialoglycoprotein entry into the liver via hASGR1. Finally, our study provides some research ideas for clini- cal doctors and researchers, as well as some references for additional research on hepatic and serum ASGR1. Overall, serum sASGR1 concentration was positively correlated with LDL-C levels. Hypertension status sig- nificantly affected the effect of sASGR1 on LDL-C levels. In addition, LDL-C levels and WBCC are independent influencing factors for serum sASGR1 concentration. Future research should focus on the function of serum sASGR1 and how it relates to CAD risk. It’s worth con- sidering whether serum sASGR1 could become a poten- tial lipid-lowering target like hASGR1. The relationship between clinical variables and sASGR1 concentrationf In addition, previous studies have shown that the bind- ing of circulating asialoglycoprotein to hepatic ASGR depends on calcium ions [24]. Our results indicate that the concentration of sASGR1 is not affected by serum calcium levels. and their causality needs to be further confirmed by pro- spective studies, although previous basic studies have revealed the causal relationship between hASGR1 and LDL-C. Second, the sample size of our single-center study is limited, so the results may not be generalizable. Study strengths and limitations Our study has some strengths. First, we investigated the variables affecting serum sASGR1 levels and the cor- relation between sASGR1 and LDL-C levels for the first time. Second, we used rigorous statistical adjustment models to minimize the unavoidable confounding factors of observational studies. Third, we observed a positive correlation between sASGR1 and LDL-C levels in CAD patients in the results of subgroup analysis rather than in non-CAD patients, which means that intervention with sASGR1 may have greater clinical significance.h The relationship between clinical variables and sASGR1 concentrationf This study showed that the single nucleotide polymorphism rs34870220 of ASGR1 was associated with the occurrence of essen- tial hypertension [20]. The authors speculated that ASGR1 may be involved in vascular endothelial injury [20]. Because there are few similar reports at present, we cannot explain why the linear correlation between sASGR1 and LDL-C can only be observed in hyperten- sive patients. Additionally, we noticed that the effect of serum sASGR1 on LDL-C showed differences in CAD status, although the P for the interaction test did not reach statistical significance. The interaction P value is related to the overall sample size and the number of stratified individuals. Therefore, the insufficient num- ber of non-CAD patients may be one of the reasons. In addition, the degree of dispersion of LDL-C levels in non-CAD patients was less than that in CAD patients (0.65 vs. 1.08 mmol/l), while serum sASGR1 may be affected by other unknown factors, showing a greater degree of dispersion in our results. This may also be one of the reasons for the lack of correlation between serum sASGR1 and LDL-C in non-CAD and nonhy- pertensive patients. However, we cannot rule out the possibility that there is no correlation between serum sASGR1 and LDL-C under healthy conditions. Furthermore, we attempted to identify clinical vari- ables that affect sASGR1 levels. Although univariate analysis suggests that a number of clinical variables are associated with sASGR1 levels, multivariate linear regression analysis shows that LDL-C and WBCC are independent influencing factors for sASGR1 levels. This is consistent with our results of setting LDL-C as the dependent variable, indicating a positive correlation between sASGR1 and LDL-C levels. WBCC is a widely recognized classic indicator of inflammation. This result seems to suggest that sASGR1 may be related to systemic inflammation. In fact, some basic studies have revealed a correlation between hASGR1, a membrane surface receptor, and inflammation [21–23]. However, Abbreviations are as shown in Table 1 Subgroup analysis and interaction tests are help- ful to evaluate the heterogeneity of populations with different characteristics. Our results showed that only hypertension status affected the effect of serum sASGR1 on LDL-C after grouping for age, sex, BMI, CAD, hypertension, and T2DM status. Through a Luo et al. Lipids in Health and Disease (2023) 22:142 Page 7 of 8 the relationship between serum sASGR1 and inflam- mation is unknown, and further research is needed. Authors’ contributions FL and QL conceived the idea. QL wrote the manuscript; JC, YS, and PW con- ducted the data collection and analysis. FL, YS, JW, and ZF read through and corrected the manuscript. All authors read and approved the final manuscript. Acknowledgements g We would like to express our gratitude to all those who helped us during the writing of this manuscript. We thank all the peer reviewers for their opinions and suggestions. We appreciate Figdraw for drawing services. Conclusion Abbreviations ASGR1 Asialoglycoprotein receptor 1 CAD Coronary artery disease hASGR1 Hepatic ASGR1 sASGR1 Soluble ASGR1 LDL-C Low-density lipoprotein cholesterol ASCVD Atherosclerotic cardiovascular disease PCSK9 Proprotein convertase subtilisin/kexin type 9 TC Total cholesterol AMI Acute myocardial infarction T2DM Type 2 diabetes mellitus TG Triglycerides HDL-C High-density lipoprotein cholesterol RC Remanent cholesterol Lp (a) Lipoprotein (a) NEFA Nonesterified fatty acid APO-A1 Apolipoprotein-A1 APO-B Apolipoprotein B WBCC White blood cell counts ALT Alanine aminotransferase AST Aspartate aminotransferase UA Uric acid Hs-TNT High sensitivity troponin T NT-proBNP N-terminal pro-brain natriuretic peptide ELISA Enzyme-linked immunosorbent assay Funding This work was supported by grants from the National Key R&D Program of China (grant number 2021YFC2500500), Hunan Provincial Natural Science Foundation of China [grant number 2021JJ40852 to F.Luo, grant number 2022JJ40675 to J.F.Chen]; National Natural Science Foundation of China [grant number 82100495 to F.Luo, grant number 8220187 to J.F.Chen]; Scientific Research Project of Hunan Provincial Health Commission [grant number 202203014009 to F.Luo, grant number B202305037231 to J.F.Chen]; Scientific Research Launch Project for new employees of the Second Xiangya Hospital of Central South University (to F.Luo and J.F.Chen); China Postdoctoral Science i There are some limitations to this study. First, this study is a cross-sectional survey; therefore, the correlation between research factors and outcomes is exploratory, Page 8 of 8 Luo et al. Lipids in Health and Disease (2023) 22:142 Luo et al. Lipids in Health and Disease (2023) 22:142 Foundation (grant number 331046 and 2023T160738 to F.Luo and J.F.Chen); and Key Research and Development Program of Hunan Province of China [grant number 2021SK2004 to Z.F.Fang]. 10. Maxwell KN, Fisher EA, Breslow JL. Overexpression of PCSK9 accelerates the degradation of the LDLR in a post-endoplasmic reticulum compart- ment. Proc Natl Acad Sci U S A. 2005;102(6):2069–74. 11. Stein EA, Honarpour N, Wasserman SM, Xu F, Scott R, Raal FJ. Effect of the proprotein convertase subtilisin/kexin 9 monoclonal antibody, AMG 145, in homozygous familial hypercholesterolemia. Circulation. 2013;128(19):2113–20. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. 12. Nioi P, Sigurdsson A, Thorleifsson G, Helgason H, Agustsdottir AB, Nord- dahl GL, et al. Variant ASGR1 associated with a reduced risk of coronary artery disease. N Engl J Med. 2016;374(22):2131–41. Declarations 13. Wang JQ, Li LL, Hu A, Deng G, Wei J, Li YF, et al. Inhibition of ASGR1 decreases lipid levels by promoting cholesterol excretion. Nature. 2022;608(7922):413–20. Consent for publication Not applicable. Consent for publication Not applicable. 15. Liu J, Hu B, Yang Y, Ma Z, Yu Y, Liu S, et al. A new splice variant of the major subunit of human asialoglycoprotein receptor encodes a secreted form in hepatocytes. PLoS ONE. 2010;5(9):e12934. Competing interests y 16. Spiess M, Lodish HF. An internal signal sequence: the asialoglycoprotein receptor membrane anchor. Cell. 1986;44(1):177–85. The authors declare no competing interests. 17. Spiess M, Handschin C. Deletion analysis of the internal signal-anchor domain of the human asialoglycoprotein receptor H1. Embo J. 1987;6(9):2683–91.f References 1. Ference BA, Ginsberg HN, Graham I, Ray KK, Packard CJ, Bruckert E, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. Eur Heart J. 2017;38(32):2459–72. 23. Liu Y, Zhu L, Liang S, Yao S, Li R, Liu S, et al. Galactose protects hepato- cytes against TNF-alpha-induced apoptosis by promoting activation of the NF-kappaB signaling pathway in acute liver failure. Lab Invest. 2015;95(5):504–14. 2. Goldstein JL, Brown MS. A century of cholesterol and coronaries: from plaques to genes to statins. Cell. 2015;161(1):161–72. 2. Goldstein JL, Brown MS. A century of cholesterol and coronaries: from plaques to genes to statins. Cell. 2015;161(1):161–72. 24. Rigopoulou EI, Roggenbuck D, Smyk DS, Liaskos C, Mytilinaiou MG, Feist E, et al. Asialoglycoprotein receptor (ASGPR) as target autoantigen in liver autoimmunity: lost and found. Autoimmun Rev. 2012;12(2):260–9. 3. Scandinavian Simvastatin Survival Study Group. Randomised trial of cho- lesterol lowering in 4444 patients with coronary heart disease: the Scan- dinavian Simvastatin Survival Study (4S). Lancet. 1994;344(8934):1383–9. Author details 1 Department of Cardiovascular Medicine, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China. 2 Research Institute of Blood Lipid and Atherosclerosis, the Second Xiangya Hospital, Central South Univer- sity, Changsha, China. 3 Reproductive Medicine Center, Department of Obstet- rics and Gynecology, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China. 4 Department of Emergency Medicine, The Affiliated Changsha Central Hospital, Hengyang Medical School, University of South China, Changsha, China. 5 Department of Health Management, The Third Xiangya Hospital, Central South University, Changsha 410013, Hunan, China. 18. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, et al. Fourth universal definition of myocardial infarction (2018). Circulation. 2018;138(20):e618–51. 19. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemi- ology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453–7. China, Changsha, China. 5 Department of Health Management, The Third Xiangya Hospital, Central South University, Changsha 410013, Hunan, China. 20. Fowdar JY, Grealy R, Lu Y, Griffiths LR. A genome-wide association study of essential hypertension in an Australian population using a DNA pooling approach. Mol Genet Genomics. 2017;292(2):307–24. Received: 24 June 2023 Accepted: 28 August 2023 Received: 24 June 2023 Accepted: 28 August 2023 21. Nakaya R, Kohgo Y, Mogi Y, Nakajima M, Kato J, Niitsu Y. Regulation of asialoglycoprotein receptor synthesis by inflammation-related cytokines in HepG2 cells. J Gastroenterol. 1994;29(1):24–30. 22. Shi R, Wang J, Zhang Z, Leng Y, Chen AF. ASGR1 promotes liver injury in sepsis by modulating monocyte-to-macrophage differentiation via NF- kappaB/ATF5 pathway. Life Sci. 2023;315:121339. Publisher’s Note 4. Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravas- tatin in patients with coronary heart disease and a broad range of initial cholesterol levels. N Engl J Med. 1998;339(19):1349–57. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. 5. Heart PSCG. Effects on 11-year mortality and morbidity of lowering LDL cholesterol with simvastatin for about 5 years in 20,536 high-risk individu- als: a randomised controlled trial. Lancet. 2011;378(9808):2013–20. 6. Fulcher J, O’Connell R, Voysey M, Emberson J, Blackwell L, Mihaylova B, et al. Efficacy and safety of LDL-lowering therapy among men and women: meta-analysis of individual data from 174,000 participants in 27 randomised trials. Lancet. 2015;385(9976):1397–405. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: 7. Ethics approval and consent to participate The study conformed to the ethical guidelines of the Declaration of Helsinki and was approved by the ethics committee of the Second Xiangya Hospital. Informed consent was obtained from each patient. 14. Chiacchia KB, Drickamer K. Direct evidence for the transmembrane orientation of the hepatic glycoprotein receptors. J Biol Chem. 1984;259(24):15440–6. Consent for publication Not applicable. 8. Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, Bhala N, et al. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta- analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010;376(9753):1670–81. 9. Dubuc G, Chamberland A, Wassef H, Davignon J, Seidah NG, Bernier L, et al. Statins upregulate PCSK9, the gene encoding the proprotein con- vertase neural apoptosis-regulated convertase-1 implicated in familial hypercholesterolemia. Arterioscler Thromb Vasc Biol. 2004;24(8):1454–9. Publisher’s Note Hoogeveen RC, Ballantyne CM. Residual cardiovascular risk at low LDL: remnants, Lipoprotein(a), and inflammation. Clin Chem. 2021;67(1):143–53. 7. Hoogeveen RC, Ballantyne CM. Residual cardiovascular risk at low LDL: remnants, Lipoprotein(a), and inflammation. Clin Chem. 2021;67(1):143–53. 8. Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, Bhala N, et al. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta- analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010;376(9753):1670–81. 8. Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, Bhala N, et al. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta- analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010;376(9753):1670–81. 9. Dubuc G, Chamberland A, Wassef H, Davignon J, Seidah NG, Bernier L, et al. Statins upregulate PCSK9, the gene encoding the proprotein con- vertase neural apoptosis-regulated convertase-1 implicated in familial hypercholesterolemia. Arterioscler Thromb Vasc Biol. 2004;24(8):1454–9.
https://openalex.org/W2937761160	https://link.springer.com/content/pdf/10.1007/s11858-019-01050-y.pdf	English	null	Mathematics identity research: the state of the art and future directions	ZDM	2,019	cc-by	15,062	Abstract Identity research in mathematics education has become increasingly prominent over the past two decades. In the last few years, there have been several reviews of identity literature in the field of mathematics education generally, or specifically focused on mathematics learner identities or mathematics teacher identity. We begin our paper by summarizing the key find- ings of these reviews, pointing to various categorizations proposed by their authors, and the key critiques raised therein. We then report on our more recent review of identity literature in mathematics education published in the top twenty mathematics education journals over the past 5 years (47 articles in total). This review enables us firstly to speak descriptively about the extent to which the field is evolving (or not) in relation to earlier findings (dominant regions, focus of research, methods and perspectives used). Secondly, following deeper analysis of the papers, we speak to the way in which the field is addressing (or not) key critiques raised by reviews. Our analysis points to several theoretical, methodological and empirical absences and challenges requiring further engagement. We further highlight key absences in the research, particularly a lack of cross- national comparative studies and studies that connect learner and teacher identities. We then provide a summary discussion and analysis of the contributions of the present special issue papers and engage with how these take the field forward. Finally, we engage with the implications of our analysis and recommendations for future directions for research. Review and introduction to ZDM Special Issue on Identity in Mathematics Education Mellony Graven1 · Einat Heyd‑Metzuyanim2 Accepted: 4 April 2019 / Published online: 13 April 2019 © The Author(s) 2019 ZDM (2019) 51:361–377 https://doi.org/10.1007/s11858-019-01050-y ZDM (2019) 51:361–377 https://doi.org/10.1007/s11858-019-01050-y SURVEY PAPER * Mellony Graven m.graven@ru.ac.za 1 Rhodes University, Grahamstown, EC, South Africa 2 The Technion - Israel Institute of Technology, Haifa, Israel * Mellony Graven m.graven@ru.ac.za 1 Introduction learning as participation; figured worlds and positioning theory; socio-political theories; cultural historical activity theory; critical race theory and commognitive theory— drawing on, for example, influential theorists such as Gee (2000); Wenger (1998); Holland, Lachicotte, Skinner and Cain (1998); Harré and van Langenhove (1999); and Sfard (2008) (Heyd-Metzuyanim, Lutovac and Kaasila 2016; Dar- ragh 2016; Radovic, Black, Williams, Salas 2018). Identity provides a powerful bridge between learning and cultural context (Sfard and Prusak 2005). In mathematics education, identity has been particularly useful in exploring why so many learners might disengage from mathematics without falling into cognitive deficit discourses. Identity research entered the field of mathematics education dur- ing what Lerman (2000) termed the social turn in math- ematics education. Indeed, most of the identity literature in mathematics education draws on socio-cultural theories of learning (Heyd-Metzuyanim, Lutovac and Kaasila 2016) and aligns with a Meadian, identity-as-action view (Darragh 2016). Popular frameworks used include socio-linguistics; The concept of identity has its origins in the work of Mead (1934) and Erikson (1968). For Mead identity was multiple (though appearing more unified to the individual), sometimes contradictory, and developing in interaction with the environment. For Erikson identity was more uni- fied developing throughout one’s life (Graven and Lerman, 2014). Darragh (2016, p. 26) notes two key views that cut across identity research, namely identity ‘as an action or identity as an acquisition’. She relates these to Meadian and Eriksonian perspectives as follows: Erikson understood identity as an acquisition, something that one has and that becomes coherent and consistent. A (0121 3456789) 3 362 M. Graven, E. Heyd‑Metzuyanim Meadian identity is an action, it is something one does, and it is multiple, contradictory and socially constituted. (Darragh 2016, p. 27). education generally, they did not conduct systematic reviews and thus do not provide descriptive statistics in the way Dar- ragh (2016) does. Darragh’s review from 1997 up to 2014 began around the start of identity research in mathemat- ics education and included 188 articles across 85 journals. Below we summarise some key findings in relation to the distribution of research over time, across mathematics edu- cation journals and across regions. We do this not only to summarize what has been established, but also to provide a basis for comparison with our more recent review of research. 1 Introduction At the most basic level, mathematics education identity research can be broadly split into two categories in terms of what the research focuses on: either learner identities (school or undergraduate mathematics learners) or teacher identities (pre- and in-service, specialised and non-specialised). We see this split emerging in recent reviews published on either math- ematics learner identity (e.g. Radovic et al. 2018) or teacher identity (Lutovac and Kaasila 2017; Lutovac and Kaasila, 2018c) and in the most recent Springer Encyclopedia of Math- ematics Education having separate entries for mathematics teacher identity and mathematics learner identity (Graven and Lerman, 2019; Darragh and Radovic, 2019a). Darragh’s (2016) review shows an increasing trend over time with less than 5 articles per annum published in the first 7 years (1997–2003) and more than 18 articles per annum in the last 5 years (2010–2014). In terms of the dis- tribution of articles across regions, Darragh found the USA dominated (36% of all articles), followed by the UK (15%), Europe (11%), Australia (10%); New Zealand (10%) and South Africa (5%). Of the mathematics education research journals, where half (94) of the articles in Darragh’s (2016) review were located, Educational Studies in Mathematics (ESM 18) and the Journal for Research in Mathematics Education (JRME 15) dominated followed by the Journal of Mathematics Teacher Education (JMTE 10); the jour- nal Research in Mathematics Education (RME 10), and For the Learning of Mathematics (FLM 8). Darragh (2016) indicates an almost even split between articles focused on teacher (45%) and learner identity (50%), further indicating research on practicing teachers (28%) being higher than that of pre-service teachers (17%). She does not however provide further breakdown in terms of the level of teachers (e.g. primary or secondary) and does not provide any breakdown of research across levels of learners. At a more complex level, identity research in mathemat- ics education can be categorized according to various theo- retical perspectives it uses (Darragh 2016) and/or the way in which identity definitions are established and operationalized (Radovic et al. 2018). These categorizations are discussed further in 2.2 below in respect to key contributions of recent reviews to developing conceptual coherence in the field. 2 Common findings and issues identified in reviews on identity research in mathematics education In the past 3 years there have been several reviews on identity research in mathematics education in general (Darragh 2016; Langer-Osuna and Esmonde 2017; Heyd-Metzuyanim, Luto- vac and Kaasila 2016), mathematics learner identity (Radovic et al. 2018) and mathematics teacher identity (Lutovac and Kaasila 2018a). We believe, these point to a ‘coming of age’ of the field as does this special issue. We begin by summarizing the common findings and critiques that these reviews point towards, so as to establish what currently seems well-known or established by those working in the field. These reviews tend to firstly provide a description of the field, often including descriptive statistics. Secondly, they critique the field particu- larly on issues of conceptual coherence. Finally, the system- atic reviews of Darragh (2016) and Radovic et al. (2018) offer various categorizations of identity research that we believe contribute significantly to establishing greater conceptual clar- ity and coherence in the field. We discuss these contributions below. In terms of perspectives used in mathematics identity lit- erature all reviews mentioned herein note that socio-cultural frames dominate in studies of learners’ and teachers’ identi- ties. Darragh (2016) identified five categories of research based on different theoretical perspectives taken, namely participative, narrative, discursive, psychoanalytic and per- formative. She did not use these categories to provide sta- tistics of the distribution of articles across these (possibly because of issues of overlap or lack of clarity on perspective used, see Sect. 2.2 below). Instead she provided descriptive statistics on the dominant theorists drawn on in the research, indicating that the greatest theoretical influence was that of Lave and Wenger (1991) and/or Wenger (1998) (41% of the articles drew on these). This was followed by the work of Holland and colleagues (1998) (20%) and Gee (2000) (18%). She further notes commonly cited literature as that of Sfard and Prusak (2005) (cited by 21% of articles), Boaler and colleagues (e.g. Boaler and Greeno 2000) (cited by 18% of the articles) and Martin (2000; 2012) (cited by 12%). The 2.1 Descriptive results of the field from various reviews While the reviews by Langer-Osuna and Esmonde (2017) and Heyd-Metzuyanim, Lutovac and Kaasila (2016) provide commentary on the field of identity research in mathematics 3 3 363 Mathematics identity research: the state of the art and future directions dominance of these theorists cohere with the findings of other reviews (e.g. Heyd-Metzuyanim et al. 2016). avoid the confusion. Critiques on the use of identity in the field of teacher education research have similarly raised con- cerns that cut across subject specific areas. For example, Beijaard, Meyer and Verloop (2004) noted in their review of teacher professional identity that there was little consensus on a working definition of identity and that it was often not defined at all. Furthermore, they noted that the literature tended to conflate the concepts of self and identity, and that more clarity was needed on the perspectives used. In terms of methods used, none of the reviews system- atically coded and counted the methods of identity studies. Darragh (2016) notes simply that the majority of articles use qualitative methods (which is consistent with other afore- mentioned reviews). She did provide some basic statistics on sample sizes of studies on which the articles are based: 58% had fewer than 10 participants and only 9% had more than 100 participants (5% of articles did not specify). Darragh (2016) notes, however, that even while studies gather data on multiple participants, in the reporting of the data ‘Many articles presented case studies from larger studies’ (p. 23). In terms of teacher identity, Lutovac and Kaasila (2018a) stated that most of the studies they reviewed used interviews, autobiographies, narratives and observations. They argue that since several studies point to differences between self- reporting and observational data, greater emphasis should be placed on observations of practice. This connects with Radovic et al.’s (2018) review that noted that research on learner identity predominantly drew on identities as narra- tives (36%) and thus on data collection through narratives. Sfard and Prusak’s (2005) work, as noted above, is widely cited in articles on mathematics identity. They firstly pro- vided a thorough critique of identity theorists prominent in the field of mathematics education, noting in particular a lack of operationalized definition in their work. Secondly, they provided an operationalized definition of identities as significant, reified and endorsable stories about a person (i.e. 2.2 Key critiques of, and contributions to, identity research in mathematics education By far, the most prevalent critique of identity research has been the vagueness of the “identity” concept and the lack of clarity around its operationalization (see Sfard and Prusak, 2005 for one of the most cited of these critiques). Critiques of the use of “identity” as a concept are not particular to mathematics education. There have been several challenges to the way in which the term is being used across fields. Brubaker and Cooper (2000) provide one such widely cited and thoroughly considered critique. They explain that in the sixties of the last century the term identity ‘proved highly resonant’ and quickly diffused ‘across disciplinary and national boundaries establishing itself in the journalistic as well as the academic lexicon’ (p. 3). They refer to “the ‘identity’ crisis in the social sciences” (p. 2) and explain it as “a crisis of overproduction and consequent devaluation of meaning” (p. 3). y gi Brubaker and Cooper (2000) provide one such widely cited and thoroughly considered critique. They explain that in the sixties of the last century the term identity ‘proved highly resonant’ and quickly diffused ‘across disciplinary and national boundaries establishing itself in the journalistic as well as the academic lexicon’ (p. 3). They refer to “the ‘identity’ crisis in the social sciences” (p. 2) and explain it as “a crisis of overproduction and consequent devaluation of meaning” (p. 3). Further, Sfard and Prusak (2005) argued that Gee’s (2000) work, which says that identity is being recognized as a certain ‘kind of person’ in a given context, and the work of Holland and colleagues, provide promising beginnings but still fall short of providing an operationalised definition. They then provide the following definition: In concert with the vision of identifying as a discursive activity, we suggest that identities may be defined as collections of stories about persons or, more specifi- cally, as those narratives about individuals that are rei- fying, endorsable, and significant. (p. 16). We have argued that the concept (identity) is deployed to do a great deal of analytic work—much of it legiti- mate and important. “Identity,” however, is ill suited to perform this work, for it is riddled with ambiguity, riven with contradictory meanings, and encumbered by reifying connotations. (p. 34). 2.1 Descriptive results of the field from various reviews the product of collective storytelling) that they argued pro- vides an analytic tool for investigating learning. In relation to the former they noted: In the current literature the use of the word identity is rarely preceded by any explanations. In the absence of a definition, the reader is led to believe that iden- tity is one of those self-evident notions that, whether reflectively or instinctively, arise from one’s firsthand, unmediated experience. The influential publications by Lave and by Wenger are representative in this respect. Although identity is one of these writers’ pivotal ideas, no conceptual preparations precede sentences such as “Learning… implies becoming a different person [and] involves the construction of identity” (Lave & Wenger, 1991, p. 53), or “The experience of identity in practice is a way of being in the world” (Wenger, 1998, p. 151). (Sfard and Prusak 2005, p. 15). 2.2 Key critiques of, and contributions to, identity research in mathematics education Albeit incredibly influential in the field of mathematics education (bearing, at the time of writing, over 1600 cita- tions in Google Scholar), Sfard and Prusak’s (2005) article also raised debate and dialogue regarding the definition of identity as “narrative”. This debate was most notable in Juzwik’s (2006) response to the paper, and Sfard’s (2006) reply to Juzwik. In a nutshell, Juzwik (2006) claimed that Brubaker and Cooper (2000) argue that alternative ana- lytical idioms, such as identification and categorization; self-understanding and social location, and commonality, connectedness, groupness would be preferable and would 1 364 M. Graven, E. Heyd‑Metzuyanim although Sfard and Prusak’s (2005) definition of “iden- tity as narrative” is useful, further consideration should be given to the concept of “narrative” itself. In addition, she cautioned that some “extra-discursive” activity, such as tone of voice, manner of dress and other para-linguistic signals may be missed in the definition of “identity as nar- rative”. To this, Sfard (2006) responded by claiming that Juzwik was, in fact, falling victim to an “ontological col- lapse” when she was differentiating between “discursively constructed” and “extra-discursively” constructed aspects of identity. She added: taken up by some of the papers reviewed (e.g. Oppland- Cordell and Martin 2015) and in our current issue. We now turn to summarizing key commonalities in the critiques of recent reviews in mathematics education with a view to setting the scene for analyzing whether, in more recent literature of the leading mathematics education jour- nals, these critiques have been addressed. The central critique of both Darragh’s (2016) review of 188 articles on mathematics identity and Radovic et al.’s (2018) review of 69 articles on mathematics learner identity is that this body of work continues to lack clear definitions of identity and often lacks conceptual coherence, including at times using incompatible theories. Other reviews men- tioned above reinforce this. For example, Langer-Osuna and Esmonde (2017) argue that researchers must more clearly communicate their theoretical frames and that better connec- tion is needed between these frames and analytic methods used. They also call for further theorizing of the concept noting: “the development of mathematics-linked individual and membership identities both need to be further theorized and studied” (p. 66). They acknowledge however, that coor- dinating individual and membership identities is methodo- logically challenging and note the need for methodological advances in this respect. 2.2 Key critiques of, and contributions to, identity research in mathematics education Lutovac and Kaasila similarly call for clearer methodological articulation in research on teacher identity (2018a; 2018c). One can hardly be surprised that the ontological col- lapse—the confusion between discursive and extra- discursive, and in particular, between what merely affects discourses and what is actually a part of them— persists in spite of the history of protests almost as long as the objectifying discourses themselves. (p. 24). In what has later evolved into the “commognitive frame- work” (Sfard 2008), Sfard insists that identities are narra- tives. Identities are not reflected or represented by narratives. There is nothing “there” underlying the “narrative” (or more generally, the discourse) which the narrative describes. Yet, from our own readings, it seems that despite the prevalence of the “identity-as-narrative” definition in mathematics iden- tity literature, few authors have engaged deeply with the full ontological and epistemological “package” of assumptions that comes with Sfard and Prusak’s (2005) theorizing of this concept. Later, we shall relate to this gap and build upon it for two points: one, the “ontological collapses” that we found in literature on identity, namely, the implicit unification of different types of identity narratives; and two, the neglect of many writers in the field to differentiate between (or take into account) various identity stories of different authors. These stories have been typified by Sfard and Prusak (2005) using the tripartite signal ABC: A being the author, B—the Protagonist (learner/teacher talked about) and C—the audi- ence (who the story is told to). Most authors, even while relying on the “identity-as-narrative” definition, focus on stories of the type IntervieweeIntervieweeResearcher and neglect to relate to (or differentiate between) other stories. Further, in our review we note that while many papers cite their work, very few papers used Sfard and Prusak’s (2005) operation- alization of identity as reified, significant and endorsable stories. In part, this may be a result of the vagueness of these qualifying characteristics of narratives. Thus, although fur- ther works by Sfard and colleagues (e.g. Heyd-Metzuyanim and Sfard 2012; Heyd-Metzuyanim 2015) offered opera- tionalizations for the “reifying” aspect of narratives, only Heyd-Metzuyanim (2015) engaged with the challenges of qualifying a narrative as “significant”. 3 Current state of identity research in leading mathematics education research journals For the present “state of the art” review, we chose to focus our analysis on identity research over the past 5 years (i.e. 2014–2018) in mathematics education published in the top twenty mathematics education research journals as identi- fied by Williams and Leetham (2017). Our rationale for this time period was firstly that Darragh (2016), who pro- vides the only comprehensive review of both learner and teacher identity, ended her review at 2014 (acknowledging that due to the time lag in publication not all 2014 articles were included). Radovic et al.’s (2018) review became available while we were writing this review and thus some overlap between our data set and that of Radovic et al. is inevitable. Their review however only includes articles, focused on learner identity, up to 2015, and again did not capture all 2015 articles in leading mathematics education journals. Secondly, since we were aiming at a “state of the art” indicating where we are now in terms of research, we did not require extensive breadth for our review. Rather we aimed at conducting an in-depth analysis of all articles considered by the community of mathematics educators to be of sufficient quality to be published in one of the top twenty peer reviewed mathematics education research journals. This search period furthermore allowed us to include recently published reviews of the field of math- ematics identity research all of which have been published from 2016 onwards. This enabled us to examine the cri- tiques and points raised by the former reviews against the current trends. As an example, ideas from Holland et al. (1998) were identified in studies that Darragh categorized as conceptualizing identities as participative, narra- tive, and positional. Using our dimensional model, it is possible to understand how Holland and col- leagues’ ideas are differently employed and opera- tionalized in these three groups. (p. 33). The focus of our review was guided by a range of research questions, some necessitating descriptive statis- tics, others requiring deeper qualitative analysis of the cur- rent field as portrayed in mathematics education research journals. The research questions relating to descriptive statistics were: What is the distribution of articles across the time period, regions where research occurred and jour- nals published in? Who is the focus of the research? What theoretical perspectives were used? What methods were used? 2.2 Key critiques of, and contributions to, identity research in mathematics education The authors provide the following example of how this contributes further clarity to Darragh’s (2016) categorization: locate and specify their identity research remains to be seen. locate and specify their identity research remains to be seen. 2.2 Key critiques of, and contributions to, identity research in mathematics education So, for example, research in the category of identity as a relationship with a specific practice was noted to operationalize identities as a sense of belong- ing/membership and to prefer surveys and interviews for data gathering. Radovic et al. (2018) further provide three dimensions for their five categories along which research can be positioned. These three dimensions are: the social/subjective; the enacted/representational and the change/stability dimension. Through these they argue the theoretical overlap between particularly the first three of Darragh’s five categories (participative, narrative and dis- cursive) can be unpacked by considering how theoretical concepts are used and operationalized. They also claimed that such analysis is more effective when one keeps in mind ‘the social/subjective, representational/enacted, and change/stability dimensions of MLIs’ (p. 26). The authors provide the following example of how this contributes further clarity to Darragh’s (2016) categorization: of definition: identity as participative, narrative, discur- sive, psychoanalytic or performative. The latter identi- fied four theoretical approaches to the study of identity, namely: poststructural theory, positioning theory, nar- rative theory, and psychoanalytic theory. More recently, Radovic et al. (2018), focusing on mathematics learner identity research, identified five broad categories of research, namely identity as: individual attributes; nar- ratives; a relationship with a specific practice; ways of acting; and identity as afforded and constrained by local practices. These reviews discuss linkages and overlaps between various broad theories, influential theorists (such as Wenger 1998; Holland et al. 1998), key literature (e.g. Sfard and Prusak 2005) and the definitions used. Radovic et al. (2018) further elaborate on the operationalization and preferred methodology of research in each of their five categories. So, for example, research in the category of identity as a relationship with a specific practice was noted to operationalize identities as a sense of belong- ing/membership and to prefer surveys and interviews for data gathering. Radovic et al. (2018) further provide three dimensions for their five categories along which research can be positioned. These three dimensions are: the social/subjective; the enacted/representational and the change/stability dimension. Through these they argue the theoretical overlap between particularly the first three of Darragh’s five categories (participative, narrative and dis- cursive) can be unpacked by considering how theoretical concepts are used and operationalized. They also claimed that such analysis is more effective when one keeps in mind ‘the social/subjective, representational/enacted, and change/stability dimensions of MLIs’ (p. 26). 2.2 Key critiques of, and contributions to, identity research in mathematics education Thus, this gap may show that Sfard and Prusak’s (2005) definition itself may have been insufficiently operational, an issue that has been In terms of key absences in the literature, Heyd-Metzuy- anim, Lutovac and Kaasila (2016) note that since the bodies of literature on student and teacher identity have developed independently (even while sharing similar theoretical back- grounds), there is an absence of research connecting these two. As Darragh (2016) indicates, only 2% of the articles reviewed studied both teacher and learner identities though, even in these articles, the research may not have focused on the relationship between these. Lutovac and Kaasila (2018a), in their review of mathematics teacher identity, note that this body of work seldom draws on general educa- tion identity research, which they believe has much to offer. They argue that there is an absence of individual or psycho- logical emphases and that research in mathematics teacher identity literature is neglecting the individual. While not- ing Darragh’s (2016) caution of identity being a “catch all” term for affect, Lutovac and Kaasila (2018a) critique the lack of attention to affect in relation to mathematics teacher identities. They also suggest a need for greater emphasis on practice as they found most studies on teacher identity were small scale, relying on autobiographical/narrative interviews. The reviews of Darragh (2016) and Langer-Osuna and Esmonde (2017), on identity research in mathemat- ics education generally, identified various categories of research according to the perspectives or definitions of identity used. The former suggested five broad categories 1 3 3 Mathematics identity research: the state of the art and future directions 365 of definition: identity as participative, narrative, discur- sive, psychoanalytic or performative. The latter identi- fied four theoretical approaches to the study of identity, namely: poststructural theory, positioning theory, nar- rative theory, and psychoanalytic theory. More recently, Radovic et al. (2018), focusing on mathematics learner identity research, identified five broad categories of research, namely identity as: individual attributes; nar- ratives; a relationship with a specific practice; ways of acting; and identity as afforded and constrained by local practices. These reviews discuss linkages and overlaps between various broad theories, influential theorists (such as Wenger 1998; Holland et al. 1998), key literature (e.g. Sfard and Prusak 2005) and the definitions used. Radovic et al. (2018) further elaborate on the operationalization and preferred methodology of research in each of their five categories. 3.2.1 Descriptive statistics The field is seemingly still growing as while only six of the 47 articles were published in 2014, in 2018, by October when our final search occurred, there were already 13 arti- cles available. The vast majority of articles were empirical (37). As discussed above, there were two large systematic reviews and one summary of 40 articles on mathematics teacher identity. There were also two commentaries (both by Langer-Osuna 2017; Langer-Osuna 2018) and four theo- retical papers (the latter sometimes drew on empirical data from earlier research studies in the field). The small number of theoretical articles is similar to Darragh’s (2016) finding of only 13% theoretical articles. The methodology of the review was guided by the Pre- ferred Reporting Items for Systematic Reviews and Meta- Analyses (PRISMA) (Moher, Liberati, Tetzlaff, Altman, PRISMA Group 2009). EBSCOhost was used to do a jour- nal-specific search in the databases Academic Search Pre- mier and ERIC with the word identi* in the fields: title OR abstract OR keyword identifying OR subject descriptors, for each of the journal included. To triangulate, the search was run through SCOPUS for each of the 15 journals that were indexed there. As the most recent articles are often not yet indexed on databases, each of the 2018 issues of the twenty journals was hand-searched on their homepages. The first round of search was performed during August 2018. The last round was performed during the 2nd week of October 2018. y Similar to Darragh’s findings, the regional distribu- tion of our empirical articles, in terms of the country that research data was gathered in, is dominated by Northern America (12) (10 from USA and 2 from Canada); UK (6); and Europe (6) with South America (Chile/Argentina) (2) South Africa (4); Australia and New Zealand (2) and Israel (2) also contributing. A single article based on research conducted in each of Japan and Turkey points to the pos- sibility of these new geographical areas entering the field. One empirical article focused on analyzing youth fiction that was not particular to a country (Darragh 2018). Only one empirical study included data from across two coun- tries, Finland and Slovenia (Lutovac and Kaasila 2014). 3 Current state of identity research in leading mathematics education research journals What sample size was drawn on for the study and in the reporting of the study? What is the balance between empirical/theoretical research over the past 5 years? The research questions aimed at deeper qualitative analysis of the current field were: Are we moving past the key critiques? What new directions/issues/perspectives are emerging? In relation to issues of conceptual coherence, Radovic et al.’s (2018) categorization, paired with the identification of one’s location along various dimen- sions within a category will potentially be helpful for researchers in providing conceptual clarity and coherence to their perspective on, and working definition of, identity used. Indeed, this intended contribution to the field is captured in the title of their review paper ‘Towards con- ceptual coherence in the research on mathematics learner identity: a systematic review of the literature.’ Since this review has only recently been published, the extent to which researchers will draw on this framework to better 1 3 M. Graven, E. Heyd‑Metzuyanim 366 or relation. Articles that did not refer in any sense to math- ematical identity (such as those that only referred to science or engineering identity) were also excluded. A few papers that had identity in the abstract were later removed from the list as thorough reading of the paper clearly indicated that the paper was not about identity, had little mention of it and did not engage with it. All papers were copied into separate journal folders and a spreadsheet was created which pro- vided summary information of each paper (such as authors, title, abstract, keywords, date) and information related to our descriptive research questions (such as where the research took place, perspectives and methods used, etc.) and our deeper analytic questions (such as How is identity defined and operationalized?). are there ontological collapses? Is identity central or peripheral to the paper? Is mathematics central or periph- eral to the paper? Is engagement with mathematics central or peripheral to the paper? What is identity being used for? What is being advocated? For each question we reflect on whether our findings point to a continuation of, or shift in, earlier findings. 3.1 Method of analysis As argued above, we included in our review all research papers published (2014–2018) on the topic of identity in the leading 20 mathematics education research journals (Williams and Leetham 2017, p. 337 ordered in relation to total citations), namely: Educational Studies in Mathemat- ics (ESM); Journal for Research in Mathematics Education (JRME); Journal of Mathematical Behavior (JMB); For the Learning of Mathematics (FLM); Mathematical Think- ing and Learning (MTL); Journal of Mathematics Teacher Education (JMTE); ZDM Mathematics Education (ZDM); Mathematics Education Research Journal (MERJ); Interna- tional Journal of Math Education in Science and Technology (IJMEST); School Science and Mathematics (SSM); Inter- national Journal of Science and Mathematics Education (IJSME); Investigations in Mathematics Learning (IML; formerly FOCUS on Learning Problems in Mathematics); Teaching Mathematics and Its Applications (TMA); The Mathematics Educator (TME); Research in Mathematics Education (RME); International Journal for Technology in Mathematics Education (IJTME); Journal of Computers in Mathematics and Science Teaching (JCMST); Canadian Journal of Science, Mathematics and Technology Educa- tion (CJSMTE); PRIMUS (Problems, Resources, Issues in Undergraduate Mathematics Studies), and The Montana Mathematics Enthusiast (TMME). Following some joint reading and analysis of eight papers in relation to our research questions and spreadsheet, some adaptations were made to the spreadsheet. We then pro- ceeded to share the remaining papers between us and ana- lyze them separately by responding to our research questions in the spreadsheet. A few papers were removed by agreement if on in-depth reading it was not seen to be about identity. This process resulted in 47 selected papers. Methods and sample size Almost all empirical studies focused on qualitative meth- ods with only two studies drawing heavily on quantitative methods (i.e. Boaler and Selling 2017; Kasperson, Pepin and Sikko 2017, focusing in their studies on 21 and 185 learners respectively). All other learner articles focused on 8 or less learners (with two and four learners being the most common number of participants reported on). While many articles drew on data from much larger studies, the data reported on in the paper focused on a small selection from the larger sample of participants. Similarly, the sample sizes of the studies focused on teachers were small, ranging from 1 to 8 teachers. A quarter of the studies focused on a single teacher. This concurs with Darragh’s (2016) finding that most of the studies had a sample of less than ten participants. As with Darragh’s finding the dominant mathematics edu- cation journals publishing identity research over the past 5 years are ESM (17), JRME (6) and JMTE (5). While Dar- ragh’s review notes significant contributions also from RME and FLM, in our review period these contribute only two and one article respectively. MERJ, however, which did not feature earlier as one of the leading contributing journals, contributed six articles—though five of these were in the 2015 special issue on the topic. This, as far as we know, was the first special issue on the topic. The remaining ten of the 47 articles were published in IJSME (1); IML (1); JMB (1); TMA (2); TME (1); TMME (1) and ZDM (3). While several empirical papers mention having gathered a broad range of data (e.g. observational/video recorded, field note, interview, documentary data) the data included and reported on in each paper is generally a selection of those data sources. The data predominantly used and reported on in articles is interview data (76%). While many papers sup- plement their interview data with another source of data gathered, particularly observational data, just under a third (32%) of the articles focused almost exclusively on interview data. Observational data or video recording transcripts was the next most reported data source (35%). Surveys, question- naires (11%) and fieldnotes (or journal entries) (14%) were used in some articles though mostly to provide background data on participants or activities. Objects of the research Like Darragh, who found that slightly more research was focused on learners (50%) than teachers (45%), our review shows a leaning of empirical research towards learners (54%) over teachers (32%). While not all articles indicate the level of learners, where an indication was given there seemed to be an almost even distribution across elementary, primary, middle school, secondary and tertiary (college or university) learners. A small number of studies followed learners through their transition from one phase to the next (e.g. Darragh 2015). Of the articles focused on teachers there was almost an even split between pre-service and in-service teachers including novice teachers and out of field teachers of numeracy. One study looked at a teacher transitioning from pre-service to beginning teaching (Losano, Fiorentini and Villarreal 2018). As in the case of learners, there was an almost even spread across levels of primary and sec- ondary, pre- and in-service teachers where stated. Single articles also focused on the identities of: adults (linked to data gathered when they were students) (Boaler and Sell- ing 2017); immigrant parents (Takeuchi 2018), a railway worker (Wake 2014), a mathematician (Solomon, Radovic and Black 2016), and a researcher/teacher educator (using autoethnography, Knapp 2017). Of interest, one paper focused on portrayals of mathematics learners and teach- ers in novels (Darragh 2018), another on text analysis of a university mathematics course (le Roux and Adler 2016) and another on advertising images (Evans, Tsatsaroni and Czarnecka 2014). These articles point to the relevance of looking at data beyond teachers, learners and the teaching context. Interestingly, our review found no article focusing on both teacher and learner identities, though in some cases Methods and sample size Learner generated artefacts such as writings or drawings were also used in a few articles (11%) while a movie that students produced was used in a single article (Pickard-Smith 2018). Another study used interview data with written autobiographies that included drawings and written reflections of pre-service teachers (Chen 2017). 3.2.1 Descriptive statistics Since earlier reviews do not mention studies working The results obtained were initially screened for whether the words containing the root identi* made reference to identity—for example that they did not refer to the “identi- fication” of something or to an identity property, function 1 3 1 3 1 3 Mathematics identity research: the state of the art and future directions 367 learner identities were analysed in relation to data gathered from teacher interviews, teacher practices or written stories about learners (e.g. Heyd-Metzuyanim and Graven 2016). across different countries it is difficult to say whether such across country research is new to the field. We believe research that includes data from different contexts, coun- tries and regions could strengthen the field. Our data con- firm that research on mathematics identity continues to be regionally clustered rather than being a global area of research, though the articles from Japan and Turkey indi- cate some engagement in new locations.i • Definitions and operationalization of identity In relation to the main critique raised by earlier reviews and commentaries, regarding the definition of identity, we see a similar picture namely that most papers either do not define the concept, or define it only partially. However, our review may add some nuances to this general observation. First, it seems that the main gap lies not in the act of defini- tion itself. Thus, many articles we reviewed included explicit statements such as “we define identity as…” or “we conceive of identity as …”. In addition, many papers allocate some significant portion of their theoretical background to review previously suggested definitions. In that sense, there seems to be awareness in the field of the importance of defining identity. Regardless of definition, an overwhelming majority of studies operationalize identity, at least in part, based on par- ticipants’ narratives. The differences lie between those that only rely on narratives, and those that attempt to incorpo- rate also “participative” aspects, mostly operationalized in the form of researchers’ observations of the participants, or stories about the participants told by other significant nar- rators (such as parents or teachers). An interesting example for a study that used an unusual assortment of “voices” (or stories) about students is Radovic, Black, Salas, and Wil- liams (2017). This study examined the construction of posi- tive mathematical identities in the case of three girls. The uniqueness of the study stemmed from the fact that a full classroom was interviewed regarding each other, including social status and groupings, as well as students’ identifica- tions of each other in terms of mathematical ability. Some studies, such as Darragh (2015), explicitly distance them- selves from the “narrative” definition of identity, claiming to adopt, a “participative” identity instead. They state: “We become a mathematics learner in a performative manner, and it is the repetition of ‘performances’ in mathematics learning contexts that generates our recognition of our- selves in certain ways as learners of mathematics.” (p. 85). However, in this work, as in others that use the identity as The weakness, instead, often lies in the operationaliza- tion of identity, namely, in stating precisely what identity is (for the researchers) and how it can be empirically studied. Some researchers choose to avoid the complexities of defin- ing identity by focusing on an aspect (or sub-concept) that is more operationalizable, in their framework. • Theoretical frameworks Although previous reviews found some works in the field relying on psychoanalytical frameworks, we found no such work in our review. In fact, it seems in the division between the “Erikso- nian” identity and “Meadian” identity pointed to by Darragh (2016), the field has chosen the “Median” identity with such a consensus that we wonder whether this movement is not leaving some potentially fruitful paths behind. For example, we found hardly any reliance on the writings of Erikson, Freud and their successors in the field of psychology, on the development of self and identity. Some exceptions could be found in Hall, Towers and Martin (2018), who drew, in part, on Markus and Oyserman’s (prominent researchers in social psychology) concepts of possible selves and self-concept. Yet these authors too, complemented their theoretical back- ground by citing socio-cultural theories. Thus, none of the reviewed papers relied solely on psychological or psycho- analytical theory. • Theoretical frameworks Similar to previous reviews, we found almost all papers relied on some form of socio-cultural frame. Since many of the works drew on multiple frameworks (though several did not identify them as “frameworks” but simply cited cen- tral works) it was hard to differentiate between sub-strands of socio-cultural theory. Still, similar to previous reviews, Holland et al. (1998), Sfard and Prusak (2005) and Lave and Wenger (1991)/Wenger (1998) figure prominently in 1 3 M. Graven, E. Heyd‑Metzuyanim 368 who they (students) are in relation to each other” (p. 81). Others define the concepts related to identity well, but leave what “identity” itself is vague and undefined. And some do define identity, yet the definition is so broad that it is dif- ficult to pinpoint the operationalization of all its subparts. For example, Boaler and Selling (2017) define identity as “a set of ideas, beliefs, and behaviors that may be performed in specific domains, such as the learning of mathemat- ics”. They claim that “Mathematical identity involves the ways in which students think about themselves in relation to mathematics and the extent to which they have devel- oped a commitment to, are engaged in, and see value in mathematics and in themselves as learners of mathematics” (p. 82). This definition includes “ideas”, “beliefs”, “ways in which the students think about themselves”, “commit- ment”, “engagement and “seeing values”—all of which are difficult concepts to operationalize. A few papers provided operationalized definitions of identity by drawing on “iden- tity as narrative” and in particular Sfard and Prusak (2005) (Andersson, Valero and Meaney 2015; Heyd-Metzuyanim 2017; Heyd-Metzuyanim and Graven 2016; Heyd-Metzuya- nim, Tabach and Nachlieli 2016; Larnell 2016). Lutovac and Kaasila (2018b) rely on the definition of “identity as narra- tive” taken from other sources (Ricoeur 1991). Some, such as Oppland-Cordell and Martin (2015), relied on Wenger’s (1998) theorizing of identity, not defining the term “identity” itself, but rather defining the concepts related to identity according to his theory, namely engagement, imagination and alignment.i the theoretical backgrounds of the papers reviewed. Some notable additional frameworks for conceptualizing identity were based on Cultural Historical Activity Theory (Joo- ganah and Williams 2016; Kaspersen, Pepin, and Sikko 2017), Archer’s Social Realism (Westaway and Graven 2019), Positioning Theory (Darragh 2018), Bakhtin’s theory of dialogism (Braathe and Solomon 2015) and Ricoeur’s (1991) social theory (Lutovac and Kaasila 2018b). 1 3 • Where are the mathematical objects? Our question What is the main storyline? was probably the most difficult to answer succinctly, since studies varied considerably in goals, questions asked and “bottom-line” answers. Also, since most studies were qualitative and based on very small samples, they rarely made generalized conclu- sions. Still, we were able to identify the following goals and functions for using the concept of identity underlying the studies we reviewed: Generally, the vast majority of the papers either back- grounded mathematics or dealt with it very tangentially. Very few (around 9) papers included specific examples of mathematical tasks or discussed mathematical content. In most of the studies, both those dealing with student identi- ties and teacher identities, mathematics was talked about as a school discipline which students/teachers relate to, talk about or are observed in, but the mathematical con- tent with which students/teachers were engaged was rarely presented or explored. In the cases where mathematics was foregrounded, it was mostly split from the discussion of identity or identity was tangential to the focus of the study. For example, Takeuchi (2018) discusses multiplication algorithms children of immigrant parents use, and parents’ involvement in the learning of these algorithms. However, the concept of identity is relatively tangential in this study, and the analysis of positional identities is not directly related to the multiplication algorithms. Some exceptions to the split between identity and mathematical content could be found in works by Heyd-Metzuyanim (2017) and Heyd-Metzuyanim and Graven (2016) who analyze identifying and mathematiz- ing processes concomitantly. These studies rely on the com- mognitive framework (Sfard 2008) to analyze identity con- struction processes as they happen in the classroom, while also foregrounding the mathematical routines followed by students. 1. Making a socio-political claim. Studies such as these were often designed to examine issues of racial ineq- uities (McGee 2015; Larnell 2016; Oppland-Cordell and Martin 2015), gender inequities (Foyn et al. 2018; Radovic et al. 2017; Solomon, Radovic and Black 2016), the need to change ways of working with students with disabilities (Goldstein 2018; Lambert 2015) or students from disadvantaged backgrounds (Heyd-Metzuyanim and Graven 2016; le Roux and Adler 2016; Takeuchi 2018). This is often a topic of studies of learner identi- ties. 2. 2. Providing a relatively holistic lens to examine learn- ers’ experiences in relation to their social context. Such studies (Black and Hernandez-Martinez 2016; Braathe and Solomon 2015; Darragh 2015; Heyd-Met- zuyanim 2017; Kaspersen et al. • Definitions and operationalization of identity For exam- ple, “positional identities” have been used by some (Foyn, Solomon and Braathe 2018; Langer-Osuna, 2015; Takeuchi, 2018), mostly based on Holland et al.’s (1998) framework. In these works, “positional identities” are mostly related to concepts of power and roles. For example, Foyn et al. (2018) define positional identities as “different roles which define 1 3 1 3 Mathematics identity research: the state of the art and future directions 369 “performance” or identity as forms of participation, what such “performance” is, and especially how it is assessed, is not clearly defined. Is it what the researcher recognizes in the interviewee’s actions? Is it what the author believes that the learner/teacher thinks himself to be “performing”? Interestingly, it is precisely with this vagueness that Sfard and Prusak’s (2005) delineation of 1st Person (1st P) sto- ries from 3rd Person (3rd P) stories, could be of help. In particular, “performances” are usually, at least in effect, the researcher’s 3rd P stories about the participants, based on his/her observations and analysis. Yet researchers aligning with the performative/participationist stance usually do not build on Sfard and Prusak’s (2005) work, likely due to the association with the identity as narrative rather than partici- pative perspective. had occurrences of ontological collapses. These could be seen in two forms: collapsing between what participants reported (their narratives) and their subjective experiences (feelings, beliefs etc.) and collapses between 1st P stories (told by the participant) and 3rd P stories (told about the participant by the researcher). Studies that relied on various sources of information regarding participants (interviews, observations, ethnographic data) were especially prone to the second type of collapse. Thus, often, researchers did not differentiate between the identity they had authored about the participant based on their observations, and the identity the participant authored in his/her interviews. • What studies on identity are saying and what identity is used for • What studies on identity are saying and what identity is used for • Where are the mathematical objects? Most of these claims relate to providing students with richer and more diverse opportunities to engage with mathematics than are com- mon in ‘traditional’ classrooms (Andersson et al. 2015; Boaler and Selling 2017; Langer-Osuna 2015; Pickard- Smith 2018; Wake 2014) as well as non-traditional assessment practices (Heritage and Wylie 2018). Some others argue for giving more attention to certain stages of transition in students’ mathematical lives such as from high-school to university (Jooganah and Williams 2016). 4. To make pedagogical claims. Most of these claims relate to providing students with richer and more diverse opportunities to engage with mathematics than are com- mon in ‘traditional’ classrooms (Andersson et al. 2015; Boaler and Selling 2017; Langer-Osuna 2015; Pickard- Smith 2018; Wake 2014) as well as non-traditional assessment practices (Heritage and Wylie 2018). Some others argue for giving more attention to certain stages of transition in students’ mathematical lives such as from high-school to university (Jooganah and Williams 2016). Interviewee Researcher Our highlighting of ontological collapses also relates to the concern, expressed by some in the field of mathematics education, of “essentializing” identity. For example, Jurdak and colleagues (Jurdak, Vithal, de Freitas, et al. 2016), state that “Without intending to do so, some research on identity in mathematics education seems to re-entrench stereotypes about what sorts of identities excel at mathematics.” (p. 15). Such essentializing—meaning the objectification of iden- tity into a mental stable entity that has a causal effect on behavior, often goes hand in hand with ontological collapses that hide the fluid, dynamic and situated nature of identity. Notably, some researchers have attempted to avoid this objectification of identity with terms relating to the process of identifying or identification (e.g. Heyd-Metzuyanim and Sfard 2012) rather than to identity as an object. However, this attempt does not seem to have been taken up by many, perhaps because the field defines itself according to the word “identity” (as can be seen, for example, in the key-words used for most reviews on the subject). Other attempts for “processualizing” the talk on identity have leaned on the term “positioning”, which is usually used as a verb and indi- cates a process (Bishop 2012; Wood 2013). Yet rarely have the links between the acts of “positioning” and the forma- tion of an “identity” been clearly defined. 3.3 Discussion Our analysis reveals that, in general, the field is slowly mov- ing towards more conceptual coherence in several aspects. First, there is unanimous reliance on social (“Meadian”) views of identity rather than on psychological ones. We note that although this may be a sign of maturation and stabili- zation of the field, some important insights from the field of psychology may be left unnoticed. Thus, there seems to be a widening gap between studies of affect (beliefs, self- concept etc.) and studies of identity (see also Heyd-Met- zuyanim 2017). This gap seems to be wider in studies of learners’ identities, which rarely build on literature from the field of affect (beliefs, emotions and attitudes) in comparison to studies of teachers’ identity, which still often connect to studies on teachers’ beliefs.i A second indicator of maturation of the field in terms of conceptual coherence can be seen in the growing number of studies that are quite explicit about their definition of identity. We note, however, that explicit definitions are not always operational. In relation to this gap, the participative/ performative types of definitions seem to be particularly challenging to operationalize. In our analysis, we have used Sfard and Prusak’s (2005) definition of identity not as “the” ultimate definition, but rather as a lens through which other definitions can be examined. This examination has revealed that although most perspectives view the audience of iden- tity-authoring as critical, many of the studies, including those that align with participative/performative conceptu- alizations, rely mostly on interviews. Thus, we found insuf- ficient reflection on how identity or identity-work would be different if interview questions were asked by others (not researchers) or simply raised in discussion in a different con- text—such as with families of the interviewees or with their other teachers. If reflecting on one’s mathematics learning • Where are the mathematical objects? This is something which Gresalfi and Hand (2019), as well as Andersson and Wagner (2019) take up in this issue. (i) New research areas of focus • Where are the mathematical objects? 2017; Langer-Osuna 2018) usually put identity at the focus of their exam- ination. Their goal is to illuminate how identity as a concept serves as a lens for examining students’ experi- ences in the mathematics classroom as multi-faceted, multi-voiced, and impacted by various factors inside and outside of the classroom. • Ontological collapses—implicit unification of different types of identity narratives 3. In studies focusing on teachers’ identities, identity is used to examine teachers’ experiences in vari- ous stages of their career (Hodges and Hodge 2017; Knapp 2017), including pre-service and in-service teachers’ change and growth (Akkoç, Yeşildere-Imre and Balkanlioǧlu 2014; Losano, Fiorentini and Villar- Studies that adopted a discursive viewpoint usually did not have ontological collapses. They were very careful to report about participants’ narratives and did not make claims about anything else. In contrast, studies that used “identity” as an eclectic (“catch-all”) term for affect often 1 1 3 370 M. Graven, E. Heyd‑Metzuyanim real 2017; Lutovac and Kaasila 2014, 2018b; Skog and Andersson 2015; van Putten, Stols, and Howie 2014), teachers’ reactions to reform (Westaway and Graven 2019), the meaning (and support) of certain teaching practices such as teaching for numeracy skills (Bennison 2015) and teaching–learning interactions in the context of pre-service education (Heyd-Metzuyanim et al. 2016) or teaching involves identity work, then understanding how this occurs differently in settings other than researcher-inter- viewee would be critically important as the importance of audience implies we are getting a very one-dimensional set of data of teacher/learner identity work as a result of the pre- sent focus on interviews. With relation to this possible skew, we also note the necessity of raising awareness of ontologi- cal collapses, to which participative/action-based definitions seem particularly prone. If identity is to be assessed by mul- tiple observations, authors and in different situations, there could be an understandable inclination for the researcher to collapse the various authored stories into one “identity” as participative/performative definitions invite. Yet such col- lapses often hide major parts of the stories-collection as they are limited to IntervieweeIntervieweeResearcher stories. real 2017; Lutovac and Kaasila 2014, 2018b; Skog and Andersson 2015; van Putten, Stols, and Howie 2014), teachers’ reactions to reform (Westaway and Graven 2019), the meaning (and support) of certain teaching practices such as teaching for numeracy skills (Bennison 2015) and teaching–learning interactions in the context of pre-service education (Heyd-Metzuyanim et al. 2016) 4. To make pedagogical claims. 4 Contributions of the present special issue The use of Nasir and Cooks (2009) ideational, material and relational resources not only provides a language of description for the what of the PD the teachers participated in, but also provides new insights into how ideational resources influence individual mathematics teacher identities (MTIs), showing both simi- larities and differences. They further argue that the insights into the interrelationship between the what and how of the PD and developing MTIs does not sufficiently explain differ- ences in MTIs—for this they include teacher motivations for PD participation that enables exploration of links between teacher learning and their practice-linked identities. Another work that stretches the realm of identity to cor- poreal aspects is that of Gholson and Martin (2019). This study connects between the micro-interactions seen on video between a Black girl and a White female assistant teacher, and connects these micro-interactions to the wider racial narratives in the US that “press” upon the Black girl and form a painful experience for her in the mathematics classroom. This work relies on Butler’s (1988) theorizing of gender as performance, connecting it with studies of Black students’ cool pose as a protective coping mechanism that enables negotiating race, class and gender in urban settings. Gholson and Martin stretch the notion of performance and relational labor to the realm of phenomenological experi- ence. They use their own positioning in the US society as means for understanding the black girl’s “cool pose”, her interpretations of the classroom situation and its connections to the physical pain (“head aches”) she feels when doing mathematics. Grootenboer and Edwards-Groves (2018), similarly to Ntow and Adler (2019), argue for a practice perspective on mathematics identity though with the emphasis on the interconnections between how developing a mathematics identity is bound up with experiences of the sayings, doings and relatings in the mathematics classroom. Drawing on the work of Kemmis et al. (2014), where learning is defined as a process of being stirred into practices, they illuminate, through transcript analysis of ‘a regular everyday-type’ Australian lesson, the way in which learner’s identities are formed through being stirred into these practices. Their analysis highlights that “one with a mathematical identity is someone who is stirred into the substantive mathematics practices (what is being learned) and learning practices of mathematics (how it is being learned)” (this issue). 4 Contributions of the present special issue In this special issue we have identified different types of contributions that advance the field and address many of the issues raised in earlier critiques of identity work in the dis- cussion above. These contributions involve: (1) new areas or foci of research (2) bringing new perspectives to mathemat- ics identity research and (3) novel methodologies. 1 3 3 371 Mathematics identity research: the state of the art and future directions As pointed to by Radovic et al. (2018), research on young students’ identities has been relatively rare. Black, Choudry, Pickard-Smith and Williams (2019) take a step forward in this direction by focusing on the emotion-cognition dialectic in the identity formation of a 6 year old boy. Black and her colleagues focus in particular on the emotion-cognition of surprise—astonishment that was expressed by the boy dur- ing a play activity where the children played ‘as if’ they were sellers and buyers in a ‘Pizza shop’. Later, the research and child watched together a video of this play activity, which elicited surprise and laughter from the boy, and identifica- tions of ‘being clever’. This contrasted with his experience of feeling ‘tired’ during classroom mathematics. This study highlights the role of the emotion-cognition experience in acts of identification in joint activity between the child and the adult. Black et al. (2019) suggest that young children’s emotional experiences in the early primary phase, which they equate with the Vygotskian concept of ‘perezhivanie’, may be key for the development of more crystalized narra- tive identities developed later on in life. productive tensions and drawing on zone theory contribute a theory of goal-directed change that is useful for understand- ing identity development and the role of individual agency within socio-cultural contexts. This theoretical advancement is useful for the field of mathematics identity more broadly than simply for mathematics teacher educators, which con- nects their contribution with (ii) below. (ii) New theoretical perspectives shedding light on math- ematics identityi Ntow and Adler (2019) bring afresh to the field Nasir and Cooks (2009) notions of practice-linked identities and iden- tity resources (ideational, material and relational) to explore how mathematics teaching identities developed in relation to the resources offered in a particular Professional Devel- opment (PD) in South Africa. Their use of this work from out-of-field is original and novel, and so is their explicit focus on linking developing mathematics teacher identities (MTIs) with resources offered in PD. 4 Contributions of the present special issue The basic concepts used from Laclau and Mouffe’s theory are “nodal points” (moments that occupy central positions in a discourse) and “floating signifiers” (nodes which remain particularly open to dif- ferent ascriptions of meaning). These are supplemented by the ideas of “hegemony” and “discursive struggle”. Using these concepts, Chronaki and Kollosche analyze one inter- view with a 15 year old German girl named Anja. The suggested theoretical framework enables the authors to show how Anja’s relation with mathematics can be under- stood as a struggle between the “proper mathematics” that retains hegemony and is described as “pointless”, involv- ing “remaining still and listening” and being “humiliated at the blackboard”, and Anja’s identity work which identi- fies her as a bodily active and a collaborative individual. Chronaki and Kollosche claim that such a discourse theory enables circumventing deficit discourses, allowing them “to avoid blaming either the students or the teacher by locating the divergent discourses from both sides in the wider discursive field” (this issue). Another framework offered in this issue for looking at teachers’ identity is that of social realism, put forward by Westaway (2019). Drawing on Archer’s (2000) stratified view of persons as including sense of self, personal identity and social identity, Westaway (2019) focuses on reflexivity, which she defines as “the internal conversation individuals have about themselves and the situations they find them- selves in”. Westaway builds on Archer’s differentiation between different modes of reflexivity—Communicative, Autonomous, Meta-reflexives, and Fractured reflexives to examine the emergence and expression of a Grade 3 math- ematics teacher identity in South Africa. Her analysis of both interview data with the teacher (named Buhle) and observational data of Buhle’s teaching practice, leads her to identify Buhle as shifting from being an “autonomous reflexive” in her early years, when she was self-reliant in her decision making and primarily concerned with upward mobility in the social order, to a “communicative reflexive” who reproduces the status quo, maintaining old teaching roles in the face of reform efforts. Like Skott (2018), Westa- way (2019) claims to “reinsert the individual”—this time, into the process of social morphogenesis and morphostasis. 4 Contributions of the present special issue The work of Goos and Bennison (2018) in the Austral- ian context focuses on analyzing the identity formation of mathematics teacher educators (those contributing to the preparation of future teachers of mathematics)—both math- ematicians and mathematics educators. They built on their earlier work, which brought a zone theory perspective to analyzing identity formation in mathematics education—of pre and in-service teachers, including out of field teachers teaching numeracy. In this paper, however, they focus on the identities of mathematics teacher educators, shedding light on this under-researched area as mathematics identity research has in the past focused either on teachers or learners or in some cases parents or mathematicians—but seldom the identities of those educating future teachers. Their analysis foregrounds the way in which teacher educators navigate Another novel theoretical approach to mathematical identity can be found in the study by Fellus (2019). Fel- lus builds on Ivanič (1998) networked identity model which differentiates between four types of identity: Authorial, Autobiographical, Discoursal and Socioculturally available identities. This framework is exemplified by Fellus through a review of 35 papers on mathematics learners’ identity. Fel- lus derives from these studies central insights gained in the past decades of study on each one of these identity types. 1 1 3 3 372 M. Graven, E. Heyd‑Metzuyanim from her students. Skott argues that PoP allows us to gain perspective on how “past and present practices and figured worlds function as generalised others” and that “The sense of agency in the notion of identity is then a matter of how Anna draws on and combines these generalized others,… when authoring an answer to the immediate situation” (this issue).f She argues that “Each of the four identity-related dimensions needs to be recognized as part of a larger system that has the potential to allow researchers as well as practitioners to have a view of the creases, wrinkles, and folds of the landscape we call identity” (this issue). Starting with the claim that research in the field of iden- tity “lacks a framework theoretically and methodologically to address the political dimension” (this issue), Chronaki and Kollosche (2019) suggest Laclau and Mouffe’s dis- course theory as a basis for extending the study of identity to the socio-political field. 4 Contributions of the present special issue By that, she is challenging “the assumption that teachers are mere products of a social system” (this issue) and aims to highlight the role of teachers’ agency in the reproduction of old systemic roles.i i An additional novel theoretical framework for looking at identities—this time teachers’ identities, is offered by Skott’s (2018) Patterns of Participation (PoP). “PoP sup- plements the notions of practice and figured worlds from social practice theory, with the notion of self in Chicago- school symbolic interactionism” (Skott 2018, this issue). Skott combines these concepts to “re-centre the individ- ual” (ibid) while keeping a participatory stance towards the study of identity. The novel theoretical framework is exemplified on the case of Anna—a teacher who was fol- lowed by Skott for 4 years, since her graduation from pre- service education. The data sources are wide in time and scale, including multiple interviews, observations, infor- mal talks, numerous observations and even interviews with Anna’s colleagues. From this vast database, Skott weaves a story about two different phases that Anna underwent during the 4 years: from being new in the Northgate school and identifying herself first and foremost as “a mathemat- ics teacher” to being a “mathematics teacher at North- gate”. The latter identification puts greater emphasis on her institutional affiliation and epitomizes other processes that Anna underwent such as working more collaboratively with her team while developing some professional distance Gresalfi and Hand (2019) is another paper with a clear theoretical contribution, this time, weaving together different strands of literature that take a situative perspective to math- ematics identity. Their proposed model connects Norms, Frames and Narratives as three different types of resources students draw on to develop their mathematical identity. The goal of this model, Gresalfi and Hand suggest, is to offer “a design heuristic” for educators who wish to develop certain identities in the classroom. They focus on two particular interventions: intervening by disrupting racial narratives, exemplified through the work of Gutstein (2016); and inter- vening through disrupting classroom norms and practices, exemplified by the work of Boaler (2008). Their proposed model affords detecting both coherence and incoherence in programs of intervention by examining how the different lev- els (narratives and norms) relate to each other and whether one level is neglected at the expense of the other. (iii) Methodological contributions (iii) Methodological contributions Several of the papers mentioned above indeed also pro- vide methodological contributions in terms of new analytical 3 3 3 Mathematics identity research: the state of the art and future directions 373 methods. Here we mention those papers whose contribution we identified as primarily methodological. certain positions and certain available identities. Their illus- trative example shows how positions and taking up certain identities can change when students turn to draw on out- of-school discourses, such as the “hobby discourse”. Spe- cifically, it enabled a breakthrough in the group’s problem- solving efforts by affording one of the students (who was previously positioned negatively in the friendship discourse) an opportunity to draw on his identity as a Rubik’s Cube player. This move changed this student’s positioning, and offered the possibility for others to listen to his solution and build upon it. i Lutovac and Kaasila (2018c) provide an original review of the methodological landscape of teacher identity research in mathematics education. Examining 52 studies published between 2000 and 2017 in peer reviewed scholarly journals they argue that the overwhelming dominance of qualitative approaches (with interviews, written reflections or narra- tives, and classroom observations) and limited analytic approaches is constraining what research is able to offer the field. They argue that future research on mathematics- related teacher identity should (1) consider mixed methods approaches that would support greater generalization of find- ings, (2) combine research on different groups of teachers (such as research into both pre- and in-service teachers or elementary and higher grade mathematics teachers) so as to facilitate the examination of the intersection of differ- ent groups of teacher identities and (3) be explicit in their writing about the methodological decisions they make (in particular, they call for the use of methodological labels). Another study, which attempts to re-construct identity from discursive actions, is Heyd-Metzuyanim and Shabtay (2019). This paper explores teacher identities in relation to their pedagogical discourse (their narratives about ‘good’ teaching) and how this links with broader pedagogical dis- courses available. They identify two key forms of pedagogi- cal discourse, namely Exploration Pedagogical Discourse (EPD) and Acquisition Pedagogical Discourse (APD). These they link with Sfard and Lavie’s (2005) commognitive dis- tinction between ritual and explorative participation. (iii) Methodological contributions EPD links with widely promoted mathematics education reform literature (and promoted in the official documentation pro- vided to the teachers in their study by the Israeli minis- try) and APD links with ‘traditional’ or teacher-centered instruction. Heyd-Metzuyanim and Shabtay’s novel focus on researching individual teacher identity in relation to the con- struct of multiple pedagogical discourses available (official and unofficial) extends the fields’ understanding and use of the notion of pedagogic identities. Their paper contributes a useful analytic frame for understanding individual teacher identities as the nexus of narratives adopted from different pedagogical discourses. While this special issue remains skewed towards quali- tative methodologies, there are some particularly novel analytic approaches that disrupt the monotony of the land- scape noted above. Darragh and Radovic’s (2019b) study on Chilean primary mathematics teacher identities after PD, while following a classic thematic analysis of semi- structured teacher interviews, introduces the innovative concept of teachers’ collective identity performances. This concept is illuminated through their combining ten indi- vidual teacher narrative analyses into a movie synopsis of a fictional protagonist ‘Tia’. Their analysis and presentation of the collective identity performances of the teachers high- light dissonance between what teachers perceived as attrac- tive identities (e.g. mother-saviour) and what was promoted in the PD (e.g. allowing learners to struggle independently with problem solving). Their innovative method of combing the data of teachers and weaving it into a fictional movie script enabled breadth and depth that powerfully illuminate how teacher identities are culturally and contextually pro- duced. This achievement turns attention to the importance of PD, noting and responding to these cultural and educational contexts. References Akkoç, H., Yeşildere-Imre, S., & Balkanlioǧlu, M. A. (2014). Examin- ing professional identity through story telling. Research in Math- ematics Education, 16(2), 204–205. https://doi.org/10.1080/14794 802.2014.918340. Andersson, A., Valero, P., & Meaney, T. (2015). “I am [not always] a maths hater”: Shifting students’ identity narratives in context. Educational Studies in Mathematics, 90(2), 143–161. https://doi. org/10.1007/s10649-015-9617-z. The contribution of papers in this special issue, we argue, has taken the field forward in terms of new and more elab- orate ways of looking at identity from action, rather than (only) from self-reports. It also contributes new connections between identity and other concepts: positioning (Andersson and Wagner), emotion (Black et al.), framing and cultural narratives (Gresalfi and Hand). We hope that these contribu- tions will assist also for questions looking at teachers and students’ identities concomitantly. This, since the exami- nation of classroom activity can provide a useful ground for studying how students and teachers enact their identities together. Andersson, A., & Wagner, D. (2019). Identities available in intertwined discourses: Mathematics student interaction. ZDM Mathematics Education. https://doi.org/10.1007/s11858-019-01036-w. Archer, M. S. (2000). Being Human: The problem of agency. Cam- bridge: Cambridge University Press. Beijaard, D., Meijer, P. C., & Verloop, N. (2004). Reconsidering research on teachers’ professional identity. Teaching and Teacher Education, 20, 107–128. Bennison, A. (2015). Supporting teachers to embed numeracy across the curriculum: A sociocultural approach. ZDM International Journal on Mathematics Education, 47(4), 561–573. https://doi. org/10.1007/s11858-015-0706-3. g Bishop, J. (2012). “She’s always been the smart one. I’ve always been the dumb one”: Identities in the mathematics classroom. Journal for Research in Mathematics Education, 43(1), 34–74. Our review of the field showed continued dominance of specific theorists and theories in the field. A key contribution of this special issue is the proliferation of theories. We see this as a strength of this issue and an indication of the field growing. However, this proliferation can also bring chal- lenges as has been noted in discussions of multiple theories in mathematics education in general (e.g. Radford 2008). The most pressing challenge would probably be, as noted also in our review above, the issue of definition of identity. Black, L., Choudry, S., Pickard-Smith, K., & Williams, J. (2019). The- orising the place of emotion–cognition in research on mathemati- cal identities: The case of early years mathematics. ZDM Math- ematics Education. https://doi.org/10.1007/s11858-018-01021-9. Black, L., & Hernandez-Martinez, P. (2016). 5 Summary and implications for future directions From the review of literature prior to this Special Issue we are able to both note some continued absences in the field as well as absences that have been taken forward by the authors herein. We note several divisions calling for researchers’ attention. For example, attention is needed to the gap between identity and mathematical objects. Unless identity is made relevant for actual learning processes in the mathematics classroom, and more particularly, in specific domains of mathematics (arithmetic, algebra, geometry etc.), it is likely that teachers will not pay much attention to it. This point echoes Jorgensen’s (2014) critique of social theories of learning in general. Jorgensen argues that mathematics edu- cators drawing on social theories and identity often render mathematics invisible. As a result, what is being advocated by these studies has made little progress in enabling greater In the domain of learners’ identity in action, Andersson and Wagner (2019) propose a novel methodology for exam- ining students’ positioning moves with relation to their lan- guage repertoires. They use an interaction among a group of grade 10 students who are engaged in a problem-solving situation to exemplify their conceptualization of identities as drawing on a “torus” of discourses that are available to a per- son. Within the available discourses, Andersson and Wag- ner identify several authority structures: Personal Author- ity, Discourse as Authority, Implicit Discursive Authority and Personal Latitude. Each one of these structures affords 1 3 1 3 M. Graven, E. Heyd‑Metzuyanim 374 access to mathematics of marginalized communities. She writes ‘the social theory has failed to make any substantive inroads into challenging the status quo in terms of equity, access and/or success. It is almost as if mathematics has become quite secondary to any analysis.” (p. 314). Jorgensen further argues that despite these theories significant con- tribution around the impact of context in a broader sense than mathematics, “such a position is limiting within this field. It could be applied to almost any field, and thus render the mathematics per se as silent.” (p. 315). Three papers in this issue foreground mathematics—Heyd-Metzuyanim and Shabtay, Andersson and Wagner, and to some extent Gholson and Martin. It is notable that these studies are also those that foreground discourse and the extraction of identity from discursive actions. 5 Summary and implications for future directions This highlights how important this building on discourse-action data is for bringing in the math- ematics, and points to the productiveness of pushing forward methodologies that extract identity from classroom activity. Fellus (2019) and Greslafi and Hand (2019) contributions in this volume, concerning the theorizing of identity from multiple viewpoints, offer ways forward for authors wishing to locate their work in the ever-growing supply of theories and methodologies for studying identity. The field of identity research in mathematics education has become established and is continuing to grow and find new creative and innovative ways of understanding and explaining mathematics complex learning and teaching pro- cesses. We hope this Special Issue contributes to pushing the boundaries of this field theoretically, empirically and methodologically. Acknowledgements This work is supported by the South African Research Chairs Initiative of the Department of Science and Technol- ogy and National Research Foundation (Grant no. 74658). We thank Pamela Vale for her assistance with the literature search. Open Access This article is distributed under the terms of the Crea- tive Commons Attribution 4.0 International License (http://creativeco mmons.org/licenses/by/4.0/), which permits unrestricted use, distribu- tion, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. g y y Another gap persists between learners’ and teachers’ identities, a gap noted already by Heyd-Metzuyanim, Luto- vac, and Kaasila (2016) and repeatedly observed, even in conference practices such as dividing sub-groups of the upcoming ICME 14 into those relating to learners’ identities and those relating to teachers’ identities (http://www.icme1 4.org/). The study of learners and teachers’ identities as they relate to each other is a curious absence, given the fact that most studies of identity rely on frameworks that see learning as becoming a participant in a certain community. It would seem probable that such a view would propel researchers to examine the central players in such a community (learners and teachers) concomitantly, yet such concomitant views are practically absent, at least in the studies we reviewed. 3 References Re-thinking science capi- tal: The role of “capital” and “identity” in mediating students’ engagement with mathematically demanding programmes at university. Teaching Mathematics and Its Applications, 35(3), 131–143. https://doi.org/10.1093/teamat/hrw016. 1 3 375 Mathematics identity research: the state of the art and future directions Boaler, J. (2008). Promoting ‘relational equity’ and high mathematics achievement through an innovative mixed-ability approach. Brit- ish Educational Research Journal, 34(2), 167–194. Boaler, J. (2008). Promoting ‘relational equity’ and high mathematics achievement through an innovative mixed-ability approach. Brit- ish Educational Research Journal, 34(2), 167–194. Graven, M., & Lerman, S. (2014). Mathematics teacher identity. In S. Lerman (Ed.), Encyclopaedia of Mathematics Education (pp. 434–438). Dordrecht: Springer. Boaler, J., & Greeno, J. (2000). Identity, agency, and knowing in math- ematics worlds. In J. Boaler (Ed.), Multiple perspectives on math- ematics education (pp. 171–200). Westport, CT: Ablex. Graven, M., & Lerman, S. (2019). Mathematics teacher identity. In S. Lerman (Ed.), Encyclopedia of Mathematics Education. Dordrecht: Springer. (in press).i Boaler, J., & Selling, S. K. (2017). Psychological imprisonment or intellectual freedom? A longitudinal study of contrasting school mathematics approaches and their impact on adults’ lives. Journal for Research in Mathematics Education, 48(1), 78. https://doi. org/10.5951/jresematheduc.48.1.0078. Gresalfi, M., & Hand, V. M. (2019). Coordinating situated identities in mathematics classrooms with sociohistorical narratives: A con- sideration for design. ZDM Mathematics Education. https://doi. org/10.1007/s11858-019-01034-y. Grootenboer, P., & Edwards-Groves, C. (2018). Learning mathemat- ics as being stirred into mathematical practices: An alternative perspective on identity formation. ZDM Mathematics Education. https://doi.org/10.1007/s11858-018-01017-5. g j Braathe, H. J., & Solomon, Y. (2015). Choosing mathematics: The nar- rative of the self as a site of agency. Educational Studies in Mathe- matics, 89, 151–166. https://doi.org/10.1007/s10649-014-9585-8. p g Brubaker, R., & Cooper, F. (2000). Beyond “identity”. Theory and Society, 29, 1–47. p g Gutstein, E. (2016). “Our issues, our people—Math as our weapon”: Critical mathematics in a Chicago neighborhood high school. Journal for Research in Mathematics Education, 47(5), 454–504. Butler, J. (1988). Performative acts and gender constitution: An essay in phenomenology and feminist theory. Theatre Journal, 40(4), 519–531. https://doi.org/10.2307/3207893. Hall, J., Towers, J., & Martin, L. C. (2018). Using I poems to illuminate the complexity of students’ mathematical identities. Educational Studies in Mathematics, 99(2), 181–196. https://doi.org/10.1007/ s10649-018-9839-y. Chen, R. J. (2017). Prospective elementary teachers’ aesthetic experi- ence and relationships to mathematics. Journal of Mathematics Teacher Education, 20(3), 207–230. Harré, R., & van Langenhove, L. (1999). Positioning theory. References Oxford: Blackwell.i Chronaki, A., & Kollosche, D. (2019). Refusing mathematics: A dis- course theory approach on the politics of identity work. ZDM Mathematics Education. https://doi.org/10.1007/s11858-019- 01028-w. Heritage, M., & Wylie, C. (2018). Reaping the benefits of assessment for learning: Achievement, identity, and equity. ZDM Mathemat- ics Education, 50(4), 729–741. https://doi.org/10.1007/s1185 8-018-0943-3. Darragh, L. (2015). Recognising ‘good at mathematics’: Using a per- formative lens for identity. Mathematics Education Research Journal, 27(1), 83–102. Heyd-Metzuyanim, E. (2015). Vicious cycles of identifying and math- ematizing – a case study of the development of mathematical fail- ure. Journal of the Learning Sciences, 24(4), 504–549.f Darragh, L. (2016). Identity research in mathematics education. Educational Studies in Mathematics, 93(1), 19–33. https://doi. org/10.1007/s10649-016-9696-5. Heyd-Metzuyanim, E. (2017). Identity as a nexus of affect and dis- course in mathematical learning. For the Learning of Mathemat- ics, 37(3), 33–38. Darragh, L., & Radovic, D. (2019b). “To Tia with love”: Chilean math- ematics teacher identities after professional development. ZDM Mathematics Education. https://doi.org/10.1007/s11858-018- 01023-7. Heyd-Metzuyanim, E., & Graven, M. (2016). Between people-pleasing and mathematizing—South African learners’ struggle for numer- acy. Educational Studies in Mathematics, 91(3), 349–373. Darragh, L., & Radovic, D. (2019a). Mathematics learner identity. In S. Lerman (Ed.), Encyclopedia of mathematics education. Dordrecht: Springer. (in press). Heyd-Metzuyanim, E., Lutovac, S., & Kaasila, R. (2016a). Identity. In M. S. Hannula, G. A. Goldin, E. Heyd-Metzuyanim, A. Jansen, R. Kassila, S. Lutovac, P. Di Martino, F. Morselli, J. A. Middle- ton, M. Pantziara, & Q. Zhang (Eds.), Attitudes, beliefs, motiva- tion, and identity in mathematics education: An overview of the field and future directions. New York: Springer Open. https://doi. org/10.1007/978-3-319-32811-9_1. Darragh, L. (2018). Loving and loathing: Portrayals of school math- ematics in young adult fiction. Journal for Research in Mathemat- ics Education, 49(2), 178–209. http://www.jstor.org.proxy.mul. missouri.edu/stable/10.5951/jresematheduc.49.2.0178. Heyd-Metzuyanim, E., & Sfard, A. (2012). Identity struggles in the mathematics classroom: On learning mathematics as an interplay of mathematizing and identifying. International Journal of Edu- cational Research, 51–52, 128–145. https://doi.org/10.1016/j. ijer.2011.12.015. Erikson, E. H. (1968). Identity, youth and crisis. New York: Norton. Evans, J., Tsatsaroni, A., & Czarnecka, B. (2014). Mathematical images in advertising: Constructing difference and shaping iden- tity, in global consumer culture. Educational Studies in Math- ematics, 85(1), 3–27. https://doi.org/10.1007/s10649-013-9496-0. Heyd-Metzuyanim, E., & Shabtay, G. (2019). Narratives of ‘good’ instruction: Teachers’ identities as drawing on exploration vs. acquisition pedagogical discourses. ZDM Mathematics Educa- tion. https://doi.org/10.1007/s11858-018-01019-3. Fellus, O. O. (2019). References Jurdak, M., Vithal, R., de Freitas, E., Gates, P., & Kolloosche, D. (2016). Social and Political Dimensions of Mathematics Educa- tion-Current Thinking. ICME13, Springer Open. Lutovac, S., & Kaasila, R. (2018c). Methodological landscape in research on teacher identity in mathematics education: A review. ZDM Mathematics Education. https://doi.org/10.1007/s1185 8-018-1009-2. Juzwik, M. (2006). Situating narrative-minded research: A commen- tary on Anna Sfard and Anna Prusak’s “Telling Identities”. Edu- cational Researcher, 35(9), 13–21. Martin, D. (2000). Mathematics success and failure among African- American youth. Mahwah: Lawrence Erlbaum Associates. Kaspersen, E., Pepin, B., & Sikko, S. A. (2017). Measuring STEM stu- dents’ mathematical identities. Educational Studies in Mathemat- ics, 95(2), 163–179. https://doi.org/10.1007/s10649-016-9742-3. Martin, D. (2012). Learning mathematics while black. Educational Foundations, 26, 47–66. Kemmis, S., Wilkinson, J., Edwards-Groves, C., Hardy, I., & Grooten- boer, P., Bristol, L. (2014). Changing practices, changing educa- tion. Singapore: Springer. McGee, E. O. (2015). Robust and fragile mathematical identities: A framework for exploring racialized experiences and high achieve- ment among black college students. Journal for Research in Math- ematics Education, 46(5), 599. https://doi.org/10.5951/jresemathe duc.46.5.0599. Knapp, M. C. (2017). An autoethnography of a (reluctant) teacher leader. Journal of Mathematical Behavior, 46, 251–266. https:// doi.org/10.1016/j.jmathb.2017.02.004. Mead, G. J. (1934). Mind, self and society. Chicago: University of Chicago Press.f g j j Lambert, R. (2015). Constructing and resisting disability in mathemat- ics classrooms: A case study exploring the impact of different pedagogies. Educational Studies in Mathematics, 89(1), 1–18. https://doi.org/10.1007/s10649-014-9587-6. Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The PRISMA Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med, 6(7), e1000097. https://doi.org/10.1371/journal.pmed1000097. p g Langer-Osuna, J. (2015). Linear equations and rap battles: How stu- dents in a wired classroom utilized the computer as a resource to coordinate personal and mathematical positional identities in hybrid spaces. Mathematics Education Research Journal, 27(1), 51–64. https://doi.org/10.1007/s13394-014-0125-8. Nasir, N. S., & Cooks, J. (2009). Becoming a hurdler: How learning settings afford identities. Anthropology & Education Quarterly, 40(1), 41–61. Ntow, F. D., & Adler, J. (2019). Identity resources and mathematics teaching identity: An exploratory study. ZDM Mathematics Edu- cation. https://doi.org/10.1007/s11858-019-01025-z. Langer-Osuna, J. M. (2018). Exploring the central role of student authority relations in collaborative mathematics. ZDM Mathemat- ics Education. https://doi.org/10.1007/s11858-018-0965-x. Oppland-Cordell, S., & Martin, D. B. (2015). Identity, power, and shifting participation in a mathematics workshop: Latin@ stu- dents’ negotiation of self and success. Mathematics Education Research Journal, 27(1), 21–49. References Connecting the dots: Toward a networked frame- work to conceptualizing identity in mathematics education. ZDM Mathematics Education (in press). Foyn, T., Solomon, Y., & Braathe, H. J. (2018). Clever girls’ stories: The girl they call a nerd. Educational Studies in Mathematics, 98(1), 77–93. https://doi.org/10.1007/s10649-017-9801-4. Heyd-Metzuyanim, E., Tabach, M., & Nachlieli, T. (2016b). Oppor- tunities for learning given to prospective mathematics teachers: Between ritual and explorative instruction. Journal of Mathemat- ics Teacher Education, 19(6), 547–574. https://doi.org/10.1007/ s10857-015-9311-1. Gee, J. P. (2000). Identity as an analytic lens for research in education. Review of Research in Education, 25, 99–125. Gholson, M., & Martin, D. B. (2019). Blackgirl face: Racialized and gendered performativity in mathematical contexts. ZDM Math- ematics Education (in press). Hodges, T. E., & Hodge, L. L. (2017). Unpacking personal identi- ties for teaching mathematics within the context of prospective teacher education. Journal of Mathematics Teacher Education, 20(2), 101–118. https://doi.org/10.1007/s10857-015-9339-2. Goldstein, D. S. (2018). Mathematics identities of competence in a middle-grades d/Deaf and hard-of-hearing classroom. Investi- gations in Mathematics Learning, 10(3), 145–158. https://doi. org/10.1080/19477503.2018.1467081. Holland, D., Lachicotte, W., Skinner, D., & Cain, C. (1998). Identity and agency in cultural worlds. Cambridge: Harvard University Press. Goos, M., & Bennison, A. (2018). A zone theory approach to analysing identity formation in mathematics education. ZDM Mathematics Education. https://doi.org/10.1007/s11858-018-1011-8. Ivanič, R. (1998). Writing and Identity: The discoursal construction of identity in academic writing. Amsterdam: John Benjamins. 1 3 3 M. Graven, E. Heyd‑Metzuyanim 376 Jooganah, K., & Williams, J. S. (2016). Contradictions between and within school and university activity systems helping to explain students’ difficulty with advanced mathematics. Teaching Mathematics and Its Applications, 35(3), 159–171. https://doi. org/10.1093/teamat/hrw014. Lutovac, S., & Kaasila, R. (2017). Future Directions in Research on Mathematics-Related Teacher Identity. International Journal of Science and Mathematics Education. https://doi.org/10.1007/ s10763-017-9796-4. Lutovac, S., & Kaasila, R. (2018a). Future directions in research on mathematics-related teacher identity. International Journal of Science and Mathematics Education, 16(4), 759–776. https://doi. org/10.1007/s10763-017-9796-4. Jorgensen, R. (2014). Social Theories of Learning: A need for a new paradigm in mathematics education. In J. Anderson, M. Cavanagh, & A. Prescott (Eds.), Curriculum in focus: Research guided prac- tice (Proceedings of the 37th annual conference of the Mathemat- ics Education Research Group of Australasia) (pp. 311–318). Sydney: MERGA. Lutovac, S., & Kaasila, R. (2018b). An elementary teacher’s narrative identity work at two points in time two decades apart. Educational Studies in Mathematics, 98(3), 253–267. https://doi.org/10.1007/ s10649-018-9816-5. References https://doi.org/10.1007/s1339 4-014-0127-6. Langer-Osuna, J. M., & Esmonde, I. (2017). Insights and advances on research on identity in mathematics education. In J. Cai (Ed.), Compendium for research in mathematics education (pp. 637– 648). Reston, VA: National Council of Teachers of Mathematics. Langer-Osuna, J. M. (2017). Authority, identity, and collaborative mathematics. Journal for Research in Mathematics Education, 48(3), 237–247. Pickard-Smith, K. (2018). Disordering mathematical identity stories through dramatic filmed parody “Math Therapy”. The Mathemat- ics Enthusiast, 15(1), 251–277. Larnell, G. V. (2016). More than just skill: Examining mathematics identities, racialized narratives, and remediation among black undergraduates. Journal for Research in Mathematics Education, 47(3), 233–269. https://doi.org/10.5951/jresematheduc.47.3.0233. van Putten, S., Stols, G., & Howie, S. (2014). Do prospective math- ematics teachers teach who they say they are? Journal of Math- ematics Teacher Education, 17(4), 369–392. Radford, L. (2008). Connecting theories in mathematics education: challenges and possibilities. ZDM Mathematics Education 40(2), 317–327. Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. New York: Cambridge University Press. Radovic, D., Black, L., Salas, C. E., & Williams, J. S. (2017). Being a girl mathematician: Diversity of positive mathematical identities in a secondary classroom. Journal for Research in Mathemat- ics Education, 48(4), 434. https://doi.org/10.5951/jresemathe duc.48.4.0434. Lerman, S. (2000). The social turn in mathematics education. In J. Boaler (Ed.), Multiple perspectives on mathematics education (pp. 19–44). Westport, CT: Ablex. Losano, L., Fiorentini, D., & Villarreal, M. E. (2017). The development of a mathematics teacher’s professional identity during her first year teaching. Journal of Mathematics Teacher Education. https ://doi.org/10.1007/s10857-017-9364-4. Radovic, D., Black, L., Williams, J., & Salas, C. E. (2018). Towards conceptual coherence in the research on mathematics learner iden- tity: a systematic review of the literature. Educational Studies in Mathematics. https://doi.org/10.1007/s10649-018-9819-2. Losano, L., Fiorentini, D., & Villarreal, M. (2018). The development of a mathematics teacher’s professional identity during her first year teaching. Journal of Mathematics Teacher Education, 21(3), 287–315. Ricoeur, P. (1991). Narrative identity. Philosophy Today, 35(1), 73–81. https://doi.org/10.1007/s10857-015-9329-4. le Roux, K., & Adler, J. (2016). A critical discourse analysis of prac- tical problems in a foundation mathematics course at a South African university. Educational Studies in Mathematics, 91(2), 227–246. https://doi.org/10.1007/s10649-015-9656-5. Lutovac, S., & Kaasila, R. (2014). Pre-service teachers’ future-oriented mathematical identity work. Educational Studies in Mathematics, 85(1), 129–142. https://doi.org/10.1007/s10649-013-9500-8. 1 3 Mathematics identity research: the state of the art and future directions 377 Sfard, A. (2006). Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Telling ideas by the company they keep: A response to the critique by Mary Juzwik. Educational Researcher, 35(9), 22–27. Educational Studies in Mathematics, 86(2), 271–290. https://doi. org/10.1007/s10649-014-9540-8. Educational Studies in Mathematics, 86(2), 271–290. https://doi. org/10.1007/s10649-014-9540-8. Wenger, E. (1998). Communities of practice. Cambridge: Cambridge University Press.l Sfard, A. (2008). Thinking as communicating. New York: Cambridge University Press. Westaway, L. (2019). The role of reflexivity in the emergence and expression of teachers’ identities in teaching primary school math- ematics. ZDM Mathematics Education. https://doi.org/10.1007/ s11858-019-01042-y. y Sfard, A., & Prusak, A. (2005). Telling identities: In search of an ana- lytic tool for investigating learning as a culturally shaped activity. Educational Researcher, 34(4), 14–22. Skog, K., & Andersson, A. (2015). Exploring positioning as an analyti- cal tool for understanding becoming mathematics teachers’ iden- tities. Mathematics Education Research Journal, 27(1), 65–82. https://doi.org/10.1007/s13394-014-0124-9. Westaway, L., & Graven, M. (2019). Exploring grade 3 teachers’ resist- ance to ‘take up’ progressive mathematics teaching roles. Math- ematics Education Research Journal., 31(1), 27–46. https://doi. org/10.1007/s13394-018-0237-7. Skott, J. (2018). Changing experiences of being, becoming, and belong- ing: Teachers’ professional identity revisited. ZDM Mathematics Education. https://doi.org/10.1007/s11858-018-1008-3. Williams, S. R., & Leatham, K. R. (2017). Journal quality in mathemat- ics education. Journal for Research in Mathematics Education, 48(4), 369–396. Wood, M. B. (2013). Mathematical micro-identities: Moment-to- moment positioning and learning in a fourth-grade classroom. Journal for Research in Mathematics Education, 44(5), 775–808. https://doi.org/10.5951/jresematheduc.44.5.0775. Solomon, Y., Radovic, D., & Black, L. (2016). “I can actually be very feminine here”: Contradiction and hybridity in becoming a female mathematician. Educational Studies in Mathematics, 91(1), 55–71. https://doi.org/10.1007/s10649-015-9649-4. Takeuchi, M. A. (2018). Power and identity in immigrant parents’ involvement in early years mathematics learning. Educational Studies in Mathematics, 97(1), 39–53. https://doi.org/10.1007/ s10649-017-9781-4. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Wake, G. (2014). Making sense of and with mathematics: The inter- face between academic mathematics and mathematics in practice. Wake, G. (2014). Making sense of and with mathematics: The inter- face between academic mathematics and mathematics in practice. 1 3 1 3 3
https://openalex.org/W1769784862	https://journals.linguisticsociety.org/proceedings/index.php/SALT/article/download/2561/2309	English	null	Modals are monsters: on indexical shift in English	Proceedings from Semantics and Linguistic Theory	2,010	cc-by	9,665	Proceedings of SALT 20: 289–308, 2010 Proceedings of SALT 20: 289–308, 2010 Keywords: indexicality, monsters, modality, attitude reports, Kaplan, de re Keywords: indexicality, monsters, modality, attitude reports, Kaplan, de re 1 Introduction On Kaplan’s (1989) classical theory of indexicality, the semantic values of indexicals like I, you and now are identiﬁed with their referents.1 This theory leads to well- known difﬁculties when indexicals occur in attitude and speech reports. Spotting a man whose hair is enveloped in ﬂames, I can warn him by saying (1) I believe that you’re on ﬁre. (1) I believe that you’re on ﬁre. Moreover, unless I also believe that my own hair is enveloped in ﬂames, I can add ∗For comments, questions, mentorship, and encouragement thanks to Fabrizio Cariani, Jennifer Carr, Irene Heim, Dilip Ninan, Orin Percus, Alejandro Pérez Carballo, Agustín Rayo, Philippe Schlenker, Yael Sharvit, Brad Skow, Bob Stalnaker, Andreas Stokke and to audiences at MIT, SPE 2009, the AAP 2009 conference in Melbourne, SALT 20 and the Arché Semantics Group in St. Andrews. p 1 Here and throughout the paper I use ‘semantic value’ to denote the kind of object that a lexical item contributes to the truth-conditional computation. Yael Sharvit, Brad Skow, Bob Stalnaker, Andreas Stokke and to audiences at MIT, SPE 2009, th AAP 2009 conference in Melbourne, SALT 20 and the Arché Semantics Group in St. Andrews. Modals are monsters: on indexical shift in English∗ Paolo Santorio Massachusetts Institute of Technology Abstract I argue, contra Kaplan, that all English indexicals are shiftable. Starting from a new puzzle involving epistemic conditionals, I propose a semantics that accounts for embeddings of indexicals under non-metaphysical modals like might or believe. The central idea is that indexicals are variables: epistemic and doxastic modals are able to shift them by shifting the value of the assignment function. So all epistemic and doxastic modals turn out to be Kaplanian monsters. I close by arguing that the monstrous account has empirical advantages over the classical semantics for attitudes in the Kaplan tradition, and in particular over the recent version of it proposed by Percus and Sauerland. (2) I believe that I am not on ﬁre. (2) I believe that I am not on ﬁre. (2) I believe that I am not on ﬁre. (1) and (2) both have a true reading even if, unbeknownst to me, I’m looking into a mirror and I am actually the man whose hair is on ﬁre. This is puzzling: if you and I in (1) and (2) denote the same individual, on Kaplan’s theory they have the (1) and (2) both have a true reading even if, unbeknownst to me, I’m looking into a mirror and I am actually the man whose hair is on ﬁre. This is puzzling: if you and I in (1) and (2) denote the same individual, on Kaplan’s theory they have the ©2010 Paolo Santorio Paolo Santorio same semantic value. Then (1) and (2) should be heard as contradictory, or at least as ascribing to me contradictory beliefs. Yet they are perfectly good reports of my (consistent) attitudes. Problems of this sort are familiar since Frege and have given rise to an extensive literature in philosophy and semantics.2 One common assumption in this literature is that the interpretation of indexicals can’t be affected by embeddings under believe. This is a straightforward consequence of Kaplan’s theory. For Kaplan only context can ﬁx the semantic values of indexicals: operators that shift indexicals, which Kaplan dubs ‘monsters’, are no part of the semantics of natural language (at least, of natural languages like English). Interestingly, recent literature has shown that the ban on monsters fails cross-linguistically.3 But this doesn’t help with the puzzle in (1) and (2): in fact it remains virtually unquestioned that common English indexicals like I, you or now are unshiftable, just as Kaplan had it. So you and I in (1) and (2) pick out, as is customary, the addressee and the speaker, and our semantics for attitude reports must work around this basic fact. Or so it is thought. In this paper I argue that in fact there is a monstrous treatment of indexicals which has empirical and conceptual advantages over traditional stories. On this account all English indexicals work as variables; modal quantiﬁers with epistemic ﬂavor, such as must, might and believe, shift the assignment function and thereby shift indexicals. So for example in a sentence like I might be lost in Main Library, Stanford. (3) I is a variable and its semantic value is ﬁxed by an assignment determined by might. 3 More precisely, it has been argued that speech report verbs in some languages (Amharic, Zazaki, Slave and perhaps, in a restricted fashion, even English) are able to produce shifted readings of indexicals. For monstrous semantic accounts of all these languages, cf. Schlenker 1999, 2003; for the Zazaki, Anand & Nevins 2004 and Anand 2006. An alternative account of the data is attempted by von Stechow 2002 and von Stechow 2003. For useful summaries of the debate, cf. Schlenker 2004, 2008. 2 In both cases the starting point for a formal treatment of the problem is Kaplan 1968; cf. also sectio 20 of Kaplan 1989. 4 I should ﬂag that this assumption is rather simplistic, and in fact is rejected by virtually everyone in recent debates about epistemic modality. This is not important for my purposes, as the puzzle can be reproduced on more sophisticated assumptions. (2) I believe that I am not on ﬁre. As a result, I is associated with a range of epistemic counterparts of the speaker, i.e. possible individuals who, for all the speaker knows, the speaker herself might be. This theory should be preferred over its competitors, which involve heavier stipulations or lead to overgeneration in a number of cases. If the foregoing is correct, it is a partial argument that the semantics for attitude and speech reports is monstrous, and that—pace Kaplan—English is replete with monsters. I proceed as follows. In section 2, I introduce new data showing that the Frege- style phenomenon generated by (1) and (2) can be reproduced with epistemic modals. The new data is interesting because, much more than its counterpart involving believe, it naturally points to a monstrous account. In section 3, I proceed to give a 2 In both cases the starting point for a formal treatment of the problem is Kaplan 1968; cf. also section 20 of Kaplan 1989. 290 Modals are monsters monstrous theory which accommodates the data. Section 4 contains the argument that this theory is superior to the traditional Kaplan-style account of attitude reports (and in particular to one of the best and latest versions of it, that put forward by Percus & Sauerland (2003)). In the conclusion, I point to directions for future research. 2.1 Data On the assumption that I invariably denotes the actual speaker, (5) gets truth- conditions: J(5)Kc,g = true iff for all w′ in Lingens’s epistemic state in wc such that the coin landed heads in w′, Lingens is in Widener Library in w′. J(5)Kc,g = true iff for all w′ in Lingens’s epistemic state in wc such But in all heads-worlds in Lingens’s epistemic state Lingens is not in Widener Library—rather, he is dead. Hence (5) is wrongly predicted to be false. Conversely, if Lauben is the speaker we get a wrong prediction for (4). But in all heads-worlds in Lingens’s epistemic state Lingens is not in Widener Library—rather, he is dead. Hence (5) is wrongly predicted to be false. Conversely, if Lauben is the speaker we get a wrong prediction for (4). In short, then, the difﬁculty is this: whatever the context of utterance is (i.e. whether we take Lingens or Lauben as the speaker), one of (4) and (5) is predicted to be false. But both conditionals are true according to the data. 5 By ‘metaphysical’ here I mean to pick out simply whatever modality will turn out to do work in typical counterfactual conditionals. 2.1 Data My starting observation is this: indexicals embedded under epistemic modalities can have prima facie ‘shifty’ readings. Consider the following scenario: Rudolf Lingens and Gustav Lauben are kidnapped. Lingens and Lauben are amnesiacs: each of them knows that he is one of the two kidnapped amnesiacs, but not which one. They will be subjected to the following experiment. First, they will be anesthetized. Then a coin will be tossed. If the outcome is tails, Lingens will be released in Main Library, Stanford, and Lauben will be killed. If the outcome is heads, Lauben will be released in Widener Library, Harvard, and Lingens will be killed. Lingens and Lauben are informed of the plan and the experiment is executed. Later, one of them wakes up in a library. He says: (4) If the coin landed tails, I am in Main Library, Stanford. (5) If the coin landed heads, I am in Widener Library, Harvar (4) (5) Here are the data: (4) and (5) are both true (or at least appropriate) as uttered in the scenario. The problem is that standard accounts of conditionals miss this prediction. Here are the data: (4) and (5) are both true (or at least appropriate) as uttered in the scenario. The problem is that standard accounts of conditionals miss this prediction. For illustration, take a basic version of Kratzer’s (1981) semantics for condition- als. Conditionals are, as usual, modal statements where the if-clause plays the role of a restrictor. The domain of quantiﬁcation (modal base) is speciﬁed by a covert variable the value of which is provided by context: for simplicity, here I assume that the modal base defaults to the epistemic state of the speaker, i.e. the set of worlds compatible with what the speaker knows.4 So the truth-conditions predicted for if p, q are, schematically: Jmust [if p] [q]Kc,g = true iff for all w′ compatible with what the speaker knows in wc and such that w′ ∈JpKc,g, w′ ∈JqKc,g. 4 I should ﬂag that this assumption is rather simplistic, and in fact is rejected by virtually everyone in recent debates about epistemic modality. This is not important for my purposes, as the puzzle can be reproduced on more sophisticated assumptions. 291 Paolo Santorio Paolo Santorio Paolo Santorio Now, consider again the amnesiac scenario and suppose that Lingens is the speaker. 2.2 Generalizing the problem (11) If the coin had landed heads, I would have been in Widener Library, Harvard A joint utterance of (10) and (11) is infelicitous, and in any case not on a par with a joint utterance of (4) and (5). (One could reply “No, that’s wrong—it wouldn’t have been you in both cases.”) On the other hand, the data can be easily replicated with other non-metaphysical modals. Following Hintikka (1962), I assume that attitude verbs also work as quantiﬁers over possibilities. For example, believe is a universal quantiﬁer over worlds which model an agent’s doxastic alternatives—worlds that, for all the agent believes, might be the actual world. Schematically: Jx believes [that p]Kc,g = true iff for all w′ compatible with what x believes, w′ ∈JpKc,g Jx believes [that p]Kc,g = true iff for all w′ compatible with what x believes, w′ ∈JpKc,g Now, suppose that, rather than (4) and (5), the amnesiac in the library scenario say Now, suppose that, rather than (4) and (5), the amnesiac in the library scenario says: (12) I believe that either the coin landed tails and I am in Main Library, Stanford, or the coin landed heads and I am in Widener Library. (12) I believe that either the coin landed tails and I am in Main Library, Stanford, or the coin landed heads and I am in Widener Library. (12) presents us with basically the same problem as the (4)–(5) pair. There is no individual that is going to be in Main in case of a tails outcome and in Widener in case of a heads outcome. So, on a simple Hintikka-style semantics for believe, (12) is predicted to be false no matter what the context is. For example, if the speaker is Lingens, its truth-conditions are J(12)Kc,g = true iff for all w′ compatible with what Lingens believes in wc, either the coin landed tails and Lingens is in Main, or the coin landed heads and Lingens is in Widener in w′. In the remainder of the paper I use might, must, and believe as my running examples. But the phenomenon can be reproduced at least for a large subclass of attitude verbs: some examples are know, suppose, imagine, conjecture, dream. So the right generalization seems to be that the phenomenon concerns the behavior of indexicals in the scope of epistemic and doxastic modals in general. 2.2 Generalizing the problem The puzzle generalizes along two dimensions: it arises both for other indexicals and for other modals. Let me look at other indexicals ﬁrst. Here are examples involving now and that: You fall asleep at around noon, and you wake up hearing the clock in your room chiming. You know it chimed once or twice, but you’re not sure which. You say: You fall asleep at around noon, and you wake up hearing the clock in your room chiming. You know it chimed once or twice, but you’re not sure which. You say: (6) If the clock chimed once, it is now one. (7) If the clock chimed twice, it is now two. (6) If the clock chimed once, it is now one. (7) If the clock chimed twice, it is now two. A group of climbers has gotten lost in the mountains between France and Italy. They are ﬁnally descending on one of the two sides, though they don’t know which. One points to a city in the distance and says: (8) If we are on the French side, that [pointing] is Lyon. (9) If we are on the Italian side, that [pointing] is Turin. (8) If we are on the French side, that [pointing] is Lyon. (9) If we are on the Italian side, that [pointing] is Turin. The problem is analogous. All of (6)–(9) are true, but on a standard Kratzer semantics one conditional in each pair turns out to be false, no matter what the actual context is. So far as I can see, analogous puzzles can be constructed for all English indexicals. The problem is analogous. All of (6)–(9) are true, but on a standard Kratzer semantics one conditional in each pair turns out to be false, no matter what the actual context is. So far as I can see, analogous puzzles can be constructed for all English indexicals. What about other modals? Interestingly, results differ depending on the kind of modality involved. Conditionals invoking metaphysical modality,5 such as counter- factuals, behave exactly as predicted by a standard Kratzer semantics. Suppose that 292 Modals are monsters the amnesiac in the original scenario is recapitulating to himself his predicament. Trying to describe the causal connections between possible outcomes of the coin toss and amnesiacs waking up in libraries, he says (10) If the coin had landed tails, I would have been in Main Library, Stanford. 2.2 Generalizing the problem In the remainder of the paper I use might, must, and believe as my running examples. But the phenomenon can be reproduced at least for a large subclass of attitude verbs: some examples are know, suppose, imagine, conjecture, dream. So the right generalization seems to be that the phenomenon concerns the behavior of indexicals in the scope of epistemic and doxastic modals in general. Let me also mention two directions of generalization that I won’t explore for reasons of space: on the one hand tense, on the other referential expressions other than indexicals. I will come back to the latter brieﬂy in the conclusion. 293 Paolo Santorio Paolo Santorio 2.3 Preview The phenomenon spotted in this section clearly belongs to a larger family of Frege- style issues concerning the interaction between indexicals and non-metaphysical modals. But it deserves speciﬁc attention because, much more than other puzzles in the vicinity, it suggests a natural solution. Consider again (4) and (5), repeated below: (13) If the coin landed tails, I am in Main Library, Stanford. (14) If the coin landed heads, I am in Widener Library, Harvard. It seems intuitive that the two conditionals work in this way: the if-clause selects a subset of the worlds among those in the modal base, and the I occurring in the consequent refers to the individual who is speaking in those worlds. It seems intuitive that the two conditionals work in this way: the if-clause selects a subset of the worlds among those in the modal base, and the I occurring in the consequent refers to the individual who is speaking in those worlds. In the next section I propose an account that endorses and implements this intuition: I in (4) and (5) works as as a variable that is bound by epistemic must. Of course, the main challenge is to allow for this kind of monstrous behavior while avoiding overgeneration. In particular, any plausible account should recapture the standard data about unembedded indexicals, counterfactuals, etc., that motivated Kaplan’s original account. This can be done, as I argue below. Why can’t I myself be my own representative in each of Lingens’s belief states? The problem is Why can’t I myself be my own representative in each of Lingens’s belief states? The problem is 3.1 Theoretical background: counterparts and information states Start from a general question: how should we evaluate indexicals embedded under non-metaphysical modals like might and believe? This question is the linguistic counterpart of a traditional question in philosophy of mind. Epistemic and doxastic modals, recall, quantify over worlds modeling information states of a subject: namely, what a subject knows, believes, etc. We standardly talk of these information states as representing speciﬁc individuals: for example, we can say that Lingens believes that I am lost in a university library. But how am I individuated in Lingens’s information state? I.e., what makes Lingens’s information state an information state about me, a real individual in the actual world? Call this the problem of individuation. The standard solution to the problem (coming from Hintikka (1969) and Lewis (1983)) is this: I have a representative in each of Lingens’s belief worlds; Lingens counts as knowing that I am lost in a university library in virtue of the fact that each of my representatives in his knowledge worlds is lost in a university library.6 What individuals are my representatives? In informal terms, individuals that resemble me 294 Modals are monsters in certain relevant respects. In more formal terms: individuals that are counterparts of me, where a relation of counterparthood is just a relation of similarity (cf. Lewis 1968, 1983): A is a counterpart of B under a certain respect just in case A and B are sufﬁciently similar in that respect. I suggest that the problem of evaluating indexicals under might or believe is essentially a linguistic version of the problem of individuation. So it can be solved with the same tools. Objects are individuated in an information state by individuating a set of their representatives, one for each world in the information state. Similarly, and in a slogan: indexicals embedded under might, believe and the like denote their referents by denoting a set of their representatives. Out of metaphor, this means that they work as bound variables and range over counterparts of the referents they pick out in ordinary linguistic environments. So e.g. in (15) Lingens believes that I am lost in a university library. I is bound and ranges over counterparts of the actual speaker. Similarly, bound occurrences of you will range over counterparts of the actual addressee; bound occurrences of that (like those occurring in (8) and (9)) will range over counterparts of the (n-th) object demonstrated; and so on. that real-world identity is ill-suited to model the ways agents are represented throughout information states. As Quine (1956) pointed out, I might enter Lingens’s beliefs under two guises: for example as that man in front of him, next to the stack of semantics books, and that man reﬂected in the mirror over there, who (for all Lingens can see) might or might not be next to the stack of semantics books. If we invariably used myself as my own representative throughout Lingens’s doxastic state we should say that Lingens has inconsistent beliefs, since in some alternatives I would both be and not be next to the stack of semantics books. But Lingens’s beliefs are not contradictory—he’s just ignorant about certain matters of fact. (For extended discussion of this problem, see Lewis 1983 and Ninan 2008) 7 For more details about acquaintance see, among others, Lewis 1979. I should ﬂag that Kaplan’s p 8 Typical examples are accounts derived from Kaplan 1968: cf. among others Cresswell & von Stechow 1982 and Ninan 2008. Interestingly, Cumming 2008 has attitude verbs quantify over pairs of assignments and worlds and bind names. There are obvious similarities between Cumming’s semantics and mine, as well as important differences. (For example the system I suggest, unlike his, enforces obligatory shift of the assignment.) idea of using acquaintance for the semantics of de re attitudes is subject to a number of objections: see e.g. Ninan 2008, especially chapters 2 and 3. I am aware of this, and suspect that ultimately we should switch to a better notion. But the issue of what counterpart relation is in place, although important, seems orthogonal to my main point here. So it seems easier to go with Kaplan on this point. 3.1 Theoretical background: counterparts and information states I have talked about similarity and counterparts, but so far I haven’t speciﬁed what notion of similarity we should use. For present purposes, it will be enough to import an idea from Kaplan’s semantics for de re reports. I assume that a subject bears a relation of acquaintance to the individuals he has attitudes about. Acquaintance is supposed to be a kind of direct epistemic relation to an object, of which perception is the paradigm example. Kaplan’s central idea is that acquaintance relations ﬁgure in the content of a subject’s beliefs about objects. For example Lingens, having seen me wander around the library, thinks of me as the man I saw wandering around the library. So in each of his belief worlds there is a man he has seen wonder around the library: just these men, who are my counterparts by acquaintance, are my representatives in Lingens’s belief state. More generally, bound occurrences of I, you, etc. range over counterparts by acquaintance of the actual speaker, addressee, etc.7 295 Paolo Santorio Paolo Santorio idea of using acquaintance for the semantics of de re attitudes is subject to a number of objections: see e.g. Ninan 2008, especially chapters 2 and 3. I am aware of this, and suspect that ultimately we should switch to a better notion. But the issue of what counterpart relation is in place, although important, seems orthogonal to my main point here. So it seems easier to go with Kaplan on this point. 3.2 Semantics: overview The semantics is based on three main innovations on classical theories. First, indexicals are variables and their anchoring to the context is determined by extra- semantic constraints on the choice of the assignment. Second, modals like might and believe quantify not simply over worlds, but rather over pairs of an assignment and a world. Third, these modals shift the value of the assignment and as a result shift indexicals. Let me present each of these points informally before turning to the formal system. First, indexicals are turned into simple variables. By ‘simple’ I mean that their lexical entries contain no reference to the context, not even in a presupposition. In a slogan, I is semantically analogous to he (at least with respect to its person features). The only difference is that indexicals are given referents not via the ordinary assignment, but by a second dedicated assignment, which I denote with ‘a’. The connection between indexicals and context is not part of the semantics proper, but takes place at the level of metasemantics, i.e. at the stage at which the dedicated assignment is selected. So nothing in the semantics of I tells us that it must refer to the speaker, but it is a constraint on the choice of a that I be invariably sent to the speaker. Second, modals are turned into more powerful quantiﬁers. might, believe and the like quantify not simply over worlds, but rather over pairs of assignments (i.e. functions from indices to sequences of individuals), and centered worlds. This move is not new. It is a natural generalization of our standard semantics for de se reports and variants of it appear frequently in the literature on de re reports.8 So far, then, my account of modals looks similar to other proposals in the literature. The innovation consists in allowing the modal to shift indexicals by shifting the assignment a. In the new system might, believe and the like shift the value of the assignment a, forcing the evaluation of the embedded clause at the alternative assignments they quantify over. Schematically: Jx believes that [φ]Kc,g,a = for all ⟨a′, w′⟩compatible with x’s beliefs ′ Jx believes that [φ]Kc,g,a = for all ⟨a′, w′⟩compatible with x’s beliefs ′ J φ K at wc, JφKc,g,a′(w′) 296 Modals are monsters This requires that the clausal arguments of modals be not propositions, but rather functions of higher type, i.e. (16) Lingens believes that I am lost in a university library. (16) Lingens believes that I am lost in a university library. the clausal argument of believe is the clausal argument of believe is λa′.JI am lost in a university libraryKc,g,a′ λa′.JI am lost in a university libraryKc,g,a′ It should be clear how, on the semantics just sketched, might and believe are able to affect the semantic values of indexicals. The idea is simple: they shift indexicals by shifting the assignment on which their semantic values depend. Notice an interesting formal analogy between assignment-shifting modals and intensional operators in intensional systems. Both manipulate a metalanguage parameter: an assignment in the former case, and the world or time parameter in the latter. To visualize the point, consider a schematic intensional semantics for must: Jmust [φ]Kc,w,g = for all accessible w′, JφKc,w′,g One feature of the shift produced in this way is that it is obligatory. The embedded clause must be evaluated at the shifted parameter (at least if no other operators are present). This point will be important for the comparison between the monstrous theory and more classical theories. So much for an overview. Now let me introduce the system. 3.2 Semantics: overview functions from (again) an assignment and a centered world to a truth-value. We get these arguments by introducing a binder on the assignment. For example, in (15) (repeated below) (16) Lingens believes that I am lost in a university library. 9 The semantics for indexicals is indebted to an idea found in Schlenker 2008, esp. pages 25–6. Schlenker attributes the crucial suggestion to Ede Zimmermann. 3.3 Semantics: formal details In line with recent literature (see, for example, Schlenker 2003), I assume a fully extensional system, which involves object language quantiﬁcation over worlds. The basic denotations of clauses are centered propositions, namely functions from worlds and individuals to truth-values. For example: JRalph loves OrtcuttKc,g,a = λw. λx. Ralph loves Ortcutt in w JRalph loves OrtcuttKc,g,a = λw. λx. Ralph loves Ortcutt in w JRalph loves OrtcuttKc,g,a = λw. λx. Ralph loves Ortcutt in w For shorthand, I’m going to use ‘i’ for a variable over centered worlds as well as for the semantic type of centered worlds. 297 Paolo Santorio Paolo Santorio 3.3.2 Composition rules Ordinary clauses denote centered propositions; at the same time, modals require a clausal argument of higher type. So there is systematic type mismatch between modals and their clausal arguments. This mismatch is ﬁxed via a new composition rule which tells us to perform lambda-abstraction on the assignment a: 3.3.1 Indexicals Indexicals are variables with a special diacritic [+c∗].9 Their semantics is analogous to the semantics of ordinary variables of the corresponding types, with the only difference that they are handled by the new assignment a. For example: JIKc,g,a = Jx1[+c∗]Kc,g,a = a(1[+c∗]) JyouKc,g,a = Jx5[+c∗]Kc,g,a = a(5[+c∗]) This makes indexicals bindable, but poses the problem of recapturing the connection between their free occurrences and the context of utterance. How to do this? This makes indexicals bindable, but poses the problem of recapturing the connection between their free occurrences and the context of utterance. How to do this? I suggest that this connection is independent of the compositional part of the semantics. Rather, it is captured at the level that Kaplan calls ‘metasemantics’, i.e. at the level at which the context of utterance determines an assignment. For present purposes, I simply assume the existence of a constraint on the selection of the special assignment a. This constraint rules out as inadmissible assignments which send indexicals into the wrong referents. I.e., An assignment a is admissible in c just in case a(1[+c∗]) = the speaker of c, a(5[+c∗]) = the addressee of c, etc. Admittedly this account is somewhat stipulative: it would be preferable to provide some independent evidence for the metasemantic story. I am not in the position to do this here. On the upside, this is the only stipulative element I need. Monstrous Functional Application (MFA) If α is a branching node and {β,γ} the set of its daughters, then for any context c and assignment g, if JβKc,g,a is a function whose domain contains λa′. JγKc,g,a′, then JαKc,g,a = JβKc,g,a(λa′. JγKc,g,a′). Notice that this is a second parallel between this system and intensional systems. In- tensional systems normally contain an extra composition rule, Intensional Functional 298 Modals are monsters Application (see Heim & Kratzer 1998), which is also triggered by type mismatch between operators and their clausal argument. 3.3.3 Modals My starting point is the semantics for epistemic and doxastic modals that is common currency in the literature on de se reports (see, among others, Lewis 1979, Chierchia 1989, Anand 2006). Modals are universal quantiﬁers over a contextually supplied set of centered worlds: JmustKc,g = λ p⟨i,t⟩. λM⟨i,⟨i,t⟩⟩. λw. ∀i′: i′ ∈((M)(i)), p(i′) = 1 Monstrous modals are different in three respects. I already mentioned two: • they quantify over pairs of assignments and centered worlds; • they quantify over pairs of assignments and centered worlds; • their clausal argument is, correspondingly, a function from assignment- centered world pairs to truth-values. • their clausal argument is, correspondingly, a function from assignment- centered world pairs to truth-values. To these I add a third: • epistemic and doxastic modals have an extra argument, which is ﬁlled by a variable ranging over sequences of counterpart functions. • epistemic and doxastic modals have an extra argument, which is ﬁlled by a variable ranging over sequences of counterpart functions. This extra argument serves the purpose of determining the assignment-centered world pairs that the modal quantiﬁes over. These counterpart functions (type ⟨es, e⟩) specify a way in which the subject of the epistemic state is acquainted with elements of the context. So f1 speciﬁes the way in which she’s acquainted with the speaker, f2 the way in which she’s acquainted with the addressee, fi the way in which she’s acquainted with the i-th demonstratum, etc. We use them to determine sequences of individuals; in turn, these sequences are used to determine an assignment. Here are the details. We start from a sequence ⟨f1, f2, ... ⟩. We plug each centered world in the modal base into all the functions in the sequence. This gives us sequences of possible individuals: ⟨f1(ik), f2(ik), ..., fn(ik)⟩ ⟨f1(il), f2(il), ..., fn(il)⟩ ⟨f1(ik), f2(ik), ..., fn(ik)⟩ ⟨f1(il), f2(il), ..., fn(il)⟩ Notice that the individuals in each sequence are the counterparts of speaker, ad- dressee, etc. in a determinate centered world. It is easy to recover an assignment function, once we have each of these sequences. We just take a function which maps the n-th index into the individual determined by the n-th counterpart function: 299 Paolo Santorio 1[+c∗] −→f1(ik) 2[+c∗] −→f2(ik) ... n[+c∗] −→fn(ik) 1[+c∗] −→f1(ik) 2[+c∗] −→f2(ik) ... n[+c∗] −→fn(ik) n[+c∗] −→fn(ik) Now, let ak be the assignment determined in this way by using centered world ik. 3.3.3 Modals The assignment-centered world pairs that a modal quantiﬁes over will be all the pairs ⟨ak,ik⟩, where ik is a world in the modal base. For shorthand, I will say that an assignment-centered world pair is compatible with a modal base M and a sequence of counterpart functions ⟨f1, f2, ... ⟩iff it can be generated from that modal base and that sequence in the way described. All this is encoded in the lexical entry of modals. In essence, modals do two things: • First, they ‘check’ that the counterpart functions provided by context are the right ones. I.e., they check that f1, if deﬁned,10 does in fact specify the way that the subject is acquainted with the speaker, f2, if deﬁned, does specify the way the subject is acquainted with the hearer, etc. • Second, they quantify over the pairs that are generated in the way described Here is the entry for epistemic must (using A as a type for assignments): • Second, they quantify over the pairs that are generated in the way described. Here is the entry for epistemic must (using A as a type for assignments): JmustKc,g,a = λχ⟨A,⟨i,t⟩⟩. λ⟨f1, f2,...⟩. λM⟨i,⟨i,t⟩⟩. λi. ( f1)(i) = (if def.) a(x1[+c∗]), ( f2)(i) = (if def.) a(x2[+c∗]) ... and ∀⟨a′,i′⟩compatible with M(i) and ⟨f1, f2,...⟩, χ(a′)(i′) = 1 JmustKc,g,a = λχ⟨A,⟨i,t⟩⟩. λ⟨f1, f2,...⟩. λM⟨i,⟨i,t⟩⟩. λi. ( f1)(i) = (if def.) a(x1[+c∗]), ( f2)(i) = (if def.) a(x2[+c∗]) ... and ∀⟨a′,i′⟩compatible with M(i) and ⟨f1, f2,...⟩, χ(a′)(i′) = 1 JmustKc,g,a = λχ⟨A,⟨i,t⟩⟩. λ⟨f1, f2,...⟩. λM⟨i,⟨i,t⟩⟩. λi. ( f1)(i) = (if def.) a(x1[+c∗]), ( f2)(i) = (if def.) a(x2[+c∗]) ... and ∀⟨a′,i′⟩compatible with M(i) and ⟨f1, f2,...⟩, χ(a′)(i′) = 1 And here is the entry for believe: And here is the entry for believe: And here is the entry for believe: JbelieveKc,g,a = λχ⟨A,⟨i,t⟩⟩. λ⟨f1, f2,...⟩. λM⟨i,⟨i,t⟩⟩. λx. λi. JbelieveKc,g,a = λχ⟨A,⟨i,t⟩⟩. λ⟨f1, f2,...⟩. λM⟨i,⟨i,t⟩⟩. λx. λi. ( f1)(⟨w,x⟩) = (if def.) a(x1[+c∗]), ( f2)(⟨w,x⟩) = (if def.) a(x2[+c∗] ( f1)(⟨w,x⟩) = (if def.) a(x1[+c∗]), ( f2)(⟨w,x⟩) = (if def.) a(x2[+c∗]) and ∀⟨a′,i′⟩compatible with what x believes and ⟨f1, f2,...⟩, χ( ′)(i′) 1 and ∀⟨a′,i′⟩compatible with what x believes and ⟨f1, f2,...⟩, χ(a′)(i′) = 1 ⟨ ⟩ χ(a′)(i′) = 1 10 Why the deﬁnedness condition? In most cases, the subject of the relevant information state won’t have a way of identifying all elements of the context of utterance. 10 Why the deﬁnedness condition? In most cases, the subject of the relevant information state won’t have a way of identifying all elements of the context of utterance. In this case I assume that the context will provide an undeﬁned function for that object. Having undeﬁned functions in the sequence doesn’t make a difference unless the function does actual work in ﬁxing the range of an indexical appearing in the clause. 3.3.3 Modals In this case I assume that the context will provide an undeﬁned function for that object. Having undeﬁned functions in the sequence doesn’t make a difference unless the function does actual work in ﬁxing the range of an indexical appearing in the clause. 300 Modals are monsters Modals are monsters Here is a sample derivation. Take (17) I must be in Main Library, Stanford. (18) λi0. must [λi1. x1[+c∗] be in Main Library, Stanford i1] (⟨f1, f2,...⟩) (M⟨i,⟨i,t⟩⟩) i0 Here is the computation (using ‘F’ as a variable over sequences of counterpart f ti ) (18) λi0. must [λi1. x1[+c∗] be in Main Library, Stanford i1] (⟨f1, f2,...⟩) (M⟨i,⟨i,t⟩⟩) i0 (M⟨i,⟨i,t⟩⟩) i0 Here is the computation (using ‘F’ as a variable over sequences of counterpart functions): Here is the computation (using ‘F’ as a variable over sequences of counterpa functions): Here is the computation (using ‘F’ as a variable over sequences of counterpart functions): Here is the computation (using ‘F’ as a variable over sequences of counterpart functions): Jλi0. must [λi1. x1[+c∗] be in Main Library, Stanford i1](F) (M) i0Kc,g,a = (by FA and abstraction rules) λi′. [Jmust [λi1. x1[+c∗] be in Main Library, Stanford i1]Kc,g[i0→i′],a (g(F)) (g(M)) i′] = (by MFA) λi′. [Jmust [λi1. x1[+c∗] be in Main Library, Stanford i1]Kc,g[i0→i′],a (g(F)) (g(M)) i′] = (by MFA) λi′. [JmustKc,g[i0→i′],a (λa′. J[λi1. x1[+c∗] be in Main Library, Stanford i1]Kc,g[i0→i′],a′) (g(F)) (g(M)) i′] = (computing the denotation of the clause) λi′. [JmustKc,g[i0→i′],a (λa′. J[λi1. x1[+c∗] be in Main Library, Stanford i1]Kc,g[i0→i′],a′) (g(F)) (g(M)) i′] = (computing the denotation of the clause) λi′. [JmustKc,g[i0→i′],a (λa′. [λi′′. [ a′(1[+c∗]) is in Main Library, Stanford in i′′]]) (g(F)) (g(M)) i′] = (by the entry for must) λi′. [JmustKc,g[i0→i′],a (λa′. [λi′′. [ a′(1[+c∗]) is in Main Library, Stanford in i′′]]) (g(F)) (g(M)) i′] = (by the entry for must) λi′. [[λχ. λ⟨f1, f2,...⟩. λM. λi. (f1)(i) = (if deﬁned) a(x1[+c∗]), (f2)(i) = (if deﬁned) a(x2[+c∗]) . . . and ∀⟨a∗,i∗⟩compatible with M(i) and ⟨f1, f2,...⟩, χ(a∗)(i∗) = 1] (λa′. [λi′′. [ a′(1[+c∗]) is in Main Library, Stanford in i′′]]) (g(F)) (g(M)) i′] = (by the deﬁnition of λ-notation) λi′. [[λχ. λ⟨f1, f2,...⟩. λM. λi. (f1)(i) = (if deﬁned) a(x1[+c∗]), (f2)(i) = (if deﬁned) a(x2[+c∗]) . . . λi′. [[λ⟨f1, f2,...⟩. λM. λi. λi′. [(g(f1))(i′), if deﬁned, is the speaker and for all ⟨a∗,i∗⟩compatible with what the speaker knows at i′ and the sequence of counterpart functions g(F), a∗(1[+c∗])) is in Main Library, Stanford in i∗] 3.4 The puzzle, solved Here are the (slightly simpliﬁed) truth conditions assigned to (4) and (5): J(4)Kc,g,a = true iff (g(f1))(ic) = the speaker of c and for all i′ compatible with what the speaker knows at ic s.t. the coin landed tails in i′, the g( f1)- counterpart of the speaker in i′ is in Main Library, Stanford in i′. J(4)Kc,g,a = true iff (g(f1))(ic) = the speaker of c and for all i′ compatible with what the speaker knows at ic s.t. the coin landed tails in i′, the g( f1)- counterpart of the speaker in i′ is in Main Library, Stanford in i′. J(5)Kc,g,a = true iff (g(f1))(ic) = the speaker of c and for all i′ compatible with what the speaker knows at ic s.t. the coin landed heads in i′, the g( f1)- counterpart of the speaker in i′ is in Widener Library, Harvard in i′. Provided that the context supplies an appropriate individual concept, this predicts that both (4) and (5) are true, exactly as we want. Although I won’t go through that, it should be clear that the semantics is equipped also to handle more traditional puzzles, like the one generated by (1) and (2). 3.3.3 Modals (f1)(i) = (if deﬁned) a(x1[+c∗]), (f2)(i) = (if deﬁned) a(x2[+c∗]) ...and ∀⟨a∗,i∗⟩compatible with M(i) and ⟨f1, f2,...⟩, a∗(1[+c∗])) is in Main Library, Stanford in i∗] g(F) (g(M)) i′] = (by the deﬁnition of λ-notation again) λi′. [(g(f1))(i′) = (if deﬁned) a(x1[+c∗]) (g(f2))(i) = (if deﬁned) a(x2[+c∗]) . . . and ∀⟨a∗,i∗⟩compatible with (g(M))(i′) and g(F), a∗(1[+c∗])) is in Main Library, Stan- ford in i∗] = (ﬁxing the value of g and a, and simplifying) λi′. [(g(f1))(i′), if deﬁned, is the speaker and for all ⟨a∗,i∗⟩compatible with what the speaker knows at i′ and the sequence of counterpart functions g(F), a∗(1[+c∗])) is in Main Library, Stanford in i∗] 301 Paolo Santorio Paolo Santorio 4.1 Kaplan-style semantics for the de re If the foregoing is right, there is a monstrous semantics for English indexicals that accounts for their interaction with might, believe and the like. But is it better than alternatives? In this section I set up a comparison between the monstrous theory and what I take to be its main competitor, i.e. the semantics for attitude reports deriving from Kaplan 1968. I focus on a recent version of Kaplan’s semantics proposed by Percus & Sauerland (2003) (see also Percus 2010). I argue that the monstrous theory has an empirical advantage: the Kaplan-style theory, but not the monstrous theory, overgenerates systematically when attitude reports are embedded in attitude reports. The monstrous theory and the Kaplan-style theory share two main features. First, they both employ relations of counterparthood by acquaintance. Second, the Kaplan-style theory, somewhat similarly to the monstrous theory, treats believe as a quantiﬁer over pairs of sequences of individuals and worlds. As a result, the truth-conditions predicted by the two are, at least for a wide range of cases, very close. So the difference lies mostly in the compositional implementation. As is well-known, most versions of the Kaplan-style theory rely on important stipulations just at the compositional level. (For example, the one presented by Cresswell & von Stechow (1982) is non-compositional, and the one used by Heim (1994) and Anand (2006) must stipulate a kind of otherwise unattested syntactic movement.) Percus and Sauerland’s version is smoother and avoids stipulations of this kind, hence it seems a good candidate for setting up a comparison. 302 Modals are monsters (i) Dom(G) = {y: x is acquainted with y at i} (ii) For all y in Dom(G), there is some acquaintance relation R s.t. ere is the deﬁnition: A function G of type ⟨e,ie⟩is an acquaintance-based concept generator for x at centered world i iff: (ii) For all y in Dom(G), there is some acquaintance relation R s.t. – x bears R uniquely to y in i, and – for all centered worlds i′ in x’s doxastic set at i, the center of i′ bears R to G(y)(i′) in i′. (19) I believe that you are on ﬁre. (19) I believe that you are on ﬁre. (ii) For all y in Dom(G), there is some acquaintance relation R s.t. – x bears R uniquely to y in i, and – for all centered worlds i′ in x’s doxastic set at i, the center of i′ bears 4.2 Percus and Sauerland’s semantics for attitude reports The central notion of Percus and Sauerland’s (henceforth, P&S) semantics is that of a concept generator. Concept generators are functions matching objects with counterpart functions (type ⟨e,ie⟩). The basic idea is that indexicals feed their semantic values into concept generators; the latter produce counterpart functions, which in turn generate descriptive truth-conditions for the sentence. Here is how this is implemented. The syntax contains variables Gi ranging over concept generators, as well as abstractors over these variables. The clausal arguments of attitude verbs are just functions from concept generators to centered worlds propositions. For example, the denotation of the that-clause in (1) (repeated below) is given in (20) – x bears R uniquely to y in i, and – for all centered worlds i′ in x’s doxastic set at i, the center of i′ bears R to G(y)(i′) in i′. This solves our puzzle. As an example, here are the truth-conditions of (4 J(4)Kc,g = true iff there is some acquaintance-based concept generator G for the speaker of c in ic s.t. ∀i′ compatible with what the speaker knows at ic and s.t. the coin landed tails in i′, ((the speaker of c)(G))(i′) is in Main Library, Stanford in i′. (20) λG0. λi1. [((G0(the addressee of c))(i1)) (is on ﬁre)](i1) The denotation of believe involves reference to acquaintance-based concept gen- erators: intuitively, these are simply concept generators that match objects with concepts that capture relations of acquaintance.11 P&S’s idea is that believe existen- tially quantiﬁes over concept generators. Here is its denotation: The denotation of believe involves reference to acquaintance-based concept gen- erators: intuitively, these are simply concept generators that match objects with concepts that capture relations of acquaintance.11 P&S’s idea is that believe existen- tially quantiﬁes over concept generators. Here is its denotation: JbelieveKc,g = λφ⟨⟨e,ie⟩,⟨i,t⟩⟩. λxe. λi: There is some acquaintance- based concept generator G for x in i s.t. ∀i′∈DOXx in i, (φ(G))(i′) = 1 This generates plausible truth-conditions for (1): This generates plausible truth-conditions for (1): J(1)Kc,g = true iff there is some acquaintance-based concept generator G for the speaker of c in ic s.t. ∀i′ compatible with what the speaker of c believes, ((the addressee of c)(G))(i′) is on ﬁre in i′. Even though P&S don’t investigate epistemic modals, it is straightforward how to extend the proposal. We can use the following lexical entry for must: 303 Paolo Santorio JmustKc,g = λφ⟨⟨e,ie⟩,⟨i,t⟩⟩. λM⟨i,it⟩. λi: There is some acquaintance- based concept generator G for the individual at i s.t. ∀i′∈M(i) in i, (φ(G))(i′) = 1 This solves our puzzle. As an example, here are the truth-conditions of (4): This solves our puzzle. As an example, here are the truth-conditions of (4): 4.3 The argument from embedded reports Despite their similarities, P&S’s theory and the monstrous theory diverge in an important respect. P&S’s theory introduces counterpart functions via variables that appear in the clausal complements of modals. By contrast, on the monstrous theory counterpart functions are arguments of the modal itself. This difference has an interesting consequence. Consider a sentence with two doxastic modals, like (21) John believes that I believe that you are on ﬁre. The monstrous theory predicts that the range of a shifted indexical is invariably ﬁxed by the closest modal. For example, the range of you in (21) must be ﬁxed by the lower believe. By contrast P&S’s theory is more ﬂexible: the range of you will depend on the coindexing patterns of concept generator variables. The argument I’m about to give is based just on this difference. P&S’s theory predicts more readings than the monstrous theory. But these readings are unavailable, so P&S’s theory overgenerates while the monstrous theory is empirically adequate. In essence, the kind of contrast at stake is the contrast between an extensional and an intensional treatment of a parameter—in this case, counterpart functions. P&S’s theory handles them via variables in the clause, while the monstrous theory handles them via an operator that affects everything falling in its scope (i.e., the monstrous modal). The argument establishes that the extensional theory overgenerates and the operator theory is correct. Start by considering (21). P&S’s theory assigns to it this LF: λi0. John believes that λG0. λi1. (G0(I))(i1) believe that λG1. λi2. ((Gk(you))(i2)) (are on ﬁre)(i2) (i1)(i0) 304 Modals are monsters The crucial issue is the index of the underlined variable. That variable may have index 0: i.e. it may be coindexed with the higher abstractor. If this is the case (21) gets (roughly) the following truth-conditions: it is true iff (a) there is some acquaintance- based way in which John thinks of the speaker (say, the F), (b) according to John’s beliefs, the F believes that the individual that the F thinks of in the way in which John thinks of the hearer is on ﬁre. And this is obviously wrong. To see this, consider the following scenario: John has seen you, an amnesiac, wander around Main Library, Stan- ford, and has come to think of you as the individual I saw wander around Main Library, Stanford. 4.3 The argument from embedded reports He also believes that I have seen some other amnesiac, different from you, wander around the library, that I’ve come to think of him under the description that I would spec- ify as the individual I saw wander around Main Library, Stanford and that I believe that he is on ﬁre. (21) is predicted to be true in the scenario described. But this is obviously wrong: my having beliefs that involve a counterpart function that is also involved in John’s beliefs is not enough to give me a belief about you. (21) is predicted to be true in the scenario described. But this is obviously wrong: my having beliefs that involve a counterpart function that is also involved in John’s beliefs is not enough to give me a belief about you. By contrast, the monstrous semantics handles (21) quite easily. Here is the LF of (21): John believes that [λi1. I believe that [λi2. you are on ﬁre (i2)] (⟨f ′ 1, f ′ 2, ...⟩)(M1)(i1)](⟨f1, f2, ...⟩)(M0)(i0) John believes that [λi1. I believe that [λi2. you are on ﬁre (i2)] (⟨f ′ 1, f ′ 2, ...⟩)(M1)(i1)](⟨f1, f2, ...⟩)(M0)(i0) (⟨f ′ 1, f ′ 2, ...⟩)(M1)(i1)](⟨f1, f2, ...⟩)(M0)(i0) What the semantics does is shift the assignment twice over. So you will be evaluated at the assignments quantiﬁed over by the lower believe. Schematically: J(21)Kc,g,a = for all ⟨a′, i′⟩compatible with John’s beliefs at ic, for all ⟨a′′, i′′⟩compatible with the beliefs of a′(1[+c∗]) at i′, Jyou are on ﬁreKc,g,a′′(i′′) J(21)Kc,g,a = for all ⟨a′, i′⟩compatible with John’s beliefs at ic, for all ⟨a′′, i′′⟩compatible with the beliefs of a′(1[+c∗]) at i′, Jyou are on ﬁreKc,g,a′′(i′′) The truth-conditions will be: The truth-conditions will be: J(21)Kc,g,a = true iff John is acquainted in the way speciﬁed by g(f1) with the actual speaker in wc, and for all i′ compatible with John’s beliefs at ic, the g( f1)-counterpart of the speaker in i′ is acquainted in the way speciﬁed by g( f ′ 2) with the the g(f2)-counterpart of the ad- dressee in i′ and for all i′′ compatible with what the g( f1)-counterpart 305 Paolo Santorio of the speaker in i′ believes in i′, the g(f ′ 2)-counterpart of the ad- dressee in i′′ is on ﬁre in i′′. which seems correct. So the monstrous theory avoids overgeneration.12 5 Conclusion I have argued that there is a monstrous theory that accounts smoothly for the inter- action between indexicals and non-metaphysical modalities. This theory is well- equipped to handle traditional puzzles in the semantics for belief, as well as a new 12 Percus (p.c.) suggests a reply on behalf of the P&S theory. We can assume that a concept generator for an individual x at a world w only yields counterpart functions that are deﬁned at the doxastic alternatives of x at w. If we do this, then (21) ends up having undeﬁned truth-conditions when the lower variable is coindexed with the higher abstractor. This is because, assuming that John and the speaker have doxastic sets that are at least partially disjoint, (Gk(the addressee of c)) is undeﬁned for some i compatible with John’s beliefs. So the problematic reading is blocked. The reply is that one can construct similar examples where the two believers have coinciding belief states. In these cases the problematic reading is not blocked, and the theory still yields a wrong prediction. Here is one example of this kind: Lingens and Lauben are amnesiacs in two different though qualitatively identical libraries. Adjust things in a way that their doxastic states (i.e. the sets of centered worlds that are compatible with their respective beliefs) are completely identical: so e.g. assume that they have identical subjective experiences. Moreover, make it the case that they’re aware that they are in qualitatively indistinguishable situations. (Say, the experimenters tell them that this is the case.) And now, suppose that I am in Lingens’s library, that Lingens sees me, and comes to think of me as the man I saw in the library. And suppose that B.J. Ortcutt (a qualitative duplicate of me) is in Lauben’s library, that Lauben sees him, and comes to think of him under the concept the man I saw in the library. And now consider the sentence Lingens and Lauben are amnesiacs in two different though qualitatively identical libraries. Adjust things in a way that their doxastic states (i.e. the sets of centered worlds that are compatible with their respective beliefs) are completely identical: so e.g. assume that they have identical subjective experiences. Moreover, make it the case that they’re aware that they are in qualitatively indistinguishable situations. 5 Conclusion (Say, the experimenters tell them that this is the case.) And now, suppose that I am in Lingens’s library, that Lingens sees me, and comes to think of me as the man I saw in the library. And suppose that B.J. Ortcutt (a qualitative duplicate of me) is in Lauben’s library, that Lauben sees him, and comes to think of him under the concept the man I saw in the library. And now consider the sentence (22) Lingens believes that Laubeni believes that I am in hisi same library. (22) The case is extremely contrived, but the judgment is still clear. According to Lingens’s beliefs, Lauben has never even seen me, hence he certainly does not believe I am in his same library. So (22) is false. But P&S predict that it has a true reading. The relevant LF is, again, the one where the lower variable over concept generators is coindexed with the higher abstractor: The case is extremely contrived, but the judgment is still clear. According to Lingens’s beliefs, Lauben has never even seen me, hence he certainly does not believe I am in his same library. So (22) is false. But P&S predict that it has a true reading. The relevant LF is, again, the one where the lower variable over concept generators is coindexed with the higher abstractor: Lingens believes that λG0. λi1. (G0(Laubenj))(i1) believe that λG1. λi2. ((G0(I))(i2)) (am in hisj same library)(i2) (i1)(i0) The point is the same as for (21). Both Lingens and Lauben think of someone (respectively, me and Ortcutt) in the way speciﬁed by the function λw. λx. the y that x saw in the library. And, as Lingens knows, Lauben thinks, of the individual that he identiﬁes in the same way in which Lingens identiﬁes me, that he’s in his same library. On the P&S theory, this is enough to give (22) a true reading. 306 Modals are monsters puzzle that was my starting data. Arguably, this theory is also superior overall to its main competitor, Kaplan-style semantics for attitude reports. Should we then reject Kaplan-style semantics in favor of the monstrous theory? Not yet. The argument I’ve given is very partial. At best, I’ve shown how to give a monstrous theory that handles the behavior of indexicals under non-metaphysical modals. 5 Conclusion But Kaplan-style semantics has a much broader coverage: it handles the be- havior of all referential expressions in modal environments, as well as quantiﬁcation into these environments. So a full comparison must wait until the present account is extended to handle names, de re occurrences of descriptions, and quantiﬁcation in modal contexts. But what I’ve said should be enough to establish that the monstrous option deserves to be taken seriously. Let me close by pointing to three topics that deserve further research: the bearing of the foregoing on the semantics of metaphysical modals; the interaction between epistemic modals and attitude verbs on the monstrous picture; the existence of independent evidence for the metasemantic account of free indexicals. References Anand, Pranav. 2006. De de se: MIT dissertation. Anand, Pranav & Andrew Nevins. 2004. Shifty operators in changing contexts. In R. B. Young (ed.), Proceedings from SALT XIV, Cornell University: CLC Publications. Chierchia, Gennaro. 1989. Anaphora and attitudes de se. In J. van Benthem R. Bartsch & van Emde Boas (eds.), Semantics and contextual expression, Dordrecht, Netherlands: Foris. Cresswell, Max J. & Arnim von Stechow. 1982. De re belief generalized. Linguistics and Philosophy 5(4). 503–535. doi:10.1007/BF00355585. Cumming, Sam. 2008. Variabilism. Philosophical Review 117(4). 525–554. doi:10.1215/00318108-2008-015. Heim, Irene. 1994. Puzzling reﬂexive pronouns in de se reports. Handout for Bielefeld conference, March 1994. Heim, Irene & Angelika Kratzer. 1998. Semantics in generative grammar. Oxford, UK and Malden, MA: Blackwell Publishers. Hintikka, Jaakko. 1962. Knowledge and belief. Cornell University Press. Hintikka, Jaakko. 1969. On the logic of perception. In Models for modalities, Dordrecht-Holland: Reidel. Kaplan, David. 1968. Quantifying in. Synthèse 19(1-2). 178–214. doi:10.1007/BF00568057. Kaplan, David. 1989. Demonstratives. In J. Almog, J. Perry & H. Wettstein (eds.), Themes from kaplan, Oxford: OUP. 307 Paolo Santorio Kratzer, Angelika. 1981. The notional category of modality. In H. J. Eikmeyer & H. Rieser (eds.), Words, worlds, and contexts: New approaches to word semantics, Berlin: de Gruyter. Lewis, David K. 1968. Counterpart theory and quantiﬁed modal logic. Journal of Philosophy 65(5). 113–126. doi:10.2307/2024555. Lewis, David K. 1979. Attitudes de dicto and de se. Philosophical Review 88(4). 513–543. doi:10.2307/2184843. Lewis, David K. 1983. Individuation by acquaintance and by stipulation. Philosoph- ical Review 92(1). 3–32. doi:10.2307/2184519. Ninan, Dilip. 2008. Imagination, content, and the self: MIT dissertation. Percus, Orin. 2010. The syntax of de re. Handout for the Goettingen workshop “A Matter of Attitude”. Percus, Orin & Uli Sauerland. 2003. On the LFs of attitude reports. In M. Weisgerber (ed.), Proceedings of Sinn und Bedeutung 7, Konstanz: Universitaet Konstanz. Quine, Willard V. O. 1956. Quantiﬁers and propositional attitudes. Journal of Philosophy 53(5). 177–187. doi:10.2307/2022451. Schlenker, Philippe. 1999. Propositional attitudes and indexicality (a cross- categorial approach): MIT dissertation. Schlenker, Philippe. 2003. A plea for monsters. Linguistics and Philosophy 26(1). 29–120. doi:10.1023/A:1022225203544. Schlenker, Philippe. 2004. Person and binding (a partial survey). Italian Journal of Linguistics/Rivista Italiana di Linguistica 16(1). 155–218. Schlenker, Philippe. 2004. Person and binding (a partial survey). Italian Journal of Linguistics/Rivista Italiana di Linguistica 16(1). 155–218. Schlenker, Philippe. 2008. Indexicals and de se reports. Manuscript available at. http://homepages.nyu.edu/~pds4/. Schlenker, Philippe. 2008. Indexicals and de se reports. Manuscript available at. http://homepages.nyu.edu/~pds4/. References von Stechow, Arnim. 2002. Binding by verbs: Tense, person and mood under attitudes. Unpublished manuscript. von Stechow, Arnim. 2002. Binding by verbs: Tense, person and mood under attitudes. Unpublished manuscript. von Stechow, Arnim. 2003. Feature deletion under semantic binding: Tense, person and mood under verbal quantiﬁers. In M. Kadawaki & S. Kawahara (eds.), Proceedings of the 33th Annual Meeting of the North East Lingusitic Society (NELS 33), . von Stechow, Arnim. 2003. Feature deletion under semantic binding: Tense, person and mood under verbal quantiﬁers. In M. Kadawaki & S. Kawahara (eds.), Proceedings of the 33th Annual Meeting of the North East Lingusitic Society (NELS 33), . Paolo Santorio Department of Linguistics and Philosophy Massachusetts Institute of Technology 32 Vassar St 32D808 Cambridge, MA 02139 USA santorio@mit.edu Paolo Santorio Department of Linguistics and Philosophy Massachusetts Institute of Technology 32 Vassar St 32D808 Cambridge, MA 02139 USA santorio@mit.edu 308

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.